21114 September 2000 *IFC Building Local Bond Markets An Asian Perspective v \ Alison Harwood Editor \ \0 Building Local Bond Markets An Asian Perspective Alison Harwood Editor R09PO ztIFC INTERNATIONAL FINANCE CORPORATION L U A V-be n f 4he WO i d 'dk G p Copyright 2000 International Finance Corporation 2121 Pennsylvania Avenue, NW Washington, D.C. 20433 USA Telephone 202-473-7711 www.i]c.org All rights reserved Manufactured in the United States of America First printing September 2000 The International Finance Corporation (IFC), a member of the World Bank Group, promotes private sector investment in developing countries, which will reduce poverty and improve people's lives. It is the world's largest multilateral organization providing financial assistance directly in the form of loans and equity to private enterprises in developing countries. IFC also gives technical assis- tance, which will promote private sector development. The findings, interpretations, and conclusions expressed in this book are en- tirely those of the authors and should not be attributed to, and do not necessar- ily reflect the views of, IFC, or its Board of Directors, or the World Bank or its Executive Directors, or the countries they represent. IFC and the World Bank do not guarantee the accuracy of the data included in this publication and ac- cept no responsibility whatsoever for any consequence of their use. The material in this publication is copyrighted. Requests for permission to re- produce portions of it should be sent to Director, South Asia Department, IFC, at the address shown in the copyright notice above. IFC encourages dissemina- tion of its work and will normally give permission promptly and, when the re- production is for noncommercial purposes, without asking a fee. Permission to copy portions for classroom use is granted through the Copyright Clearance Center, Inc., Suite 910, 222 Rosewood Drive, Danvers, Massachusetts 01923, U.S.A. This book is based on the South Asia Regional Debt Market Symposium held on October 6-8, 1999, in Sri Lanka. Distributed by the World Bank LIBRARY OF CONGRESS CATALOGING-IN-PUBLICATION DATA CIP applied for ISBN 0-8213-4819-1 Contents Foreword v Editor's Note vii INTRODUCTION & FRAMEWORK 1. Building Local Bond Markets: Some Issues and Actions 1 Alson HIarit'od) 2. A Strategic Priority for Emerging Markets 39 Addaad Jabre 3. Strengthening Sri Lanka's Financial Sector 43 A.S. Jaya i'czrden!t 4. The Risk Management Benefits of Bonds 47 Michael Pettsi 5. The Benefits of Supranational Issuance for Local Bond Markets 59 fainztlz Shah COUNTRY CASE STUDIES 6. Korea's Bond Market Following the Onset of Financial Crisis 63 Yong6qeonz Kiu 7. The Development of a Government Bond Market: The Australian Experience 85 Jo/rn Bro7adbent CON 'EN TS 8. The Development of a Corporate Bond Market: The Malaysian Experience 97 Rztznit Alit Suzgh 9. India's Debt Market: A Review of Reforms 105 UII2a Thorat 10. The Importance of Local Bond Markets in Financing India's Infrastructure Needs 125 Rakewh Afohan 11. India Survey: Issues in Local Bond Market Development 129 John Leonardo 12. Pakistan Survey: Issues in Local Bond Market Development 167 John Leonardo 13. Sri Lanka Survey: Issues in Local Bond Market Development 207 Lennart KtYntgson ci Malin NVystranzd 14. Bangladesh Survey: Issues in Local Bond Market Development 253 7/Jlkae/ KvibYck 15. Nepal Survey: Issues in Local Bond Market Development 267 zlfikae/ Kvzi6ck Contributors 285 * -; Foreword uilding local currency bond markets is an important topic for emerging market countries. This book is part of IFC's efforts to assist countries in South Asia and other parts of the world to identify their need for local bond markets, the impediments to de- veloping them, and how those impediments might be removed. The book is based on papers presented at the South Asian Debt Market Symposium held in Sri Lanka in October 1999. That Symposium brought together key regulators, stock exchange presidents, and mar- ket participants from the five South Asian countries, along with rep- resentatives from other parts of Asia, the Middle East, Australia, and the U.S. Regulators and market participants from South Asia shared and exchanged ideas on how to build their local markets and benefited from the experiences of countries a few steps ahead of them. We hope the case studies included here on South Asian and selected other countries, and the general views on how to build bond markets, will provide valuable insights to emerging market nations wrestling with these same issues today. The Symposium and Symposium papers were financed by Trust Funds provided by the Governments of Sweden and of New Zealand, and the IFC Trust Funds. The project was managed jointly by IFC's South Asia Department and the Financial Markets Advisory De- partment. I wish to thank all those who worked so diligently at making the Symposium a success and in bringing the lessons learned to light through this publication, and the participants whose contributions enlivened and deepened the discussion. -Ra9ad Khal)danzy Directol, South AdItz Department Editor's Note ALISON HARWOOD Since the Asia and Tequila crises of the late 1990s, a growing number of emerging market countries have focused on devel- oping local bond markets to lock in local-currency, fixed-rate, and long-term funding, and help governments and corporations better manage their financing risks. International organizations from Washington to Southeast Asia are pushing bond market develop- ment, to reduce global instability by improving domestic risk man- agement. Both crises accentuated the severe problems borrowers can face when they rely on short-term, foreign-currency funding and take sizable interest rate, refunding, and foreign-currency risk. When the Thai baht's downfall sparked a cascade of falling currencies through- out the Asian region, loans were defaulted on, others were not re- plenished, and many countries had to contend with a far more serious crisis than warranted by any economic problem in the country. The fault lay in large part with the way borrowers Financed themselves. If corporations had funded with bonds, funds and rates would have been locked in, and corporations could have weathered the crisis better. The Asia crisis also showed how quickly problems can esca- late and be transmitted throughout an economy and to other parts of the globe. And it showed that this could happen even when coun- tries have solid macroeconomic fundamentals. These crises should not diminish the other beneFits of local bond markets, such as diversifying financial sectors, allocating capital more effectively, and increasing financial sector competition. Growing ALISON HARWOOD economies need to support major trends such as infrastructure de- velopment, privatization, securitization, and the rise of new institu- tional investors requiring long-term assets to match long-term liabilities. Bonds will play an important role in meeting those needs. To determine the possible role of bond markets in emerging market countries, impediments to their development, and possible ways to remove those impediments, a symposium was held in October 1999 on developing bond markets in five South Asian countries: India, Pakistan, Sri Lanka, Bangladesh, and Nepal. The goal was to get the full range of regulators and market participants from South Asia to discuss these issues together, along with representatives from other Asian emerging market countries a few steps ahead to gain insights into how they dealt with similar concerns. Representatives from dif- ferent parts of the market within the same country and from differ- ent countries throughout South Asia and other parts of the globe exchanged views, learned from one another, and discussed how best to move forward to reach their respective goals. This book is a collection of papers prepared for the symposium and reflects the discussions held at the symposium. It is divided into two parts. Part I addresses why and how to develop bond markets. It begins with a chapter describing a framework for evaluating bond market development (which was used to produce the five South Asian country surveys in Chapters 11-15), and summarizes key points from the symposium papers and the South Asian situation. It is followed by presentations on the importance of bond markets from IFC Vice President Assaad Jabre and Sri Lanka's Central Bank Governor, A.S. Jayawardena, then presentations on the risk-management ben- efits of bonds and the benefits of supranational issuance. Part II consists of country case studies. The first three are from outside South Asia-on developing corporate bond markets in Malaysia, creating more market-oriented government and corporate bond markets in Korea, and how Australia created a highly liquid government bond market. The remaining chapters cover bond market development in India, Pakistan, Sri Lanka, Bangladesh, and Nepal. They include two chapters on India by senior Indian officials, the first by Usha Thorat of the Reserve Bank of India on actions taken to promote India's markets, followed by the five South Asian country surveys. :Vi:I EDITOR S NOTE The book is a start at looking at the key issues in an increasingly important topic. More work is needed to suggest how countries might deal with the various problems they face in developing their bond markets -how to move forward and transition from an emerging to an emerged market, what works best in what types of situations, and how to evolve in a continually changing technological and glo- bal environment. These papers reflect the views of their respective authors and not necessarily the views of the International Finance Corporation (IFC). I would like to thank all the Symposium participants, all of whom actively contributed as speakers, roundtable panelists, and/ or working group members. Together, they created an engaging, informative Symposium and the insights and materials for this book. I would also like to thank Peter Taylor from IFC for his excellent assistance in getting this book through the publication process. - ix CHAPTER 1 Building Local Bond Markets: Some Issues and Actions ALISON HARWOOD D eveloping local currency bond markets has become a much- talked about topic, particularly since the recent Asia crisis. A growing number of emerging market countries around the globe are looking into the prospects of building local bond markets to reduce the currency, interest rate, and funding exposures that precipitated the Asia crisis. This chapter provides a basic framework for evaluating the ma- jor impediments to developing local corporate bond markets and some suggestions on how those impediments might be removed. This approach was used to evaluate the situation in the five South Asian countries, and the chapter briefly reviews the main conclusions of those surveys (Chapters 11 through 15). The paper points out the many benefits of bond markets and the importance of developing them. But it also shows that building bond markets is difficult and takes a lot of time, and not every country will be able to develop active markets. Markets grow from partici- pation by issuers, investors, and intermediaries -not just from build- ing market infrastructure. Participation results when a fairly comprehensive range of economic, technical, as well as political and "behavioral" factors come together. Unfortunately, many of these factors, by definition, are not well developed or are inappropriate in emerging markets and take considerable time to get into place. Many of them are controlled by parts of the government that may care ALISON HARWOOD little about developing bond markets, which further complicates the process. In addition, as is true when building any financial market, these factors develop at different paces, leaving countries to try to grow their markets in less then ideal conditions. The paper notes that coun- tries likely will have to introduce their market in stages-starting with privately placed transactions and graduating over time to pub- lic primary markets and then to secondary markets as conditions improve. Many countries will never be able to develop active sec- ondary corporate bond markets and many may have only limited primary markets as well. But borrowers and investors in these coun- tries still will gain many benefits from access to some form of bond market. Regulators and market participants need to evaluate this com- prehensive set of factors to determine which ones constrain their market's growth and how to deal with them. Market development will be accelerated if regulators who are interested in the market's development work closely with market participants to identify prob- lems and solutions, and with other regulators to persuade them to address problems under their control. Although there is no one way to build a market, countries should draw on the experiences of other emerging market countries for guidance on how to progress from "emerging" to "emerged" and on what works best in what types of conditions along the way. All five South Asia countries could benefit from having long- term, local-currency, fixed-rate financing for their local corporations. As might be expected, they differ in their ability to develop such financing. India is farthest along on the path toward active primary and secondary markets, followed by Sri Lanka. Both countries are actively moving to develop their markets. Pakistan too is now tak- ing steps forward. Bangladesh's securities markets are still relatively undeveloped, as are those of Nepal.' l The surveys are based on information available in mid- to late- 1999. Ex- change rates on 31 August 1999 against US$1 were: Indian Rupees, 43.5; Paki- stan Rupees, 51.8; Sri Lanka Rupees, 71.9; Bangladeshi Takas 49.5; Nepalese Rupees, 68.7. BUILDING LOCAI BOND MARKETS Developing local corporate bond markets is a relatively new activity for many emerging market countries, and insights from ex- perience are limited. This paper offers some general suggestions on how to deal with key problems. More work is needed to distil spe- cific ideas on how countries with particular circumstances might move through the development process. BENEFITS AND GOALS The current emphasis on local-currency bond markets stems mainly from their risk-management benefits, as highlighted by the Asia and the Tequila crises. Issuing bonds can reduce the types of interest rate, foreign exchange, and refunding exposures that created those crises and can help ensure that emerging market borrowers have more shock absorbers-more tools-to limit the impact of those exposures. Foreign investment is clearly a plus for economic development but it does create certain risks. Since financial sector crises will never be eliminated, and, at least for many years to come, flows into emerg- ing markets will be large in relation to the markets in which they are investing, any rapid outflow will create serious problems for the bor- rowing country. Emerging market countries must find ways to man- age the risks, and hence benefit from international capital flows. They need to be able to reduce exposures to foreign-currency borrowing and also absorb the associated shocks and volatility, so that small problems will not escalate into broadly based social catastrophes, harming people who were in no way directly involved in the markets. As Michael Pettis discusses in chapter 4, local-currency bonds dampen the effect of crises created by international capital flows by locking in interest rates and local-currency funding. This allows borrowers to hold on to their funds and positions and work their way through a crisis. But, as happened in Asia, many borrowers want to rely on short-term, foreign-currency funding because when their economy and local currency is strong, such borrowing creates a double benefit to their net worth: the borrower's liabilities fall while its assets and revenues rise. The flip side is that when times turn bad, borrowers get a double hit on their net worth: liabilities rise 3 ALISON HARWOOD and assets fall, causing strains and in some cases defaults. The solu- tion to this problem is to use funding structures that have a neutral effect on net worth, as in the case of bonds. The difficulty lies in convincing borrowers that good times may turn bad, and in getting them to incur the potential opportunity cost from locking in stable funds and rates. Chapter 6 by Yongbeom Kim and chapter 8 by Ranjit Ajit Singh make clear that the Asia crisis was a major impetus for developing local-currency bond markets in Malaysia and Korea. Korea had a local-currency corporate bond market for years, but it did not play a stabilizing role during the crisis because it was not used exten- sively enough. Corporations relied mainly on bank loans. When Korean banks were hurt in the crisis, says Kim, the resulting credit crunch precipitated corporate bankruptcies. Korea is now working to develop a larger, more market-oriented, corporate bond market, because the government recognizes how necessary that market is to financial and economic health. It is also aware that the government securities market must be active and market oriente(d to support the corporate bond market. Actually, emerging market countries have been developing local bond markets for the past 5 to 10 years, though at a slower, less focused pace. Any country that is liberalizing and growing economi- cally needs diversified financing tools beyond just banks and equity markets. Banks often cannot provide the size or structure of financ- ing needed. In many countries today, banks are increasingly con- strained from financing longer-term, large-scale projects because they are trying to improve the quality of their operations. Pressure from international agencies to contain credit extension (for example, via legal lending limits or provisioning for nonperforming assets) is lim- iting lending as well. Issuing new equity is not always an option, as it is costly and dilutes ownership. Local bond markets also support major trends that stem from economic and financial sector growth. For issuers, infrastructure development is creating demands throughout Asia and other parts of the world for large-scale, longer-term funds that banks cannot often provide. Privatization, securitization (particularly for housing finance), and decentralization of governments are all creating new financing demands. On the investor side, many countries are now 4 BUILDING LOCAL BOND MARKETS rich enough for insurance and social security and are creating insti- tutional investors that need long-term assets. They want to keep their interest rate (fixed), reinvestment (long term), and local-currency risks to manageable levels. With macroeconomic stability increas- ing in many countries, issuers and investors alike are more willing to lock in rates. Local bond markets also strengthen the financial sector by en- couraging greater transparency, pushing companies to disclose in public markets and forcing them to better understand themselves and in turn improve their management (as is the case in equity mar- kets, too). Bond markets create competition with the local banking sector, which can reduce lending rates. Ideally, countries should try to build both primary and second- ary markets for bonds. Primary markets reduce the three risks noted; secondary markets, by adding liquidity and broadening the investor base, help reduce funding costs. As discussed below, many coun- tries will not be able to create secondary markets, and some will find it hard to develop public primary markets. Whatever the situation, reducing one or two of the three financing risks is worthwhile. Get- ting local-currency, fixed-rate, long-term funds in a private place- ment may cost more than a publicly traded issue but it might be all that a country can do, and will reduce the issuer's risk, and allow the investor to lock in an asset. BUILDING BOND MARKETS Equity market history offers a few important lessons about how fast a country can build its bond markets. In several cases -despite op- timistic expectations, bursts of activity, and establishment of a new equity market infrastructure-little or no issuance or listings fol- lowed, or there were listings but no trading. One reason for this is that many persons involved in building capital markets have operated as if creating markets is a technical, top-down, infrastructure-building exercise. You set up regulations and regulators, incorporate exchanges, introduce trading and clear- ing systems, and give regulators and market participants some edu- cation. Once the systems are plugged in and doors opened for business, ~~5 ALISON HARWOOD a market will spring to life.2 But this "build it and they will come" approach often does not work. Doors are opened, systems are turned on, and markets do not operate. Instead, there is often "no prod- uct"-no issuers-and no investors or intermediaries to transact. No market participation means no market. Some General Issues Generally speaking, issuers, investors, and intermediaries will par- ticipate in a market if they see an economic benefit (better costs, better structures), are willing (have the right attitudes) and able (have the skills, regulations), and are structured right as an industry to participate (see table 1). Conversely, there are lots of reasons why they might not participate. Clearly, the three elements of need/ben- efit, willingness, and ability drive one another. If the benefits are clear and significant, participants will be more willing to do "costly" activities like disclosing information. If better skilled, they will be less fearful and more willing to enter the market. Market participation cannot be declared or forced, but it can be encouraged by an enabling environment, and it can be discouraged by an "unabling" environment. The environment consists of a range of interactive factors around the market, across other parts of the financial system, and inside the market (see figure 1): Aroundthe Market. An enabling environment consists in part of macro and political stability, including economic growth that generates a sufficient number of issuers, inflation and interest rate structures that are not too high or volatile, tax policies that do not disadvan- tage use of bonds, and a broader legal framework (securities laws, bankruptcy codes, and the like) that supports bond markets. Acroas Other Parts of the Financial System. Ideally, a government securities market is present or in the making and is helping to build a benchmark yield curve and a dealer community. The equity mar- ket is relatively well developed, with a stock exchange and clearing 2 Building market infrastructure is a foundation for any market, but it is just a foundation. 6EW- BUILDING LOCAL. BOND MARKETS Table 1. Encouraging Market Participation Category Comments Instrument Attractiveness Need Issuers and investors get new features they cannot get elsewhere such as larger volumes, quicker access, and better maturities Intermediaries get new business line. Economic benefit The product provides a tinancial benefit such as reduced costs for issuers, higher returns for investors, new profit sources for intermediaries. Individual Participant Ability Market Participations can participate in the market (i.e., Features investors are not unduly constrained by statutory liquidity requirements, investment directives, or an inability to take positions in the market; intermediaries by unnecessarily high capital requirements that reduce protits, inability to tinance or hedge their positions, inability to take positions; issuers by too high and costly disclosure requirements). Issuers, investors, intermediaries have the skills needed to perform the business Willingness Issuers are willing to disclose because they see an economic benefit to being in the market. Investors are willing to take risks and trade. Industry Structure Economy provides issuers and investors that are diversified, large enough to support the market, not so big that they dwarf the market, in sufficient number to create competition but not so many that they make it impossible for any one firm to make money (i.e., the number of entities in the market fits with the size of the market). corporation, and securities firms that are familiar with securities markets. The banking sector can contribute to the market as inves- tors, issuers, and intermediaries. Indide the Market. Regulators support and are committed to devel- oping the market. Trading, clearing, and settlement systems exist for equities and might be modified to support bonds. Issuers form a large enough core, are profitable enough to attract investors, and are willing to disclose information. Intermediaries are present who are capable of dealing in securities. There is a growing institutional investor base that is not captive in other markets (or any captivity that is present can be removed). Privatization programs are in place that create issuers who need bond markets. 7 ALISON HARWOOI) Figure 1. Bond Market Environment Around *Political *Macro Lims & RegfXt atlons Regulators * ~~~~~~~~~~~Taxation Central Market Infrastructure X Trading CSO tt *Flexibl Wey ll CDst Effective * Reliable Accesible * Cost Effective tt Issuers } ~~~~~Investors , All these factors influence the market's attractiveness to an is- suer, investor, and intermediary and hence need to be considered from the market participant's point of view-that is, from the bot- tom up. In addition, because the factors interact in different ways, they need to be examined simultaneously. Finally, because these are emerging markets, different components will be developing at dif- ferent rates. They will rarely be in place as an entire set. That means the ultimate evaluation is something of a balancing act. Sometimes one important element is strong and another is weak. Leaders will have to weigh the impact of different factors. Recognizing that market participants are effected by such a broad range of factors makes it easy to see why developing financial mar- kets is so difficult and time-consuming. Moreover, many of the fac- tors go beyond technicalities and involve attitudes, cultures, and politics3. Some matters take years to get right, such as corporate 3Developing new attitudes and behavior is a main reason why developing markets take so long. People need to learn new ways of thinking about how and with whom they do business. New cultures need to be created -for taking and managing risk, disclosing information, and maintaining quality operations and accounts, among other areas. 8 BUILDING LOCAL BOND MARKI'TS governance, which influences an investor's willingness to invest in a company, particularly its bonds, since they involve credit risk and debt servicing. Equally important, many of the factors affecting market partici- pation are outside the control of entities that may want to build the market, such as the local Securities and Exchange Commission (SEC). Tax policy may be run by the Ministry of Finance, macro and bank- ing policy by the central bank, insurance and pension rules by their respective regulators. The assumption that building market infra- structure builds markets is appealing because the parties who want to build it can control that process. It is a technical activity, involves a limited number of parties, and so is relatively easy to undertake. By contrast, an SEC that wants to develop the corporate bond mar- ket may be faced with a central bank that insists on keeping interest rates high to maintain foreign exchange rates, which prevents bond issuance. Indeed, market building will take a long time and will be "noisy." With so much going on and so many entities involved, market activ- ity and capabilities will get out of balance, causing growing pains. Leaders will constantly be challenged to balance activity and capa- bilities to avoid explosions that destroy market confidence and cre- ate "participation scars" that are hard to erase in new markets. This is a reality of the activity. Remember, too, that developing bond markets can be more com- plicated than developing equity markets. Bond markets need sup- porting pricing infrastructure. They operate best when they have money market and longer-term benchmarks. Most emerging mar- kets lack these benchmarks. The issuer's credit risk is another ma- jor concern. The issuer has to service and repay the bonds, whereas with equity the issuer can be "incubated" from payments as it grows. Investors need to make sure issuers have the cash flow to make interest payments and redeem principal. Bond markets simply can- not grow as quickly as equity markets can. Furthermore, bond mar- kets need more sophisticated market participants. Issuers need to be able to manage their cash flow to make repayments. Bond mar- kets typically need dealers and market makers, which means creat- ing a new class of intermediaries who can take positions and manage their risks. 9 ALISON HARWOOD All the pieces of this puzzle are important, but a few are "key factors for success" while others are "second-level" success factors. (Although market infrastructure is critical to the process, it is as- sumed here that the infrastructure is functioning eff'ectively or can do so.) Key Success Factors Around, Across, and Inside the Market Whether a market can be built depends on four key factors-two from "inside" the market and two from "around" it. The "inside" factors are suitable and appropriate issuers, investors, and to a lesser extent intermediaries, and a committed government. The "around" factors are macrostability and taxation. For most countries, the lack of appropriate issuers and investors is the main stumbling block to developing the market. MarketParticipantd. Clearly, a market needs issuers and investors. As obvious as this sounds, a surprising number of countries have pushed forward to build markets despite the lack of these players. An active primary and secondary market needs a diversified is- suer base with varied credit risk representing different economic sectors. Potential issuers can include corporations, financial institu- tions (banks, housing finance), infrastructure projects, and munici- palities. Financial institutions (FIs) are often the biggest nongovernment issuers in the early stages.4 As noted, issuers need to see some economic benefit to the product, must be willing to use it, and must be able to use it. Investors should be diversified and composed of institutions such as pension funds, insurance companies, mutual funds, and other FIs, interested in different credit risk and economic sectors, and not so large that they dwarf and dominate the market but large enough to take positions and risks. They too need to see economic benefits, such as higher returns to compensate for longer-term investments, and instrument structures or maturities that better match their li- abilities than other products. They need to be willing to be in the 4 Tadashi Endo, "Corporate Bond Market Development," unpublished paper, 2000. 10 _ - BUILDING LOCAL BOND MARKETS market (take risks) and able (through skills and regulations). How diversified, sizable, and capable the issuers and investors are will determine how fast and large the market can grow. Intermediaries are also needed to bring issuers and investors together.5 They need to make money from the business, be willing to take and manage the risks (being profitable will help), and be able, through regulations and skills, to do the business. They need to make enough money in good times to support the ups and downs of the market, which can be accentuated in emerging markets. As for the industry, there should be enough firms to create competition but not so many that no one firm can make enough money. The industry should not be dominated by banks but should include several inde- pendent securities firms since banks can constrain the operations and perspectives of their securities affiliates. Government Commitment. "Inside" regulators (such as the SEC) must be committed to building the market. In many cases, the govern- ment needs to take the lead in getting the process under way by bringing together key market players to build systems and by lob- bying "outside" regulators and leaders. In Korea and Malaysia. the government made bond market development a priority and took steps to make it happen.6 The level of commitment will determine how fast the market grows. Without "inside" government commitment, the market is not likely to grow. Macroeconomic Stability azd Credibility. Bond markets require stable macro and political environments to grow. Economic growth must be strong enough to generate appropriate issuers and investors; in- Development of local bond markets is causing many emerging market coun- tries to introduce dealers and market makers to their capital markets, because the bond markets need firms that can take positions with their own capital rather than just broker for others. See the discussion on liquidity below. 6 The Malaysian government gave several tax incentives to encourage market development. It waived the stamp duty and exempted from tax the interest earned by individuals on corporate honds and then the income earned by unit trusts and closed-end investment funds. These efforts helped the market move forward. Such measures are not always needed (and are often not even suggested since they can create distortionary tax structures that are harder to remove later). 11 ALISON HARWOOD flation and interest rates cannot be too high or volatile. Without sufficient GNP growth, savings and investment rates, and per capita GNP, the economy might not provide the issuers and investors needed (see table 2). Korea, Thailand, and Malaysia, three countries that are promoting their local bond markets, have sizable GNP growth rates and very high savings rates (34%, 36%, and 44%, respectively). High interest rates can slow issuance by creating high and unaffordable costs. If short-term rates are 25%, what company can pay 40% for 5- or T O-year money? Companies may prefer or be forced to take interest rate and refunding risks. Volatile rates will stop issu- Table 2. Macroeconomic Statistics for the United States, United Kingdom, Southeast Asia, and South Asia (percent) Savings rate Budget defcitP Inflationd Interest Country Real GOP growtl? (%of GOP) (% of GDP) (CPi) fate South Asia Bangladesh 5.7 14.7 563 8.3 8.26 India 6.0 20.0 4.9 10.5 8.61 Nepal 2.3 10.0 4.1 10.0 3.40 Pakistan 4.3 10.4 7.9 6.2 8.37 Sri Lanka 4.7 17.3 4.5 9.4 12.61 Developed France 2.6 19.7 3.5 0.6 2.98 Germany 1.3 22.4 1.4 0.6 2.85 Italy 1.3 22.3 3.1 1.6 2.90 Japan 1.3 30.5 1.5 -0.3 0.03 United Kingdom 1.8 15.1 5.3 1.5 4.86 United States 3.9 16.0 0.3 2.2 4.70 EmergingAsia Korea 8.0 34.2 1.4 1.2 4.81 Malaysia 5.5 44.4 -3.0 2.9 3.10 Thailand 5.5 35.7 0.9 0.2 1.48 a. Real GDP growth for South Asia is for l998 from the Economist Intelligence Unit. Real GDP growth for the developed countries and for emerging Asia are JPMorgan forecasts tor 1999. a. Gross domestic savings for 1997 (1996 for the six developed countries) from World Development Indicators. c. Overall budget deficit including grants, as a percentage of GDP for 1997 (1996 far the United Kingdom; 1993 for Japan) from World Development Indicators (the deficit for Bangladesh is trom Bangladesh Bank). Numbers for Bangladesh are not available. The negative figure for Malaysia indicates an overall budget surplus. d. Inflation for South Asia is consumer price index (CPI) inflation for 1998 from the Economist Intelligence Unit. Inflation for the developed countries and for emerging Asia are JPMorgan forecasts of CPI inflation for 1999. e. Figures are from international financial statistics, except those for India, which are from the Reserve Bark of India, and for Bangladesh, which are from Bangladesh Bank). They represent short-term treasury bill rates (except in the case of Pakistan, Japan, Korea, and Thailand, where they are the money market rates). 1,2 BUILDING LOCAL BOND MARKETS ers and investors from locking in rates. Both will hope that over time rates will move in their direction. At present, short-term rates in the industrialized countries and in Southeast Asia are generally below 5% (see table 2). Governments also cannot crowd out the private sector from lo- cal and foreign investment. Deficits to GNP combined with sav- ings-investment gaps provide some indication of whether this is a problem. Taxation. Taxation is a well-known potential market destroyer, di- recting financial flows by changing relative costs of different prod- ucts. Bonds do not need preferential treatment, but they cannot operate at a disadvantage compared with alternative products such as bank loans and equity or they will not be ablc to compete. Close attention must be given to the effect of stamp duties, transaction taxes, and income taxes on the cost of issuing, investment returns, and intermediation profits. Second-Layer Success Factors Some countries may have a sufficient issuer and investor base, macroenvironment, and government commitment but are constrained by the lack of development "across" the financial system, in the gov- ernment securities market, banking sector, and equity markets. These factors will affect how fast and how far the market can grow. Government Securities Markets. The government securities market is an important foundation for the corporate bond market. It pro- vides a benchmarkyield curve and helps promote a class of dynamic, profitable fixed-income dealers.! A government benchmark is ideal As noted in several chapters in this volume, to create a benchmark, the government ideally must issue at market-oriented rates, have large issues with representation across the yield curve, maintain current coupons by issuing fre- quently or through trading, have broad distribution of bonds among investors, and have frequent public announcements about issuance plans so investors stay informed about the supply and demand in the market. 13 ALISON HARWOOD as few alternatives usually are available in emerging markets. Inter- est rate structures tend to be skeletal, particularly after one-year maturities, and few high-quality credit alternatives exist. Government markets also provide dealers with experience trad- ing fixed-income securities (as these markets are likely to have some trading), and a chance to earn profits and build credibility as an intermediary, which helps in obtaining better structured and priced financing from banks and financial markets.' Brokers in emerging markets can have difficulty getting bank funding for various rea- sons, which they need since other typical funding tools such as re- purchase agreements (repos) are often limited. Equity and Money Markets. The existence of an operating equity market is important for bond market development because it im- plies that the country has a "capital markets culture" with support- ing institutions, issuers with disclosure experience, and investors with some understanding of what it means to invest in securities. Money markets can provide short-term pricing benchmarks and offer deal- ers less risky trading experience (because potential losses on a money market trade will be smaller than with longer-term paper). Banking System. Banks support bond issuance indirectly, when their lending is constrained (by capital adequacy, legal lending limits, NPAs, and so on), and directly by acting as issuers, investors, and interme- diaries.9 They often dominate bond issuance in a market's early stages. They are frequent investors, though they can stall market develop- ment by buying and holding securities. Because they are the best- capitalized financial institutions in emerging markets, banks tend to Manyemerging market countries-including Malaysia, Korea, Hong Kong, Thailand, India, Sri Lanka, and Egypt- are introducing primary dealers in their government securities markets. These dealers often receive funding benefits which can help support other business lines directly and indirectly. 9 In many countries, the Bank for International Settlements (BIS) capital adequacy requirements, which were not intended for emerging markets and often were implemented rapidly, left banks scrambling to reduce loans and substitute low-risk weighted assets such as government securities to meet the requirements. Corporations needed to find new funding sources. 014-- BUILDING LOCAL BOND MARKETS act as bond market intermediaries."0 Their ability to deal in the mar- ket is important since their affiliates are usually the primary dealers in the government securities markets, which trains them to deal in corporate bonds. Banks can seriously hurt bond market growth if they are given preferential tax or regulatory treatment to protect them from bond market comnpetition. Credit-RatingAgencies. "Inside" the market, credit-rating agencies (CRAs) face problems that can also have an adverse impact on bond market growth. As is well known, CRAs need to be credible, inde- pendent, and able to obtain information if they are to function prop- erly. They also need to be profitable or they will not survive." This means CRAs need enough deal flow to earn profits or they will have to charge high fees, which will deter bond issuers. Because they need deal flow, but also need to encourage new issues, there is often a question of when to introduce a CRA. Creating Market Liquidity To this point, the discussion has focused mainly on primary mar- kets. Secondary markets are important for well known reasons, but they do not necessarily emerge from primary markets and in most cases do not. Active trading is difficult to get. Outside the United States, corporate bond trading is not sizable. In many countries, whatever trading exists is usually concentrated among a few larger issues, ratlher than spread across a range of issues. Yet mar-kets need enough trading for price signaling and to attract a broad investor base. Dealers and market makers can play a critical role in promot- 10 Some emerging market countries are requiring that the bond business be done in a separately capitalized subsidiary, to ensure business focus and commit- ment. Banks often own and capitalize those entities. l According to Standard & Poor's, to sustain a credit-rating system, a market needs a critical mass of issuers (for research, default prediction, investor aware- ness of credit risk differentials, range of investment opportunities), a professional financial intermediary industry, a broad institutional investor base sensitive to credit risk and pricing, and a sound regulatory environment and market infrastructure (that is, accounting, other information, market-based benchmarks). See unpub- lished report concerning the Egyptian Capital Markets, September 1999. 15 ALISON HARWOOD ing market liquidity, and many emerging markets are introducing these kinds of intermediaries when creating bond markets. But dealers alone cannot create liquidity. They need the right trading tools, an underlying investor base, and the ability to make money, finance, and hedge their positions. Experience to date has shown that bond markets usually work best with quote-driven trading, particularly when the bonds are not liquid, because bonds are traded on the basis of many variables - interest rate levels and payment structures, maturity dates, and the like. Negotiations often are needed to find the bond with the right characteristics. In addition, issues should be sizable and issued fairly frequently. Dealers need financing to buy securities, finance securities held in inventory, and help settle brokered transactions-mainly short-term financing, but in sufficient amounts and at reasonable prices. Fi- nancing that is too difficult, unreliable, or expensive constrains the dealer's ability to trade. In more developed markets, dealers typi- cally finance their activities with long-term capital, repurchase agree- ments, and bank lines. Dealers also need to be able to hedge their positions, particu- larly if they are official market makers required to quote "selling" positions to make a market. Liquid treasury bonds, repurchase agree- ments, short selling, bond-lending facilities, swaps, forwards, futures, and other derivatives are all potential hedging mechanisms. Problems and Solutions Some of the problems emerging markets face are technical and can be addressed if the government is willing to tackle them. Others are more structural in nature and often involve elements that are out- side the control of the "inside" regulators, such as constraints on the issuer and investor base and macro-instability. These are more diffi- cult to resolve and take more time to fix. Each country will have a different combination of pluses and minuses to address. Key Succeds Factors. The key to success is to build market participa- tion, secure government commitment, ensure macrostability, and eliminate tax disadvantages. 16 BUILDING LOCAL BOND MARKETS Buiilduig niarket particittionz. Issuers can be deterred from participat- ing in the market by a lack of sufficient economic benefits, often due to regulations. The disclosure process may be too costly and onerous, the approval process may take too long (the SEC may ask for too much information and lack the skills to evaluate it), registration may be expensive, and/or listing costs may be high. When economic ben- efits are limited, issuers may be unwilling to lock in long-term rates. While locking in rates reduces interest rate risk, it has potential in- terest rate opportunity costs that issuers may be unwilling to "pay." Many of these problems can be dealt with by changing regula- tions in ways that do not sacrifice prudential standards. Regulators can keep regulations simple to make them easier for the regulated entity to comply and reduce approval delays caused by the regulator's lack of experience. Removing unnecessary information disclosure requirements and consolidating approval processes under one regu- lator can speed up the issuance process, as can shelf registration by allowing companies to update existing information when making new issues. 12 A more difficult issue, though, is the lack of sizable and profit- able issuers. Equity markets can also face this problem,13 but it can be worse in bond markets because bonds need to be serviced and repaid, which means issuers have to be of a higher credit quality. Not every firm can meet periodic and ultimate obligations of a bond. In most countries, issuers will have to come from the top credit category. Dropping below that level to build the issuer base can be a problem if the country does not offer enough diversified instru- ments and credit risks or some form of credit guarantee for the bonds. Local investors will be assuming risk they cannot diversify because of limited investment opportunities. Some argue that issu- 12 Singh (chapter 8) notes that approval in Malaysia was taking four to six months because the process was fragmented across a number of entities-the cen- tral bank, registrar, and stock exchange if the bonds were listed. That is being rationalized and will likely end up under the SEC. 3Many equity markets have hundreds of listings (usually from privatizations) and no capital raising or trading because the listed companies are not attractive to investors-because of poor corporate governance, lack of profits, lack of trans- parency, or other management and economic reasons. 17 ALISON HARWOOD ers should be allowed to issue with ratings that indicate their lower credit quality and let the investor decide/beware. This approach may work in particularly more experienced countries, but often investors buy the higher risk to get the higher return without fully understanding the risk/return trade-offs. Other times, they refuse to buy lower-quality paper, regardless of the return. It may take time to get to the point where lower-quality paper can be sold to knowing investors.'4 In the meantime, countries will have to bal- ance the desire to grow the market with the credit quality constraints of the issuer base. Issues by foreign firms and supranationals can provide better quality paper to the market and help diversify credit quality. They also can create a demonstration effect (see chapter 5 by Mamta Shah), especially for foreign investors, and help introduce international stan- dards and best practices to the market. They help establish a risk- free benchmark and provide a foundation off which other issuers can extend the yield curve. Shah argues that supranational issues will not necessarily crowd out local issues, but will likely enlarge the pie, allowing investors to diversify their portfolio risk, creating an appetite for local lower-rated bonds, and increasing investor com- fort with local-currency investments. However, not all countries will have the option of letting supranationals issue in their local-currency markets, mainly because issuers will not be able to swap out of the currency. Developing the investor base is one of the biggest impediments to market growth. Many countries do not have institutional inves- tors, particularly pension funds and insurance companies, or are just introducing them. Many countries only have a handful of very large, conservative, and relatively inexperienced state-owned financial in- stitutions, such as employee provident funds and insurance compa- nies, that dwarf the market. In addition, as seen throughout South and Southeast Asia, often institutional investors do not buy corporate bonds because their money 14 Malaysia dealt with this problem by requiring issues to be rated (to create a credit-rating culture) and then only allowed investment-grade firms to issue. Now that its market is maturing somewhat, it is considering removing this restriction. BUILDING LOCAL BOND MARKETS is tied up in government securities, which they buy and hold be- cause of statutory liquidity requirements, or because they are afraid of taking credit risk. They may lack incentives to buy corporate paper because portfolio managers are penalized for incurring losses but are not rewarded for making gains; they are not judged against the market (that is, by whether they provide the "best" return); or there is no marking to market and unrealized gains and losses are not rec- ognized. In other cases, whole ranges of investors are not serviced because distribution channels are weak or intermediaries do not know how to cultivate investors. Some of these problems can be dealt with by providing investors with incentives to buy and trade securities, such as marking securi- ties to market and evaluating portfolio managers against performance- based measurements. Requiring investors to mark the securities to market reduces incentives to hold paper, but may cause sizable losses for some investors, and so may have to be phased in over time. Again, many of these problems depend on where a country is in developing its institutional investor base. Unless some move has been or is being made in this direction, market growth will be severely constrained. Retail investors alone usually cannot support the market's growth. Countries often take a two-pronged attack to building this base. They improve operations at existing (often state-owned) insti- tutions, implementing rules and incentives to encourage more pro- fessional management. At the same time, they encourage new private sector, professionally managed mutual funds and pension and insur- ance firms as this change usually has a quicker and stronger impact then the other. Of the problems intermediaries face, one of the most difficult is how to make sufficient profits in the market, because capital require- ments or financing costs may be high and revenues low (from low fees/spreads or low volume). Too many firms may be competing for too little business, splitting volumes and fees so no one can make money. Intermediaries may lack the skills to attract issuers and in- vestors, and the tools to hedge market-making positions. Some coun- tries allow only separately capitalized subsidiaries to conduct bond market activity. Some can only handle government securities, or corporate and government securities together, but not mixed with equity. This approach is thought to ensure focus, commitment, and 19 ALISON HARWOOD sufficient capital for the fixed-income business. But it can increase costs and make a business unprofitable. Regulatory changes by "inside" regulators can help overcome such problems. The economic benefits of the business can be im- proved by having minimum and net capital requirements that meet prudential standards but are not unduly onerous. Many countries use a minimum capital requirement to "select" better capitalized firms for the business, but the amounts are not too high, and then a net capital rule that allows capital to grow with business activity. Some- times the cost side is fine but the revenue side-underwriting fees, trading spreads, other fees-is too low. (Market makers need to be able to make spreads that compensate them for their risk.) Potential problems can be identified by creating an income statement for a typical securities firm to determine a firm's ability to profit in the business, and then for the entire industry to deterrnine how many firms can survive given revenue and costs structures and anticipated market-wide volumes for different business lines. Another issue affecting economic benefits concerns how bond market intermediaries should be structured. A diversified securi- ties firm that can conduct all facets of the securities business -pri- mary and secondary markets for all debt and equity products, is most efficient since it reduces overhead costs and lets firms use their limited skilled personnel across a range of businesses. Most developed countries have diversified securities firms. Moreover, the skills and profits earned in the government securities business helps support the corporate business, and the two should not be sepa- rated. Mixing in the equity business can be handled with strong Chinese walls and appropriate supervision. When supervisors are weak, regulators can opt for dedicated capital for the fixed-income business. As discussed later, these structural problems-which boil down to having enough of the right types of issuers and investors-take time to resolve. Countries may find that they simply have to grow slowly while these areas are being addressed. Goi'ern,nent comnnit,nent. Often the "inside" government is not con- vinced that building corporate bond markets is important, or it may want to build the market but cannot persuade the tax authorities or 20 BUILDING LOCAL BOND MARKETS the central bank to change policies that impede the market's growth. In some respects, lack of government commitment from "inside" regulators, and perhaps "outside" regulators as well, is less of a prob- lem today because the crises in Asia and Latin America have moti- vated more countries to develop these markets. To develop a bond market, market participants (and advisors) will have to convince "inside" officials that the markets are needed. If "outside" regulations create problems, "inside" regulators will need to find ways to get around those regulations or work with "outside" regulators to change them. More and more emerging markets are setting up cross-regulator working groups for this reason (i.e., Ma- laysia and Korea). Many countries are creating a single regulator, which may eliminate some of these problems by establishing a broader- based regulatory consensus. AMacrofhz6ality. Though macrostability is increasing in many emerg- ing markets, several still suffer from macro volatility or lack of cred- ible policymaking. Crowding out remains a problem in many countries, as does political instability. A country that has significant political or macro instability and high interest rates may have one option for its bond market: local-currency floating-rate notes of medium maturity. Issuers and investors will be unwilling to lock in Fixed rates or to commit to floating rates after a certain time period for fear that rates will move too much against them. Severe crowd- ing out can altogether stop the market's growth. Tzaxatwn Many emerging market countries employ stamp, transac- tion, and income taxes, usually in ways that are disadvantageous to the bond market. The most striking examples are duties that favor bank products and are set by government bodies whose first prior- ity is tax revenue, not bond market development. These impediments need to be eliminated, or they will prevent the market from ever starting. A neutral tax environment is preferable. This means re- moving a tax that favors another product rather than introducing a new tax incentive for bonds. It also means convincing the tax au- thorities to make these changes, perhaps by showing them how the market's growth can generate other kinds of taxes (such as interest income) that may be less distortionary. 21 ALISON HARWOOD Second-Layer Solutions. The main issues "across" the financial sys- tem concern building benchmarks and compensating for weak or dominant banks. Governnzetzt securitdej nzrket.s. Most emerging market countries have little in the way of developed government securities markets. On the whole, these markets lack auction schedules, fungible issues, market pricing, and yield curves-often because governments are reluctant to create a market that will reduce their direct control over their funding costs and volumes. Secondary markets are usually il- liquid. Governments often influence interest rates to control debt- financing costs. But if rates are set low, the securities will not trade because the underwriters (the primary dealers) will have to take a loss selling the bonds in the market. 15 Dealers are not developed. As a result, the corporate bond market, and the dealer community, lack an important foundation. It is generally accepted that government securities are the best benchmark for corporate bonds. Developed markets such as those in the United States have other alternatives, like top-quality corporates and swap rates. Some emerging market countries may have a few good-quality local or foreign corporate and supranational issuers. Even if these entities only issue bonds periodically, the issues can still serve that purpose, especially if they are large enough and have enough trading to create a current coupon. Other countries will not have these options. They may be able to use shorter-term treasuries (like three- to six-month paper) and other money market products that may be available to price medium-term, not longer-term, matu- rities or floating rate notes. Bank,i. Many banking systems are weak and unable to support the corporate bond market, or they dominate the financial system, as issuers or intermediaries, in ways that hurt bond market develop- " In chapter 6, Yongheom Kim provides a useful overview of the reasons why the Korean government securities markets did not perform this benchmark role effectively and how nonmarket determined rates seriously hindered the market's development. 22 BUILDING LOCAL BOND MARKETS ment. A good number buy and hold securities to meet statutory li- quidity requirements. Others do so because net returns on securi- ties are higher than on bank loans. Some use their political might to push for tax policies that protect them against bond markets. If banks are weak, other Firms must be available to issue, invest, or intermediate. If banks dominate the financial sector, their securi- ties activities might be put in a separately capitalized subsidiary, and independent securities Firms could be licensed. To get banks to in- vest in corporate bonds, marking to market and asset-liability man- agement (ALM) requirements can be introduced. Special tax breaks should be avoided along with policies that drive banks from loans to bonds, as banks may see the bonds as loan substitutes to buy and hold, not portfolio products. Banks should be active, not wholly passive, investors. Credit-ratilnq agencieo. In many emerging market countries, new credit- rating agencies have difficulty making profits and becoming eco- nomically viable. Some countries (such as China and Pakistan) make ratings mandatory in some form or another to ensure economic sustainability and to build a ratings culture. Having credit-rating agencies do various lines of business, such as rating the issuer as well as the issue and doing broadly based market research on indus- tries and companies, also helps provide a livelihood. The question of how many agencies to have depends on local circumstances. Com- petition is good, but so is survival. If the volume of business cannot support two agencies, then rely on one. It is important not to over- estimate a rating agency's impact on market growth. A credit-rating agency will not spur market growth significantly. Its absence is an obstacle but its presence will not in and of itself be a spark. Creating Liquidity. Creating liquidity is a serious problem in emerg- ing markets. Some constraints stem from market infrastructure, par- ticularly the trading system used, but the main ones relate to the size, structure, and capabilities of issuers, investors, and intermediaries. As for trading systems, many emerging market countries want to use their existing stock exchange to list and trade bonds because they do not want to create redundant infrastructure. But stock ex- changes usually have order-driven systems; over-the-counter (OTC) 23 ALISON HARWOOD markets usually rely on quote-driven trading. Countries that do not want to make the stock exchange redundant can combine listing on the exchange with trading OTC. This is often done in developed markets when investors such as pension funds can only invest in listed securities for prudential/disclosure reasons." Or the exchange may be allowed to have a separate bond-trading system. If countries are willing to have an OTC market but are concerned about not having enough regulation, they can have a regulated OTC market, with listing, membership, and trading rules. Several countries, Aus- tralia is one, are working to develop centralized trading systems for OTC trades to increase market transparency and regulation. On the issuer side, issuers rarely meet the required criteria for liquidity. Most corporate bonds are not issued frequently or in large sizes, and so have limited trading ability.17 As for investors, dealers need a diversified and sometimes active investor base to buy and sell with. They cannot just trade among themselves. Again, marking to market and judging portfolio managers against the market's perfor- mance can encourage reasonable trading by investors. Many emerging markets lack financing and hedging tools or dis- courage their use, sometimes through taxation. According to Singh, trading in Malaysia is constrained by limited RP facilities, many of which are restricted to purely principal dealers. That, in turn, limits wider market participation and increases the cost of risk manage- ment and transactions. Kim notes that Korea does not have repur- chase agreements (RPs), partly because of tax problems. Dealers rely on high-cost call market funds. There also is no short selling, so 16 A similar issue concerns whether to use existing clearing, settlement, and depositories (CSDs) and add a bond module. The trend worldwide, and in many emerging markets, is to have one CSD that clears and settles for a range of mar- kets, rather than following the U.S. multiple-CSD model, wherein one or two ex- changes have their own clearing corporations. Consolidation has important benefits: notably, reduced overhead and better netting. Indeed, the major CSDs for gov- ernment and equity securities recently merged in the United States. Malaysia is considering consolidating its clearing and settlement systems into one. One prob- lem is that CSDs are often owned by the local exchange, so the questions of where to trade and where to clear and settle become intertwined. Some countries are separating the CSD from the stock exchange so the CSD can serve more markets. 7 Endo. 24, BUILDING LOCAL BOND MARKETS dealers must own bonds for sale and finance them, which raises fi- nancing costs and impedes trading. (Both constraints may be re- moved in the near future to increase liquidity.)"8 Bank lines are often expensive, partly because brokers are less-known institutions and not so highly regarded. These problems are common in many emerging markets but can be rectified. RPs are frequently allowed, even in many emerging markets. (Activities such as RPs and short selling are usually not regulated or taxed in ways that distort the market.) They can start with government securities, since these are the most liquid instru- ments, and need mechanisms such as a central depository to ensure that securities are not double-pledged as collateral, and standard- ized agreements. Banks will be active in the business, so RPs/Re- verse RPs (RRPs) should not be considered a loan or deposit, to avoid expensive capital or reserve requirements. Most emerging market countries worry about introducing short selling because it can promote speculation. To address this concern, some countries - among them, Hong Kong - only allow dealers/market makers to short- sell to make sure they can provide liquidity. Note that derivatives require sufficiently active underlying cash markets and skilled deal- ers, which will influence when they should be introduced. As noted earlier, a local government securities market can help develop dynamic and profitable dealers, because primary dealers (PDs) for government securities often get preferential financing, profits, skills, and credibility. These benefits illustrate why it makes sense for the government and corporate bond businesss to be con- ducted under one roof (see the earlier discussion on intermediaries). DEVELOPING THE MARKET IN STAGES Most countries will have numerous pieces missing for building their bond markets and will want to know what steps they should take to " RPs are less common in the corporate market, mainly because corporate bonds are less liquid and not eligible for reserve requirements, which limits their attractiveness to many buyers. But countries rarely have explicit regulatory bans on corporate RPs. 25 ALISON HARWOOD build their market, not just where they need to end up. Steps are hard to map out, since no two countries will have the same combina- tion of pluses and minuses and political and behavioral conditions. Few emerging market countries provide extensive examples yet. Below are some very general guidelines. To start, each country will have to ask itself: What are my prob- lems and how and when can they be solved? That question should be asked repeatedly over the years as it will help determine what types of activities can be supported at different points in time. The structure, abilities, and attitudes of market participants will influ- ence significantly what can be built and when. If issuers are limited in number, size, and profitability, the chances of building an active secondary market are slim. The same is true for the investor base. The goal then should be to create a well-functioning primary mar- ket. It could be a public or a private placement market. In either case, issuers should have to disclose sufficient information for an investor to evaluate the issuer's credit quality. The macro environ- ment and the state of the government securities market, particularly whether pricing benchmarks exist, might influence the type of prod- ucts that can be introduced-that is, longer-term, fixed rate securi- ties or medium-term, floating rate paper. Stages of Change By and large, countries might move through three general stages. Stage I consists of doing one-off transactions -where an issuer needs local-currency, long-term funding or an investor needs a long- term investor and they get matched up in a limited private place- ment market. It occurs in a country with limited issuers and investors, relatively unskilled intermediaries, and relatively undeveloped capi- tal markets. Government securities markets are essentially undevel- oped. The banking system is likely to be quite weak but dominant. This country cannot support a strong primary, much less a second- ary, market in the immediate to medium term. Intermediaries would be needed to broker the deal along with such things as contract law, private placement and basic disclosure regulations, and mechanisms to register ownership and pay for funds. Setting prices would be a problem, as there will be no benchmarks. The issuer and investor C2 BUILDING LOCAL BOND NMARKETS may want to limit the maturity and use floating rates because both parties are afraid to lock in rates without benchmarks. The transac- tion may be more expensive than a public offering but may be the only way to obtain longer-term local currency for the time being- thereby reducing some refunding and currency risk but retaining interest rate risk. Stage 11 involves building a good primary market for public as well as privately placed issues. It takes place in a country that has several attractive issuers, a growing but still limited investor base, developing capital markets, and fairly good macro-political environ- ment. The country would need public company and disclosure regu- lations, a credit-rating agency, and OTC arrangements to support trading that might be done with the public issues. Having a bench- mark, even a limited one, would be useful and important for pricing slightly longer maturities. Stage III involves adding on secondary markets. The country must have sufficient issuers and investors and skilled intermediar- ies, along with a supportive macroenvironment. A government se- curities market is being built that provides a foundation for pricing new issues, as might other elements of the financial system. The credit- rating agency must be able to handle a larger number of issues. Though issuers will still be of higher credit quality, some of slightly lower credit quality will be able to enter the market as well. Efforts should continue to build institutional investors, and to ensure that they are able to invest in bonds by removing unnecessary constraints on in- vestment patterns. Disclosure rules may have to be strengthened as the investor base may broaden beyond the largest institutional in- vestors. Considerable training is needed to educate all persons in- volved (regulators as well as market participants) about the benefits and risks of the product, and of participating in the market. Accelerating the Transition Process As noted earlier, some problems are not so difficult to address; they just need to be identified. Three steps can be taken to assist the pro- cess. First, regulators can work closely with market participants to understand their perspective on the pluses and minuses of the mar- ket. Walking through a transaction with an issuer or an investment 27 ALISON HARWOOD with an investor can help clarify problems-whether they concern fees, capital, the lack of incentives, attitudes, or skills. Prudential goals need not be sacrificed for business attractiveness. In some cases, the government may just need to recognize that its requirements are unnecessarily strenuous and make the business unattractive. Work- ing groups should be set up to facilitate this exchange, and to strategize about how the market should move forward. Second, "demonstra- tion" bond issues by domestic and/or foreign entities should be en- couraged, since they powerfully demonstrate how to conduct a transaction and the benefits and difficulties it entails. Third, "inside" regulations should work closely with other regulators to get them to solve problems under their control. Cross-regulator working groups and dialogues can facilitate "outside" government commitment to the process and build support for changing policies and regulations around or across from the bond market. Many emerging market countries are setting up these cross-regulator working groups. Many are also creating a single regulator, which may help eliminate some of these problems and establish more similar goals. Finally, all countries should look at the experience of others with emerging markets for informa- tion on how to solve problems and what works best in different types of conditions. This experience is accumulating with time."9 Some countries provide considerable tax incentives to increase the economic benefits. We would suggest identifying the full range of obstructions to market participation before "pushing" on the eco- nomic benefits, since they can distort tax and other structures in ways that are difficult to remove later on. These distortions may be avoided, if other problems are addressed. THE SITUATION IN SOUTH ASIA This "around, across, and inside" approach was used to identify is- sues in developing bond markets in South Asia. Not surprisingly, IS Often changes need to be taken opportunistically, given that so many fac- tors are involved. It is difficult if not impossible to follow a predetermined se- quence of steps, and countries that try may miss key opportunities for moving forward. 28 BUILDING LOCAL BOND MkARKETS each country of the region could benefit from bond financing, par- ticularly India, Sri Lanka, and Pakistan, but also Bangladesh and to a small extent Nepal. Also not surprisingly, each country is in a dif- ferent stage of development with different problems, needs, and mixes of pluses and minuses, as discussed in the South Asia country sur- veys in Chapters 11 through 15. India, Sri Lanka, and to a lesser extent Pakistan, are actively working to build these markets. Bangladesh has taken a few steps. Each country has different reasons for wanting a bond market, but in each case some need stems from the banking system's inability to support funding requirements. Bank lending in Sri Lanka, Pa- kistan, Bangladesh, and Nepal is constrained, to varying degrees, by portfolio-quality problems. At the same time, demand for funds is rising from the private sector and state-owned enterprises, be- cause the government cannot go on funding the entities out of their budgets. India has large infrastructure financing needs, and its banks can- not support the structures and volumes required. Sri Lankan banks dominate the financial system and competitive alternatives are needed. Financial sector diversification is a goal (almost all of the country's investible funds are in bank deposits and government securities), as is corporate risk management though it is a weaker force than in Southeast Asia (but it is indeed a prominent concern in Sri Lanka). More generally and to varying degrees, the countries are shifting to more market-oriented economies and must increase private sector funding sources. Except for Nepal, each of the five countries has some form of corporate bond market, but only India has one of any size. Trading is virtually nonexistent across the region. Several of the countries, including India, have only private placement markets. Many issuers are state-owned enterprises, even in India. Nonetheless, markets in India, Sri Lanka, and Pakistan are growing. If we evaluate the countries by their success factors, based on the preliminary surveys included in this book, we can say that they fall into one of the three stages identified earlier. India is in Stage III. It is on the way to developing active primary and secondary markets. The government and private sector are 29 ALISON HARWOOD working to identify and overcome impediments, particularly techni- cal ones that constrain issuance and liquidity. Significant steps have been taken to build the government bond market, and the corporate market as well. Recent removal of the stamp duty on bond issuance and trading is expected to heighten corporate bond market activity. Sri Lanka and Pakistan may be in Stage II. The development of the bond market, particularly the government securities market, is a priority in Sri Lanka, which is currently focusing on improving rules and regulations, central market infrastructure, and a dealer commu- nity. The markets will likely be small, because of the small size of the issuer and investor base and the overall economy, and hence take time to develop. Pakistan's recently reactivated corporate bond market is tiny, may be constrained by extensive government borrowing and political instability, lacks key market infrastructure, and has a lim- ited issuer and investor base. But the government is taking mea- sures to make the market more attractive. Overall, Pakistan might focus on creating a solid public primary market. Bangladesh and Nepal are in Stage I. Bangladesh needs new funding sources, as bank financing and foreign aid are declining, but as Kviback describes in Chapter 14, market growth will be seri- ously constrained by a limited issuer and investor base and a weak banking sector. Bangladesh might create a small private placement market. Nepal lacks market participants and has extensive macro and broader financial sector problems. It might be able to create a very limited private placement market over time. The Status of Key Success Factors India clearly has the strongest range of key success factors, and supplementing factors as well. Sri Lanka comes second, followed by Pakistan. MarketParticipantd. Except for India, the countries have very lim- ited sizable issuers and investors; in addition, intermediaries need to develop their skills in bond market transactions. As a result, corpo- rate bond markets in each of those countries are likely to grow slowly. India does not seem to have a problem with potential issuers (see chapter 10 by Rakesh Mohan). The economy is growing, creat- 3300 BUILDING LOCAL BOND IMARKETS ing new and stronger companies. Sri Lanka's issuer base is small, but K6nigson (see chapter 13) sees a sufficient range of companies to support a small market. Bangladesh and Nepal as well are con- strained because their economies have a lot of very small agricul- tural and/or garment companies that are not appropriate for bond issuance. All of the countries have slow privatization programs, which limit the supply of new Firms to the market. Most of the countries suffer from regulatory burdens that con- strain issuers (issuance costs are high, approval processes are long, and stamp duties remain in effect). India, Sri Lanka, and Pakistan are working to fix these impediments. India and Pakistan are reduc- ing obstacles such as long approval processes. India plans to have a 144A-type private placement market for its corporate bond market, which means less onerous disclosure. The Karachi Stock Exchange (KSE) recently organized a committee of market participants and KSE and SEC officials to identify and address issuance-related prob- lems and plans to reduce costs. A similar story can be told for the investor side. Market par- ticipants buy and hold a lot of government securities because of investment regulations and lack of trading experience and incen- tives. This is not surprising, since each country is emerging from a situation of high government borrowing financed mainly by push- ing paper into state-owned financial institutions. India, Sri Lanka, and Pakistan are taking concrete measures to create more dynamic investors. India has by far the most diversified and sizable range of institu- tional investors. It has many large, government-owned financial in- stitutions, but private sector mutual funds, insurance companies, and pension funds are growing, which will generate more funds to be invested and traded actively. Sri Lanka's investor base is also driven by large buy-and-hold institutional investors who have been captive in government securities. K6nigson reports that the government is liberalizing investment policies-for instance, for long-term inves- tors such as insurance companies and for unit trusts-and is open- ing new markets for pension funds. Pakistan's institutional investors are limited to mainly banks, and there is modest knowledge of debt markets at both the retail and wholesale levels. Leonardo (chapter 12) notes that several obstacles 31 ALISON HARWOOD remain, most notably, disadvantageous tax policies for provident funds and a history of corporate bond defaults. But, like its neighbors, Pakistan is looking to liberalize investment policies (for provident funds), and mutual funds and portfolio management skills are im- proving somewhat. Both Bangladesh and Nepal are less developed in this area. For intermediaries, the government securities market in India and Sri Lanka is helping to create a class of skilled dealers that could help support the corporate bond market.20 To varying degrees, they still need to improve their skills in fixed income securities, particu- larly as markets like India's become more sophisticated. The ability to make profits may be a problem in Sri Lanka because the govern- ment is splitting the government and corporate securities businesses. (Primary dealers cannot engage in the corporate securities business while non-PDs cannot trade government bonds.) Financing and hedging is another problem, as explained in the section on second- layer factors. Go,vernment Commitment. The government's commitment to change and its actions in this regard are among the most positive features in India and Sri Lanka, and to a lesser but increasing extent in Paki- stan. Bangladesh is interested in developing bond financing and is looking at steps to help develop the primary market. Nepal is also interested but in many respects has more pressing issues on which to focus. India, Sri Lanka, and Pakistan in particular have created working groups between the government and market participants and with other regulators to identify problems and solutions. Macroeconomic Environment: The macroeconomic and policy envi- ronment in South Asia supports bond market development in some respects and in some countries. To varying degrees, each of the five countries has been shifting to a more market-oriented, less govern- ment-controlled economy, a shift that has been slower than in many other parts of the world. 20 India has a solid core of intermediaries affiliated with major foreign firms such as Goldman Sachs, Merrill Lynch, and ABN Amro. 52 BUILDING LOCAL BOND MIARKETS The real GDP in the countries as a group has been growing at a healthy pace and is projected to continue climbing at about 5 to 6 percent a year between 1997 and 2006. Savings rates are relatively high (see table 2), except in Nepal and Pakistan. In addition, infla- tion and exchange rates are stable, and interest rates have been de- clining over the past several years. At the same time, the countries have relatively high deficits as a percentage of GDP, and crowding out may be a problem-reflecting the strong role played by the gov- ernment in each of the countries. This, too, is expected to decline as the governments reduce their involvement in commercial activities. Not surprisingly, the statistics for financial sector structure show considerable room for developing corporate bond markets as an- other financing vehicle. The size of bond financing relative to GDP is tiny compared with other countries in the world and compared with other financial resources in each country. (Sri Lanka and Paki- stan also suffer from some political instability -the war in Sri Lanka and a fairly recent coup in Pakistan.) Taxation. Several of the countries have been reducing stamp duties and other taxation that negatively effects bond markets. Second-Layer Factors The strongest point here is that India and Sri Lanka, and now Paki- stan, are working to develop their local government bond markets in ways that will help support development of the local corporate bond market. Government Secnritiei Market. All five countries have some sort of government bond market, which is an outgrowth of the strong role played by the government in each country. India's market is by far the largest in absolute tcrms and is furthest along in its develop- ment. India and Sri Lanka have been strengthening their primary mar- kets in the past two to three years to make them more market ori- ented, competitive, and liquid. India's primary market is large, and a primary dealer community has been introduced, along with new in- struments, a wider range of maturities, and an expandingyield curve 33 ALISON HARWOO[) Table 3. Fixed Income Markets to GDP X vernment Securities Corporate Securities Country Outstanding/GDP (%) Outstanding/GDP (%) India 18.5 (NA) 8.30 (8.0) Pakistan 9.9 (5.2) 0.5 (0.5) Sri Lanka 17.0 (5.0) 0.6 Bangladesh 6.0 (1.0) 0.2 Nepal 14.9 (9.3) NA Note: Numbers in parentheses break out bonds with maturities greater than one year * Sri Lanka has outstanding government securities worth 44% of GDP, but only the amount noted above is negotiable. Source: South Asia country sarveys and symposium presentations. (see chapters by Thorat and Leonardo). Today's focus is on improv- ing liquidity (trading value is high but volume is low) with better trading, financing, and hedging mechanisms. India also has new regu- lations requiring banks to mark securities to market and meet new asset-liability management guidelines that will encourage banks to manage their portfolios in ways that will contribute to market activ- ity. Sri Lanka also is strengthening its primary dealers but, as Konigson notes, needs to build a more market-oriented yield curve and to im- prove trading, clearing, and settlement infrastructure. The govern- ment has rescinded stamp duties on RPs to encourage trading. The other three markets have relatively undeveloped primary markets and essentially no trading. Pakistan's market is fragmented, with a wholesale and a retail segment and sizable rate distortions between the two. Interest rates are high, often not market deter- mined, and there is no real yield curve or market makers. Trading is limited (retail securities, the bulk of offerings, cannot be traded) and is done primarily by local banks for liquidity management.23 Bangladesh's market is small and undeveloped, with essentially no yield curve and intermediaries with limited skills and profits. It con- sists mainly of long-term savings certificates with high interest rates. Only rates for T-bills are market oriented. Nepal's market has market-oriented T-bills, but all other securities are controlled by the central bank. A small network of dealers does marketing, but not market making. Banking Sector. As might be expected, banks dominate the financial systems in these countries, so their ability to be lenders, intermedi- BUILDING LOCAL BOND MARKETS aries, investors, and issuers affects market development. Where markets exist, the banks are significant investors but tend to buy and hold. Banks are sizable issuers, particularly in Sri Lanka, where they have been the largest bond issuers in the past few years to ob- tain Tier 2 capital, and they dominate as intermediaries. Bangladesh and Nepal both have weak banking systems and limited numbers of other issuers, investors, and intermediaries to take their place, which will constrain market growth. Central Market Infradtructure. India, Pakistan, and very recently Sri Lanka, have established credit-rating entities. India has four agencies, one that is viewed as dominant and Pakistan has two CRAs(introduced prematurely, perhaps). There are no rating agen- cies in Bangladesh or Nepal. Creating Liquidity Liquidity is a problem in the corporate bond markets of all five coun- tries because of the structure of the issuer and investor bases, lim- ited issuance, and lack of financing and hedging tools for intermediaries. These elements are being worked on in India and Sri Lanka, and to a lesser extent in Pakistan. In terms of trading mechanisms, Sri Lanka and Pakistan (Bangladesh as well) are promoting bond listing and trading on their order-driven stock exchanges, particularly through special tax treat- ment, which might constrain market liquidity. India, on the other hand, does not want to use the National Stock Exchange's order matching system to trade bonds, but instead plans to use the NSE as a reporting system-to quote bids and offers and then confirmed trades. The structure of the issuer and investor bases was discussed earlier. In India, the stamp duty has been a constraint on corporate bond trading but was recently removed. Prospects for trading are good in India because there are many large-scale issuers who may need funds on a regular basis. Prospects in Sri Lanka may be more limited ow- ing to the small size and number of the corporations. Again, the fact that government markets are growing in India and Sri Lanka will help create a more active dealer community that 35 ALISON HARWOOD will learn about trading. Financing and hedging tools are a constraint in all of the markets, but India, Sri Lanka, and to some extent Paki- stan are working on these. More work needs to be done if the deal- ers in the government and the corporate markets are to perform well. India has an active RP market, has introduced forward rate agreements and interest rate swaps, and is considering introducing short selling. Sri Lanka plans to introduce short selling. Konigson notes, however, that dealers in the corporate bond market may find it difficult to hedge their transactions because they are not allowed to also trade government securities. Pakistan does not have short selling, but does have RPs. CONCLUSION Emerging market countries can benefit from having local-currency fixed-income markets. These markets can reduce risks and help pre- vent or absorb crises caused by international financial flows and increased volatility that comes from having more open, liberalized financial markets and economies. But active markets can only de- velop in certain conditions: above all, they need suitable issuers and investors, and an environment that makes those parties interested in transacting in the market. This chapter briefly touched on the key elements needed. It suggests that the conditions needed have yet to appear in many emerging market countries. Although such coun- tries will still want to benefit from the risk reduction and diversifica- tion that bonds offer, they may have to do so in limited ways to start, expanding the market over time as conditions improve. These conditions, it must be emphasized, are not only technical. Many require a change in attitude and behaviours. Others are con- trolled by government bodies that are not interested in building bond markets. These factors complicate the process considerably. Clearly, "inside" regulators and market participants need to look at the over- all situation and realistically assess the problems that exist and how to address them. In this way, they will better gauge what can be built and when. The five South Asian countries span the spectrum of stages of development: at one extreme, India is on the way to developing ac- t36 - BUILDING LOCAL BOND MARKETS tive primary and secondary markets; at the other extreme, Nepal has a considerable way to go on the path of development before it can really consider creating bond markets. The other three coun- tries-Sri Lanka, Pakistan, and Bangladesh-lie in between. The country surveys in this volume offer specific recommendations con- cerning what might be done to move forward in each country. It is encouraging that many of the "inside" regulators in South Asia are working with market participants and with other regula- tors to identify obstacles to market development and how to fix them. That is a useful exchange of information, not only for bond markets, but for overall economic change. 37 CHAPTER 2 A Strategic Priority for Emerging Markets ASSAAD JABRE To introduce this symposium, I would like to share with you a few of the painful lessons that IFC has learned from the recent crises in emerging markets. These lessons have made it abso- lutely clear that the development of local debt markets is more ur- gent and critical than ever. However, we also need to discuss the complexity and difficulty of this undertaking. Last, I would like to say a few words about how IFC can work with you and help you in this process of developing local debt markets. Things are getting better in the emerging markets, but I would be very cautious about declaring victory. We have yet to see investors and lenders coming back to the vast majority of these markets in what I would call a meaningful way. Some of the best companies in Latin America and Indonesia are still struggling to gain access to the inter- national markets. Unfortunately, some of the reform programs that are under way may never be seen through to completion. This will only increase the vulnerabilities that we have seen in those markets. The first lesson that we have learned from this crisis is that for- eign exchange exposures can be devastating, not only for the corpo- rations and the financial institutions that have taken those exposures, but also for the rest of the financial system. Second, the recent crisis has shown us that foreign capital flows can be extremely volatile. The risk of liquidity and maturity mismatches can be very acute and is a source of systemic risk. In 1997 the net capital flows to the de- ASSAAD JABRE veloping countries reached about US$300 billion. By 1998, one year later, these were down to US$200 billion, with the bulk of the reduc- tion coming from lenders calling back their short-term loans. Indi- rect investments made by strategic investors remained stable at about US$150 billion. This shows that short-term loans had been financ- ing long-term investments, and when lenders decicled to withdraw those loans, they caused havoc in the markets. I do not think any market would have been able to survive this type of withdrawal, which demonstrates that maturity risk is very real. The third lesson concerns the future financing of some new sectors that have been privatized and liberalized, namely, the infrastructure sectors. I have no doubt that such liberalization and privatization can bring-and has brought-substantial benefits to the countries concerned, to the consumers in those countries, and has significantly improved the competitiveness of those economies. However, it would be very na- ive to believe that the huge and unmet demand in those sectors can be met by foreign capital flows alone. Prudent risk management re- quires the development of the financial sector. The development of the financial sector goes hand in hand with the need for a substantial mobilization of local savings, with a key role for the local debt mar- ket and local stock market. The same conclusion can be drawn for another very important sector in the social area, the housing sector. Another reason for working on developing local debt markets is to meet the growing needs of pension funds by providing long-term, local currency liabilities. The recent crisis highlights the need to accelerate the development of local debt markets, but progress has been excruciatingly slow in many countries. They face many chal- lenges in this regard. They need to create a supportive stable macro- economic environment and not crowd out the private sector with government borrowing. They must also build a prudent and trans- parent market with proper regulation and efficient and reliable in- frastructure; in addition, they must create active and capable market participants. We have all learned that building market infrastruc- ture is not the same as building a market. It does not ensure market participation: issuers, investors, and intermediaries need to see an economic benefit to being in these markets, and they need to be al- lowed by regulation to be in these markets. They also need to be able to operate capably, prudently, and effectively. 40 A STRATEGIC PRIORITY FOR EMERGING MARKETS These are difficult challenges that require time and effort in or- der to be able to meet them. They require active input and commit- ment from policymakers, regulators, and market participants, who together must set the goals and gain consensus on how to move for- ward and implement the steps that must be taken to achieve these goals. Those addressing these challenges can also benefit from learning from each other: sharing each other's experiences, successes, frus- trations, and solutions across countries and across regions. Many emerging markets have already made some progress in developing their local debt markets. Representatives from two countries out- side the South Asia region, from Korea and Malaysia, will share with us their experiences. For IFC, the whole question of local debt markets is a strategic priority. We shall do our best to continue assisting both regulators and market participants in the development process. We stand ready to help countries decide on the most appropriate instruments for them to develop while bearing in mind that our member countries are not all at the same stage of development and that their needs may be quite different. IFC will continue to provide technical assis- tance on policies and regulations and on the establishment of regu- latory bodies, helping the relevant authorities better understand the market participants' perspective. Through our investments in credit- rating agencies, in securities firms, and in institutional investors such as mutual funds, insurance companies, and pension funds, we will bring not only capital, but also best practices to our clients. This will also have an effect, we hope, on the development of other institu- tions and other companies in the relevant markets. We can also help support the development of local bond markets by issuing our own paper. We could also work on credit-enhancing bond issues and do- mestic bond issues, and on helping to invest in securitization ve- hicles. We hope that the presence here at the symposium of both regulators and market participants will help each country achieve the purpose of the symposium, which is to develop a road map for removing the obstacles that are constraining the development of their local debt markets. Perhaps all of you can become the cornerstone of a regional coa- lition for development of local debt markets. Someone may want to create a web site on which to exchange views, answer questions, 41 ASSAAD JABRE share problems, and measure progress as you move ahead in this important effort. If this works, then a coalition such as this could later be expanded to other regions, such as Latin America, the Middle East, and Africa. 421 CHAPTLR 3 Strengthening Sri Lanka's Financial Sector A.S. JAYAWARDENA It is essential for us in Sri Lanka to keep up with Financial devel- opments taking place outside the region, particularly in the bond and fixed-income securities markets. This is needed to guard against marginalization and the kinds of severe crises that hit many parts of Asia. Sri Lanka has been liberalizing its economy for many years now by moving away from a public sector to a private sector orien- tation. We are very pleased that we are being assisted in this process by the International Finance Corporation (IFC), the World Bank Group's private sector lending institution. We started opening up the Sri Lankan economy in 1977, long before many other countries did so. Transforming a highly state- dominated and controlled economy to an essentially private sector led economy is an extremely difficult task, however, and should not be underestimated. As pioneers in the Field, we had no well-defined path or ready-made plan to follow. Our First major obstacle was to liberalize trade, domestic transactions and the real sector of the economy, and to get the private sector involved in activities earlier dominated by the government. Although the real sector of the economy was important, in retro- spect we could have done much more by concentrating equally, and from the beginning, on the financial sector. We did not think that the financial sector was a high priority and put most of our emphasis on improving the banking system. We were trying to make the banking A.S. JAYAWARDENA sector more efficient, more competitive, and less directed by the state. However, to reduce the government's role in the banking sector, and reduce directed lending, turned out to be a major task. But today, as a result of this effort, the government plays a much smaller role in the commercial banking sector, while the private sector is being en- couraged to take on increasing responsibilities. As we slowly moved ahead with our banking reforms, we also tried to restart a rudimentary stock market. With modern technol- ogy, this project moved quite fast, although for a number of rea- sons the stock market has remained small. Recently, we hosted a seminar for international fund managers in Colombo, to identify the next steps required to develop equity markets in the country. An important aspect of all these reforms -of moving away from a state-dominated economy to a private economy-is the underlying macroeconomic policies. As we quickly learned, downsizing a gov- ernment that plays such an overwhelming part in the economy is a tricky job, particularly when the general public has come to expect the government to provide it with extensive services. The slow, steady progress toward liberalization received a jolt after the recent economic crisis. NWe began to realize that we could no longer take our time in making our financial markets more effi- cient. We have serious problems with capital inflows and outflows alike. Fortunately, we managed to survive this crisis, but the fact of the matter is that we are now a feature of the worlcd economy. All this time we have been complacently depending on official bilateral markets. One lesson we have to learn is that we have to improve the risk management and efficiency of our institutions. At present, Sri Lanka is incapable of adequately intermediating in these markets. The problem is that so many things need to be done simultaneously. It is mind-boggling for those of us who have to carry out these re- forms. How do we sift through this vast array of reforms to sequence them? How should we move forward? Building market infrastructure does not automatically create a market, just as putting up a school building does not necessarily mean the children in it will get a good education. So much more needs to be done and we are still not sure how to go about doing it. As I just mentioned, how does one sequence these reforms? How does one take into account objective realities on the ground? What ;44 STRENGTHENING SRI LANKA S FINANCIAL SECTOR is the best way to create a secondary market in public debt? We have created many market makers or market dealers, but we still find that the market has yet to take off. I think that we as regulators and as promoters have not properly understood the problem. In Sri Lanka, we had outside experts look at our markets, but the traditional public sector fears these people or does not under- stand the language they use, or has a defensive attitude that makes it difficult to move ahead. In the past, we have difficulty getting our public and private sector to talk with one another, but something like a dialogue is just beginning on how to improve the private mar- kets, particularly debt securities. In the public sector, we have regu- lators and promoters who are extremely competent in their own spheres of micro or macro management, but who find it difficult to relate to the market players, and who are rooted in a different disci- pline. As a result, many people do not understand the market. I am sorry to confess that even my own people are reluctant to "talk to" the market, and when they do talk, sometimes their understanding of the issues is poor, partly because we come from a long tradition of having the central government operating in the markets. We were lord of the market, and the private sector came to us on bended knees to get things done. That old master-slave relationship is no more. I suspect that the public sector is in some measure still inimi- cal to market development. That is why we need to create a climate, a culture in which the public and private sectors can interact in an efficient and mutually rewarding way. In other words, the attitudes of regulators and other people in the public sector need to change. However, the public sector is not the only one failing to talk to the market. The private sector, too, has an extremely negative atti- tude. With the exception of one or two firms that set the standards and provide some intelligent observations, financial analysts in the private sector seem to have little understanding of the markets. A further problem we face is the overhang of huge social secu- rity funds in the market. In the past, the government once dictated what happened in our markets. Today the social security fund es- sentially decides what the market should be. Is that acceptable? We say that we have a competitive market, but who is actually behind the decisions on the social security funds? What should be done about the thinness of this market? We need to reform the social se- 45 A.S. JAYAWARDENA curity funds and we need to reform the pension funds. The govern- ment has committed to such reform -in fact, we just had a few semi- nars on the subject, and we are working out a plan or road map for pension reforms-but we need to move forward with this quickly. We also need to look at why we do not have mutual funds, and why we still perpetuate rent losses introduced more than 50 years ago. There are indeed many impediments to market development in this country: the laws, the conventions, the customs, and even our cultural and other attitudes to public and private sector develop- ment. The public sector must realize that it has to create the condi- tions conducive to the development of markets and the private sector must enter into a close dialogue with the public sector because the regulator will be there whether they like it or not. Financial markets fail for a variety of reasons, so somebody has to be there to guard the system. Remember, it is not only the private markets in the fi- nancial sector that can fail. The government can also fail. I would like to see Sri Lanka put together a "road map" for de- velopment of our fixed income markets. It may differ from the mar- kets in India and Pakistan, but there are considerable similarities in the areas we need to address. Malaysia and Korea provide us with striking examples. This country information will be helpful for the work that our central bank, securities commission, stock exchange, and many others are now trying to do in our markets. Their ideas are spreading throughout Sri Lanka. However, their ideas and ef- forts need to be coordinated because the markets are interrelated. One market should not be developed faster than another; rather, there must be a holistic approach to these problems. CHAPTER 4 The Risk Management Benefits of Bonds MICHAEL PETTIS The importance of local bond markets has become much clearer in the wake of the international financial crises of the past three to fouryears, following the Mexican crisis of 1994/95. Like similar events dating back to the nineteenth century, these crises can be divided into two basic types. One is simply a good old-fashioned insolvency crisis in which a country has borrowed far too much and is unable to pay its foreign lenders. To take a current example, Ec- uador has an external debt of approximately 90% of GDP, which is clearly not repayable. The second type of financial crisis is more like a classic bank run. Although it may reflect policy mismanagement, lack of reforms, capital controls, crony capitalism, and other causes, its roots lie in unstable capital structures. This type of crisis is a problem of corporate finance theory more than a problem of eco- nomics, although the economic consequences are huge. This presentation covers four topics. I look at the current state of the local bond market, specifically medium-term or long-term fixed-rate local currency bond markets, among the developing coun- tries. Many of these markets have short-term or floating-rate local currency instruments or may have instruments that are indexed to another currency, such as U.S. dollars. However, these are radi- cally different instruments from the point of view of risk manage- ment and the development of local economies. Therefore the emphasis here is on a local currency yield curve that extends beyond one MICHAEI. PETTIS year. Second, I consider some of the corporate finance benefits that local bond markets bring to developing countries. The underlying assumption here is that corporate finance theory applies not only to corporations but also to sovereigns and, indeed, to any entity with a capital structure. Third, I discuss the implications of local bond markets for risk management, corporate finance, and the stability of capital structure. And fourth, I identify some of the main impedi- ments to the development of local bond markets. THE STATE OF BOND MARKETS Most entities obtain their financing from three major sources: (1) equity markets, (2) capital markets (specifically bonds), and (3) bank debt. The following numbers give a sense of their relative impor- tance: in the developed world, the United States has at least $12 trillion in traded equity; $13 trillion in the bond market; and about $4 trillion in bank debt. As for their scale, the value of equity mar- kets is equal to approximately 140% of U.S. GDP, bonds are equal to 160%, and bank debt 50%. Other large developed countries re- port a relatively similar breakdown. In Japan, equity accounts for about 50% of GDP, bonds about 110%, and bank debt 150%. In the United Kingdom, equity, bonds, and bank debt account for 150%, 80%, and 260% of GDP, respectively; and in Germany the respec- tive figures are 40%, 100%, and 170%. Although these numbers vary greatly among countries, the split in these three types of financing is fairly even throughout the developed world. On average, bonds rep- resent about 110% of the developed world GDP, with a total of $US24 trillion of capital markets instruments. Equity represents 80% and bank debt represents 150%. In each case, most of the bond markets are denominated in the home currency of the borrower. Debt markets in developing countries can be divided into exter- nal and domestic markets. In East Asia, domestic debt amounts to 10% of GDP and external debt 3% of GDP. For the emerging mar- kets of Europe, domestic debt accounts for 10-15% of GDP and external debt 10%. In Latin America, the corresponding figures are roughly 20% for domestic debt and 20% for external debt. In South Asia, domestic debt is fairly negligible, while external debt runs at 48 RISK MANAGErIENT BENEFITS about 20% of GDP. These numbers lead me to conclude, First, that local bond markets play a much greater role in the rich countries than in the poor countries; second, and more important, that rich countries have mixed funding structures whereas poor countries depend heavily on bank debt. Over the past few years, there has been an increased focus on diversity as an important ingredient of the funding structure. How- ever, emerging market countries tend to lack such diversity in their funding structures and to rely heavily on bank debt. This pattern has serious risk management implications: because banks tend to have certain systemic features, problems tend to crop up across the entire industry. When the banking system is the central means by which capital is allocated and is experiencing strains, those strains can shut down the whole system. That can put additional pressure on the banks and create further problems. Fortunately, local bond markets in emerging market economies have experienced fairly rapid growth over the past decade. At the beginning of the decade the numbers were almost negligible, even in Latin America, which has traditionally had a reasonably large local bond market. The South Asian region is an exception in this regard, with no significant growth. BENEFITS A long-term local bond market can have several economic benefits. First, it allows a more efficient allocation of savings as it matches the borrowers and savers directly. Hence it reduces the role of banks in the investment process, and reduces the amount of political interfer- ence in the allocation of credit since banks are subject to regulators such as the central bank. Also, in view of the especially important role banks play in the payment system, any threat to a credit market that focuses entirely on the banking system is likely to have marked effect on domestic liquidity and on the payment system. Local bond markets can help to separate these links. Second, local bond markets allow borrowers to use capital that is tailored to their assets and operations. Such tailoring may occur in many ways, the most important of which concerns maturity. Banks typically like to lend fairly short because their funding sources are - 49 MICHAEL. PETTIS very short, but projects are not necessarily short term. That is why maturity mismatches have traditionally been one of the biggest sources of domestic market problems. Third, economic benefit of long-term local bond markets is that they provide retail and institutional investors with several high-quality and liquid domestic saving vehicles. Bonds have many of the char- acteristics that allow savers to choose their risk and maturity, and to develop investment funds and pension funds. Traditionally, the most important traded security in new emerging markets, dating back to those of the United Kingdom in the eighteenth century and the United States in the nineteenth century, has always been government debt. Even when countries have a fairly large liquid government debt market, they often lack a liquid corporate bonds or liquid equity market. Yet it is important to the local bond market to develop pen- sion funds, mutual funds, and more efficient ways of rounding up the savings of the country. Incidentally, more and more banks are also talking about trad- ing bank loans. This activity should be encouraged for a number of reasons, the primary one being that banks need to liquefy their port- folios. Such trading also enables market participants to obtain an enormous amount of information from secondary market pricing. Market reactions to outside events are immediately apparent in the way bonds or loans trade. One can work out exactly how the mar- kets are measuring specific credit risk and other types of risk. At times, of course, such information may annoy bankers because it can make clear the volatility in the portfolio. Fourth, local government bond markets create monetary policy instruments. Government bonds also provide a pricing benchmark for the corporate sector. In fact, a corporate bond market should probably not be considered until there is a government bond market, since the government is the only borrower big enough to develop a very liquid market. Furthermore, since government debt tends to be the lowest risk asset within a country, all other assets can be priced off this asset. In countries with currency boards, however, the gov- ernment borrower may not be the least risky borrower in the cur- rency, since it will no longer have the option to monetize the debt but will have to raise the money just like any other borrower. Raising dollars may be extremely difficult for heavily indebted governments. 50 RISK MIANAGEMENT BENEFITS The benefits of securitization also merit a few words. Securitization refers to the pooling of different assets and the issu- ing of new securities backed by those assets. In principle, these as- sets can be any claims that have predictable cash flows, such as real estate, financial assets and loans, or even future receivables. One of the most famous securitizations of recent years involved the future royalty payments on records sold by the rock musician David Bowie. As long as there is some predictability in his royalty payments, fu- ture payments can be packaged and sold today to investors. Why is securitization useful? Like loan trading, securitization allows banks to take a collection of illiquid assets and make them liquid, and therefore make them marketable enough to be bought and sold. This enables banks to adjust their asset position to fit their capital needs, their risk appetite, and their risk profiles. It is also a way of converting and monetizing future earnings so that they can be used as a source of capital in the current period. This can be done with rent payments on a building, for example. If a building owner is earning rent over time but needs to make a capital investment today, securitization will allow that owner to borrow in the current period. The future rent payments that are owed can be used as a source of capital in the current period. Securitization also facilitates development of a mortgage market, which may be politically impor- tant because it assists and encourages the banking system to pro- vide low-cost loans for individuals to buy homes. Furthermore, securitization provides an incentive to developing local bond markets because securitized assets cannot be evaluated and priced without a liquid "plain vanilla" asset in the market. "Plain vanilla" assets are straight bonds characterized by bullet payments, fixed coupons, and, in contrast to securitized assets, single obligors with clean credit structures and payment structures. Once a market for plain vanilla securities is in place, it is possible to build complex markets such as securitization on top of it. RISK MANAGEMENT BENEFITS As mentioned earlier, liquid markets incorporate all the information that investors care about and embed it in the prices of every asset 51 MICHAEL PETTIS that is traded. Such information helps mitigate uncertainty, which significantly reduces the prices of financial assets and in many cases prevents investors from buying into them. A local bond market gen- erates a huge amount of useful information. A yield curve, for ex- ample, allows for calculation future interest rates, such as forward rates. This information is very useful for an investor or a trader be- cause it gives investors or traders a sense of market expectations concerning liquidity premiums, inflation and the cost of hedging. This "information discovery" mechanism helps stabilize the national capital structure and permits development of derivative instruments. It also allows one to derive estimates of expected inflation from gov- ernment yield curves and to calculate currency forwards, which are not necessarily priced on the expectation of the future exchange rate, but simply the interest rate differential of the two currencies. That is why, in examining a particular market, I generally spend a great deal of time looking at the way yield curves shift over time. Note, too, that ayield curve and a local bond market can be used to figure out corporate credit spreads. To illustrate, suppose a gov- ernment is considered a risk-free borrower and is borrowing for five years at 13%, while a corporation is borrowing for five years at 15%. In other words, according to the market, the risk of default-the credit risk component-of that corporation implies roughly 2% a year of additional risk. This provides a basis on which to assess com- parable credit risk in different borrowers. I believe that local bond markets can contribute a great deal to the management of financial crises and risk. Economist Guillamo Calvo has asked why the Asian crisis meted out such a huge punish- ment for such small mistakes. In his new book, Paul Krugman calls this the most important question that has come out of the Asian cri- sis. What happened to Korea, Malaysia, and Indonesia, in 1997 and 1998 seems all out of proportion to what actually went wrong. I strongly believe that, from an economic policy point of view, prob- ably not that much was done wrong-at least not enough to cause the market collapses. Rather, the explanation lies in corporate fi- nance issues relating to the national capital structure. In corporate finance, the capital structure is not simply a way of raising money, but more a way in which the borrower indexes itself 52 RISK MANAGEMNIENT BENEFITS to market risks. Money can be raised in many ways, ranging from equity to different types of debt and all their permutations. The critical feature of the capital structure is how the repayments are indexed. That is what can transform a small shock into a disaster or keep it a small problem. How does this process work? An external shock affects both sides of the national balance sheet: the asset side, which reflects a country's ability to make money, generate revenue, and so on; and the liability side, which reflects the way repayments are indexed in the future. Most of the Asian countries had what I would call an inverted capital structure-one in which the performance of the as- set is inversely correlated with the liability sides of the national bal- ance sheet. An inverted capital structure, in other words, is one in which improvements in underlying conditions (the "asset" side) will lead automatically and immediately to lower borrowing costs (the "liability" side); conversely, deterioration in external conditions will lead to higher borrowing costs. A typical example of inverted funding is short-term debt denomi- nated in dollars (and by "dollars" I mean any foreign currency). As the economy improves, the local currency strengthens in real terms and along with this strengthening, presumably, asset values and lo- cal revenues increase. Because the borrowings are in U.S. dollars, whose value in real terms is declining if the local currency is strength- ening, the real cost of the borrower's existing debt continuously drops. With a short-term debt, the financing spreads decline quickly as the debt comes due since, presumably, the borrower's credit quality is improving, and every time the debt is refinanced it is refinanced at a lower rate. But this type of inverted capital structure carries a substantial cost, in the form of increased volatility. If external conditions dete- riorate, debt-servicing costs on existing loans automatically increase for the same reason that they decrease in improving conditions: a deterioration in global conditions can cause the currency to weaken and asset values and revenues to decline. Debt costs, however, be- cause they are fixed in dollars, increase. This adds instability to the balance sheet by "doubling up" the effect of both good events and bad events. In domestic U.S. markets, debt and equity analysts are extremely sensitive to these types of unstable capital structures, and 53 MICHAEL PETTIS they will penalize a company whose balance sheet incorporates too much market risk. But this does not seem to be the case for emerg- ing market countries, even though the consequences there can be much worse. In a stable capital structure, changes on one side are matched by offsetting similar changes on the other side. If there is a currency collapse, for example, domestic revenues may go down for whatever reason; ideally, the cost of the liabilities will go down simultaneously. That is, a deterioration on the asset side will be matched by an off- setting improvement in the liability side. In the case of a corpora- tion, if assets and liabilities move together, the value of equity never changes- and the structure remains very stable. This is the key factor that determines whether external shocks are, or are not, transformed into crises. When the economy of an entire country is deteriorating, this, by definition, means the asset side is deteriorating. "Asset" is used here as a generic term to repre- sent the value of its future revenues. If its liabilities are structured in such a way that the value of the liabilities also decrease-in other words, that the liability side improves as the economy deteriorates - then net worth will not be affected. This is a correlated capital struc- ture. An unstable inverted capital structure, which is characteristic of most emerging markets, will experience the opposite result, as demonstrated in every country that underwent a crisis between 1994 and 1999. It is the instability of the capital structures that turns a small shock into a big shock. Countries should therefore aim for a structure that minimizes volatility in net worth. However, most emerging market countries face the reverse situation: when economic circumstances improve, their existing cost of financing decreases, whereas when circumstances grow worse, the cost of financing increases. A prime example of this can be seen in Thailand's rather large short-term borrowing in dol- lars. As long as things went well in Thailand, more capital came into the country, asset values boomed with the growing economy, and the currency strengthened. For Thai corporations as well as the Thai government, the dollar cost of borrowing declined as the currency grew stronger in relative terms. In consequence, revenues increased and liabilities decreased. As long as it can be guaranteed that things will stay good forever, this is a wonderful capital structure. I4 RISK MANAGENIENT BENEFITS The problem is that there can be no such guarantee. When things turn around and the economy does badly, the currency weakens and the real value of dollar borrowings goes up. Although initially net worth may remain positive, in a deepening crisis it will collapse be- cause both sides of the balance sheet will do the wrong things at the same time. Why was this structure ever created? If a country be- lieves that things are going well and will continue to do so indefi- nitely, for example, because of a whole set of reforms that are eagerly being implemented, then it makes sense to choose the structure in which the cost of liabilities is always going down. However, things cannot go well forever, so it is better to have a capital structure that protects against these risks. What bearing does all this have on local bond markets? As it turns out, long-term and fixed-rate local currency bonds are prob- ably the most stable type of borrowing that a country or corporation can engage in. Emerging markets, to the extent that they rely on local-currency bonds at all, favor short-term bonds almost exclu- sively, and if they do extend beyond a year, they are almost always floating. In Latin America, both Mexico and Brazil have large local currency bond markets, but Mtexico has nothing beyond one year and Brazil little beyond ayear, much of which reprices every day, in any case. This repricing means that the interest rates will keep chang- ing, even for purely external reasons, such as the collapse of Russia or Brazil. When investors are nervous and there is some external liquidity contraction, it is normal for them to pull their money out. As inter- est rates go up, however, debt servicing costs change immediately because this is all short-term debt. Unfortunately, commodity prices are also likely to fall, since they tend to drop whenever global li- quidity contracts. Remember that when real interest rates go up in emerging markets, they do not go up by 50 or 100 basis points. In the case of Brazil, nominal rates rose from 20% to about 50% with inflation remaining at 1% to 2%. Faced with such a huge increase in the real rates, a legitimate legal enterprise will be unable to earn that type of return on its capital. Therefore, credit quality will plum- met. As credit quality deteriorates, investors become nervous. At the same time, corporations may begin defaulting on bank loans, and the banking system may get into trouble. Interestingly, when _______ ____ ______ ___________ ____ ____ ____ ________ 55 MICHAEL PETTIS conditions are good, the same sort of cycle occurs in reverse. When interest rates come down, credit quality improves, confidence in- creases, and interest rates come down even further. The trouble is that the process is uncontrollable -external shocks can quickly send it in the opposite direction. Now consider how this process works when local currency debt is fixed at a 5-year or 10-year rate, for example. When there is a crisis and interest rates go up, inflation often increases. Brazil and Mexico are good examples of this pattern: until their currency broke, inflation was low, but afterward it shot up. As inflation goes up following the crisis and the value of the currency declines, the real cost of the fixed rate debt goes down. The payment is fixed, and as it goes down in real terms owing to inflation, the debt burden shrinks. An automatic stabilization mechanism comes into play here, be- cause as things get worse on the asset side, the liability side shrinks enough to reduce or even eliminate the debt burden. This is a radi- cal and elegant concept: fixed rate local currency bond markets act automatically to stabilize markets both when times are good - although perhaps low volatility is not desirable when times are good -and, most important, when times are bad. Borrowing in dollars or in short-term local currency does the exact opposite. It makes good times even better, but it causes bad times to spin out of control. ANT IMPEDIMENT TO CREATING A BOND MARKET Of the various impediments to creating a bond market, one of the most troubling is the common misperception among government fi- nance officials that rates today are always too high and that they will always be lower tomorrow. In consequence, these authorities tend to assume that development of the local bond market should be delayed. The Mexican government, for example, recognizes the need for a long-term market but keeps saying that rates are too high. Officials think that if they wait 6 or 12 months, rates will be half the levels that they are today. In reality, of course, rates are as likely to move in an unexpected direction as they are to do what they are supposed to. i566:- RISK MANAGEMENT BENEFITS At the beginning of 1998, Mexico could have gone ahead with 20% rates for five-year peso bonds but decided that this was too high and chose to wait two quarters for rates to come down. Two quarters later, one-year rates climbed past 30%. In the case of Bra- zil, in mid- 1997 when its rates were 20%, it was told that there would be a strong demand for five-year fixed-rate debt at 17% or 18%. Brazilians refused to accept this, saying that in six months rates were going to be around 12%, so why should they borrow at 17%? In fact, six months later Brazilian rates were at 40%. The problem with always wanting to borrow at a cheaper funding cost is that timing the market is inherently speculative. Needless to say, developing a bond market should not be a speculative trading activity. Rather, it is a long-term program with bonds being issued no matter where interest rates happen to be, and the market has to find this pricing over time. Such insistence on "gaming" the market has often been the biggest impediment to developing markets in Latin America and other developing countries. CONCLUSION Credibility is largely about the ability to manage volatility. A coun- try with low credibility is one that is perceived to be unable to main- tain the integrity of fiscal, monetary, or exchange rate policies in the face of shocks. Because the policies of small countries already suffer from low credibility, and because an unstable national balance sheet is the most common cause of financial collapse, sharp increases in volatility can seriously threaten their stability. This problem must be met with a disciplined approach to national liability management because financial instability carries an enormous cost. Even when countries are able to figure out appropriate development policies, history argues on the side of caution, in view of the repeated failed attempts to develop and follow through on earlier policies for eco- nomic growth. Moreover, almost no government can be expected to maintain the integrity of its economic policies over the long run when faced with periodic market collapses. A local currency long-term bond market cannot prevent market shocks, but it can certainly help absorb them better. It is an impor- 57 MICHAEL PETTIS tant way for borrowers-corporate and government alike-to en- sure that investors and users of capital share risks appropriately. For low-credibility countries, liability management must be geared toward reducing volatility and preserving the integrity of economic policies. Their overall capital structure should be designed to pass volatility onto investors. It is the misunderstanding of the role of liability management-rather than crony capitalism, bad banks, or any of the other culprits usually blamed for financial crises-that explains the speed and surprise with which markets have collapsed in the past. The problem is not that global financial markets are too volatile or free capital flows too dangerous, but that sovereign capi- tal structures are not usually designed with this volatility in mind. In the end, an optimal capital structure is not enough to ensure that an emerging economy will develop rapidly. The wrong capital struc- ture, however, can guarantee that the economy will break down be- fore it can achieve its goals, regardless of overall policy. ;U58: CHAPTER 5 The Benefits of Supranational Issuance for Local Bond Markets MAMITA SHAH In each of the past three years, IFC has raised about US$4 billion to US$5 billion to finance operations and aims to raise the same amount this year. IFC has issued bonds in 28 currencies from both established and emerging markets. Currencies in which IFC has borrowed include the South African rand, Polish zloty, Czech koruna, Hong Kong dollar, Singapore dollar, and the Philippine peso. IFC has been the pioneering first issuer in many emerging markets, most recently issuing in the Israeli shekel and Singapore dollar. In Asia, it was the first institution to make a Philippine peso issue. Many are interested to know how IFC funding promotes devel- opment of the local debt market, what benefits accrue to a country when it opens its doors to responsible creditworthy issuers such as the IFC, and how the issuance of bonds by nonresident institutions helps to develop the local debt market. Such bond issues have a huge development impact on local bond markets. An obvious ben- efit is the demonstration effect by an issuer like IFC. This activity also helps integrate local marketing practice with the best interna- tional standards. Although every country has its own ideas about how to develop its debt market, it can learn a great deal from other markets. Even developed nations offer some useful lessons as they are constantly looking for ways to improve market practices, which include the structuring of issues, standardization of documentation, and depository, clearing, and settlement operations. MAMTA SHAH Introduction of best market practice is by far the most impor- tant benefit. There is usually huge resistance the first time a change is made to normal practice because it is difficult to anticipate the advantages of doing things differently. Once a change is implemented, however, it paves the way for subsequent issues. IFC's counterparts in London, New York, and Hong Kong that are working in the emerging markets have also found it difficult to persuade people in local markets to consider a different approach. Some regulators claim that nonresident issuers crowd out local issuers, but there are three problems with this argument. First, in IFC experience, issuance in these markets plays a catalytic role in encouraging more activity, both by issuers who are in the market and by new issuers. Second, none of these markets has an oversupply of debt securities. Third, over the long run, the size of' the investor pie becomes larger. Both Standard & Poor's and Moody's rate IFC bonds AAA, which is several notches higher than the sovereign bond rating in any of these markets. This allows investors to diversify their port- folio risk and hence create an appetite for lower-rated bonds. Foreign investors who are not keen on taking both the credit risk and the currency risk at the same time are more comfortable with IFC bonds because they do not have to worry about the credit risk. Over time, as they become comfortable with the currency risk, they tend to prefer bonds of the same currency, namely, government bonds and the corporate bonds of the local currency. lFC uses a standard information memorandum for all its bond issuance in various markets. This prospectus complies with all the requirements of the 28 markets that IFC borrows in and provides good disclosure and total transparency of IFC's financial statements. Not all jurisdictions require this information in full, but for consis- tency IFC makes it available across the markets. IFC's Treasury Department also has a web site at which it posts any significant or material change that may have an impact on its portfolio. IFC makes every effort to be prudent about keeping the various jurisdictions that it borrows in completely informed. For example, when it did the first Singapore dollar bond issue, IFC decided to make a bigger provision in its portfolio for Russia and Brazil. Singapore was hav- ing problems at that time. IFC informed its investors and syndicate members of the bond issue. After the changes to the portfolio were 60 THE BENFF ITS OF SLPRANATIONAL ISSUANCE explained, not a single investor or a syndicate member backed out of the investment because of the strong AAA rating of IFC bonds. Another argument to counter the alleged crowding-out effect is that regulators can always turn off the tap if they do not wish to have further issuance from nonresidents. During the run on the Hong Kong dollar equity market, for instance, the authorities thought there were a lot of speculators in the market. Therefore they asked all the supranationals not to do any borrowing with a maturity of less than three years. IFC does not do any borrowing without government consent. An additional benefit of IFC bonds is that they help to establish a risk-free benchmark in countries without a benchmark issue. Many markets lack a well-developed yield curve, and swaps are not nor- mally available for longer maturities. As the yield curve and swap market develop over time, however, the maturity of the bonds also increases. After IEC inaugurated the Singapore dollar bond issue in October 1998, which had a three-year maturity, other highly rated nonresident issues followed, with maturities of up to 10 years. In the case of the South African rand, IFC has done 12 issues with matu- rities ranging from 1 to 25 years. IFC manages its risk exposure to various currencies by swapping them to U.S. dollars, which is the currency of IFC's balance sheet. Since this is the currency of IFC's balance sheet, regulators fear that it draws out valuable foreign cur- rency. However, a structure has been worked out to ensure that the currency does not leave the country. Therefore opening markets to the nonresident issuers also fosters the development and deepening of the swap market. While one issue alone does not give rise to a cross-currency swap market, a series of issues does and over time helps shift the market from one customized back-to-back swap to a system of swap books among key market players, which greatly increases the overall effi- ciency of the financial markets. This has been demonstrated in many markets, including Hong Kong and South Africa. The importance to local corporates of having deep and liquid stock markets should not be underestimated. Such a market enables them to hedge their foreign exchange exposure on internationally competitive terms and to compete globally on a level playing field. The market structure also needs to be in place. 61 MAMTA SHAH Such standardization not only provides a great deal of comfort to credit rating agencies when rating a sovereign bond issue, but over a period of time it also reduces the cost of borrowing for all issuers, be it the government or the corporates. It also increases market discipline on the part of issuers as investors hold issuers to global standards of transparency and disclosure. A successful bond issue by an organization such as IFC is a strong validation of a country's market infrastructure to the international financial community, par- ticularly with respect to depositary, clearing, and settlement opera- tions. IFC has often been the first issuer doing the spade work, with other issuers subsequently entering the market. IFC has led the way for several reasons. First, it has a very strong AAA rating. IFC re- ceived its first rating in 1984 and since then has been rated AAA. Second, IFC is a strategic borrower rather than an opportunistic borrower. It does not pursue the lowest available rate without con- cern for the market, the investor, or the country image. IFC is risk- averse and intent on maintaining that image in the eves of all parties, including investors and regulatory authorities. Third, its accounting procedures comply with internationally accepted accounting prin- ciples. Fourth, IFC believes in total transparency and the highest standards of disclosure. :62 CHAPTER 6 Korea's Bond Market Following the Onset of Financial Crisis YONGBEOM KIM KOREAN BOND MARKET AT A GLANCE Primary Market In Korea, bonds and other forms of fixed-income securities are important financial instruments for all sectors of the econom,y. On the issuer side, manufacturing firms use them alongside risk capital and bank loans; infrastructure projects obtain much of their long-term funding from bonds; and local authorities are often ac- tive issuers in the bond market. The government also depends on bonds to finance its budget deficit. On the investor side, bonds are the dominant asset in the balance sheets of banks and investment trust companies. Korean debt securities are classified into five major categories according to the type of issuer: government bonds, municipal bonds, financial debentures, special (law) bonds, and corporate bonds (see figure 1). Fixed-income securities are issued through public offer- ings, private placements, or, in the case of government bonds, through auctions. As of July 1999, the value of outstanding listed bonds was 356 trillion won, an amount larger than the stock market capitalization. Government bonds accounted for 15% of the total. The outstanding amount of government bonds increased 10.1 trillion won from the level of 1998. Corporate bonds were worth 122.7 trillion won, or YONGBEOM KIM Figure 1. Outstanding Bonds, by Type, July 1999 Mrnetary Stabi zat on Corpoaate 344% Special ec a 24.2°.h Financial enm n 9 5% Municipal 14,a% 244% Total outstanding amount = 356.6 trillion won Source Korean Stock Exchange, "Stock" (August 1999). 34.4% of the total. And the proportion of non-guaranteed corporate bonds to corporate bonds skyrocketed from 15% in 1997 to 68% in 1999. Secondary Market Bonds are traded on the exchange and in the over-the-counter (OTC) market but mostly in the latter, even though most bond issues are listed on the exchange (see table 1). Member firms of the exchange Table 1. Market Capitalization of Bonds and Turnover Ratio (billions of won) Excbange-traded OTC-traded Totalamount Outstanding Turnover Year Amount % of total Amount % of total traded (A) amount (B) ratio (A/2B) 1990 6,433 15.3 35,623 84.7 42,056 51,117 0.33 1994 2,362 1.4 162,085 98.6 164,447 102,492 0.55 1997 6,875 2.4 279,280 97.6 286,155 224,116 0.64 Source: Securities Sapervisory Board, Securities Markets Yearbook, 1998. T64 - KOREA S BOND MIARKET are required to concentrate orders for small-lot government bonds and convertible bonds in the exchange market. The financial sector holds about 82% of Korean bonds issued (see tigure 2). The major bondholders among the financial institu- tions are banks and investment trust companies, which tend to hold the bonds to maturity. The investment trust company (ITC) and investment trust management companies (ITMCs) account for about 33% of the total outstanding amount of Korean bonds. Banks ac- count for about 32% of the total outstanding amount of Korean bonds. Securities companies have relatively small bondholdings. They invest in bonds for speculative purposes and hold on to their invest- ments for only short periods. Insurance companies, mostly those in the life insurance business, account for 8% of the total outstanding amount of Korean bonds. Because their liabilities extend over the long term, life insurance companies tend to hold their bond invest- ments until maturity. Until May 1998, local bonds had exceptionally high yields in response to the government's contractionary monetary policy after the financial crisis. Market interest rates quickly regained stability, however, once the government took steps to lower interest rates. The yield on three-year corporate bonds fell from an annual rate of 18.28% at the end of March 1998 to 8.0% in mid-1999. Figure 2. Bondholdings by Economic Sector, 1997 Financial 82% LBank 32n%/h \In vestment Trust 33% _nsLeance Ind vidual |8% 8% Pension Corporate Ler 6% 3% 8% Government L 2% -65 YONGBEOM KIM HISTORICAL BACKGROUND The Financial System in Economic Development In its earlier stages, the Korean economy, unlike the capital market- dependent economies of the United States and the United King- dom, relied heavily on the banking system to channel savings to industrial investments. By regulating financial institutions with credit rationing and interest rate ceilings, the government intervened heavily in the allocation of financial resources for developing strategic in- dustries. Financial policies implemented to develop the economy included regulating interest rates and credit control through "policy loans." Policy loans became an important means of providing funds to specific sectors, industries, and even enterprises. As a consequence of the government's credit control, banks played a significant role in the country's economic development, especially as an instrument of national industrial policy. The capital market played only a supplementary role. Meanwhile, the banking system failed to promptly provide sufficient financial resources for the grow- ing and increasingly complex economy of the 1970s and 1980s. Development of the Bond Market Policymakers realized that the capital market needed to be devel- oped to cover the deficiencies of the banking system. The corporate bond market was seen as a most favorable alternative source of fi- nancing for corporations. In the 1970s, the government introduced various measures to create a favorable environment for the issuance of corporate bonds, most notably a guarantee system for corporate bonds and some revisions to the taxation system. With the establish- ment of the Securities Investment Trust System and the underwrit- ing system, Korea's corporate bond market started taking shape. In the 1980s, the government deregulated the issuance limit on guaranteed corporate bonds, introduced Repurchase Agreements, allowed self-regulation on issuance of corporate bonds, and permit- ted issuance of floating rate notes. These measures helped pump up the volume in the corporate bond market more than in the stock market and pushed corporate bonds to the forefront of the Korean KOREA S BOND MARKET bond market. What is unusual, however, is that guaranteed corpo- rate bonds dominated the market. Meanwhile, the government bond market remained stagnant because the government basically viewed it as just a place to raise necessary funds. It did not recognize the way government bond market structure and operations affects growth of the corporate bond market. The government focused mainly on developing the corpo- rate bond market and, unfortunately, failed to recognize that the government bond market, still being immature, hampered develop- ment of the entire bond market, including the corporate bond mar- ket. Consequently, the Korean bond market experienced disproportional and fragmented development, despite a significant increase in size. Following the financial crisis, the government's attitude toward the bond market changed fundamentally. Faced with the need to raise substantial sums through the bond market, the government saw that the bond market could play an important role as a provider of long-term bonds. A well-functioning bond market would help lower funding costs and minimize the impact on the financial market of large-scale issuance of government bonds. Deregulation and Market Opening Starting in early 1998, with the financial crisis in full bloom, the Korean government implemented some comprehensive economic reforms. At the same time, it prepared a full-fledged deregulation plan for the financial market that would allow banks to have au- tonomous authority and self-regulation in their business activity. In addition, the government decided to lift completely all ceilings on bond investment by foreign investors from January 1998 and also increased the limit on foreign ownership of domestic stocks from 18% to 55% of outstanding stock. In May 1998, the ceiling on for- eign ownership of stocks was abolished. Despite this full opening of the market to foreign investors, that group has shown little interest in the Korean bond market; their share remains below 1%. It is believed that foreign demand for Korean bonds is undermined by four factors: exchange rate risk, default risk, liquidity risk, and high trading costs. Among these, exchange risk 67 YONGBEOM KIM and default risk can be mitigated as the Korean economy recovers from its worst recession on record. However, the high trading costs and liquidity risk stem primarily from the underdeveloped bond market. To cope with these kinds of structural problems, the govern- ment needs to undertake full-scale reforms in the securities market. UNDERLYING PROBLEMS IN THE BOND MARKET Underdeveloped Government Bond Market As noted, underdevelopment in the Korean bond market is partly due to immaturity of the government bond market. No government securities function as a benchmark in the bond market. Several fac- tors account for this lack of a government benchmark. First, the fiscal deficit of Korea is smaller than that of developed countries. Between 1981 and 1993, the average ratio of the fiscal deficit to current GNP amounted to only 1.2%. Since the govern- ment was reluctant to increase the fiscal deficit, government bond issues had to be small and limited. The ratio of outstanding govern- ment bonds to GNP in 1996 was about 6.7%, which is significantly lower than the 68.8% of the United States, 55.8% of the United King- dom, 53.0% of Japan, and 20.7% of France. The main obstacle to Korea's bond market development is the lack of truly market-determined rates in the public bond auctions. In the past, most public bond issues were allocated to captive syndi- cates and sold through tender offers, public sales, or compulsory sales to individuals and firms in connection with permit applications and administrative registrations. Since November 1993, all market- able government bonds have been issued using a British auction system. However, the government set a maximum rate for the auc- tion that was generally below the market interest rate; and the high- est rate had to be lower than the minimum rate set by the government. As a result, the total amount awarded was lower than the intended government offering. The unsold portion of bonds was purchased by the government bond underwriting syndicate (consisting of 102 financial institutions) at an interest rate of 0.2% lower than the av- erage bid-awarded rate. 68 KOREA S BOND MARKET Furthermore, too many issuers of government bonds are spread out among numerous special accounts and funds, making it diffi- cult to standardize the issuance of government bonds. Korean gov- ernment bonds are given different names depending on their specific purpose. Consequently, the outstanding amount per issue is too small to maintain sufficient liquidity in the secondary market, de- spite the growth of the government securities market in recent years. In addition, the irregularity of the issue cycle makes it difficult for investors to anticipate future issuance. This, in turn, has an ad- verse effect on the marketability and liquidity of government bonds. The short-term maturity of most government bonds pushes up their cost of management. This has also prevented the development of a meaningful yield curve that could be used in pricing issues of corporate bonds. In addition, the fact that bonds are issued in greater concentration toward the end of the year undermines the stability of the bond market. Bond prices are often set by a select group of large institutional investors, and the liquidity of government bonds remains low. Hence transaction costs are high and individual investors are reluctant to participate in bond trading. Dominance of Short-term Bonds Most Korean bonds have short maturities, averaging less than three years, because investors are concerned with inflation. With the preva- lence of short-term bonds, it has been impossible to develop a mean- ingful yield curve. Corporations are forced to rely heavily on short-term borrowing, putting them into an unstable state, with maturity mismatches between their assets and liabilities. Lack of Liquidity in the Secondary Market The turnover rate in the bond market is relatively low. Illiquidity in the secondary market can be traced to the following factors: most purchasers hold bonds until maturity, no real-time information is available on bond prices and quantities, there are no market makers for bonds, and the credit rating system is underdeveloped. - 69 YONGBEEOM KIM Predominance of Guaranteed Corporate Bonds With investors averse to credit risk and a price determination mecha- nism that malfunctions, guaranteed corporate bonds have come to dominate the Korean bond market. "Guaranteed" here means that guarantees for the payment of principal and interest are made by financial institutions, such as local commercial banks, local merchant banks, local guarantee insurance companies, the Credit Guarantee Fund of Korea, or foreign banks with branch offices in Seoul. Being under close government control, financial institutions have relied on the government to ensure their soundness. The conven- tional wisdom has been that the government would surely give a hand to any faltering or failed financial institutions. Indeed, until 1997, no financial institution was allowed to fail. The implicit gov- ernment guarantee created moral hazard among financial institu- tions. They guaranteed corporate bond issuers without a prudent examination of the credibility and financial soundness of such insti- tutions. Although the guarantee system helped create a leading po- sition in the Korean bond market, it passed the credit risks of an industrial sector on to the financial sector without any filtering. Eventually, the burden of restructuring these industries was trans- ferred to the public. UNDERDEVELOPMENT OF THE BOND MARKET AND THE FINANCIAL CRISIS Many economists have attributed the financial crisis in Korea to the following structural problems: overvaluation of the currency and pro- longed imbalances on the current account, excessive investment in risky and low-profitability projects, moral hazard effects of implicit and explicit government bailout guarantees for banks, and accumula- tion of short-term foreign-currency debt. The overexposure of firms to short-term debt is undoubtedly one of the key causes. The mis- match between their liabilities and their investments made firms vul- nerable to both market and structural risk. Once the market conditions turned against the highly leveraged corporations, the firms experi- enced serious liquidity problems that forced many into bankruptcy. 70 KOREA S BOND MIARKET Indeed, the financial crisis of 1997 was triggered by a series of bankruptcies involving cI,ae60ls, large business groups, that had bor- rowed heavily to finance their investment projects. The spate of bankruptcies started in January 1997 when Hanbo Steel, the four- teenth largest cbaebol, sought court receivership. Sammi Steel, Jinro Group, and Kia Group followed suit in March, April, and July, respectively. In 1996, 20 of the 30 largest conglomerates in Korea showed a rate of return on invested capital lower than the cost of capital. The excessive mismatch between borrowed and invested funds could have been alleviated by making active use of the bond market as a key funding source. In the absence of such a market, firms turned to bank loans. As of the end of 1997, Korea's manufacturing sector was financing only 17.7% of its outstanding debt by issuing bonds. Relying so heavily on bank loans, Korean firms were par- ticularly hard hit when the crisis evolved into a credit crunch in the banking sector. The shock could have been greatly softened had indirect financing not been the only funding route for the business sector, and had bond issuance been operating as an alternative fund- ing channel.' As for the moral hazard problem, implicit government guaran- tees make it difficult to properly assess risk when selecting projects. In the Korean economy as a whole, too much attention was given to investment and too little to risk. The problem would not have been so severe with an active bond market, which would have acted as a competitor to the banking business and thus would have curtailed l For example, if the investment projects driven by Hanbo Steel had been profitable, and if the bond market in Korea had been functioning well, the corpo- ration could have raised a large amount of long-term capital at a lower cost with- out the added worry of frequent repayment on principal and interest. If investment projects were risky or not profitable enough to cover the cost, the corporation would have had difficulty in raising funds and would have given up the investment project owing to high costs. This investment decision mechanism based on capital market response was also applicable to the investmitenit projects driven by all the other corporations. That meant the financial crisis in Korea could have been avoided, or at least softened, had the bond market functioned properly as a provider of long-term capital. 71 YONGBEOM KIM the banks' power in the financial sector. This, in turn, would have reduced the adverse effect on the local business sector caused by the banks' moral hazard behavior. MEASURES FOR RESTRUCTURING THE BOND MARKET In February 1998, on the recommendation of the World Bank, Ko- rea formed a task force to investigate how to improve the govern- ment bond market system and activate the secondary market. With input from a workshop sponsored by the Bank and Korea's Minis- try of Finance and Economy (MOFE), the task force prepared a restructuring plan for the Korean bond market. The government agreed to adopt several of the plan's key policy measures to restruc- ture the primary and secondary markets, along with the infrastruc- ture for the bond market: Developing a Benchmark Rate through the Government Bond Market The problem of low liquidity and lack of a benchmark rate is one of the major problems in Korea's government bond market. In the ab- sence of liquidity, no benchmark rate could be formed on which all fixed income pricing depends. The absence of a benchmark dis- courages fixed income transactions, creating a chronic vicious cycle between liquidity and a benchmark rate. To address this, the gov- ernment is taking several steps to make government bonds a reli- able benchmark. First, it has decided to promote three-year bonds as the benchmark debt instrument, making them much like U.S. Treasury bonds. The proportion of three-year government bonds is soon scheduled to increase to a predominant share of the entire market. In line with a downward stabilization of interest rates, the emphasis will be on expanding long-term bonds with a maturity of five years or more. Second, the government is simplifying the type and kind of gov- ernment bonds. National Debt Management Fund bonds have been renamed "Treasury bonds." Grain Security bonds are being merged with National Debt Management Fund bonds. The government will 72 KOREA S BOND MARKET also examine introduction of a fungible issue system to increase the outstanding volume of a benchmark bond by matching the terms and coupons of new issues with those of existing ones. Enhancing Transparency of the Issuance Process In an effort to reform the process of issuing government bonds, Korea has discontinued the policy of fixing a preset rate for the auction. Instead, free market forces are being allowed to determine the rate. Furthermore, to raise the efficiency in bond offerings and transac- tions, the government has introduced electronic (that is, paperless) tenders at auction. Introducing a Primary Dealer System Under the proposed system, a financial institution that is named a primary dealer (PD) will be given exclusive rights to purchase a portion of sovereign bonds for resale. This measure is designed to make the domestic bond market more competitive and to resolve the liquidity problem in the secondary bond market. Under such a system, the government will be able to issue government bonds pe- riodically. It will also be able to simplify fragmented bonds and es- tablish government bonds as the benchmark for domestic interest rates. As a preparatory measure, the government has allowed every commercial bank to deal in the government securities market as a principal under its own name and without going through a broker. A total of 66 financial institutions -22 domestic banks, 27 securities houses, 8 merchant banks, and 9 foreign bank branches in Seoul - have applied for the primary dealerships. About 24 applicants are to be chosen as primary dealers after their performances are evaluated during the test period. Qualificationsfor Candidacy. Candidate status is restricted to syndi- cate members who have been licensed for dealing in government securities. This limits potential PD candidates to the commercial banks, merchant banks, and security houses duly authorized by MOFE and supervised by the Financial Supervisory Commission 73 YONGBEOM KIM (FSC).2 This group includes branches and subsidiaries of foreign in- stitutions authorized to deal in government securities. The pooI of can- didates from which the first batch of PDs was selected at the end of the test period consisted of those who had submitted letters of intent. Minimum capital requirements and other financial criteria for prudential regulation purposes were not imposed at this stage. Such restrictions might unfairly eliminate some of the institutions, par- ticularly those in the middle of restructuring. After candidates were identified, they were asked to undergo a test period of close monitoring for about three months. At the end of this test period, MOFE invited candidates to apply formally for pri- mary dealership. The first group of PDs was selected from the ap- plicants on the basis of their performance in specified areas during the test period and other qualification criteria. Selection Criteria. During the test period, from March 29 to the end of June, candidates were asked to demonstrate their potential market- making capacity, and their performance was monitored. Test scores were based on the arithmetic average of two subparts, one for the pri- mary market activity and the other for the secondary market activity. Performance in the primary market was assessed primarily in terms of the actual volume taken, not the volume of bids at the auc- tions. The secondary market was divided into the centralized elec- tronic dealer market organized by the Korean Stock Exchange (KSE) and the over-the-counter (OTC) market. In each section of the sec- ondary market, performance was monitored and evaluated on the basis of the trading volumes for the benchmark issues specified as the most recent issues of each type of government securities offered through auctions. Preliminary primary dealers had to make markets at the KSE's inter-dealer bond (IDB) market, providing daily bids and offers greater than 1 billion won on benchmark issues accord- ing to their maturity. Selection of PDs. It was agreed that the selection of PDs needed to reflect the diversity of the institutions (commercial banks, security 2 As a result of government restructuring, the authority to grant financial ser- vice licensing will be removed from MOFE and transferred to the FSC. f74;; KOREA S BOND MARKET firms, and merchant banks) active in the market. To this end, 50% of the predetermined number of primary dealerships to be licensed were distributed among the different types of institutions in propor- tion to their presence in the total applicant pool. Within each cat- egory of institution, PDs were selected on the basis of the quantitative performance index. The rest of the PDs were selected solely on the basis of performance criteria, regardless of the type of institution. In order to alleviate the effect of adverse selection and excessive competition, additional members consisting of up to 20% of the pre- determined number of PDs were nominated as PDs by the minister of MOFE, on the basis of such factors as their prospective contri- bution to the government securities market, amount of capital, and reputation. Obligations of Primary Dealerd. Over time, PDs in the primary mar- ket must meet a 2% minimum underwriting requirement. Their main secondary market obligation is to make a two-way market in govern- ment securities, normally for a specified minimum volume and a maxi- mum spread. The two-way prices are made on the KSE electronic trading system. In addition, a minimum trading requirement similar to that of the primary market is also imposed on 24 PDs. That is, each PD is required to account for at least 2% of the total secondary market trading between dealers in all government securities. The other main obligation is that the PD must provide information to the debt manager. In the primary market, PDs could provide information about client interest ahead of each auction. PDs must also provide daily information about their proprietary trading and own positions. Prii'keged of Primary Dealers. The greatest privilege is the prestige that comes with the status of being a PD. Although it is hard to put a value on being a PD, this is one of the most important marketing tools in fixed-income markets. T'he government does not grant PDs exclusive access to auctions, but it allows them to enter noncompeti- tive bids. The share of the noncompetitive auction is set at 20% of the total volume offered. PDs undertake bidding on behalf of cli- ents. If the share of noncompetitive auction is not sold off, PDs are exclusively entitled to bid on the remainder. Meanwhile, PDs will be granted a standing borrowing facility with the Korea Security - 75 YONGBEOM KIM Financing Company. It is believed that access to such a standing (borrowing) facility could be useful to PDs in the early stages of market development in Korea. Finally, through regular consulta- tions with the debt manager, they could influence government deci- sions on structural issues relating to the market, such as regulatory procedures and rules of conduct, and on more operational issues such as types of the securities to be issued. Implementing the Primary Dealer System. The PD system officially went into effect at the beginning of August 1999. The first PDs to be appointed consisted of 11 securities houses, 13 commercial banks, and 1 merchant bank. So far, the market activity of these PDs has been invisible because of adverse market conditions: for example, no additional supply of treasury bonds has been available since May, and market yield has begun an upward trend owing to the fast eco- nomic recovery and possible dumping of ITC holding bonds caused by Daewoo's credit crunch. Thus it will take time for the PDs to play an effective role in the government bond market. Participation of IndividuaInveetors in the Bidding. Along with in- troducing a primary dealer system, the government has adopted a "noncom" bid system for individual investors. The government is assigning 20% of newly issued treasuries securities to individual investors and allowing them to bid. Under the revised bidding pro- cess, an individual investor can subscribe for a minimum of 1 mil- lion won and a maximum of 1 billion won through 24 primary dealers or financial institutions. A person making a noncompeti- tive tender will get them at the average price of the competitive bids accepted. Improving the Infrastructure of the Secondary Market The government is setting up a system of disclosing major quotes in real time. Financial institutions having an exclusive government bond dealership (primary dealership) must submit bid-ask orders to the stock exchange, where, in turn, they are announced in real time. To enhance the role of credit rating agencies, the government is en- couraging foreign credit-rating agencies to enter into the domestic 76 KOREA S BOND MARKE T market (or into joint ventures). The settlement system for bond trans- actions will be improved with introduction of a delivery-versus-pay- ment (DVP) settlement for OTC bond transactions; this will be established through a linkage between the Korea Securities Depository's (KSD) securities settlement system and the payment system of the Bank of Korea (BOK). On November 15, 1998, in an effort to activate the secondary market and to improve the transparency of trust fund management, the government introduced marking-to-market of bond portfolios held by institutional investors. Before then, financial institutions were permitted to swap bonds among the different funds they operated to guarantee, more or less, stable yields at maturity for their custom- ers. The new regulation no longer allows fund managers to transfer bonds into different funds. Instead, they are able to trade bonds held in trust funds in the secondary market without the worry of invest- ment loss. However, marking-to-market is required only for exist- ing trust funds. It will not be required for newly established funds until the middle of 2000. Diversifying the Bond Market To diversify market structure and provide hedging tools, the gov- ernment is introducing asset-backed securities and mortgage-backed securities, and is establishing an interest rate futures market, among other measures. To create demand for bonds, particularly long-term bonds, the government is expanding the investor base by increasing the num- ber and size of financial institutions. As a first step, beginning in October 1998, it allowed mutual funds to be established. Now, it is also considering introducing corporate pension funds and open-end trust funds, as well as providing attractive tax benefits to foreign institutions investing in domestic fixed-income securities. NEW FEATURES OF THE KOREAN BOND MARKET The bond market has experienced significant structural changes since the government began promoting greater activity in this 77 YONGBEOM KIM field. The restructuring process appears to be headed in the right direction, if the following developments are any indication: gov- ernment bonds are growing in importance; structurally, there has been a clear movement from guaranteed corporate bonds to non- guaranteed bonds; the issuer base has expanded from the "Big Five" chazebols to other corporations; and the shape of the term structure has changed. Growing Importance of Government Bonds As mentioned earlier, the Y rrean bond market has long been domi- nated by corporate bonds, even when government bonds were not functioning as a benchmark instrument. Before August 1998, mar- ket participants used three-year guaranteed corporate bonds as benchmark instruments of debt and from then until recently used non-guaranteed corporate bonds. Corporate bonds became pre- dominant in part because government bonds were in insufficient supply. In general, the government bond market was strongly in- fluenced by the government's fiscal policy. Efforts to maintain fis- cal soundness hampered the development of the government bond market. Things are changing drastically now. The government injected about 60 trillion won into restructuring the nation's financial and industrial sectors, and in the process produced a huge fiscal deficit. To finance this deficit, the government issued bonds in various forms. The public issuance of government bonds totaled 15.1 trillion won in 1998, 76% greater than in 1997. Meanwhile, large volumes of government bonds are issued on a continuous basis, and the various types of government bonds are unified into a single treasury bond. The share of treasury bonds in gross government bonds increased by 11 trillion won to 17.3 trillion won, which improved the market- ability and liquidity of treasury bonds. The increased liquidity of treasury bonds can be seen from the changing market share held by government bonds in the secondary market, which jumped from 1% in 1997 to about 20% in 1999. As corporate bonds slipped from their benchmark position, market participants began looking for more reliable benchmark instruments. Since 1998 government bonds have reflected this change in market 780 KOREA S BOND MARKET mood to one favoring government bonds, with yield spreads between treasury and corporate bonds widening since October 1998. With the government's continued reform efforts, treasury bonds will be a reliable benchmark in the near future. Structural Change in the Corporate Bond Market Before the financial crisis, the vast majority of corporate bonds were issued as fixed-rate coupon bonds, which were guaranteed by local financial institutions such as commercial banks, merchant banks, and securities companies, for example, Daihan Fidelity and Surety Co., and Hankuk Fidelity and Surety Co. In addition, most corpo- rate bonds had a three-year maturity and were redeemed at matu- rity. Since the financial crisis, the corporate bond market has been diversified in several aspects. Moveenentfronz Guaranteed to Non-guaranteed Bonds. After the worst economic crisis in the country's history, nonguaranteed corporate bonds became the major fixed-income instrument in the Korean bond market. Concerned about their financial soundness, financial insti- tutions were reluctant to guarantee the payment of corporate bonds. In addition, as the number of corporate bankruptcies grew during the financial crunch, the guarantee insurance companies, the only available guarantors for corporate bond issuers after the financial crisis, also experienced serious financial insolvency. Consequently, few financial institutions were able to guarantee payment on inter- est and principal. Corporations now have to issue bonds based on their own credit, without any credit enhancement from other finan- cial institutions. Consequently, there has been a significant increase in the issu- ance of nonguaranteed bonds in recent months. The ratio of nonguaranteed corporate bonds to listed corporate bonds skyrock- eted from 10% at the end of 1997 to 97.2% in July 1999. In addition, the market price of corporate bonds came to be based on the credit rating of each corporation. Hence the yield on corporate bonds now reflects the real difference in default risks among issuers. As credit ratings take on more importance, various measures are being taken to improve the function of credit-rating agencies and to 79 YONGBEOM KIM increase the credibility and power of discrimination of credit rat- ings. To this end, one domestic rating agency recently established a joint venture with a prominent U.S. credit-rating agency. Spread of the Issuer Base from the Big Five Chaebols to Other Cor- porations. Since non-guaranteed bonds were the only fixed income instruments available to bond investors, institutional investors sel- dom exposed to credit risk grew more cautious with their bond investments. To lower the possibility of default, investors purchased only highly rated bonds, most of which were issued by the Big Five c/aebo!s. Therefore large corporations raised funds freely and easily by issuing bonds. Thus the chronic problem of excessive capital among the Big Five worsened, pushing other corporations into the worst financial hardship in their history. Corporate bonds issued by the Big Five, estimated at 11.08 trillion won in 1996, accounted for 37.5% of the total. As a result, their rate of 'mo- nopoly' in the bond market soared from 56.9% in 1997 to 70% in 1998. To weaken the chaebols' grip on the domestic money market and to help ease the credit crunch for smaller firms, the govern- ment imposed a ceiling, on October 28, on the holdings of bonds issued by the Big Five. That is, domestic banks and insurers were barred from owning corporate bonds issued by the Big Five in excess of 10% of their total bondholdings. The comparable ceiling for investment-trust firms will be 15%. The corporate bondholdings above the law-stipulated level should be disposed of by the end of 2000. Since the implementation of these measures, the Big Five's share of new issuance in the corporate bond market has dropped, from 68.7% in 1998 to 33.8% in 1999. Whether or not this may be just a voluntary change in the investment pattern, investors are buying non-guaranteed bonds issued by corporations other than the Big Five, with relatively less resistance to credit risk. This is thought to be a sign that the corporate bond market is beginning to work properly as a provider of long-term capital, not only for corporations with high credit rating but also for low-rated corporations without any credit enhancement from other financial sectors or institutions. 80 KOREA S BOND AMARKET In summary, since the financial crisis the market benchmark has changed as follows: in the first quarter of 1998, corporate bonds were bank guaranteed; up to the end of August 1998, they were insurance company guaranteed; after that, they were non- guaranteed, with a credit rating over A+. Change in the Shape of the Term Structure In the face of a chronic shortage of investment resources and persis- tent high inflation, short-term yields were usually higher than long- term yields. In addition, bond maturities were abnormally short, averaging only about three years, mainly because of investor aver- sion to interest rate risk. For the most part, the yield curve sloped downward until April 1998, when the government began focusing on stabilizing the foreign exchange rate, and opted for a contractionary monetary policy. In response to the lack of cash demand by businesses for fresh investment, a large trade surplus, surging capital inflow, and sluggish domestic consumption, the foreign exchange market became more stable and the government began lowering interest rates to boost the sagging economy. Subsequently, market interest rates regained their stability faster than expected. In 1999, the yield on three-year corporate bonds fell from an annual rate of 18.28% at the end of March to 9.55% on November 25, with a temporary rise to 10.59% on September 18. The drop in the short-term money market interest rate was more dramatic than that for longer-maturity money. On September 18, the overnight call rate plummeted from 22.19% to 4.77%. As a result, the yield curve began an upward trend, augmented by the financial re- sources being circulated only within the financial institutions, owing to the persistence of credit risk and uneven performance of industry. Providing higher returns to long-term bonds not only makes them more attractive but creates new demand for them. In general, inves- tors typically expect a higher return for securities with high risk, under the liquidity hypothesis of the maturity structure of interest rates. With short-term interest rates being lower than long-term rates, the Korean bond market is able to extend the maturity spectrum to longer terms and develop a meaningfulyield curve for use in pricing corporate bond issues. 81 YONGBEOM KIM AGENDA FOR THE FUTURE Implementing a Delivery-versus-Payment (DVP) System Since most bonds are traded in block by financial organizations, the OTC market has been the major trading arena. In the OTC bond market, however, settlements take place on a free-of-payment basis, in the absence of regulations regarding cash settlement. This exposes trading parties to settlement risks. DVP will remove settlement risk, by ensuring that cash is paid only when the securities are available, and vice versa, so that there is no risk of giving away one asset (cash or securities) without receiving the other in return. Normally, this would be achieved by linking the wholesale payments system (Bank of Korea (BOK) wire) with the electronic book-entry settlement system (Korea Securities Depository (KSD)). The BOK and KSD are currently making steady progress in their discussion of a DVP system. Introduction of Interdealer Brokers Now that the government securities market is becoming more di- versified, a primary dealer system has been introduced and a pro- gram for restructuring bonds is being implemented, attention turns to creating a new type of intermediary facility, known as the interdealer broker (IDB). As yet, there are no clear guidelines or legal definition for IDBs. Their role in the bond market, business characteristics, business area in which they will participate, and su- pervisory system therefore need to be clarified. Activate Bond-lending and Repo Market Repos can be a very powerful tool for dealer financing. They can facilitate management of bond portfolios by allowing for short posi- tions. At present repo trading and bond lending face several ob- stacles, such as tax problems, legal issues related to bankruptcy law, and the lack of a viable transaction system. Without a repo market, short-term borrowers have no choice but to rely on high-cost financing in the call market. 82; KOREA'S BOND MARKET Along with repo trading, the bond-lending system can improve liquidity in the bond market. Bond lending and borrowing is a prin- cipal means for hedging in an open and deregulated financial mar- ket. In addition, bond lending and borrowing activates arbitrage transactions between the cash and futures market. This may give institutional investors a new way of managing assets and a new source of financing. A few working groups sponsored by the gov- ernment are investigating the problems entailed in developing these suggestions. Prepare a Demand-stimulation Plan The investor base for bonds needs to be widened to promote a more competitive and transparent bond market environment. If only a few financial institutions are allowed to hold most of the outstand- ing government bonds, those few investors will be able to set bond prices, liquidity will decrease, transaction costs will rise, and indi- vidual investors will be kept out of the bond markets. The fact that a handful of institutional investors (such as invest- ment trust companies and commercial banks) dominate the Korean bond market is one of the key causes of the market's high volatility and low liquidity. To attract more bond investors, the government should consider removing various restrictions on the operation of their Financial assets and giving them more preferential tax treat- ment. The National Pension fund invests almost 70% of its assets in the public sector, in effect giving loans to government departments. Converting these loans to government bonds would create a new demand for government bonds. A new study seeking ways to en- courage the demand for bonds also suggests focusing on tax incen- tives for bond investment, such as deferring the tax levied on interest when bonds are borrowed or returned or exempting the withhold- ing tax on the interest income levied on mutual funds. 83 CHAPTER 7 The Development of a Government Bond Market: The Australian Experience JOHN BROADBENT The gradual evolution of a liquid and efficient government bond market in Australia can be traced largely to the opening up of its financial system in the early 1980s. The evolution has been gradual, with innovations stemming from a recognition that the market needed to operate differently, if the system were to become more efficient. Though deep and liquid, this market remains small compared with those of many other countries, both in absolute terms and as a proportion of GDP (see table 1). In absolute terms, it is around US$60 billion and represents less than 20% of GDP. Its relatively small size is a reflection of Australia's generally low level of government debt-we have not gone out of the way to promote a government bond market for any perceived advantages that it may provide to the financial system more generally. Yet, our bond market rates quite well internationally in terms of liquidity, as indi- cated by the quite tight spreads generally seen in the secondary market. Figure 1 plots the size of 24 major government markets against a typical bid/ask spread for a trade in a liquid linel of stock. Each market in the sample deals in over US$10 billion and has unrestricted access to foreign investors. Australia's typical spread is lower than in markets of comparable size and lower than in many other larger markets. JOHN BROADBENT Table 1. Government Bond Markets in Selected Countries Country Billions of U.S. dollars Percentage of GDP Italy 971 85 Belgium 191 79 Greece 69 57 Denmark 90 53 South Africa 61 47 Sweden 102 45 Japan 1,855 44 United Kingdom 459 41 France 484 35 Canada 210 34 United States 21741 34 Spain 178 33 Finland 40 33 Germany 653 31 Portugal 29 28 Malaysia 19 27 New Zealand 11 16 Singapore 14 16 Australia 58 15 Switzerland 27 10 Hong Kong 13 8 Korea 23 8 Norway 19 8 China 66 7 Source. Salomon Smith Barney, APEC Survey, end 1997. Looking back over the past few decades, I think that several factors stand out as having fostered liquidity in our market: Austra- lia accepts market prices by tendering government securities, it has consolidated outstanding bonds into benchmark lines of stock, and the central bank has been actively involved in the market. TENDERING Since the early 1980s, bonds have been sold through a tender auc- tion system. The Treasury determines the amount of debt to be sold and the market sets the price via competitive bidding. The amount of bonds issued at tender in any Financial year depends on the size of 860 THE AUSTlRALIAN EXPERIENCE Figure 1. Bid/Offer Spread in Government Bond Markets Basis Basis Points Points 20 - - 20 15 - 15 10 _ 10 * Australia 5 _ *5 0 * 10 25 100 2500 Bonds on issue (US$b); log scale Source JP Morgan. the Commonwealth government's budget deficit and the volume of maturing bonds that need to be financed. An estimate of the amount that needs to be issued is provided to the market at the start of the year, and there is a fairly regular issuance calendar that spreads the task throughout the year. What I have described is, of course, based on the notion that the government fully funds its deficit by borrowing from the public at market prices. It is the bedrock of sound financial management, with the government no longer tempted to cover its revenue shortfall through central bank financing. Prior to the introduction of tenders, however, arrangements for issuing and marketing government se- curities were not well suited to the large volumes involved. The gov- ernment set the yields and accepted whatever quantities were offered at those yields. Under those conditions, investors were mainly inter- ested in "captive market" arrangements, which required or encour- aged certain financial institutions to hold minimum quantities of government securities. There was very low activity in bond markets at this time (figure 2). Initially, the tendering process did not provide either the central bank or the Treasury with a particularly easy ride. Figure 3 shows 87 JOHN BROADBENT Figure 2. Government Bonds $B $B Average Daily Turnover (RHS) 80……- - ___------ _ - -…4 60…- _ __________ _ ___-_-_-_-/-- -- -- 3 Stock outstanding (LHS) 40 …2 20 -- __ - _ -- _ 1 0 0 82/3 84/5 86/7 88/9 90/1 92/3 94/5 96/7 98/9 Figure 3. Bond Tenders: Spread Basis Basis Points Points 120 - -- - - - - - - - - - - 120 100 -t - __ - - - __ __ - __ - - - - - 100 80 - - - } --- - --- ----- 80 60 … - - - - - - - 60 40 - - - - - - -…-- - - - - - - - - - - - - - - - - - - - - - - - - - - - 40 20 - - - - - - - - - - -V - - -e - - - 20 20 -j-20 10 20 30 50 70 90 110 130 150 Tender No. that spreads at tender were very wide: the average spread in the first 10 tenders was about 60 basis points. This partly reflected the nature of the market at the time. There was little demand from in- vestors because they were coming from a captive market environ- ment, market-making intermediaries were unaccustomed to such large THE AUSTRALIAN EXPERIENCE price fluctuations, and there was an issuer that needed to move a large volume of securities. But these pressures settled down fairly quickly as market liquidity improved. In the past decade, the range of bids in tenders has averaged about 4 basis points. Markets quickly adapt to new opportunities when they present themselves. But the central bank made a few changes to the tender- ing process so as to reduce uncertainty in the market and to encour- age further development. First, it cut back significantly the period between bidding and announcing the results. At the first tender, bids closed at noon on Thursday, and the results were not announced until 5 P.M. on Friday, after the market closed. Participants could not adjust positions until markets opened on the Monday. By the eighth tender, this gap had been cut significantly, so that results were available when markets opened the following day. With the intro- duction of electronic tendering, results are now announced just an hour and a quarter after the tender closes. The speedier response to tenders has helped narrow the range of bids by significantly reduc- ing market risk between the time that a bid is entered until the re- sults are known. Second, the bank tried to improve trading in the secondary mar- ket by encouraging more participants to make a market. It did so by altering the counterparties it would be willing to deal with in the government securities markets. Typically, a central bank is a fairly large player in its own market by virtue of the size of its balance sheet. In the mid-1980s, for instance, Australia'a central bank in- creased the number of counterparties that it would be willing to undertake transactions with in government securities, on the basis of the level of activity that each of these dealers undertook in the markets. Dealers with relatively high levels of turnover (more than I% of total market turnover) were accorded the status of 'reporting bond dealer,' which gave them the opportunity to undertake trans- actions in longer-dated government securities with the central bank. CONSOLIDATION OF SECURITIES ON ISSUE Australia has also sought to improve liquidity in the market by building up benchmark lines of stock. This has not only limited the number - 89 JOHN BROADBENT of individual securities on issue, but has allowed outstandings of each stock to rise to a critical mass that promotes liquidity in that issue. Investors feel confident that they can undertake reasonably large transactions without moving the price too far against them- selves. Market-makers can quote two-way prices knowing that an adequate level of stock is likely to be available in the market to fill orders. The issuer, the Commonwealth, shares in the benefits by being able to finance its spending more cheaply than would otherwise be the case. In the past financial year, Australia has built up several lines of stock at tender (figure 4). There were seven tenders during the year with a total gross issuance program of around $4 billion to cover maturing Commonwealth Government Securities (CGS). (Because the government is running a fiscal surplus, net issuance was actually negative.) The June 2011 bond issue2 was a new line of stock with an initial tender of $600 million sold to the public. Being a new line, an additional $200 million was also taken up by the Reserve Bank for its portfolio. (This is shown by the lightly shaded area of figure 4.) Over the remainder of the year, further tenders added another $2'/2 billion of stock to the market, bringing the total amount issued to over $3 billion. Successive tenders will probably endeavor to build Figure 4. Bond Issues in 1998-99 $M $M 800 800 * Sep 2009 MalurtyL 700 F] Jun 2011 Maturity 700 600 600 500 500 400 400 300 300 200 200 100 100 0 - 0F---- T 25 Aug 22 Sep 10 Nov 19 Jan 9 Mar 13 Apr 25 May THE AUSTRALIAN EXPERIENCE Figure 5. Bonds Issued $B $B 6 6 5 5 4 4 1 1 0 IL0 0000 0iC a , L C, , O CNoo,oC C_)- , o C the outstanding amount of this bond line to around $5 billion to $6 billion. The other line of bonds offered were maturing in September 2009. In 1999, $800 million was offered, bringing the total on issue to over $5 billion, a level that we see as being large enough to ensure that there is good liquidity in the market. The stock on issue has been consolidated over a number of years. Currently, there are 15 bench- mark lines; they total about $70 billion and make up over 95% of the bonds outstanding (figure 5). In marked contrast, in 1985 when the bond market was still in its infancy, there were 128 separate series of Treasury bonds on issue, with an aggregate face value of about $36 billion. This represented an average of just over $280 million per line of stock. Australia has been consolidating the lines over a number of years, though adjustment has speeded up since it began retiring some lines, so that liquidity might be further enhanced in the benchmark lines. Most often, this has been undertaken by the Reserve Bank in the normal course of managing its own portfolio. In turn, the bank sells the stock to the Commonwealth for retirement. Less frequently, the Commonwealth has also undertaken reverse tenders in an attempt to draw in selected lines from the market. 91 JOHN BROADBENT ACTIVE CENTRAL BANK INVOLVEMENT IN MARKETS The central bank has always sought to promote efficient bond trad- ing, thereby adding to market depth and improving liquidity in the government securities market. It has done this by improving market infrastructure, seeking an appropriate form of dealer arrangement, providing timely data on the market, and maintaining a relatively high profile in the repo market. Improving Market Infrastructure During the 1990s, we set out to strengthen settlement arrangements for government securities to ensure they continued to meet interna- tional best practices. Securities are now settled on a delivery- versus-payment (DVP) basis in real time and in electronic form. Seeking an Appropriate Form of Dealer Arrangement Initially, the Bank offered special arrangements to a select group of Financial institutions, the so-called authorized money market deal- ers. These dealers were obliged to make markets in short-term Com- monwealth government securities and in repurchase agreements based on Commonwealth government securities. In return for these and other repository facilities for the banking system, they were granted access to liquidity support at the central bank and the right to be a counterparty to the Reserve Bank's domestic open market opera- tions. The central bank encouraged market-making for longer-dated securities with only with a limited range of counterparties that were active in these securities. These arrangements served the market well in the early stages of its development. Over time, however, changes in the institutional arrangements, growth of the market, and con- cerns about perceptions of the status accorded to these dealer net- works suggested that it was no longer necessary to give these groups special recognition. As a result, the bank decided to deal with any counterparty that was a member of the settlement system for Com- monwealth Government Securities. (At the time, the Reserve Bank would only accept Commonwealth Government Securities as col- lateral in its repo transactions with the market.) This greatly ex- 92 THE AUSTRALIAN EXPERIENCE panded the range of counterparties that the Bank was willing to deal with. It encompassed about 150 institutions, including the banks, those who were previously authorized dealers or reporting bond dealers, and other significant holders of Commonwealth securities, such as insurance companies and nominee companies. Providing Timely Data to the Market The Reserve Bank publishes data on activity in the bond market on a daily basis (table 2). These include a survey-based average of bid/ offeryields, turnover, and the amount on issue for each of the bench- mark stocks. The data provide an objective measure of the yield curve as well as an indication of the relative levels of activity in selected stocks. Maintaining a Relatively High Profile in the Repo Market The repo market is essential in promoting liquidity in the govern- ment securities market. As well as encouraging participants to fol- Table 2. Daily Information on Government Bonds Released by the Reserve Bank of Australia Indicative yield Turnover Total on isswue Day mid rates (millons of dollars) (miHlons of dollars) maturity 28-Sep 27-Sep 27-Sep 24-Sep 28-Sep Apr-00 4.73 4.70 220 55 5,995 Jul-00 4.70 4.66 117 174 2882 Jan-01 5.10 5.04 134 73 5,199 Nov-01 5.28 5.21 114 358 2638 Mar-02 5.36 5.29 595 668 3,904 Oct-02 5.50 5.44 54 191 4,497 Aug-03 5.63 5.57 496 417 5,290 Sep-04 5.75 5.69 388 72 5,712 Jul-05 5.88 5.82 394 276 5,502 Feb-06 5.97 5.91 345 244 4,102 Nov-06 6.02 5.96 280 315 6,103 Oct-07 6.12 6.06 229 344 4,207 Aug-08 6.16 6.11 444 508 4,494 Sep-09 6.18 6.12 953 1074 5,309 Jun-11 6.25 6.20 256 264 3,395 Source: Reserve Bank of Austra ia Daily Press Release. 93 JOHN BROADBENT low best market practices (such as using the Public Securities Asso- ciation/International Securities Markets Association Master Agree- ment for transactions), the Reserve Bank is an active participant in this market (figure 6). The bank carries out most of its market op- erations in the repurchase market. This provides it with the flexibil- ity needed to manage cash conditions and to implement monetary policy effectively. We are also reasonably active in the securities lending market using repurchases. Being a fairly large player in the domestic mar- ket (the bank holds more than 10% of total Commonwealth securi- ties on issue), our portfolio can provide a source of specific lines of stock in times of market squeezes. We are not "on the frontline" of the securities lending market, but rather we are an active partici- pant if we think that a squeeze in a specific line of stock may not be in the best interests of the market in general. CONCLUSION A deep and liquid government bond market is a worthwhile addi- tion to a nation's financial infrastructure. While there are no golden Figure 6. Turnover in Repurchase Agreements on Government Bonds *a BP) $B (0) Total Turnover 10 _-- - - - - _-- - - - - - - - -__- - _-- - -_-- - - -_-10 8 ___- -- -- - -- - - - - - - - - - - - - - - - 8 2 - - - - - - - - - - - - - - - - - - - - - - _ _ _ _- - - 2 RBA Transactions 0 0 84/5 86/7 88/9 90/1 92/3 94/5 95/7 98/9 94X1 1'HE AUSTRALIAN EXPER[ENCE rules that ensure a deep and liquid market, some basic elements need to be present if a market is to flourish. At a minimum, systems must have clear and unambiguous rules and procedures, must be legally robust, and be subject to effective regulation. Once these conditions are met, some factors that I have alluded to can further encourage market development. In Australia, we are faced with a declining market in govern- ment securities as the government continues to run fiscal surpluses. In line with the reduction in bonds outstanding, turnover has also declined. One of our challenges over the next few years, therefore, is to continue on the present course and to maintain a deep and liq- uid market for government securities. 95 CHAPTER 8 The Development of a Coporate Bond Market: The Malaysian Experience RANJIT AJIT SINGH The development of a corporate bond market is a key strategic priority for capital market development in Malaysia. Although the government has made significant strides in this direction since 1990, the development of the corporate bond market still has some way to go. The need to continue developing the Malaysian corporate bond market was further emphasized by certain weak- nesses in the financial system highlighted during the recent Asian economic crisis. Nevertheless, there are some useful insights that can be gained from Malaysia's experience, by looking at the historic role of the bond markets in financing Malaysia's economic growth, the development efforts so far, the present structure of the corpo- rate bond market, and some of the key issues that still need to be addressed. Malaysia has experienced tremendous economic growth in the past decade, averaging 7% or more, despite the effects of the 1997- 1998 Asian crisis. Over this last decade, the economy has shifted from a principally agricultural base to a manufacturing base. Some aspects of this growth are particularly noteworthy. There has been a This chapter reflects the views of the author and not necessarily the views of the Malaysian Securities Commission. RANJIT AJIT SINGH strong pro-business environment. The government of Malaysia also initiated policies to ensure that most of the funding for economic growth was carried out by the private sector in line with its privatization efforts in the 1980s. Malaysia has had considerable infrastructure development needs, which require more long-term financing. Projects such as the Kuala Lumpur International Airport and the North-South Expressway, among others, have provided a strong impetus to the market. As a result of these policies, as well as the recent infrastructure development efforts, Malaysia's financing needs have risen sharply. According to the Seventh Malaysian Plan for the period 1996-2000, long-term financing needs are around US$140 billion. Although a substantial amount of these financing needs had been funded by bank loans, the corporate bond market also began to play an increasing role, growing from 1 .5 billion ringgit of private debt securities outstanding in 1989 to approximately 47 billion ringgit in 1997. Furthermore, liberalization of the financial markets, which en- sured the diversification of funding sources, as well as Malaysia's high savings rate, which has averaged above 40% of GDP for the last few years, have provided additional catalysts for the significant growth of the corporate bond market during this period. Therefore it is important to note that the need to develop the bond market has always been recognized by the government, and the growth over the last decade can, in fact, be traced to the Malaysian government's strong strategic push for the market's development. One of the specific steps taken to develop the bond market dur- ing this period focused on the regulatory framework. The central bank introduced basic guidelines on the issuance and trading of cor- porate bonds, as well as criteria to be used in evaluation of corpo- rate bond proposals. These issuance guidelines provided a degree of transparency and certainty to the market at the initial stage of the market's development. In addition, several amendments were made to legislation to encourage greater investor participation, such as allowing unit trusts to invest in bonds with a BBB rating and insur- ance companies to invest in unsecured bonds with an A or P2 rat- ing. Another key development was the incorporation of Cagamas Berhad (the Malaysian National Mortgage Corporation) by the Malaysian government in December 1986 with the objective of pro- W98 T'HE MALAYSIAN EXPERIENCE moting the development of a secondary market in Malaysia by pur- chasing mortgages from primary lenders and issuing debt securities. Cagamas began operations in October 1987 with a purchase of RM120 million in housing loans from commercial banks and the issue of RM100 million of mortgage loans. Another notable effort was the establishment of the Rating Agency of Malaysia (RAM) in 1990, to provide an expert and independent view of the credit quality of private debt securities (PDS). The rat- ing agency was set up as a private limited company with equity par- ticipation by the financial institutions in the country. This was followed by the establishment of a second rating agency, the Malaysian Rat- ing Corporation (MARC) in 1995. In addition, from 1996 all un- listed PDS issues were made scripless and trade through the electronic inter-bank funds transfer and scripless book-entry system (SPEEDS), a computerized scripless securities trading system operated by the central bank. This made the secondary market trading environment for unlisted corporate bonds more efficient. One significant development in terms of benchmark securities was the issuance of Khazanah bonds. As Malaysia moved towards a diminishing government deficit financing requirement, there was a fall in the level of government securities being issued. This lack of supply became a concern to the government in its attempts to sus- tain an efficient benchmark yield curve. Khazanah is the Malaysian government's investment corporation, and these bonds were designed to provide an alternative to Malaysian government securities as the benchmark bond. The Khazanah bond was introduced in 1997 to provide a regular issue of paper along the lines of the successful exchange fund notes issued by the Hong Kong Monetary Authority, which provide a good benchmark proxy for bond issues in Hong Kong. Furthermore, in an effort to enhance secondary market trans- parency, a Bond Information and Dissemination System (BIDS) was established in October 1997, which allowed a range of market price information to become available to market participants. On the fiscal side, efforts focused on reducing the transaction costs of investing in these markets. A key issue that emerged was the level of stamp duty on the issuance of corporate bonds. There was a waiver for Malaysian government securities, but a similar waiver did not - 99 RANJIT AJIT SINGH exist for corporate bonds. By way of a fiscal incentive, the govern- ment extended the wavier to corporate bonds: a tax exemption was given for the interest earned by individuals on corporate bonds listed on stock exchanges; in 1993 nonlisted bonds also became subject to the exemption, as long as they were rated by the rating agency. As a further incentive, some exemptions were granted for income received by unit-trust and closed-end funds. Credit-risk issues were examined also. In 1991-92, a compul- sory system of rating bonds was introduced. Investment grade re- quirements were established for issuers to ensure that paper coming into the market was of high quality. In addition, the government decided that ratings for bank-guaranteed issues should reflect rat- ings of individual guarantor banks. However, although the invest- ment grade restriction, for instance, worked well in the initial stages of bond market development, it may be having an adverse effect on the diversity and supply of paper today. Malaysia is therefore look- ing closely at allowing non-investment grade paper onto the market in the future. Although the market has seen some growth and vari- ous initiatives have been introduced, the corporate bond market still has some way to go to match the average financing profile of devel- oped countries. This can be seen from the composition of the finan- cial markets in 1998: the equity market accounted for about 134% of GDP, banking about 160% of GDP, and the bond market, particu- larly the corporate bond market, only about 25% of GDP. The Asian crisis certainly helped to further focus attention on the underdevelopment of bond markets. Over-reliance on the bank- ing sector brought tremendous strains because of exposure to po- tential credit withdrawals. Maturity mismatches created another problem. The limited opportunity to diversify portfolio risk argu- ably aggravated capital flight. These difficulties created a further impetus and prompted the Malaysian government to push ahead with greater urgency towards the development of the bond markets. In order to appreciate the relevance of initiatives to be taken to develop the corporate bond market in the future, it is essential to understand the current structure of the market today. In terms of the regulatory structure, several authorities play a role in the issu- ance process for listed bonds. First, the central bank evaluates issu- ance proposals based on a set of PDS guidelines as well as for monetary THE MALAYSIAN EXPERIENCE policy concerns. Second, the Securities Commission is involved in evaluating the issuance of listed bonds from a disclosure and utiliza- tion of proceeds perspective. Third, Malaysia still has a registrar of companies looking at prospectus content issues. Fourth, the rating agencies have to rate issues to ascertain whether they meet invest- ment grade requirements. Fifth, the stock exchange also comes into the picture for listed bonds with its listing guidelines. All these insti- tutions create a regulatory structure that is quite fragmented and cumbersome to the issuer. As a result, it can take four to six months for a corporate bond to become listed on the stock exchange. The secondary bond market structure today can be divided into the unlisted, over-the counter (OTC) market, which is run by the central bank, and the listed exchange-traded market which is regu- lated by the Securities Commission. It should be noted that the un- listed corporate bond market is much more active than the listed market. Market makers on the OTC market are, however, restricted to financial institutions that are governed by the Banking Act or specifically approved by the central bank. This therefore excludes securities houses from participating in this market. This fragmented trading structure, as well as restrictions on the number and scope of participants in the over-the-counter market, has caused the second- ary market for bonds to be relatively illiquid. In terms of the range of instruments available in Malaysia, one major issue needs to be addressed. Many issues that have come onto the Malaysian market have been corporate bonds with equity "sweet- eners" such as Transferable Subscription Rights (TSRs) and Irre- deemable Convertible Unsecured Loanstocks (ICULs). These hybrids are extremely popular, particularly because the Malaysian market typically has a strong appetite for equity exposure. Moving forward, certain critical issues have been identified that need to be addressed in relation to the overall development of the corporate bond market. One of the key concerns is that Malaysia does not have an appropriate liquid benchmark yield curve. In or- der to establish an appropriate benchmark yield curve, the bench- mark security should be fungible, default-risk free, sufficient in size and frequency of issuance, and also span the maturities required. Malaysian government securities, Khazannah bonds, and Cagamas bonds are among the viable alternatives being considered as the 101 RANJIT AJIT SINGH benchmark security. However, whether the establishment of a liq- uid benchmark yield-curve requires the rationalization of these al- ternatives or whether several benchmarks can coexist is an issue that needs to be addressed, as well as issues related to issuance in a surplus budget environment for government securities. As for the secondary markets, Malaysia needs to free up some of the restrictions that have been imposed that create a captive demand for government securities, which by its nature has the most market acceptance as the benchmark security. Because of certain invest- ment, statutory reserve, and liquid asset requirements, much of the government paper issued is locked in. For example, the Employees Provident Fund (a statutory fund), which has significant investment capacity, is required by statute to have a very large amount of its funds invested in government securities. As this fund holds on to these government securities until maturity, partly because of the shortage in the supply of these securities, there is a lack of second- ary market liquidity in government securities and this has inhibited the development of an efficient yield curve. Repurchase agreements (repos) or, more generally, securities lending and borrowing provide efficient mechanisms for financing positions and generally are considered to facilitate secondary mar- ket liquidity. Repos enable market intermediaries to take long and short positions in a flexible manner, buying and selling according to customer demand on a relatively small capital base. While a repo market currently does exist in Malaysia, it is not sufficiently deep to support liquidity in the government securities market. Therefore, the range of participants in the repo market will need to be looked at to allow other selected participants (such as institutions licensed under the Securities Industry Act and fund managers) to conduct repo transactions with the banking institutions. Liquidity and efficient price discovery in secondary markets are also hampered by the lack of an organized futures market offering interest rate futures and options on government securities. The regulatory framework is also another priority in terms of current reform efforts. The entire process of obtaining approvals of all the relevant authorities currently can take about 4-6 months from the date of making the first submission to the rating agency. The problems that may arise within this regulatory process include ex- 102 THE MALAYSIAN EXPERIENCE posure to interest rate movements, regulatory overlap in exercising the functions of the three respective regulatory bodies (the Securi- ties Commission, Central Bank and Registrar of Companies) and the preference for bank loans, which in contrast take no more than two weeks to process. The government of Malaysia has taken steps towards establishing a more efficient issuance process by announc- ing that all the necessary laws and regulations should be amended to provide the Securities Commission with sole authority over the cor- porate bond market and that a shelf-registration system should be introduced, where only a one time approval is necessary for the reg- istration of the prospectus with the Securities Commission. Further- more the current process of moving towards a disclosure-based regulatory framework is also likely to facilitate issuance. Among the proposals identified to help diversify and widen the issuer base are the setting up of a Financial Guarantee Insurer, as a substitute for bank guarantees, which might reduce costs substan- tially, and allowing non-investment grade paper on to the market, which would let more Malaysian companies issue debt as an alter- native source of financing. Furthermore, promoting policies and a framework for asset securitization would allow the increasing amount of credit card, hire purchase and housing mortgage loans to become tradable. The key reform needed on the investor side is to increase the involvement of pension funds in bond investments. Malaysia will also be promoting retail participation by encouraging the establish- ment of more bond funds. In general, a lot of effort has gone into developing collective investment vehicles and ways to overcome the impediments to their increased participation in the bond market. In terms of developing the current market microstructure, the issue is one of determining the most appropriate alternative to facili- tate secondary market liquidity while ensuring its integrity. Several factors must be considered when choosing the trading platforms. First is the type of investor. It has been argued that OTC markets are more suited to institutional activity while exchange-traded mar- kets are more conducive to retail participation. Next, the issue of competition needs to be reviewed. Having several competing sys- tems may drive the market to adopt the most cost-effective and effi- cient trading platforms available and lower the cost of trading. However, this must be weighed against the possibility of market frag- 103 RANJI'I' AJIT SINGH mentation where the existence of several platforms can unduly frag- ment the market at the cost of efficient price-discovery. Further- more it might also be argued that the integrity of the clearing and settlement system requires it to be independent of regulators. Hence, there is a need for a complete review of the existing structures with a view to determining the most appropriately organized and central- ized trading, delivery and settlement system for the corporate bond market. One of the major government initiatives taken since the Asian crisis has been the establishment of the National Bond Market Com- mittee (NBMC) to provide policy direction and to rationalize the regulatory framework for development of the bond market. As an initial step to rationalize the regulatory framework, NBMC an- nounced that the Securities Commission would be the single regula- tory authority for the supervision and regulation of the corporate bond market. The members of NBMC consist of the Ministry of Finance, Economic Planning Unit of the Prime Minister's Depart- ment, Central Bank, Securities Commission, the Registrar of Com- panies, and the Kuala Lumpur Stock Exchange. 104 CHAPTER 9 India's Debt Market: A Review of Reforms USHA T1O0RAT I ndia's macromanagement of debt has gone through three phases: from 1950 to 1986, from 1986 to 1992, and from 1992 onward. Phase 1 was characterized by a captive market and the absence of debt management. This was the consequence of relying on reserve requirements and a policy of directed investment coupled with an emphasis on raising cheaper resources to finance the government's developmental activities. Phase 2 saw a gradual shift toward passive debt management because of concern about expected future devel- opments and the potential danger of continuing with automatic monetization. These efforts culminated in the active management of debt in phase 3. As an integral part of a comprehensive program of financial sec- tor reform, India's government securities market (GSM) has under- gone a huge transformation since 1991/92. The main objective of this active debt management policy has been to moderate liquidity growth, contain inflationary pressure, and conduct debt manage- ment in a cost-effective manner. Considering India's macromanage- ment and micromarket structure, the developments since 1991/92 can be divided into two subphases. During the first subphase (1991-96), reforms were engineered to facilitate "market-borrowing," with price discovery through auc- tions; restrict automatic monetization by fixing a cap on the dur- ing-the-year and end-year amount; develop appropriate instruments; USHA THORAT introduce a delivery-versus-payment system (DVP) in order to miti- gate the settlement risk; promote greater transparency of prices and volumes traded through daily publication of transactions in government securities; and introduce a system of marking to mar- ket in the valuation of government securities held by commercial banks. The second subphase (post-1996) has focused mainly on the market microstructure. Here, the emphasis has been on introduc- ing a system of primary dealers (PDs) and satellite dealers, with underwriting or bidding commitment for 100% of the issue; intro- ducing various tenors in treasury bills and publishing a half-yearly calendar for the issuance of treasury bills; establishing ways and means advances (WMAs) to the central government to bridge tem- porary mismatches in its receipts and payments; strengthening the government of India's (GOI) cash management system to facili- tate government borrowing in a cost-effective manner; permitting foreign institutional investors to invest in government securities including Treasury bills, both in primary and in secondary mar- kets; expanding the players in the repos market by allowing non- bank participants to borrow and lend in that market; switching over from yield-based to price-based auctions, to facilitate finer bidding; passively consolidating debt by reopening or reissuing existing securities, and thereby helping to create benchmarks and to improve price discovery; making price auctions uniform in the case of 91-day treasury bills on an experimental basis; fostering state government borrowing through the auction system; estab- lishing 100% gilt funds to promote retail holding of government securities; and issuing long-dated securities of up to 20 years to serve as a benchmark for private debt and to elongate maturity for the government. In addition, a high-powered Standing Technical Committee on money and government securities markets has been set up under the deputy governor to advise the Reserve Bank on various matters relating to money and debt markets. Significant reforms have also taken place in the nongovernment debt market. On the regulatory front, the Securities and Exchange Board of India (SEBI) was established in 1992 to regulate the pri- mary issuances in capital and debt markets other than government 106; INDIA S DEBT MARKET: A REVIEW OF REFORMS securities and to ensure sound trading practices in the secondary market through stock exchanges. SEBI has introduced a number of measures in the areas of investor protection, listing, and disclo- sure norms. Depositories have been set up for equity and debt in- struments to facilitate dematerialization as well as transfer and settlement. A wholesale debt market (WDM) segment has been set up by the National Stock Exchange (NSE) where trading in corporate and government debt can take place through screen-based systems. Currently, however, it functions as a reporting system as the market is not yet comfortable with anonymous trading through electronic brokering. New additions to the WDM segment in the form of mu- nicipal bonds, bonds of statutory corporations, corporate infrastruc- ture bonds and units of mutual funds have widened the scope of debt market trading. IMPACT OF REFORMS IN THE GOVERNMENT SECURITIES MARKET India's GSM has undergone significant transformation in recent years in terms of policy initiative and market evolution. Emergence of a market-responsive yield curve and the shifts therein have started reflected the changing economic scenario and give an idea of the future course of economic events. The auction system has helped market interest rates converge in the sense that in a highly liquid market, secondary market yields should anticipate yields in pri- mary issues emerging from time to time. The sovereign yield curve, now more reliable, has emerged as a benchmark for pricing non- government loans. With less reliance on reserve requirements, open- market operations (OMO) have emerged as a major tool of monetary policy. In the face of stiff competition from the government's small savings schemes, which offer tax incentives, the retail investor base has been tapped through gilt funds. The move toward price discovery through a price-based auction sys- tem and reopening has helped financial intermediaries develop bidding skills. These various benefits have also brought certain constraints, however. 107 USHA THORAT CONSTRAINTS The main constraints have to do with the magnitude of government borrowing, seasonality in credit needs, shrinkage in the maturity profile of debt, fragmentation of loans, and the role of the Reserve Bank of India (RBI) as monetary authority and debt manager. Magnitude of Government Borrowing Now that the government has decided to raise resources at market- related prices, the composition of the fiscal deficit and maturity pro- file of the central government debt has changed. Gross market borrowing, which was only about Rs 8,000 crore (Rs 80 billion) in 1990/91 shot up to Rs 93,453 crore (Rs 935 billion) during 1998/99. With the abolition of automatic monetization and introduction of WMAs, the pressure on the government's market borrowing pro- gram has increased immensely. The government is forced to resort to market borrowing when the magnitude of WMAs surpasses the mu- tually agreed limit between the central government and Reserve Bank (Rs 11,000 crore or Rs 110 billion during the first half of the fiscal year and Rs 7,000 crore or Rs 70 billion during the second half). Furthermore, any increase in the budgeted borrowing requirement (fiscal deficit) is met through additional market borrowing. This is evident in the trends of market borrowing since fiscal 1997/98, when the budgeted amount significantly increased in the actual outcome. The market borrowing program has increased 10 times since 1990/ 91, and the statutory liquidity ratio (SLR) prescription has been brought down from 38.5% to 25% over the same period. Further- more, banks are holding excess SLR securities to the tune of Rs 60,000 crore (Rs 600 billion). Assuming that deposits in the bank- ing system grow about 15% to 20% during 1999/2000, which trans- lates to an increase of about Rs 130,000 crore (Rs 1,300 billion), India would have to maintain an SLR of about Rs 32,500 crore (Rs 325 billion), which is less than the level of excess SLR holding by banks. This would mean that in the present circumstances the SLR prescriptions do not in reality obligate banks to buy fresh govern- ment paper. Hence the need for a diversified investor base is greater than ever before. 108 INDIA'S DEBT MARKET: A REVIEW OF REFORMS Seasonality in Credit Needs During the first half of the fiscal year (April-September), when the demand for credit from the nongovernment sector is low (slack sea- son), the government usually tries to complete most of its borrow- ing requirements. The demand for credit from the nongovernment sector picks up during the second half (October-March), called the busy season. This seasonal demand for credit from both the govern- ment and nongovernment sector makes it difficult to spread the government's borrowing program evenly over the fiscal year. Fur- thermore, in the absence of cash management, owing to heavy reli- ance on automatic monetization, little information is available on the quantum and timing of borrowing requirements. This, in turn, makes it hard to take advantage of market timing to optimize the cost of borrowing, with the result that short-term borrowing may at times cost more than long-term borrowing within the span of the same fiscal year. To some extent, the problem is now being addressed through periodic meetings of officials from the government and the Reserve Bank of India through a forum called the Cash and Debt Management Committee. Shrinkage in the Maturity Profile of Debt With the switchover to the system of borrowing at market-related rates and the growing borrowing requirements, there has been some need to shorten the maturity structure so as to reduce the cost, apart from making the tenor of government securities attractive to inves- tors. Consequently, the weighted average maturity was drastically reduced from about 16years in 1990/91 to about 7years in 1997/98. As a result, redemptions are expected to balloon into the foreseeable future. With greater net borrowing required to meet the fiscal defi- cit, the gap between the gross and net borrowing program has grown and prompted serious deliberations on the need to stretch the matu- rity profile of the government debt. Some have argued that short-term borrowings would reduce the relative cost of borrowing for the gov- ernment, but this short-sightedness, however can result in a rise in costs, as demonstrated in 1995 when interest rates rose as the gov- ernment tried to refinance its repayment obligations. A higher in- 109 USHA THORAT terim cost for the government in terms of periodic coupon payments, while borrowing long-term, has to be regarded as compensation for the long-term sustainability of government borrowing. Fragmentation of Loans The size of the gross borrowing requirement and the market's, ab- sorptive capacity at any particular point in time has put constraints on the size of individual issues and accordingly has increased the frequency with which the government enters the market to raise resources. Efforts were made to accommodate a series of new issu- ances within a narrow I O-year maturity band in order to prevent a bunching of repayments, but loans became fragmented as a result, and this impinged upon the liquidity of the government securities. Loans therefore need to be consolidated to improve the fungibility and liquidity of securities and to pave the way for introducing STRIPS in the GSM. The reopening of existing stocks marks a first step to- ward passive consolidation. RBI's Role as Monetary Authority and Debt Manager During 1998/99, yields in the government securities market were greatly influenced by the degree of effectiveness in coordinating debt management policy and monetary policy, with the former focusing on the timing of the loan for deriving maximum benefit of liquidity conditions and the changing maturity mix of borrowings. Apart from these factors, the slack in demand for bank credit and the surplus liquidity conditions within the banking system contributed to the high absorption of government securities. To achieve effective debt management in the medium term, however, government borrowing from the market needs to be reduced substantially and fiscal policy needs to dampen inflation expectations. EXPERIENCE Since 1991/92, debt management has been tackled through a judi- cious mix of policy measures and market timing, and the develop- 110 INDIA S DEBT MARKET: A REVIEW OF REFORMS ment of markets has become the main focus. This section deals with that experience, highlighting primary issuance, open-market opera- tions, market absorption and the monetization of fiscal deficit, sec- ondary market transactions, and the performance of the primary dealer system. Primary Issuance Primary issuance can be discussed from the perspective of the cen- tral government, state governments, elongation of the maturity pro- file, and initial subscription by the RBI. Central Government. The central government's market borrowing program comprises the issuance of dated securities and 364-day trea- sury bills. The central government also borrows through 182-, 91-, and 14-day T-bills. The amount mobilized through other treasury bills, though added to the total borrowing requirements, do not form part of the market borrowing program. The surge in the market borrowing program since 1997/98 can be traced to the abolition of automatic monetization, which made the program a residual source of financing for the government. Note, too, here have been some deviations from the budgeted program on account of the increase in the fiscal deficit. Such high levels of borrowing, particularly during the past few years, made it necessary to resort to primary issuance of dated securities on 24 occasions (including private placement with the Reserve Bank on 8 occasions) during 1998/99, as against 13 oc- casions (including private placement on 3 occasions) in 1997/98. Of these 24 occasions, 11 took place during April-July 1998, when the government ran into overdrafts with the Reserve Bank. During the first half of 1999/2000, the central government has already borrowed more than three-fourths of the budgeted amount by entering the market on 15 occasions (including private placement on 7 occasions). State Governments Over the years, the state governments have also borrowed increas- ing amounts under the market borrowing program, rising from Rs 2,569 crore (Rs25.69 billion) in 1990/91 to Rs 12,114 crore (Rs - 111 USHA THORAT 121.14 billion) in 1998/99. According to the constitutional provi- sion, the state governments cannot borrow without the approval of the central government, and are not permitted to borrow exter- nally. At present, the state governments are borrowing through dated securities and there are no treasury bills borrowings by state gov- ernments. In the move toward a system of auctioning state loans, state gov- ernments have been allowed to raise resources through auction to the extent of 5% to 35% of the allocated borrowings. A few state governments have exercised this option, at an interest rate slightly below the rate on the conventional jointly floated loans for all states for a similar maturity. Elongation of Maturity Profile. As indicated earlier, growing bor- rowing requirements coupled with redemptions on account of huge short-term borrowings between 1992/93 and 1997/98 posed serious problems for refinancing the maturing loans, and it was essential to issue securities with tenors above I Oyears. Though the idea of stretch- ing the maturity profile has been in the cards for some time, it was only during fiscal 1998/99 that a decision was made to test the mar- kets with long-term securities. Consequently, after a gap of 7years, longer-dated securities with tenors of 11, 12, 15, and 20years were issued in fiscal 1998/99. With this change, the weighted average term to maturity of the dated securities rose from 5.50 years in 1996/97 to 6.58 years in 1998/99 and 7.71 years in 1998/99. In line with the trend, more than 60% of the gross borrowing requirements have so far been met through issuance of dated securities with term-to-maturity of above 10 years in the present fiscal year. As a result, the weighted average maturity of the loans issued during the current fiscal year has so far been about 13years as against a weighted average maturity of 7 years in the previous year. Price-badedAuctionsf and the Reopening of Existing Loans. A note- worthy aspect of the internal debt management operations during the current fiscal year was the shift from yield-based auctions to price-based auctions. While this helped participants achieve more aggressive and finer bidding, the reopening of existing issuances also sharpened price discovery. Of the total 20 loans issued in the 112 INDIA S DEB'I' MARKET: A REVIEW OF REFORMIS current fiscal year, only 3 were new issues whereas the rest were reissues. Initial Subscription by Reserve Bank of India. In view of the grow- ing market borrowing requirements as well as the market's absorp- tive capacity at any particular point in time, the Reserve Bank decided in Fiscal 1998/99 and the current financial year to take securities on its book on a private placement basis and subsequently unload them through active open-market operations. Of the total amount of gross borrowing raised through dated securities, private placement with the Reserve Bank and devolvement on the Reserve Bank during 1998/99 was Rs 38,205 crore (Rs 382 billion), or 45.6%, compared with Rs 13,028 crore (Rs 130 billion), or 30% in 1997/98. So far during 1999/2000, initial subscriptions by the Reserve Bank in terms of private placement has totaled Rs 23,500 crore (Rs 235 billion), or about 42% of the total dated securities issued. During fiscal 1998/99, there was also a devolvement on the Re- serve Bank on account of treasury bills. The devolvement with re- spect to 364-day T-bills totaled Rs 1,047 crore (Rs 10.47 billion), or 21% of the gross issues. Similarly, the devolvement on the Reserve Bank on 91- and 14-day T-bills reached Rs 451 crore (Rs 4.51 bil- lion) and Rs 260 crore (Rs2.6 billion), or 13% and 23% of the noti- fied amount, respectively. During the first half of the current fiscal year, the devolvement of 364-day T-bills amounted to Rs 1,692 crore (Rs 16.92 billion) whereas for 14-, 91-, and 182-day T-bills the devolvement as a whole has been Rs 1,757 crore (Rs 17.57 billion), or 26% and 27.5% of the notified amounts, respectively. Open-Market Operations With less reliance on reserve requirements, more emphasis is being placed on indirect instruments of monetary control. This, coupled with the need for large market borrowings, has paved the way for open-market operations (OMO). During fiscal 1998/99 and the current financial year, the Reserve Bank of India resorted to active OMO to neutralize a large amount of private placement of government securities with itself and to maintain an orderly flow of liquidity in the market. An important 113 USHA THORAT feature of the open market operations during 1998/99 and at this time was the inclusion of treasury bills; up to September 1999, the net sales under OMO amounted to Rs 25,750 crore (Rs 257.50 bil- lion). During 1998/99 the net sales under OMO were Rs 29,669 crore (Rs 296.69 billion). Because of large-scale OMO by the Reserve Bank, its holdings of central government securities declined substantially. Consequently, the net RBI credit to the center was placed at Rs 11,800 crore (Rs 118.0 billion) in 1998/99 as against Rs 12,914 crore (Rs 129.1 bil- lion) in the previous year. During the first half of the current fiscal year, net RBI credit to the center was Rs 3,510 crore (Rs 35.10 bil- lion) as against Rs 1,857 crore (Rs 18.57 billion) in the previousyear. Ways and Means Advances to Central and State Governments India's experience with WMAs can be seen at both the central and state levels. Central Governmtent. During fiscal 1999/2000, the limit for WMAs has been fixed at Rs 11,000 crore (Rs 110 billion) for the first half (April-September 1999) and Rs 7,000 crore (Rs 70 billion) for the second half (October-March 2000). From April 1, 1999, the inter- est rate was revised to the bank rate on WMAs and to the bank rate plus two percentage points on overdrafts beyond XVMAs. A transi- tion period of two years that provided for implementation of the Overdraft Regulation Scheme came to an end on March 31, 1999. According to the provisions of the agreement dated March 26, 1998, between the central government and the Reserve Bank, overdrafts beyond 10 consecutive working days will not be allowed from April 1, 1999. The minimum balance to be maintained by the central gov- ernment with the Reserve Bank has also been revised from not less than Rs 50 crore (Rs 0.5 billion) to Rs 100 crore (Rs 1 billion) on Fridays, and from not less than Rs 4 crore (Rs 0.04 billion) to Rs 10 crore (Rs 0.1 billion) on other days. The government's recourse to WMAs and overdrafts during fis- cal 1998/99 revealed the pressure on liquidity management. This was in marked contrast to the trend of 1997/98 when the central government ran a surplus for most of that year. The year-end level 114 INDIA S DEBT MARKET: A REVIEW OF REFORMS of WMAs was placed at Rs 3,042 crore (Rs 30.42 billion) in 1998/ 99, as against Rs 2,000 crore (Rs 20 billion) in the previous year. So far during fiscal 1999/2000, though the pressure on liquidity contin- ued in April 1999, thereafter the recourse to WMNAs has been lower, and as on September 30, 1999, the WMA outstanding level was Rs 3,856 crore (Rs 38.56 billion). State Governments. Recognizing the difficulties faced by state gov- ernments and following the recommendations of the Informal Advi- sory Committee on Ways and Means Advances to State Governments, constituted by the Reserve Bank of India, the Bank has revised the amount of WMAs in effect from March 1, 1999. According to the revised scheme, the limits of WNlA have been increased by 65% to Rs 3,685 crore (Rs 36.85 billion). The limits for special WMAs were liberalized. The overdraft regulation scheme has been made more stringent. The minimum balance has been revised upward. The re- vised scheme has been in effect from March 1, 1999. The monthly average peak of normal WNIAs, special WMAs, and overdrafts relied on by state government during 1998/99 amounted to Rs 2,147 crore (Rs 21.47 billion), with the peak aver- age ranging from Rs 536 crore (Rs 5.36 billion) in May 1998 to Rs 4,818 crore (Rs 48.18 billion) in March 1999. During 1998/99, 17 states resorted to overdraft, and 11 of them did so frequently. Five states could not clear their overdrafts with the Reserve Bank within the stipulated time limit, and consequently the Reserve Bank had to stop payments on their behalf. In its role as debt manager to state governments, RBI has been consulting more and more with the state governments in order to improve cash and debt management. Its responsibilities in this re- gard include the constitution of Consolidated Sinking Fund, a flex- ible approach to market borrowing, and sensitization to the growing problem of contingent liabilities, especially guarantees. Retailing of Government Securities One way to reduce the cost of borrowing is to widen the investor base for government securities. Considering the tax incentives given to other competing instruments of similar quality, promoting a retail 115 USHA THORAT market for government securities implies ensuring liquid markets for the retail investor. This would require custodial facilities to be dis- persed with fully automated connectivity to the central settlement agency. With the promotion of gilt funds, India took the first steps toward indirectly establishing a conduit for retailing government se- curities. These funds have been made eligible for liquidity support from RBI to the extent of 20% of the outstanding government secu- rities in their portfolio. The first such fund was set up toward the end of 1998. Since then the number of gilt funds has been growing steadily. Enlistment of More Primary Dealers To promote greater competition, eight more primary dealers, 75% of which are subsidiaries of foreign banks/security houses, were in- ducted during 1998/99 and 1999/2000. Thus the total number of PDs functioning in the government securities market has risen to 14. Enhanced underwriting options up to 100% have been introduced for dated securities and compulsory bidding commitments so as to absorb 100% of the primary issue in the case of treasury bills. These measures are aimed at reducing the possibility of devolvement on the Reserve Bank of India and the consequent monetization. An analysis of the performance of the six PDs for fiscal 1998/99 revealed that whereas the bidding commitment was Rs 16,350 crore (Rs 163.50 billion) and Rs 26,900 crore (Rs 269 billion) in treasury bills and central government dated securities, respectively, the ac- tual bids tendered by PDs were Rs 32,134 crore (Rs 321.34 billion) and Rs 25,024 crore (Rs 250.24 billion). Of these, the bids accepted were Rs 15,652 crore (Rs 156.52 billion) for treasury bills and Rs 13,252 crore (Rs 132.52 billion) for central government dated secu- rities. The amount of devolvement to PDs was Rs 2,991 crore (Rs 29.91 billion) in dated securities and Rs 2,367 crore (Rs 23.67 bil- lion) in treasury bills. Trading turnover by PDs accounted for 19% of the total market turnover. Secondary Market Transactions As the government securities market deepened, the aggregate vol- ume of transactions (outright as well as repos) in central and state 116 INDIA S DEBT NIARKET: A REVIEW OF REFORMS government dated securities and treasury bills moved significantly higher, from Rs 1,85,708 crore (Rs 1857.08 billion) in 1997/98 to Rs 2,27,228 crore (Rs 2,272.28 billion) in 1998/99. The volume of trans- actions in state government securities was marginal at Rs 1,544 crore (Rs 15.44 billion). The bulk of transactions, amounting to Rs 1,87,531 crore (Rs 1875.31 billion) (83%) were conducted on an outright basis, and the balance by way of repos. Total turnover in govern- ment securities during fiscal 1998/99 amounted to Rs 5,33,850 crore (Rs 5,338.50 billion), as against Rs 4,20,655 crore (Rs 4,206.55 bil- lion) in 1997/98. The outright turnover came to Rs 3,75,062 crore (Rs 3,750.62 billion) as compared with Rs 3,22,179 crore (Rs 3,221.79 billion) in 1997/98. Yields on Primary Issues Notwithstanding the large-scale borrowing program, the interest rates remained fairly stable during the course of 1998/99, as they have so far in the current fiscal year. The weighted average yield on dated securities declined from 12.01% in 1997/98 to 11.86% in 1998/99, which was very close to the level of 1991/92. Despite significant elongation in maturity to an average of 13 years during the current fiscal so far, compared with about 8 years during the previous year, the weighted average yield on dated securities has increased mar- ginally by 11 basis points, to 11.97%. Primaryyields on government securities fell nearly 66 basis points over a 10-year period, dropping from 12.25% in December 1998 to 11.59% in the recent auction on August 27, 1999. FUTURE AGENDA In the wake of a substantially large market borrowing program, the Reserve Bank recently adopted a strategy of funding the government's fiscal deficit and at the same time reducing the pressure on interest rates and inflation expectations. Market development was an im- portant component of the agenda in recent years and will receive concentrated attention in the future. The issues that need to be ad- dressed in this regard are outlined in the following paragraphs. 117 USHA THORAT Auction Systems Despite the controversy surrounding the relative merits of the uni- form-price auction system and the multiple-price auction system, the Reserve Bank of India has been using the uniform-price auc- tion method for the issuance of 91-day treasury bills since Novem- ber 1998. This has been done on experimental basis. Though it is true that the uniform-price method reduces the uncertainty and the incidence of the winners' curse, in the absence of empirical evi- dence regarding the possible cost to government borrowing, it has not been possible to test this suggestion. In a stable and deep mar- ket, expectations as to the yield/price are very close to the yield! price that emerges at the auction. Furthermore, in such markets the price difference between the uniform-price and multiple-price auction systems are not so significant. The difference could be sig- nificant, however, in the case of developing markets such as those in India. Apart from this, concerns such as collusion and the state of development of the market could favor the multiple-price auc- tion system. Introduction of a "When-Issued" Market Although the auction system has contributed to a price discovery mechanism, such a mechanism is not fully efficient without a "when- issued" (WI) market. In other words , a WI market complements the efforts of an efficient auction system through the price discovery mechanism. At present one cannot trade in securities prior to the issue because forward trading in securities is prohibited. A consen- sus is emerging in favor of developing a WI market as part of the overall development of the debt markets. To a certain extent, reissuances have facilitated price discovery, as it is possible to trade in the issue to be auctioned. Issuance of Auction Calendar At present there is an auction calendar for the issuance of treasury bills, announced half-yearly. It has been argued that such a calendar imparts a sense of certainty to the market in regard to the volume, 1-18 INDIA S DEBT MIARKET: A REVIEW OF REFORMS timing, and tenor of securities, but the issuer is forced to sacrifice flexibility and to sharpen the skills used in predicting market changes. With government finances following a random pattern, giving the least indication of its borrowing requirements, both in terms of vol- ume and of timing, the issuer cannot but retain this flexibility. So far, this constraint has not permitted the announcement of an auc- tion calendar for government securities. Repos Market As is now widely recognized, repo markets facilitate better cash and debt management and also provide liquidity to securities markets through securities lending and borrowing. Because repos are short- term collateralized instruments, repo markets are intimately con- nected with money securities and derivative markets. Central banks find them a useful monetary policy instrument and source of infor- mation on market expectations. Recognizing that a repo market con- tributes to a vibrant secondary market in debt instruments, a subgroup of the Technical Advisory Committee on Government Securities Markets has recommended changes in the existing legal system as well as the kind of infrastructure needed to develop the repo mar- ket. The proposals include giving RBI regulatory powers over repos; facilitating the electronic transfer of securities; withdrawing of gov- ernment notification dated June 27, 1969, under the Securities Con- tracts Regulation Act (SCRA) of 1956, which prohibits forward trading in securities; and introducing of over-the-counter and tri- partite repos. The group has also recommended a uniform documentation code of conduct for market participants and uniform accounting policies to reflect the content rather than the form of repo con- tracts settlement through the clearing corporation for participants other than the central bank constituents. The risks inherent in repo transactions can be minimized by ensuring timely mark to market of securities repoed, by issuing prompt margin calls and haircuts, and by ensuring adherence to the discipline of the exchange. The RBI is committed to develop the repo market with adequate safe- guards and the recommendations of the group are being put into operation. 119 USHA THORAT Development of STRIPS Introducing STRIPS in GSM would hasten the process of diversi- fying the investor base and also help boost secondary market activi- ties in government securities. This would also give rise to a zero yield curve. The preconditions for such a market are fungibility of interest payments under different loans so as to constitute a critical minimum mass, and active repos/reverse repos market programs coupled with fully automated trading and settlement systems to handle the enormous amount of bookkeeping that STRIPS entail. As a pre- cursor to STRIPS, efforts are under way to consolidate the loans. Regulatory Concerns The Indian debt market has a huge private placement market in debt instruments that is unregulated. At present, private placement does not come under the usual "disclosure and protection of investors' interest" norms prescribed by the Securities and Exchange Board of India (SEBI), even though most of the issues are rated by credit- rating agencies. The risk of default to investors could be a problem, one that could pose systemic risks if the investors involved happened to be large financial institutions. Hence, this market needs to be brought under direct regulation, apart from indirectly doing so through banks and institutions that are the major informed inves- tors in these markets. Since state government guaranteed bonds con- stitute a significant segment of this market, any assessment of risk would require much greater transparency and disclosure in the ac- counts of the state governments. Another committee has been set up specifically to look at the kind of disclosures that should be made in the accounts. Legal Concerns A move has begun to replace the now outmoded Public Debt Act with a new Government Securities Act, but it still needs approval of the central and state governments. The new act seeks to recognize electronic modes of transferring the title of government securities, would facilitate pledging of securities without actual transfer, and 120 INDIA S DEBT MIARKET: A REVIEW OF REFORMS would recognize depositories other than RBI for paperless transfer. It is hoped that the bill will be enacted soon. There is a proposal to amend the Securities Contract Regulation Act of 1956 to provide RBI with regulatory power over the securi- ties markets. This would ensure that RBI regulates the repos mar- kets and securities markets insofar as they impinge on monetary and debt management policies. The proposed amendment would also fa- cilitate introduction of indexed futures. The Indian Stamp Act is another piece of legislation that inhib- its development of debt markets. The government has received a number of recommendations for modifying the stamp duties so as to exempt securities held in dematerialized form. It is hoped these, too, will be acted upon soon. Diversification of Investor Base India urgently needs to diversify the investor base and encourage active trading in secondary markets. Provident funds, pension funds, insurance companies, and mutual funds are the typical investors in government securities. In addition, large corporates in the private and public sector at times have large surplus cash balances. Active management of cash is usually discouraged by public sector under- takings in view of the perceived risks involved. Provident funds and insurance companies are typically buy-and-hold investors. There is no system of marking to market and capturing total return for these funds and no incentives to improve performance. Some corporates have active treasuries but are deterred by the lack of liquidity in the markets. The challenge before the RBI and the primary dealers is how to improve liquidity in the markets. Encouraging individual investors in the face of other more at- tractive measures, such as tax-free mutual funds and small savings, is well-nigh impossible. The National Stock Exchange proposes to set up a retail market segment that will have settlement guaranteed by the clearing corporation. The scheme will work if the PDs and other large holders of securities put up two-way quotes for small lots for screen trading and are able to easily transfer and retransfer stocks from RBI's public debt office to the Clearing Corporation and vice versa. 121 USHA THORAT Short Selling and Forward Transactions Closely linked to liquidity is the issue of whether to permit market participants to undertake forward trades in securities and also go short on a net basis. Among the questions that need to be addressed here are (1) whether development of the repo market should proceed by introducing securities lending and borrowing first and then forward trading and short sales next, or whether both be introduced simulta- neously; (2) whether only primary dealers should be allowed to go short or whether other large players-especially banks in the sec- ondary markets should also be permitted to do so; (3) to what extent short selling should be permitted, whether the restrictions should apply to the period and/or the quantities, and whether such quantities should be duration weighted or based on some other measure of risk such as Value At Risk (VAR); and (4) whether restrictions should be placed on the specific securities that can be shorted, taking care to allow se- curities for shorting that have large floating stock and are widely held. Primary Dealer System The RBI has been looking at a number of issues associated with instituting a primary dealer system. The PD gets certain benefits in return for a commitment to bid in the primary auction, provide two- way quotes for securities, and achieve a certain turnover. The ben- efits include an underwriting commission, permission to participate in interbank money markets as borrower and lender, and also li- quidity support from RBI. The support is linked to bidding commit- ments in dated securities and treasury bills and turnover in secondary markets. The questions of particular concern here are as follows: (1) What is the most equitable way of providing liquidity support while ensuring that the PDs provide continuous two-way quotes for specified securities? (2) Should there be any restrictions on other activities of PDs in the securities markets? Considering that they are first and foremost dealers in government securities, should there be any turnover requirement for government securities in relation to total turnover in all securities? Should there be any other require- ment to ensure that PDs are predominantly dedicated to govern- ment securities? (3) From a prudential point of view, how should 122z INDIA S DEBT MARKET: A REVIEW OF RlFORMS capital be allocated to the various activities? How should minimum capital be calculated and maintained, and how can multiple leverag- ing be avoided? (4) To what extent can the principles for prudential regulation in developed markets be applied in nascent markets? (5) What kind of relationship should PDs have with subsidiaries? Should PDs be allowed to float subsidiaries? How should the risk on the parent company be assessed? (6) Should subsidiaries of foreign com- panies be allowed to set up subsidiaries for undertaking PD busi- ness? (7) In the context of universal banking, should the concept of keeping banking and securities business separate be reexarnined? Trading and Settlement Systems Transparent and on-line trading systems provide better dissemina- tion of prices in the market and narrow bid-offer spreads. Although, as indicated earlier, the experiment of matching trades electronically through NSE's WDM did not take off, the trades are being reported with minimurn delay. The RBI is trying to get a dealing system in- stalled that will be able to disseminate to the market the details of all trades undertaken by members of the system as soon as the deals are done, without disclosing the identity of the entities. A project for the complete automation of the RBI's Public Debt Office has just begun. Its main objectives are to provide for connec- tivity between different offices, undertake electronic transfer of title or pledge, and operate a securities lending and borrowing scheme if required. Collateralized intraday liquidity support will also be pro- vided to facilitate settlement of transactions in the DVP system. Such support would also be required once the Real Time Gross Settle- ment (RTGS) is introduced. It is hoped that the project will be com- pleted for a trial run in a year's time. In order to undertake repos through clearing corporations, it would be necessary to have a sys- tem that would ensure prompt margining/haircuts, for example. This issue is currently being addressed. Securitization Securitization is considered an effective way to liquify the assets of financial intermediaries and make them marketable. The RBI has 123 USHA T HORAT set up an expert group to go into all the legal and regulatory prob- lems involved in promoting securitization and it is expected to sug- gest solutions, including recommended changes in the legal framework. Derivatives Although derivatives can prove useful in risk management and in unbundling liquidity market and credit risk, market making in de- rivatives is possible only when there are deep and liquid cash mar- kets to facilitate pricing, at least in the early stages, until the momentum of the derivatives market takes over. In response to re- quests that market participants be allowed to conduct interest rate swaps (IRSs) and forward rate agreements (FRAs), the RBI allowed IRS and FRAs in the current year, with certain reporting require- ments for market makers. Only banks and primary dealers have been allowed to undertake market making, while others-including fi- nancial institutions and corporates -are to use the products for hedg- ing their balance sheet risks. The IRSs and FRAs have not taken off, mainly because of the lack of a term money market. Such a market will develop only when banks and financial institutions observe strict Asset-Liability Management (ALM) discipline for positive and nega- tive mismatches, and these are rigorously monitored and enforced by the supervisors. Some also have suggested that T-bill futures should be introduced. Given the poor liquidity in the cash market, India would be ill-advised to allow futures at this stage. A critical step for the development of a derivatives market, then, is to direct all effort toward the development of the underlying cash markets. 124 CHAPTER 10 The Importance of Local Bond Markets in Financing India's Infrastructure Needs RAKESH MOHAN A n expert committee on infrastructure finance that I chaired re- cently looked at India's infrastructure requirements and the fi- nancing and debt needed to meet them. The committee con- cluded, first of all, that the public sector would have to continue investing at roughly the same proportion of GDP, which means an annual increase of 6% to 7% if the economy is growing at 6% to 7%. On top of this, private sector investment in infrastructure would have to increase substantially. None of that would happen in either the public sector or the private sector unless India could develop a major debt market. If such a market does emerge, then the subcontinent, and India in particular because of its size, could have a huge bond market in the next 5, 15, or 30 years. In 1991 India had 23 cities with a popu- lation of over 1 million and about 300 cities with over 100,000. The number of issuers of bonds could be huge if each of these cities, and also each state in India, were to issue bonds. The issuers would come from bodies such as urban development authorities, special purpose vehicles, and public-private partnerships that would develop the in- frastructure needed. In other words, if the bond market develops, there will be no shortage of issuers, but rather hundreds of them. On the supply side, the National Council of Applied Economic Research has been studying the consumer behavior of more than 300,000 households across India since 1985, in both rural and urban RAKESH MOHAN areas. In its survey, the council is examining what people are buying to get an idea of how the markets are developing, and, secondarily, how household incomes in India have changed in the last 13-14years. On the basis of these changes from low to middle and higher in- comes, the council can predict with some confidence what will hap- pen over the next 10 years. One dramatic finding so far, which presumably can be applied to other countries in the region, concerns the number of households in the low income category. The Council roughly defines the low income level as the income level below which people do not buy much beyond their essential food and subsistence requirements. As of 1995, the approximate proportion of families in India in this low- income category was 50%. With 6% to 6.5% growth in the next 10 years, that proportion is projected to fall to about 15%. The importance of this finding in the context of debt market development is that in the early to mid- 1990s only half of the Indian population had savings, particularly financial savings, of any kind. As that proportion goes up to, say, 85%, Financial savings will see an enormous increase. In 1995, at the high income level group which has the highest savings rates, comprised about 5% of total house- holds. Within 10years, this proportion could increase to about 15%. If all these predictions are correct, then the overall savings rate in India would go up from about 25% to 35% over the next IOyears, with similar changes across the region. All these savings require in- vestment instruments because the savings will not go up unless such instruments are available. This is where the bond nmarket comes in: all these developments -economic growth, infrastructure investment, increased savings rates-will not happen without the market infra- structure and avenues by which people can invest. In turn, this is where pension funds, mutual funds, insurance funds, and invest- ment banks come in. For a country at India's stage of transition, "safe" debt instru- ments are far more important than in developed countries. As fami- lies begin to move into the savings mode, they must not suffer big losses as a result of unsafe debt instruments; if they do, they will withdraw from the market. This is what happened in the Indian equities market in 1993-94: as new investors got burned, they with- drew. The same could easily happen in the bond market. Therefore 126 FINANCING INDIA'S INFRASTRUCTURE NEEDS it is very important to make these instruments safe for the investor, by enhancing the credit quality of the issuers. This will require a lot of work in terms of improving the accounts and increasing the infor- mation of municipal bodies, state government, parastatals, and pri- vate sector issuers. 127 CHAPTER 11 India Survey: Issues in Local Bond Market Development JOHN LEONARDO I ndia has debt markets for government securities, corporate bonds, and short-term bank and commercial paper. The Government securities market is far the largest market, it has expanded con- siderably since 1991, as has the range of available maturities and secondary market activity in both short- and longer-term maturi- ties. Since 1997, when banks were permitted to hold corporate debt securities, the market for these securities has grown as well, although it is dominated by paper issued by state-owned enterprises rather than private sector entities, and trading is limited. India's debt markets as a whole have grown steadily since 1992, and many efforts are under way to support further development. However, policymakers face a number of major challenges to fur- ther market development particularly by providing a more support- ive economic, legal, and regulatory policy environment; upgrading market infrastructure; authorizing the use of new debt instruments; and improving education services for debt market participants. If the impediments to growth identified in this study can be re- moved, secondary market trading in government and corporate debt securities should increase significantly. In the short term, secondary market trading in corporate debt securities is likely to show the stron- gest growth, once the planned stamp duty reforms are implemented.' ' Since this report was prepared in mid-late 1999, the Indian Parliament has JOHN LEONARDO Table 1. India's Economic Performance, 1997-99 (percentage change from previous year) Indicator 1997 1998 199 GDP 7.8 5.0 6.0 Agricultural production 9.3 -5.6 7.6 h Industrial production 5.6 6.6 4.0 Inflation (WPI) 6.9 5.3 4.8 Provisional 1998-99figures. Estimated whoiesale price index calculated on a point-to-point basis. Source: RBI (1998-99), appendix table 1.1. In the medium term, further growth in secondary market activity in government securities is also anticipated. This chapter discusses the status of the debt markets in India, some impediments to their development, and recommendations for removing those impediments and building the markets further. First, an overview of the Indian economy and the government's fiscal position is presented. Next, the government securities market is discussed followved by an overview of the Indian corporate bond market. The penultimate section discusses impediments to the de- velopment of the Indian debt markets and assesses their relative priority from a remedial action perspective. The chapter concludes with recommendations on how to remove the identified impedi- ments. ECONOMIC OVERVIEW Economic growth in India is currently on an upward trend. Real GDP growth in fiscal 1998/99 was about 6%, and it is expected to be about 6% to 6.5% in fiscal 1999/2000 (RBI, 1998-99). Overall economic growth during 1998-99 was materially assisted by a marked improvement in agricultural output. India's recent economic perfor- mance is set out in table 1. Continuing favorable growth in the agri- passed legislation which allows bonds to be dematerialized which in turn will elimi- nate the stamp duty problem. (See later discussion.) INDIA SURVEY Table 2. India's Debt, 1996-2000 (billions of rupees) Debt parameter 995"96 199697 1997-98 1998- 1999-200 Domestic debt 5,549.83 6,214.37 7,229.62 8,199.65 9,344.29 (% ot total debt) (91.5) (92.0) (92.9) (936) (94.3) External debt 512.49 542.39 553.32 559.60 561.34 (% ot total debt) (8.5) (8.0) (7.1) (6.4) (5.7) Total debt 6,062.32 6,756.76 7,782.94 8,759.25 9,905.63 Domestic debt as % of GDP 45.6 44.0 46.2 46.6 46.7 External debt as % of GDP 4.2 3.9 3.6 3.2 2.8 Total debt as % of GDP 49.8 47.9 49.8 49.8 49.6 Note: Figures in brackets represent percentages to GDP at current market prices based on new series with 1993-94 as base year Revised estimates Budget estimates Source: RBI (1998-99), appendix table IV.7. cultural and industrial sectors is likely to be the major source of the economic improvement forecast for 1999-2000. During 1998-99, the central government's gross deficit rose from Rs 889.37 billion to Rs 1,037.37 billion, following a marked in- crease in 1997-98 (RBI, 1998-99). Budget estimates for 1999-2000 show the deficit shrinking, from Rs 1,037.37 billion to Rs 799.55 billion. The revised estimate for 1998-99 was 5.90% of GDP, com- pared with 5.70% in 1997-98. This revised figure should be treated with caution, however. Gross domestic savings (GDS) as a pro- portion of GDP remained a modest 24.4% in 1996-97 and 23.1% in 1997-98. As mentioned earlier, India is a heavily indebted nation. Revised RBI estimates indicate the value of total government debt was equiva- lent to 49.8% of GDP during 1999. Trends in national debt as a percentage of GDP (see table 2)2 suggest that India has made little progress in curbing the level of debt. To complicate matters, debt-servicing costs have been rising steadily. Between 1991-92 and 1998-99, interest payments as a ra- 2 Based on a new series of GDP data using 1993-94 as a base year. 131 JOHN LEONARDO tio of government revenue receipts increased from 40.28% to 49.00% (RBI 1998-99, appendix table IV.2). Since 1997-98, this ratio has stabilized at around 49%. Details of the composition of the funding of the government's deficit are set out in table 3. Most of the deficit is financed inter- nally; government securities represent the largest source of domes- tic funding. Between 1996 and 1999 the fiscal deficit as a proportion of GDP rose steadily to just under 6%. The RBI relies on several mechanisms to manage domestic li- quidity. During 1998-99, the principal ones were open market op- erations and repo transactions. GOVERNMENT SECURITIES MARKET Financial reforms introduced by the RBI since 1992 have brought some major changes to the banking sector. Significant changes have also been made in the government securities market. As the RBI is pursuing a deliberate policy of gradual reforms in the financial sec- tor, the government securities market is steadily evolving. The RBI actively seeks the views of market participants to proposed policy changes that affect the government securities market. This section describes recent trends in the government securities market and the key impediments to market development. Table 3. Net Financing of India's Gross Fiscal Deficit, 1996-2000 (billions of rupees) 1998 93 1999-2010 1996-7 1997-91 Revised Estimates Budget Estimates Deficit to be financed 667.33 889.37 1.037.37 799 55 Financed through: (a) External Resources 29.87 10.91 9.10 8.45 (b) Internal Resources 637.46 878.46 1.028.27 791.10 (i) Market Borrowings 200.12 324.99 649.11 574.61 (ii) Other Liabilities 305.50 562.57 346.82 216.49 (iii) 91-day Treasury Bills 131.84 -9.10 32.34 0 Fiscal deficit as % of GDP 4.7% 5.7% 5. 9/, Not available Source. RBI (1998-9), appendix table IV4. 132 INDIA SURVEY Market Size The main points to consider here are the sector's outstanding debt, secondary market turnover, and yield curve. Outs4tanding Domestic Debt. At present, the government of India is- sues securities only to wholesale investors when raising debt finance. The composition of outstanding domestic debt on March 31, 1999, is shown in table 4; these figures are based on unaudited accounts supplied by the Ministry of Finance. Wholesale government debt instruments comprise treasury bills and dated securities purchased by wholesale investors for invest- ment and/or statutory ratio purposes. These instruments are sold either through auctions conducted by the RBI or in the secondary market. In April 1999, the RBI announced the introduction of 182- day treasury bills. Auctions of 14-day and 91-day treasury bills are now held weekly, while auctions of 182-day and 364-day treasury bills are held fortnightly. The Indian government does not offer retail securities at present, as the RBI's offices are not fully computerized. However, Post Of- fice savings bonds issued by the Ministry of Finance and relief bonds issued by the RBI are de facto retail government securities. These securities carry a 9% tax-free interest rate, but market participants do not consider tax to be a major issue influencing retail investment. Banks act as agents for the Ministry of Finance in the distribution of Post Office savings bonds. The distribution of domestic debt, based on the figures presented in table 4 (provided by the Ministry of Finance), is shown in table 5. As of March 31, 1999, domestic debt consisted mainly of dated secu- rities and securities issued to the RBI for eventual sale to the mar- ket. Although there was only a modest amount of treasury bills on issue on that date, material use is being made of these securities. Nationalized banks and insurance companies appear to be the largest holders of government securities. No detailed breakdown of the distribution of ownership of treasury bills and dated securities is readily available. The 182-day treasury bill was one of several steps taken in 1999 to increase the available range of treasury bill instruments. The RBI - 133 JOHN LEONARDO Table 4. Outstanding Domestic Debt, March 31, 1999 (billions of rupees) Kind of debt Amount Market loans (i.e., dated securities) 2,855.9 91-day Treasury Bills 19.0 Securities held by RBI 1,228.6 14-day Treasury Bills 71.5 364-day Treasury Bills 98.0 Securities held by International Financial Institutions 215.2 Other 123.7 Total Internal Debt 4,611.9 also authorized interest rate swaps (IRS) and forward rate agree- ments (FRAs) and extended the number of organizations able to participate in repurchase transactions in government securities. SecondaryMarket Turnower. Between 1997-98 and 1998-99, the value of outright turnover in government securities rose from Rs 3,221.8 billion to Rs 3,750.6 billion. Over the same period, the value of repos in government securities increased from Rs 984.8 billion to Rs 1,587.9 billion. Total turnover in government securities amounted to Rs 5,338.5 billion in 1998-99, compared with Rs 4,206.6 billion in 1997- 98. Average daily trading volume in government securities, inclu- sive of repos, is approximately Rs 20 billion per trading day. Turnover figures confirm comments from market participants that the overall level of trading in the government securities market is rather mod- est, and the percentage of stock traded is small in relation to the amount on issue. The primary dealers (discussed below) have contributed to a material increase in daily secondary market trading volumes since Table 5. Distribution ot Domestic Debt, March 31, 1999 Debt Distribution (%) Treasury Bills 4.09 Market loans (i.e. dated securities) 61.92 Securities held by RBI 26.64 Other 7.35 Total Internal Debt 100.00 INDIA SURVEY Table 6. Growth in Trading Volumes 1996-98 (March years) (billions of rupees) NSE Transactions SC L Transactions NSE Share of SGL 1996-7 1997-98 1998 1996-97 1997-98 1998 1996-97 1997-98 1998 June qtr Jaine qtb June qtr Government Securities 272.3 842.3 247.4 832.5 1,411.3 383.0 33% 60% 65% Treasury Bills 109 0 188 7 24 0 366 5 464 2 70.5 30% 41O/% 34% Total 381.3 1,031.0 271.4 1,199.0 1,875.5 453.5 32% 55% 60% Source. NSF (1998), Table 38 1996. The Rs 20 billion average daily trading volume is consider- ably higher than the Rs 2 billion to 3 billion of three years ago. Table 6 shows details of the volumes traded on the NSE and total transac- tions recorded through the RBI's Subsidiary General Ledger (SGL) system during the period April 1, 1996, to June 30, 1998.2 (Note that the published Cigure for SGL transactions in 1997-98 is materi- ally less than the turnover figures published in RBI's 1998-99 an- nual report). Primary dealers account for 25% of secondary market trading, foreign banks 25%, state-owned banks 40%, and others 10%. There is a marked difference in the relationship between shares of gov- ernment security volumes and shares of banking assets. Foreign banks hold 7% of banking assets yet generate 25% of turnover. By contrast, Indian banks hold 93% of banking assets and generate only 40% of the volume. This demonstrates the potential for growth in trading volumes with a more competitive and efficient banking sector. Details of volumes traded on the NSE and transactions recorded through the RBI's system during June 1999 are set out in table 7. Interestingly, before the scandal surrounding fraudulent trading in government securities erupted in 1992, volumes were 30 to 40 times current levels. Trading volumes did go up in 1999, however, owing to a reduction in the corporate spread over the sovereign rate, sur- plus liquidity in the banking system, and limited opportunities for prime lending on the part of the banks. Trading received another boost from the RBI's mark-to-market requirements introduced on March 31, 1999, under which bank investments in government se- 135 JOHN LEONARDO Table 7. Volumes of Secondary Market Trading in Government Securities,1999 (billions of rupees) April May June June % traded (NSE) (NSE) (NSE) (R1I:S6.) at NSE Treasury Bills 13.1 5.3 9.9 27,0 36.86% 14-day Bills .03 .4 54 7.52% 91-day Bills 2.0 1.5 1.1 5.4 20.14% 182-day Bills .2 364-day Bills 11.1 3.8 8.3 16.2 51.03% Dated Government Securities 179.5 186.0 144.1 246.5 58.45% Converted 42.0 25.1 16.3 21.0 77.95% Regular 130.5 154.0 120.8 183.6 65.79% Zero Coupon 4.2 5.5 5.2 7.3 71.36% Repo 2.8 1.3 1.7 34.6 5.00% Total Government Securities 192.6 191.2 154.0 273.4 56.33% curities must be valued at no less than 70% of the market value. This requirement is to increase progressively to 100% by March 31, 2002. On balance, the state-owned banks have little understanding of opportunity cost concepts. India's two major insurance companies (General Insurance Corporation and Life Insurance Corporation of India, LIC) recently started trading to improve yields, and a num- ber of the state-owned banks are now following suit. This trend has been aided by a change in management since the initiation of tight controls on secondary market trading after the 1992 scandals. There are only a few genuine long-term investors in government securities at the moment. The Yield Curvefor Government Securitied. Table 8 shows secondary market yields for government securities. As can be seen, the market is relatively active in terms of issuance, but trading on the whole is limited. Figure 1 presents a graph of the current yield curve in gov- ernment securities based on the yields in table 8. Dealers in the Government Securities Market The RBI has established a system of primary and satellite dealers to help promote development of the wholesale and retail government securities markets. 136 INDIA SURVEY Table 8. Yield Curve for Government Securities, 31 July 1999 (percenf) Tatter 'treasuryBtilk Bated secorflies 14 days 8.11 91 days 8.98 182 days 9.91 364 days 10.31 1 year 10.48 2 years 1C 71 3 years 10.97 4 years 11.08 5 years 11.25 6 years 11.33 7 years 11.45 8 years 11.56 9 years 11.66 10 years 11.75 12 years 1194 14 years 12.10 Source: For treasury oills. SOi Gilts Lirrnited, tor dated securOres ICICI Secutities and Finance Company Linited. Figure 1. India Government Securities Yield Curve as at 31 July 1999 14.00D% - . _ f2O% - ~......... I..... .-. .... .... ------- - ---- -- --- 129 6.00% - - --- - ---- 2~400%"----- 0.00%6 P P~~ ~ Maturity 157 JOHN LEONARDO Structure. Primary dealers were introduced in March 1995 to acti- vate wholesale primary and secondary markets; satellite dealers were established in December 1996 to provide a mechanism for purchas- ing government securities from primary dealers for subsequent re- sale to retail investors. By September 30, 1999, there were 14 primary dealers and 4 satellite dealers, as well as 9 gilt funds (discussed later in this section). Primary dealers are required to bid for prespecified amounts of treasury bills and dated securities in RBI auctions each financial year. In addition, they must achieve a minimum success ratio of 40% for treasury bills and 33.3% for dated securities. Although the primary/satellite dealer system has contributed to growth of the secondary market since 1995, trading volumes remain very low compared with those before the onset of the 1992 scandal. As discussed below, this is due in part to weaknesses in the support- ing infrastructure and policy framework for the government securi- ties market. Increasing Hedging and Financing Mechanidnm for Dealers. In July 1999, the RBI issued guidelines covering interest rate swaps and for- ward rate agreements. This action has allowed banks, primary deal- ers, and financial institutions to hedge their interest rate risks and represents the first step in the development of the fixed-interest de- rivatives market. In August 1999, the RBI increased the number of nonbanking entities authorized to undertake reverse repo transactions in treasury bills and some dated government securities. In addition, it allowed these parties to undertake ready forward (repo) transactions. Reverse and ready forward repo transactions must be settled through the RBI's SGL accounts. No sales are allowed if the securities are not in the seller's portfolio -that is, no short selling is permitted. Central Market Infrastructure for Government Securities The notable aspects of this infrastructure are its trading mechanisms and clearing and settlement practices. Trading Mechanidsm. Trading in government securities is conducted through two channels: by telephone, on an over-the-counter basis; 0138 INDIA SURVEY and through the wholesale debt market segment of the NSE, initi- ated in June 1994. The WDM has an automated trading facility, but it is only for clients who are not major dealers in debt securities. Most secondary market trading in government securities is trans- acted over the telephone, with only a small proportion going through the NSE. This demonstrates market participants' preference for tele- phone trading rather than screen trading. All the major institutional players in the government securities market trade are registered on the NSE, which allows them to moni- tor the market via inquiry screens, establish exposure limits for other "participants" in the NSF market, provide settlement details to the NSE for distribution to the market, and transact repo trades through the WDM. As of October 1998, there were 79 registered non-NSE members participating in the WDM and trading through NSE brokers. Trading in the WDM is not fully on-line as participants have to enter transaction details separately. These details are disseminated to the market when trades are concluded/reported in the NSE sys- tem. The transactions are settled privately. At present, the RBI releases secondary market trading infor- mation daily, drawing on transactions actually settled through the SGL on any given day. Reported transactions may not necessarily have been concluded on the day they are reported. The SGL trans- fer form has to be lodged within 24 hours after the contract has been executed, but the lodgement date may not coincide with the date of the transaction. Consequently, the SGL system allows gaps to arise between the deal date, contract date, and actual settlement date. Clearing and Settlement. Ownership records relating to govern- ment securities are maintained on the RBI's SGL accounts. Over- the-counter transactions for those parties with RBI bank accounts (banks and primary dealers) are settled through the RBI's SGL and clearinghouse facilities using the RBI's delivery-versus-pay- ment (DVP) system, which synchronizes the transfer of securities with purchase and sale settlements. Other over-the-counter trans- actions are settled privately, with transfers effected through the RBI's depository. The volume of market-sensitive information released 139 JOHN LEONARDO by the RBI is increasing steadily; most of this information is avail- able on its website. Issuers and Investors in Government Debt As previously mentioned, government securities are issued only to wholesale investors. No retail market exists owing to constraints in the RBI's MIS. On the investor side, nationalized banks and insur- ance companies appear to be the largest holders of government se- curities, with provident funds and mutual funds having only minor holdings. Direct investment by retail investors is negligible. Gilt mutual funds specializing in government securities ("gilt funds") are a recent phenomenon. With two new funds established in October 1999, their total now comes to nine, and their value to about Rs 11.2 billion. The growth in gilt funds has been assisted by the declining margin of corporate bond yields over long-term gov- ernment securities, which fell from about 3% to 1.5% over the past year. Gilt funds are a useful addition to the government securities market. With growth, they will bring a longer-term and more varied perspective to the current single-dimensional market and provide an attractive option for retail investors. At the same time, this group should also be allowed to invest directly in retail government secu- rities if they wish. India has six money market mutual funds, with total assets of Rs 6.3 billion. As of October 31, 1999, the total value of funds they manage was in the neighborhood of Rs 869.5 billion. While the total amount of mutual fund investment in government securities is not readily available, it is unlikely to be more than Rs 400 billion. Organizations Formed by Market Participants. Market participants monitor developments in the government securities market through three organizations: the Primary Dealers Association, the Fixed Income Money Markets Dealers Association (FIMMDA), and the NSE's Committee for the Development of the Debt Market. The Primary Dealers Association has considerable interaction with the RBI. According to several primary dealers, the foreign securities firms with newly appointed primary dealers have brought an inflow tlAG INDIA SURVEY of expertise into the government securities markets. The FIMMDA is composed of banks, some of which are primary dealers. The NSE committee has been very active in raising matters affecting the WDM. One of its major achievements has been to establish an NSE bench- mark interest rate for the country's debt markets. Following the launching of the NSE-MIBID/MIBOR rate (NSE Mumbai inter- bank bid/offer rate) in June 1998, the NSE introduced 14-day, 1- month, and 3-month benchmark rates for the term money market. The NSE-MIBID/MIBOR rate has been used as a benchmark for some IRSs, FRAs, floating rate debentures, and term deposits. CORPORATE BOND MARKETS Public companies have been issuing term debt securities on a regu- lar basis ever since India's independence in 1947. Throughout this period, the primary market has been active, whereas the secondary market has experienced a marked lack of liquidity. Market volumes remain relatively low, and the secondary market yield curve is not well developed. In 1985-86, following some debt market reforms, state-owned enterprises (referred to as "public sector undertakings" or PSUs) began issuing term debt instruments known as "PSU bonds." Size of the Market Most bonds issued in the corporate bond market are issued as pri- vate placements and mostly by public sector corporations. Primary Market The total value of outstanding private sector debentures as of March 31, 1993, was Rs 98.4 billion (this figure excludes the value of zero coupon/deep discount bonds issued by private sector borrowers, which amounted to less than Rs 17 billion). In contrast, the total value of outstanding PSU bonds (bonds issued by public sector cor- porations) was over Rs 150 billion. During 1992-93 the annual value of new PSU issues amounted to approximately Rs 50 billion, while the figure for private sector bond issues was between Rs 10 billion 141 JOHN LEONARDO and 20 billion. The corporate debt market has grown considerably since 1996, mainly because of relaxed RBI controls on bank term lending and vMinistry of Finance amendments to the investment rules for LIC and provident funds (as discussed below). According to figures presented in the NSE's 1997-98 review of India's securities market, the total value of bonds raised by private placement amounted to Rs 181 billion in 1996-97 and Rs 309.4 bil- lion in 1997-98. Details of debt funds mobilized in the fiscal 1997- 98 are presented in table 9. Bonds raised by financial institutions and private sector corporations represented 56% of total placements in 1996-97 (Rs 101.4 billion) and 64% of total placements in 1997- 98 (Rs 198.1 billion). Remaining placements were made by federal and state government agencies. Bonds raised by private sector cor- porations represented 10% of total placements in 1996-97 (Rs 18 billion) and 25% of total placements in 1997-98 (Rs 76.2 billion). Owing to the generally sluggish economic conditions, few leading Indian private sector corporations are making issues at present. Bonds raised by Indian financial institutions represented 46% of total placements in 1996-97 (Rs 70 billion) and 39% of total place- ments in 1997-98 (Rs 121.8 billion). The 1997-98 figures include new bond issues by the Industrial Development Bank of India (IDBI) (Rs 15 billion in two bond issues) and the Industrial Credit and Investment Corporation of India (ICICI) (Rs 16 billion in five bond issues). Most corporate bonds continued to be issued as private place- ments during 1998-99. Between 1997-98 and 1998-99 private place- Table 9. Mobilization of Debt Funds in Fiscal 1997-98 (billions of rupees) Issuer Type Private placemest Public Total Public sector units 40.1 1.2 41.2 All India financial institutions/banks 121.8 18.1 140.0 Private sector 76.2 0 76.2 State undertakings and state financial institutions 71.3 0 71.3 Totals 309.4 19.3 328.7 Source: NSE (1998). E1.a2 INDIA SURVEY ments by banks, financial institutions, and private and public sector companies rose from Rs 301.0 billion to Rs 496.6 billion (RBI, 1998- 99). New issue activity continued to be dominated by public sector entities rather than private sector entities (see table 10). According to the RBI, total outstanding corporate debt inclu- sive of bonds issued by public financial institutions in September 1999 is estimated at Rs 1,400 billion. By contrast, the value of out- standing domestic debt was nearly Rs 5,000 billion (see table 4). The public sector is crowding out the private sector in the bond markets: the funding of the relatively higher-risk state government issues is increasing only because these public organizations are able to tap funds from other state-owned entities through regulatory controls. Although the range of debt instruments in the Indian corporate debt market is limited, it is steadily increasing. Authorization of FRAs in August 1999 is particularly encouraging. The corporate bond market as a whole, however, faces a weak supporting infrastructure. Details of corporate bond interest rates prevailing in late July 1999 are set out in table 11. A graph of the current yield curve in corpo- rate bonds based on yields in table 11 is presented in figure 2. Secondary Market Trading Volumes. The average total daily turn- over in private sector corporate bonds is low, about Rs 400 mil- lion. The total annual secondary market turnover continues to be less than the value of private sector bonds issued. Trends in annual Table 10. Mobilization of Debt Funds in Fiscal 1998-99 (billions of rupees) Issuer type 19g8-9 1997-98 Private sector: 169.8 92.0 Financial 121.7 43.2 Nonfinancial 48.1 48.8 Public sector: 326.8 209 Financial 203.8 96.6 Nonfinancial 123.0 112.4 Totals 496.6 301.0 Source. RBI (1998-99), table 5.3 143 JOHN LE.ONARDO Table 11. Corporate Bond Yields, July 31, 1999 (percent) Maturity Primary market Secondary market 1 year Not available 11.25% 2 years Not available 11.75% 3 years 12.25% 5 years 12.50% 7 years 12.75% Source: I C iCI (1 999). traded volumes at the NSE WDM (see tables 12 and 13) reflect modest trading in private sector corporate bonds at the NSE dur- ing 1997-98 (less than Rs 160 million per day). Between 1997-98 and 1998-99, the traded volume of corporate debentures (i.e., corporate bonds) fell from Rs 11.5 billion to Rs 8.8 billion (RBI, 1998-99). As a result of the low trading volume in private sector corporate bonds, secondary market retail and whole- sale yields differed by as much as 300 to 400 points (three to four percentage points) during 1999. No real secondary market yield curve has emerged to date. Although more institutions are now dealing in the secondary bond market, trading volumes remain very low. The increased trading in Figure 2. Indian Indicative Corporate Bond Yields as at 31 July 1999 13.00% 12.50% -./o --____ --_ ------_ ----------_ ---___-__-__- ! 12.00% /o .___-----------------___ --_ ----- > 11.50%- _ 11.000/ _____ _____ _ ___ _ __ __ ____ 1 Year 2 Years 3 Years 5 Years 7 Years Maturity INDIA SURVEY Table 12. Traded Volumes at NSE WDM, by Security Type, Fiscal 1995 to Fiscal 1998 (billions of rupees) 1994-95 1995-96 1996-97 1997-98 1998 June (9 months) Quarter Dated government securities 30.3 77.3 273.5 847.2 247.4 PSU bonds 7.6 10.0 19.7 25.2 3.5 Financial institutional bonds 0.7 1.5 8.0 15.3 9.0 Treasury bills 26.3 22.6 109.6 188.7 24 0 Bank bonds & CDs 1.0 1.4 5.0 11.9 2.2 Corporate bonds & CPs 1.9 5.9 7.0 24.3 3.5 Total 67.8 118.7 422.8 1,112.5 289.6 Source NSE(1998) Table 13. Distribution of Traded Volumes at NSE WDM, by Security Type, Fiscal 1995 to Fiscal 1998 (percent) 1994-95 1995-9C 1996-97 1997-98 1998 June (9 months) Quarter Dated government securities 44 66 64 77 85 PSU bonds 11 8 5 2 2 Financial institutional bonds 1 1 2 1 3 Treasury bills 39 19 26 17 8 Bank bonds & CDs 2 1 1 1 1 Corporate bonds & CPs 3 5 2 2 1 Total 100 100 100 100 100 Source: NSE (1998). government securities is having a modest impact on trading of cor- porate bonds, but a significant proportion of corporate bonds are still not traded. The planned removal of stamp duties on transac- tions in securities held in dematerialized form is expected to boost secondary market volumes considerably. Recent deregulation of the insurance sector (and perhaps the provident sector in due course) should also encourage increased secondary market trading in the medium to long term. Although RBI has authorized repos in corpo- rate bonds, providing the securities are held in dematerialized form, such transactions remain negligible for the time being. 145 JOHN LEONARDO Developments in Corporate Bond Infrastructure The two notable features of the corporate bond infrastructure are its trading and settlement practices and its depositories. Trading and Settlement. Like government securities, corporate bonds are traded over the counter and through the NSE. However, the over-the-counter market has no depository and clearinghouse ar- rangements, such as those provided by the RBI for the government securities market. PSU bonds are transferred by endorsement and delivery, with no stamp duty payable. Corporate bonds are transferred by a deed, a process that usually takes at least six weeks to complete. Sales cannot be made if the scrip has not been issued. It is also difficult to trade partial holdings. Settlement is undertaken privately, and trad- ers who do not exercise caution face potential settlement risks. In the absence of a real-time gross settlement system, the risk associ- ated with private settlement will remain an impediment until elec- tronic banking services are provided nationwide. Corporate bond transactions undertaken in the over-the-counter and private placement markets tend to lack transparency, as they are reported only if a broker is involved. There is no real-time price discovery mechanism. Depositories. India has two depositories: the National Securities Depository Limited (NSDL), established in November 1996; and the Central Depository Services (India) Limited, established in February 1999. During fiscal 1999, the market value of demateri- alized securities principally equities, held by NSDL increased from Rs 227 billion to Rs 1,150 billion. This jump represents significant progress in expanding the use of dematerialized trading. In the same year, the number of companies whose securities had been admitted for dematerialization at NSDL increased from 191 to 375. These companies represented about 80% of the market capitaliza- tion of all listed companies in India. Four government stock matu- rities and one state government issue are available for dematerialization at NSDL. No data were found for the opera- tions of the Central Depository Service Limited. 146 INDIA SURVEY Factors Influencing Demand for and Supply of Corporate Bonds Issuers. The issuers of corporate debt securities include private and public financial and nonfinancial entities. Paper is issued mainly by public sector financial and nonfinancial entities rather than private sector entities. During the past two years, private sector corpora- tions have made limited new bond issues, owing to the uncertain eco- nomic outlook for the medium term. The overall supply of private sector corporate bonds is kept in check by the significant size of India's state-owned enterprise sector. Increased privatizations will help the private sector corporate bond market grow in the medium term. According to the RBI, most securitization deals transacted to date have consisted of direct purchases of receivables by institutions and large nonbanking finance companies. There has been little reported use of special purpose corporation structures. InveAtors. Institutional investors -principally banks, insurance com- panies, mutual funds, and provident funds - hold approximately 80% of corporate bonds, with the balance (20%) held by retail investors. A relaxation in investment regulations for banks and institutional investors has created additional investment interest in corporate bonds and led to strong growth in new corporate bond issues since the 1997 June quarter. The RBI first allowed scheduled banks to invest in corporate bonds (and shares) in October 1993, up to a ceiling of 5% of their incremental aggregate deposits of the previousyear. In October 1996, banks were allowed to purchase bonds (and shares) in the second- ary market within the existing ceiling. The regulations on prefer- ence shares and bonds were removed in April 1997. The banks' investments in bonds have increased materially in the two years since the removal of investment controls in 1997. Ac- cording to the RBI (1999), bank investments in bonds/debentures/ preference shares issued by the private corporate sector increased from Rs 23.5 billion in March 1997 to Rs 94.75 billion in March 1998. During fiscal 1999, these investments further increased to Rs 178.8 billion. Between March 1997 and February 1999, the banks' investments in bonds/debentures/preference shares issued by PSUs increased from Rs 118.6 billion to Rs 231.3 billion. 147 JOHN LEONARDO In 1997-98, the LIC's ceiling for market-related investments rose from 25% to 60%, with the balance invested in public sector securi- ties. In 1998-99, nongovernment provident funds were allowed to make investments in private sector bonds/securities for the first time (up to a ceiling of 10% of total investments in securities with invest- ment-grade rating from at least two credit rating agencies). At the same time, government provident funds are not allowed to invest in private sector securities, although up to 40% of their investments can go into bonds/securities issued by public sector organizations. The current investment regulations for nongovernment provident funds are heavily skewed in favor of public sector securities. Retail interest in private sector corporate bond investment re- mains low because of the limited use of fixed-interest investment in India. Retail investors are almost absent, except for those holding issues made by the two state-owned banks (the Industrial Develop- ment Bank of India and the Industrial Credit and Investment Cor- poration of India). Intermediaried. The major intermediaries in the Indian debt markets are domestic and foreign banks and the primary dealers. Although detailed information on turnover by different intermediaries in the government securities and corporate bond markets is not readily avail- able, turnover statistics at the wholesale debt market at the NSE pro- vide some broad guidance (for details of turnover in the period April 1, 1996, to June 30, 1998, see table 14). IMPEDIMENTS TO BOND MARKET DEVELOPMENT The corporate bond market in India faces several impediments, primarily the stamp duty, which hampers trading. Several exter- nal, macroeconomic, and internal factors also create problems for this market.3 External Impediments External constraints consist mainly of political, macroeconomic, and legal factors, along with weaknesses in the broader financial system. 148 INDIA SURVEY Table 14. Wholesale Debt Market Volume on the National Stock Exchange, 1996-98 (billions of rupees) 12-month 12-month 3-month Investors Match31, 1996/97 March 31, 1997/98 AprIlune 1998 Rs % Rs % Rs % Foreign banks 314.0 37.13 504.1 22.65 139.8 24.13 Indian banks 253.8 30.01 917.7 41.24 236.1 40.75 Primary dealers 136.0 16.08 269.6 12.12 88.2 15.24 Financial institutions and mutual funds 25. 3.01 85.2 3.83 18.4 3.17 Trading members and corporates 116.4 13.76 448.7 20.16 96.6 16.68 Total 845.6 100.00 2225.3 100.00 579.1 100.00 Around the Bond Markets Political &ldues. Political developments slowed the pace of economic reform in India during 1998-99. The previous government was un- able to implement several planned financial reform measures owing to limited political support during its term in office. The newly elected government of Prime Minister Vajpayee, which took office in Octo- ber 1999, has undertaken to implement some outstanding financial reforms and has already passed the long-awaited legislation deregulating the insurance sector. The powerful financial sector trade unions create additional obstacles to improving the viability of the state owned banking sector. The first major concern is the slow pace of economic reform. Although there is apparent widespread acknowledgement in India of the need to implement further economic reform measures (such as addressing the federal governments's fiscal situation and privatization of government-owned corporations (including the public sector banks), Indian politicians have been slow in recent years to implement recommended measures. Unless substantial economic reforms are initiated in the short to medium term, subse- quent governments may be faced with implementing unpopular measures, which will be resented all the more if the economy moves into an unexpected downturn. A second concern is that recent governments have, on balance, been reluctant to let the financial markets operate in a conventional 149 JOHN LEONARDO market-oriented manner, partly because they consider some finan- cial instruments and practices to be highly speculative, especially after the 1992 scandal surrounding government securities. This stance creates further obstacles to the development of efficient debt mar- kets. Other obstacles stem from limited understanding of debt mar- kets and the desire in some quarters to retain the inefficient public sector banks. A third problem is the government's inability to implement ma- jor financial sector reforms. Despite efforts of the Ministry of Fi- nance and the RBI, the outgoing Parliament was unable to pass legislation exempting from the stamp duty debt market transactions in securities that are held in dematerialized form. (To do this, the government would have to amend the Securities Contract [Regula- tions] Act and replace the Public Debt Act.3) The proposed legisla- tive changes are critical to the development of secondary market trading in the corporate debt market. Such changes would also pro- mote the use of dematerialized (electronic form) government secu- rities, introduction of new debt instruments to broaden the debt market, and creation of an effective regulatory agency for this mar- ket. The merits of introducing derivative products for the debt mar- ket cannot be considered until after the government has determined which organization (either the RBI or SEBI) will be responsible for regulation of the debt markets. Political instability is another, more broadly based impediment. Politicians supporting proposed banking and insurance reforms face a possible political backlash from some voters. Political instability frustrates the drafting of legislative changes, such as deregulation of the insurance sector. In that case, it took more than four years to complete the draft legislation. Even though all major political par- ties in the outgoing Parliament supported the proposed insurance reforms, they were not passed. Macroecononu2cpprobLenw. One major impediment to the growth of debt securities markets is India's extensive government borrowing, which crowds out the private sector, or has the potential to do so. Large 3 See Footnote 1. 15C N0 - INDIA SURVEY government spending and low government revenues have put the fiscal deficit at a level that impedes the development of both the private sector and the domestic corporate debt market. Yet the gov- ernment continues with its spending preferences, such as its large ongoing subsidies. The borrowing requirements of the public sec- tor (including public sector undertakings) retard private sector de- velopment by crowding out private borrowing. The government's excessive spending is also due to its continued support of numerous poorly performing state-owned corporations and rising government debt-servicing costs. With government revenue still relatively low, it is all the more difficult to reduce the fiscal deficit. Broader Legal and Regutlatory Environment In the broader legal environment, the major problem is India's com- mercial law, which does not provide a supportive legal framework for debt market development. Sweeping changes are required here, especially to enable bond investors and trustees to enforce commer- cial contracts. Extensive legal and tax policy changes are also re- quired to build an active market in securitized paper. Acrosd the Financial System The problems relating to the financial system can be divided into those arising within the banking sector and those in the government securities market. The Banklingqsector: The state of the banking system is impeding de- velopment of the local corporate bond market. Notwithstanding the banking reforms undertaken since 1992, India's banking sector will not fully comply with international prudential banking stan- dards until 2002. Numerous other problems remain that need close RBI monitoring. Special attention should also be given to the ratio- nalization of the public banking sector in order to increase compe- tition. The weakness of public sector banks puts additional strain on the development of the secondary market in government securi- ties and to a lesser degree, the secondary market in corporate debt securities. 151 JOHN LEONARDO Th9egqoierlnzentsiecuilties market. As noted earlier, the government se- curities market is increasingly providing a solid foundation for cor- porate bond market development. But despite this gradual strengthening, the government securities market will not be able to operate at full tilt unless something is done to improve primary dealer relationships with RBI; to introduce short selling; to facilitate trad- ing, clearing, and settlement; and to cultivate the investor base. With regard to primary dealer relationships with RBI, primary dealers are concerned about the increased proportion of dated se- curities that the government is placing directly with the RBI, ow- ing to its growing deficit. In order to limit pressure on market bond yields (and therefore maintain stability in the money mar- ket), the government placed significant volumes of its securities with the RBI during 1998-99 -the amount came to almost Rs 290 billion, out of a total of Rs 700 billion. The RBI subsequently re- sold these securities in the secondary market when short-term con- ditions were more favorable. On some occasions when the RBI has taken dated securities into its portfolio, it has resold the securities to primary dealers within half an hour of purchase from the government. This practice re- duces the profit of primary dealers. According to the RBI (1998- 99), total commission/underwriting fees paid to primary dealers in 1998-99 amounted to Rs 833.1 million, compared with Rs 1,100 million in 1997-98. With regard to short selling, current RBI guidelines prohibit banks from going short in securities. This makes it difficult for primary dealers to actively provide two-way quotes in government securities, which significantly impedes market making. Although these guide- lines do not extend to other market participants, it has become cus- tomary for them to be in the Indian debt markets. There are no legislative prohibitions, however, if the RBI wished to approve short selling of securities by the banks. Currently, there is no real short- term incentive for major dealers in government securities to seek primary dealer status. Increased market making in government securities markets is unlikely to change appreciably until short selling is approved. The RBI is currently considering such a move, although probably not in the short term. The RBI's unwillingness to approve short selling 152 INDIA SURVEY reflects a cautious approach to implementing further changes in the government securities market and historical concerns emanating from the problems experienced in the government securities market after the 1992 scandal. Another point the RBI should consider is that short selling would encourage primary dealers to increase their use of competi- tive two-way quotes. Before short selling is approved, reasonable volumes of stock would have to be available in individual maturi- ties to prevent a short squeeze from developing. The RBI should make every effort to attend to these and other details that would help create a climate for the introduction of short selling of gov- ernment securities. As far as the central market infrastructure is concerned, some market participants believe that the largest impediment to the de- velopment of the government securities market is the lack of on-line trading and real-time settlement facilities. Because of delays in the distribution of market information, distortions occur on occasion. These anomalies tend to benefit banks with large market shares. A sophisticated on-line trading system would alleviate many of these problems. The RBI would like to promote on-line trading with accompa- nying real-time settlement, especially because it would be beneficial for price discovery, market transparency, and efficiency. However, the RBI does not have the computer resources to provide real-time trading and settlement capabilities, and legislative impediments also stand in the way of such a move. As a result, secondary market trad- ing in government securities is showing limited growth. With a bet- ter supporting infrastructure, trading volumes could increase by an estimated 20% to 30%. Without real-time systems, there are delays in the dissemination of secondary market trading information. MIS systems also need to be expanded before a retail market in govern- ment securities can develop. These matters are currently being ad- dressed by the RBI, which foresees changes in these areas by June 2001. In June 1999, India inaugurated a major initiative to facilitate interaction within the financial sector, known as the Indian financial network (INFINET). INFINET will provide the communication backbone of the proposed integrated payment and settlement sys- - 153 JOHN LEONARDO tem (IPSS) for the financial sector and facilitate the establishment of a real-time gross settlement system. The RBI (1998-99) has indi- cated that it plans to introduce a real-time gross settlement system "in about one and a halfyear's time" (i.e., by June 2001). Full screen trading seems unlikely to occur for at least two years. The RBI is also working to establish a system for disseminating details of all trades undertaken by members of the trading network as soon as the deals are done, though the traders would remain anony- mous. The RBI has been discussing this idea with the Primary Dealers Association. Barring satisfactory progress in developing on-line trad- ing and settlement systems, a useful alternative would be to con- sider integrating the NSE and RBI systems. Since the RBI is unable to service current trading volumes ef- fectively, some questions arise about its role as a clearinghouse and a depository. While the RBI's dematerialized registry system is satisfactory, some market participants suggest trading could in- crease by 20% to 30% with a better settlement system. They also wonder how quickly the RBI will be able to improve settlement procedures. Other concerns relate to market participants. Some worry about the difficulty of developing the local retail investor base and the lack of supporting market infrastructure. Thus it is not surprising that satellite dealers have encroached on the primary dealers' market. At the same time, primary dealers have entered the retail market. Because of various problems with the system, there are only four satellite dealers so far. They have slim profit margins because the RBI has not given them access to call money, they can only borrow up to 50% of the government paper they own, and they face significant capital constraints. However, the RBI's planned computerization of its branch offices will provide the foundation for the development of a retail market in govern- ment securities. Another problem is that few investors are aware of government securities. This problem can be addressed through a major educa- tion initiative. Although most of the scheduled banks have a sound understanding of the government securities market, this is not the case among the other banks, the insurance companies, and the provi- dent funds. INDIA SURVEY The RBI is well aware of the concerns of market participants and has taken several steps to improve market infrastructure, while also continuing to evaluate other possible changes. This is in line with its policy objective of progressively developing the government securities market. Market participants remain skeptical, however, that the RBI will be able to implement pending major changes in market infrastructure within proposed time frames. Priorities it addressbVq external impedimentf. The priorities that should be set in dealing with such impediments can be ranked as high, mod- erate, and low according to the degree to which their removal would contribute to substantial increases in corporate bond trading. As tables 15 and 16 show, external impediments are all significant, and none fall at the low end of the scale. Internal Impediments Significant impediments exist within India's debt markets themselves. Particularly notable are the weaknesses in the supporting regula- Table 15. External Impediments of High Priority Political Slow pace of economic reform Past reluctance to deal with financial sector reform measures Government's historic inability to implement agreed financial sector reforms Opposition to economic and financial reform measures Macroeconomic Crowding out of private sector by excessive government borrowings Retention of a large number of inefficient public sector undertakings that should be privatized Significant regulation of organizations controlting long-term savings, particularly provident funds Broader financial system Without rationalization of public sector banking, banking sector not competitive National debt markets not fully developed Legal environment Ineffective commercial law ______ _____ _____ _____ _____ ___ _ _ _ - 155 ,JOHN LEONARDO Table 16. External Impediments of Moderate Priority Macroeconomic Low government revenue, limits government's ability to reduce its tiscal deficit Low savings rates, limit resources available for investment in corporate debt securities tory frameworks, in some sections of the central market infrastruc- ture, and in the overall level of knowledge of significant institu- tional investors. Some steps have already been taken to address these weaknesses. Regulators and Regulations India's government and corporate debt securities markets face a number of regulatory problems. One is the uncertainty regarding which apex organization will be responsible for the debt markets. Another is the delay in implementing planned stamp duty changes, which is limiting secondary market trading in corporate debt secu- rities. Yet another is the need for some minimum regulation of the corporate debt private placement market. These issues can be sub- divided into those concerning regulators and those having to do with regulations. Regulztor,s. The two main questions that need to be addressed here are how can regulators improve their knowledge and how should the RBI and SEBI split responsibility for regulation of the local bond markets? The regulatory parties (Ministry of Finance, RBI, and SEBI) do not always have the technical knowledge required to deal with more complex issues relating to the development of the debt markets. As for the question of whether RBI or SEBI is officially responsible for supervising the corporate debt markets, this matter will be resolved when the Securities Contract (Regulations) Act is amended. The uncertainty surrounding the question for the time being is hindering market development. Regulation. Wholesale investors have expressed support for some (limited) regulation of the private placement market. This matter I;6f INDIA SURVEY will inevitably be considered by RBI and SEBI in light of the present inquiry into the performance of numerous trustees for corporate bond issues and the findings arising from a recent HLC subgroup study of the debt markets. The Ministry of Finance clearly needs to promote legislation exempting transactions in securities in dematerialized format from stamp duty.4 This change was outlined in the 1999 budget but was not implemented by the outgoing Parliament. This delay has put a serious constraint on secondary market trading. The government also needs to improve policies for supervising institutional portfo- lios, especially in the case of insurance companies (and provident funds), which are currently not required to value fixed-interest in- vestment portfolios at market yields. This is expected to change when the planned insurance reforms are implemented. Rules that apply to provident funds are also in need of reform. For instance, the legislative requirement for a 12% guaranteed re- turn on provident funds should be abolished. This policy helps to prop up interest rates, impedes development of retail fixed-interest security markets, and acts as a disincentive for provident funds to trade in the government security and corporate debt secondary markets. The guarantee should be eliminated prior to the introduc- tion of retail government securities. Wholesale investors should not be allowed to invest in retail government securities when these in- struments are eventually introduced to the market. The regulations governing provident fund investments should be progressively re- placed with rules based on international prudential norms. Impediments Relating to Central Market Infrastructure The trading, clearing, settlement, and information infrastructure for the corporate bond market is underdeveloped. There are no plans for any supporting infrastructure for over-the-counter transactions in corporate debt securities that are executed outside of the NSE's WDM facility. There is only limited support for full screen-based trading among market participants, in spite of its advantages over See Footnote 1. 157 JOHN LFONARDO the current over-the-counter trading practices. There is no principal entity for settlement of transactions in corporate bonds. Settlement may be made through the NSE's facilities for WDM participants. Since transactions are reported only when brokers are involved, certain market information is not disseminated. Although India has four credit-rating agencies, they need to become more proactive. Market Participants The present low trading volumes and poor market liquidity in India have little to do with the number of issuers or investors. Rather, the problem is the stamp duty, on one hand, and the fact that most large investors know little about debt portfolio management techniques, on the other. A serious concern of institutional investors is the lack of regulation of the corporate debt private placement market. Isjuers. The main problems on the issuer side are limited develop- ment of the corporate debt secondary market, which reduces financing options for issuers, and the lack of retail interest in corporate debt securities. In spite of the higher costs associated with making public corporate debt issues, prospective issuers would like to see retail investors having a healthy interest in corporate debt securities. Inveator,. One problem on the investor side is the poor liquidity in the secondary market for corporate debt securities, which will be partly alleviated when the promised stamp duty reform occurs. Another impediment arises from deficiencies in the private place- ment market for corporate debt securities. Wholesale investors are concerned about the performance of trustees, low disclosure stan- dards, lack of transparency in trading information, and the need for some minimum standard practices in the corporate debt private place- ment market. A further concern is the limited knowledge of debt market portfolio management. Until recently, nationalized banks, insurance companies, and provident funds knew little about debt management techniques. Hence these organizations have tradition- ally followed a cautious approach to managing their portfolios, es- pecially in the aftermath of the 1992 scandals. These institutions are now taking a greater interest in the market, and some have even INDIA SURVEY begun trading in the secondary markets. Limited participation by long-term investors in the secondary market generally results in the market having a one-sided view. As a result, market yields tend to become inflated during periods of tight liquidity, when overnight interest rates drive the market's pricing of long-term securities. For wholesale investors, a main concern is the need for more debt mar- ket education services, at reasonable cost. Intertnediari,es. The main constraints for intermediaries operating in the market today are the inability to short-sell and thin secondary mar- kets, which restrict their profit-making opportunities. Priorities in Addressing Internal Impediments Priorities for addressing these impediments can again be ranked on a scale of high, moderate, and low. As tables 17 and 18 show, most of the impediments here are in the high category and only one is in the moderate category. None are considered to be of low priority. RECOMMENDATIONS FOR REMOVING IMPEDIMENTS This section outlines specific recommendations for removing exter- nal and internal impediments to the development of the debt mar- kets in India. Removing External Impediments The following recommendations apply both around the bond mar- kets and across the financial sector. Around the Bon9dMarkets. The government should prepare new eco- nomic and financial sector reform programs. These should focus on reducing government expenditure, increasing government revenue, reducing the government's borrowings, and accelerating the privatization of commercial activities now carried out in the public sector. In addition, the reforms should accelerate the privatization of public sector banks and gradually deregulate the provident fund - 159 JOHN LEONARDO Table 17. Internal Impediments of High Priority Rules, regulatiorns, and regulators Lack of sophisticated understanding by regulators Uncertainty as to the apex organization responsible for supervising the corporate debt markets RBI's tolerance of weaker prudential banking policies, which delays improvements in efficiency in the banking sector Poor monitoring of organizations managing long-term savings Inadequate standard accounting practices Limited overall knowledge of the role of the debt market in private sector development, especially in the public sector Central market infrastructure No real-time infrastructure for tran in government and corporate debt securities undertaken outside the NSE system Limited distribution of market information for secondary market transactions in government and corporate debt securities RBIs limited MIS systems Primary dealer structure not achieving RBI's obJectives owing to the manner in which the RBI is undertaking government debt management operations Outgoing government's failure to implement proposed stamp duty reforms, which is seriously limiting secondary market trading in corporate debt securities Credit rating agencies not very proactive. Market parficipants issuers Limited secondary market trading Low retail interest Market participants: investors Continuing low level of investor confidence in debt securities, owing to the 1992 scandal in government securities Deficiencies in the corporate debt private placement market Nonbank wholesale investors relatively inactive in the secondary markets Limited reasonably priced training facilities Continuing poor secondary market liquidity Weak portfolio management skills Market participants: intermediaries Limited market-making capability Thin secondary markets 160 INDIA SURVEY Table 18. Internal Impediments of Moderate Priority Central market inf rastructure Credit rating agencies not very proactive sector, through the progressive elimination of guaranteed investment returns by provident funds and a gradual relaxation in provident fund investment restrictions. It is also essential to develop and imple- ment a strategy for reducing the fiscal deficit. Multilateral agencies should provide the incoming government with continued assistance. On the broader legal and regulatory front, the Ministry of Fi- nance, RBI, and SEBI should encourage the incoming government to promptly enact the debt market legislation that the outgoing gov- ernment did not pass before it left office. This legislation would have exempted from stamp duties those debt market transactions in secu- rities that are held in dematerialized form; would have amended the Securities Contract (Regulations) Act to allow the introduction of new debt instruments to broaden the debt market and creation of an effective regulatory agency for this market; and would have replaced the Public Debt Act to allow the electronic transfer of government securities. In addition, the government should enact legislation au- thorizing deregulation of the insurance sector. The RBI has indi- cated that it expects amendments to the Securities Contract (Regulations) Act and the Public Debt Act to be passed by the end of 2000. Finally, the government should initiate an extensive review of commercial legislation that is outdated and a barrier to achieving a major improvement in corporate governance. This matter is now under review and should be given the government's full support. Also important, accounting practices for some elements of debt market transactions should be brought in line with generally accepted international standards. The Ministry of Finance (together with the RBI and SEBI) should encourage the Institute of Chartered Ac- countants of India to amend its standard accounting practices as soon as practicable.4 Acrods the Finanecal Sector. To guard against unforeseen setbacks, the RBI should ensure that all banks implement the planned bank- 161 JOHN LEONARDO ing sector reforms by 2002. To strengthen the government securi- ties market, the RBI should review the present primary and satel- lite dealer structure in government securities and determine whether any changes to current policy are appropriate; complete the planned introduction of a national branch office settlement system to help develop a retail market in government securities; and review the present timetable for the planned introduction of a real-time gross settlement system for government security transactions. The RBI should also consider integrating its depository and clearinghouse systems with those of the NSE if the implementation of a real-time gross settlement system does not appear feasible within, say, two years. Removing Internal Impediments These recommendations can be divided into three groups: those pertaining to rules, regulations, and regulators; central market in- frastructure; and market participants. Rules, Regulations, and Regulators. As soon as the government has determined whether RBI or SEBI is to be responsible for regulating the corporate debt market, the organization selected should review the regulatory requirements for the corporate debt private place- ment market to determine whether any regulation of this market is necessary. From time to time, the RBI, Ministry of Finance, and SEBI need help in understanding technical issues associated with the development of the debt market and in formulating related poli- cies. That means the Indian government should increase its finan- cial support to the ministry and SEBI to allow them to play larger roles in promoting the debt markets. Central Market Infrastructure. The regulatory agency for the cor- porate debt market should monitor clearinghouse arrangements and act as a catalyst for development of additional infrastructure, if nec- essary. Current trading in corporate debt securities undertaken out- side the NSE structure is limited because the appropriate trading and settlement infrastructure is lacking. This matter must be ad- dressed in the long term. The immediate concern should be to de- 162 INDIA SURVEY velop the government securities market infrastructure, which is slated within the next two years. Market Participants. All market participants need to learn more about debt markets. Mechanisms are required to provide long-term educational services for the different groups participating in the wholesale debt markets. Likely participants would include profes- sionals employed in the accounting, actuarial, banking, corporate, financial, insurance, legal, and regulatory sectors. The first kind of technical assistance provided should enable interested Indian par- ties to evaluate possible educational mechanisms and prepare pre- liminary cost estimates, taking into consideration international experience. A draft strategic plan should subsequently be prepared for the mechanism of choice. Some possible mechanisms would be to expand the role of the present FIANMA organization, create a new entity to service these needs, or add this activity to an existing private sector organization that has a marked focus on education. To judge by the apparent interest in these mechanisms, there should be little difficulty obtaining a private sector sponsor for this exer- cise. With increased access to such educational services, portfolio management skills in the private sector should improve greatly over the long term. Investors urgently need to see improvements in the management of private sector portfolios. Present governance requirements for organizations managing long-term savings are rudimentary and should be revised. CONCLUSIONS India's debt markets have experienced considerable growth since 1992. The RBI has helped the government securities market expand by introducing new debt instruments and creating market structures to boost secondary market trading. These steps have been very benefi- cial, as can be seen in the increasing volume of securities traded. Even so, market volumes in government securities remain below the levels prevailing before the 1992 scandal emanating from fraudulent tradcing in securities. 163 JOHN LEONARDO Activity in the primary corporate debt market has also increased, following the liberalization of investment regulations for banks, in- surance companies, and provident funds. To date, the increased pri- mary activity has not had much effect on secondary market activity owing to the stamp duty on secondary market debt transactions. Government securities and corporate debt markets currently face a number of impediments to growth. Many can be easily surmounted. Others will require a firm commitment by the parties affected. The RBI, Ministry of Finance, and SEBI must work together to remove these obstructions as soon as practicable. The Regulators (particu- larly the RBI, Ministry of Finance, and SEBI) should adopt a flex- ible approach to changing market infrastructure and formulating supporting policies because it is essential to achieve material im- provements within a relatively short period of time. The opportu- nity is ripe: the new government has indicated that further economic and financial reforms are likely to be introduced during its term in office. REFERENCES Association of Mutual Funds in India. 1999. "AMFI Monthly" (October). Issue IV (Internet version). Banks, Erik. 1994. The Einegying Fixed-Incomne Ala rketi 'in A.s,ia:A C'ountry-by- Country Guzide toz the Strutcture, Practice anzd Pl/zyerl of the World :, Fastest Grotv1Zng Debt A'farketm. Chicago, Ill.: Probus. ICICI Securities and Finance Company Ltd. 1999. "Mark to Market," for the week ending July 31. India, Government. 1998. "Economic Survey 97-98, Capital and Money Markets." . 1999. "Economic Survey 98-99, Capital and Money Markets." International Finance Corporation. 1999. "South Asia Debt Market Sym- posium October 1999: Informal Meeting Notes." Washington, DC. National Stock of Exchange India Limited (NSE). 1998. Inidial Securities Ahirket: A Rei'eui' (October). Reserve Bank of India (RBI). 1997-98. "1997-98 Annual Report." . "1998-99 Annual Report" (Internet edition). . "Bulletin May 1999." 164 INDIA SURVEY . 1999. "Debt Market in India: Reforms, Impact and Future Agenda." Speech delivered by Usha Thorat, RBI Central Office, to South Asia Debt Market Symposium (October). SBI Gilts Limited. 1999. "Market Analysis-Review of Fortnight ended 31 July 1999" (Internet version). ENDNOTES 1. While official Indian publications use the figure crore (= 10,000,000), numbers in this discussion have been converted into billions. "Billion" is defined as one thousand million (= 1,000,000,000). Thus Rs. 88,937 crore = Rs. 889.37 billion. The Indian Central Statistics Office's "Quick Estimates," as reported in RBI (1998-99). 2. The SGL is where the RBI records who owns the government secu- rities. 3. For the purposes of this discussion, external impediments are deemed to be influences outside the capital markets, but include forces within the country. Internal impediments are related specifically to the capital mar- kets and include regulators, market infrastructure, issuers, investors, and intermediaries. This definition has been suggested by I FC. External and internal impediments to the development of the bond market in India are discussed later in the chapter. 4. SEBI should be proactive in promoting efforts by the Institute of Chartered Accountants of India to adopt international accounting stan- dards on all accounting issues associated with fixed-interest securities. Matters of concern to wholesale investors in corporate debt securities include the determination of accrued interest income, the valuation of investments, and accounting for repo transactions. 165 CHAPTER 12 Pakistan Survey: Issues in Local Bond Market Development JOHN LEONARDO Pakistan's fixed-income securities markets are only partly devel- oped by developed country standards. They currently com- prise markets in government securities and corporate bonds. There are no real markets for short-term bank bills, certificates of deposit, or commercial paper in Pakistan. Nor are there any futures exchanges. Furthermore, the government securities market does not provide efficient market signals for pricing corporate debt securi- ties. Hence the challenge for policymakers is to provide a more sup- portive and consistent economic policy environment, improve governance standards, improve the rule of law as it applies to the capital markets, introduce new debt instruments and improved market practices, establish real-time market infrastructure, and improve edu- cation services for debt market participants. The Pakistani debt markets show good promise. The State Bank of Pakistan (SBP) and Securities Exchange Commission of Paki- stan (SECP) have demonstrated a steady commitment to reform. Continuing progress in addressing identified impediments will as- sist in the development of the Pakistani debt markets and pay huge dividends in encouraging capital formation and efficient allocation of capital in the development of the economy. This review describes the state of these markets, key impedi- ments to their development, and some possible ways of moving forward. The discussion opens with an overview of the Pakistan JOHN LEONARDO economy and government securities market, followed by a descrip- tion of the government bond market. This market has failed to pro- vide adequate signals to the private sector regarding the cost of "risk-free" sovereign debt, especially for borrowings of medium and long tenors. The factors contributing to the absence of markets in certain other debt instruments in Pakistan are also discussed, along with the Pakistan corporate bond market. Attention then shifts to impediments to the development of the Pakistan debt markets and their relative priority in the framework of remedial action. The chapter concludes with some recommendations on how to remove the impediments. ECONOMIC OVERVIEW Fiscal 1998/99 was a very difficult year for Pakistan. Its economic performance, especially in the first half of the year, was adversely affected by economic sanctions and the suspension of multilateral funding. These measures came about as a direct result of the May 1998 nuclear tests. Following these setbacks, however, the economy is expected to recover steadily in 1999/2000. Pakistan's recent eco- nomic performance is set out in table 1. The 1999 federal budget, presented in June 1999, shows a slight increase in the deficit forecast for 1999/2000. Inclusive of miscel- laneous financing items, the deficit is expected to increase from Rs 118 billion to Rs 135 billion in fiscal 1999/2000, which is equal to 3.3% of forecast GDP, somewhat less than the previously fore- cast 3.4% of GDP. Gross revenue is expected to increase from Rs 502 billion to Rs 561 billion and total expenditure from Rs 570 billion to Rs 642 billion. However, a federal budget deficit of around Rs 146 billion (4.4% of GDP) is considered a more likely out- come. I Increased government spending, higher exports, and lower in- terest rates are likely to be the major contributors to the expected I The figures in this paragraph are from Jardine Fleming Research (1999), fig. 12. 168 PAKISTAN SURVEY Table 1. Pakistan's Recent Economic Performance, 1997-2000 (showing percentage change from previous year) 1997 1998 199 2000(Gvtp 20111(F)b GDP 1 3 5. 4 3. 0 5. 0 4. 5 Agricultural production 0. 1 5 9 0. 4 4. 3 4. 5 Industrial production 1. 0 6. 2 4. 1 5. 8 4. 7 Services 2.1 4. 8 3. 6 5. 1 4 4 Intlation(CPI) 11.8 7. 8 6. 0 6. 2 7.0 a Government target figuLres. b. Jardine Fleming forecasts. Source:Jardine Fleming Research (1999). recovery in 1999/2000. The government is forecasting a reduction in the current account deficit from US$2,416 million to US$1,771 million on the basis of an assumed 18% growth in exports.2 This assessment may be too optimistic, however. Savings as a proportion of GDP remains relatively low, with an expected increase from 11% in 1998/99 to 15.3% in 1999/2000. The size of the informal sector in Pakistan remains significant and may well be larger than the official economy. Pakistan is a heavily indebted nation. In 1997/98, the value of total government debt was equivalent to 91.2% of GDP. Trends in the government debt as a percentage of GDP are set out in table 2. Since at least 1997/98, the government's fiscal deficit has been financed principally from internal (non-banking) sources (see table 3). Overall domestic liquidity continues to be managed principally through open-market operations conducted by the SBP, the central bank. During the past year, the SBP has maintained stable condi- tions in the money markets despite the difficulties arising from the May 1998 nuclear tests. Currently, the government is introducing numerous initiatives to improve long-term economic performance, including a program to develop the capital market, which is discussed later in this chapter. 2Merrill Lynch (1999). ' Merrill Lynch (1999). 169 JOHN LEONARDO Table 2. Pakistan's Government Debt, 1993-98 (billions of rupees) Debt parameters 1993-94 1994-95 1995-96 199697 1997-98 Domestic debt 702. 0 798. 6 908. 9 1041. 9 1151. 4 (percentage of total debt) (48. 4) (50. 4) (48.9) (48. 0) (45. 7) External debt 749. 4 785. 1 951. 0 1127.3 1366. 9 (percentage of total debt) (51. 6) (49. 6) (51. 1) (52. 0) (54. 3) Total debt 1,451.4 1,583. 7 1,859.9 2,169.2 2,518.3 Percentage of GDP Domestic debt 44.6 42. 4 42. 0 43. 3 41. 7 External debt 47. 6 41. 7 43. 9 46. 9 49. 5 Total debt 92.2 84 1 85.9 90.2 91.2 Total debt servicing as 45. 8 44. 6 47. 6 56. 8 54. 5 percentage of current expenditure Source. State Bank of Pakistan I1997/98), tabie 7.1. Table 3. Funding of the Pakistan Government's Deficit, 1996-99 (billions of rupees) 1996-97 1997-9 1998-99 provfsional actual revised budget budget esrimates estimates Deficit to be financed 156. 6 147. 4 143. 2 Financed through: External resources (net) 27. 6 35. 1 14. 7 Internal resources 129.0 112.3 128.5 Domestic nonbank 56.5 81.6 89.3 Banking system 72. 5 30. 7 39. 2 Source: State Bank of Pakistan (1997/98), table 4.1. PAKISTAN GOVERNMENT SECURITIES MARKET An analysis of the securities market must take into account its size, infrastructure, and participants. 170iC PAKISTAN SURVEY Market Size Market size may be examined from the perspective of outstand- ing government debt, trends in the issue of government securi- ties, secondary market turnover, and the yield curve. Outstanding Government Donmetic Debt. The government of Pa- kistan currently issues a range of securities. Trends in the volume of outstanding Pakistan government domestic debt are shown in table 4. Table 4. Outstanding Pakistan Government Domestic Debt, 1996-98 (mil/ions of rupees) June 30, 1996 June 30, 1997 June 30, 1998 (provisional) Permanent debt 294,705 296,283 289,707 Market loans (including provincial 23,007 22,970 17,454 government loans) Federal government bonds 9,471 10,336 10,231 Special government bonds for 7,304 9,135 10,346 State Life Iisurarice Corporation Bearer National Fund Bonds 21,681 21,681 21,680 Federal investment bonds 165,809 156,525 146,296 Bearer certificates 16,358 15,014 12,567 (foreign exchange/currency/U.S. dollar) Prize bDnds 51,000 60,547 7T,058 Other 75 75 75 Floating debt 361,298 433,833 473,849 Ad hoc treasury bills 61,478 61,500 61,524 6-months treasury bills 299, 807 0 59,953 3-months treasury bills 13 13 13 Short-term federal bonds 92,720 105,340 Short-term federal bonds for replenishment 279,600 247,020 Unfunded debt 252,892 311,783 387,889 Defense savings certificates 105,163 136,568 162,619 Specia savings certificates 88,508 102,144 116,123 Regular income certificates 14.119 30,592 72,860 Specia saving accoLnts 20,984 24,834 26,494 Other 3,806 2,920 2,820 Savings accounts 20,312 14,724 6,972 Total debt 908,895 1,041,899 1,151,445 Source. State Bank of Pak stan (1997/98), annex table 7.5. 171 JOHN LEONARDO Wholesale government debt instruments comprise treasury bills and longer-term federal investment bonds (FIBs). The issuance of short-term federal bonds (STFBs) was phased out during 1998/99. The former two types of securities are purchased by financial insti- tutions for investment and/or liquidity ratio purposes. Short-term instruments are purchased either through auctions conducted by the SBP (the central bank) or in the secondary market. Long-term debt instruments, FIBs, are sold at par, with a fixed yield, rather than via auction in a manner that allows bids at discounts and pre- miums to par. Banks can sell these securities in the secondary market to in- crease their cash positions. The use of repurchase agreements ("repos") is also well established, by which banks can obtain liquid- ity, and a ready market exists for them. In addition, the 46 licensed banks, designated "scheduled banks," are able to reliquify by enter- ing into three-day repo transactions with the SBP for both STFBs and FIBs. The SBP provides this repo facility only to its 46 sched- uled banks. All other financial institutions are excluded from the SBP repo window. As from November 15, 1999, non-banking fi- nancial institutions were allowed to enter into repo transactions with licensed banks. Debt instruments designed for the retail market comprise a va- riety of government savings certificates. Although these securities cannot be traded in the secondary market, an early redemption fa- cility is provided, but with attached penalties. Wholesale non-bank- ing financial institutional investors, such as large insurance companies and provident funds, are also able to invest in these retail instru- ments; this position has prevailed for some time. The distribution of domestic debt as of June 30, 1998, based on the figures presented in table 4, is shown in table 5. It is divided into the three types of domestic debt on issue in Pakistan: (1) "per- manent debt,"principally longer-term FIBs and prize bonds;4 (2) "Prize Bonds" pay a nominal rate of interest, but holders are eligible for cash prized awarded after drawings. New Zealand has a similar product called "Bonus Bonds," which were once offered by the Post Office Savings Bank and are now continued by ANZ Bank. 172 PAKISTAN SURVEY Table 5. Distribution of Pakistan Government Domestic Debt, 1996-98 (percent) June 30, 1996 June 30, 1997 June 30, 1998 (provisiollal) Permanent debt 32. 42 28. 44 25 16 Floatingdebt 39.75 41. 64 41.15 Unfunded debt 27. 83 29. 92 33. 69 Total debt 100. 00 100. 00 100. 00 "floating debt," comprising treasury bills and STFBs; and (3) "un- funded debt, " principally government savings certificates. This latter category is so called because neither this source of funds nor its repayment appears in the budget. The data in tables 4 and 5 indicate three significant trends in recent government debt raising during 1996/98: (1) reduced raising of long-term debt from wholesale investors; (2) increased use of STFBs funded by wholesale investors; and (3) increased use of un- funded debt from retail and wholesale investors. The interest rates set recently for the retail-designated govern- ment savings certificates are significantly higher (as much as 300 basis points, that is, three percentage points) than those prevailing in the wholesale markets. As a result, non-bank wholesale institu- tional investors are joining individuals in investing in this market. The government seems to have moved away from lower-cost bank deposits to more expensive "non-funded" financing. The SBP's current approach to the wholesale section of the gov- ernment securities market namely, reduced reliance on wholesale securities and greater reliance on retail securities that have higher coupon rates -may well be creating unnecessarily higher interest rates for the country than might otherwise prevail. Consequently, additional debt costs are being borne by the government and (as a result) by private sector borrowers. This is not only more expen- sive, but it lacks transparency, as unfunded debt acts to hide gov- ernment borrowing from the budget. Action is required to foster use of the wholesale section of the government securities market. There must be a reasonable supply of securities to meet the needs of whole- 175 JOHN LEONARDO sale investors in government securities and a pricing mechanism to keep them out of the unfunded (retail) sector. Wholesale investors should be participating fully in the wholesale section of the govern- ment securities market and not diverting part or all of their demand for securities into the retail section of this market. To date, the SBP has not established a primary dealer structure for transactions in government securities. Trends in Isdue of Government Securities. Excluding securities held by the SBP itself, the majority of treasury bills are held by the 46 scheduled banks. No detailed breakdown of the distribution of own- ership of treasury bills and FIBs is readily available. However, the value of outstanding STFBs and treasury bills held by market par- ticipants is known to have increased from Rs 92.7 billion in 1997 to Rs 130.8 billion in 1998, which is equivalent to an increase of 41%5. By the end of April 1999, however, it had reached only Rs 131.4 billion, an increase of less than 0.5%. In July 1998, the State Bank of Pakistan reintroduced 3- and 12-month treasury bills and ceased offering STFBs. As a result, the composition of short-term government securities on issue has changed considerably during 1998/99, as investors have taken up these new short-term instruments. Details of outstanding Pakistan government short-term securities as of April 30, 1999, are presented in table 6. Market participants welcomed the change from STFBs to trea- sury bills because STFBs were issued at par and carried varied in- terest rates. As a result, each issue had a limited volume, which reduced their appeal in the secondary market. Treasury bills, by contrast, are issued at a discount and in larger tranches. This increases their appeal to institutional investors. Federal investment bonds are longer-term instruments with maturities of 1, 2, 3, 4, 5, and l0years. These are fixed-return bonds sold at par, which cannot be bid for on an auction basis at a dis- count. The majority are held by the 46 scheduled banks and the balance by institutional investors such as insurance companies and provident funds. The total amount of outstanding FIBs has been State Bank of Pakistan (1997/98), annex table 7.5. 174 PAKISTAN SURVEY Table 6. Pakistan Government Short-Term Securities, April 30, 1999 (millions of rupees) 3-mih 6-mtb 12-mth STFBs Total T- bills T-bills T-bills June 30,1997 92,720 92,720 June 30,1998 25.453 105,340 130,793 April 30,1999 6,699 50,106 74,629 0 131,434 Sources State Bankof Pakistan (1997/98), annextable 7 5, State Bankot Pakistan (1999), table4. 6. declining since 1995/96, as old bonds mature faster than new ones are issued. Table 7 summarizes changes in the value of outstanding FIBs during 1997/98. This decline from Rs 154 billion to Rs. 144 billion represents a 6.7% annual decrease in FIBs outstanding. The decline continues. The stock on issue amounted to Rs 141.5 billion at the end of April 1999, a further decline of 1.4% over 10 months. During 1998/99, the State Bank of Pakistan was unwilling to accept bids for FIBs in auctions at yields similar to those being of- fered on national savings certificates (about 19%), a practice that has impeded development of long-term markets for government and corporate debt securities. Few new government securities that can be traded on the secondary market are being issued for maturities greater than one year. In recent weeks, no FIBs whatsoever have been sold by the SBP, and this has caused a significant reduction in the supply of securities for trading in the secondary market. No detailed breakdown of the distribution of ownership of na- tional savings certificates is readily available. However, anecdotal evidence suggests that provident funds and insurance companies have material holdings of these higher-yielding securities, at the expense of FIBs. Table 7. Activity in Federal Investment Bonds, 1997-98 (billions of rupees) Outstanding, New issues, Maturities, Outstanding, June 30, 1997 1997/98 1997-98 June 30, 1998 FIBs held by market 153.8 3. 7 13.9 143.5 participants Source State Bank ot Pakistan (1997/98), p. 68 175 JOHN LEONARDO In 1997/98, the government was unable to raise funds through other long-term debt instruments such as foreign exchange bearer certificates and foreign currency bearer certificates. It raised some additional overseas debt in fiscal 1998/99 using the latter instruments. Secondary Market Turnover According to market participants, the annual value of turnover in government securities in the secondary market is approximately Rs 2,500 billion6. Average daily volumes of repo transactions are Rs 9.5 billion, while average daily outright volumes (that is, transactions in physical securities) are Rs 0.5 billion. Trading is undertaken through the 12 licensed money market brokers authorized to trade in the interbank money market. There is no vigorous secondary market in physical securities. Because of this and long-term investors' limited participation in government securities, secondary market prices of FIBs are frequently discounted during periods of tight liquidity in the banking sector. Transactions in government securities undertaken by wholesale investors seem to be driven largely by liquidity rather than portfolio management requirements. The SBP receives, but does not distrib- ute, any information concerning secondary market government secu- rities trading. On balance, market participants do not believe that these practices materially reduce transparency in the government se- curities market. Short selling is not allowed. The SBP is not likely to consider allowing short selling without considerable changes in the government securities market, such as a more effective market in long- term securities and increased trading volumes across all maturities. The Yield Curve Current secondary market yields for government securities exhibit a normal (although slightly hooked) positively sloped yield curve This estimate is based on estimated average daily trading volume of Rs 10 billion over 250 trading days peryear, inclusive of repo transactions. This amounts to approximately nine times the combined average value of FIBs, treasury bills, and STFBs on issue in 1997/98 (Rs 280,430 million). 176; PAKISTAN SURVEY Table 8. Government Securities Yield Curve, July 26, 1999 (percent) Tenor Treasury bills Repos FIBs Government and STFBs savings certificates 1-month 7. 00 6. 35 3-month 6. 90 6 60 5-month 7. 25 6-month 7 15 1-year 9.10 12. 00 2-year 9,10 14. 02 3-year 13. 00 14. 47 4-year 13. 00 14. 52 5-year 13.90 14. 87 10-year 14. 25 15. 97 Source: Khadim Al Shah Bukhari & Co. (see table 8 and figure 1). However, present policy is not generating a well-developed yield curve. At the short-term end of the market, the SBP's decision last year to cancel new issues of STFBs was helpful in the context of developing the market in short-term government securities. This market seems well supported. Central Market Infrastructure At present, Pakistan has no on-line system of trading, clearing, settle- ment, and registration for government securities. As noted, trading is undertaken through the 12 licensed money market brokers autho- rized to trade in the interbank money market. Indicative bid/offer rates for repo transactions are published by all scheduled banks on Figure 1. Pakistan Government Securities Yield Curve, July 26, 1999 108.00% 16.00% - _______________________________ 14.00% - - > 10.00% -he 40.00% - 6.00% 4.00% -___________________________________ 2.00%i 0,00%1 1 1 1 1 3 6 1 2 3 4 5 10 Month Months Months Year Year; Years Years Years Years Maturity 177 JOHN LEONARDO their Reuters pages. However, no indicative bid/offer rates for other securities are published. The repo rates are used as a basis for out- right quotes on physical securities. All transactions in treasury bills and FIBs are settled through manual book-entry transfers submitted to the SBP's Public Debt Office. Transactions and associated transfers are not settled simul- taneously. The SBP maintains the register of government securities using a manual system known as the "Securities General Ledger" (SGL) account, a cumbersome system that is unlikely to be comput- erized for four to five years. Likewise, the National Savings Organi- zation, part of the Ministry of Finance (MOF), uses a manual system to administer the government savings schemes. These systems do not have the capacity to readily produce data on the distribution of ownership of government savings certificates. Market Participants Market participants comprise intermediaries, issuers, and investors. Intermediariev. There are no true market-makers in the wholesale government securities market. However, the current development of the secondary market in government securities is constrained more by the limited tenors on offer and limited position-taking by partici- pants than by the absence of market-makers. Capital is not a major constraint for intermediaries at present because trading volumes are limited and interest rates are falling. If this situation were to change, capital constraints might well arise. IJsaerJ. As described earlier, the government issues permanent debt (principally FIBs), floating debt (treasury bills), and unfunded debt (national savings certificates). The values of outstanding debt for these different types of instruments during 1996/98 period are pre- sented in table 4. Investors. Scheduled banks hold the majority of treasury bills and FIBs. Although ownership data are not available for national sav- ings certificates, the distribution of ownership is wider than for trea- sury bills and FIBs and cuts across all investor categories. Some 178 PAKISTAN SURVEY non-banking institutions have major holdings. Life insurance com- panies and provident funds, which hold long-term positions in gov- ernment securities, are not active secondary market participants in part because trading expertise is limited, and in some cases, such as State Life, because the volume of government securities available in the secondary market is also limited. For some years now, the banking sector has been faced with high levels of non-performing assets. To remedy the situation, the SBP has been promoting new prudential guidelines and regulations and facilitating changes in the management and ownership of state- owned banks. As a result, the banking sector is beginning to show signs of improvement, which should eventually be reflected in in- creased secondary market trading in government securities. CORPORATE BOND MARKETS The issuing and trading of corporate debt securities is not a new phenomenon in Pakistan. During the 1960s and early 1970s, be- fore the nationalization of financial institutions in 1 972, corporate debentures issued by Pakistani companies were listed on one or more of Pakistan's three stock exchanges (Islamabad, Karachi, and Lahore), and there was some secondary market trading in these securities. The government is conscious of the importance of developing the term debt market and has pursued a series of initiatives since 1995 of assistance to market development. Most notably, it has in- troduced moderating policies on stamp duties, the taxation of in- vestment income for individuals from investment in term finance certificates (TFCs) (income received by individuals from TFC in- vestments is now exempt from taxation), and the withholding tax. The government also established SECP, which has taken over the role of the Corporate Law Authority. The corporate debt market is expected to benefit from the Capi- tal Market Development Program (CMDP) funded by the Asian Development Bank (ADB) and being implemented under a pro- gram loan. Initial planned outputs of this program include the es- tablishment of a national clearing and settlement system and an 179 JOHN LEONARDO over-the-counter debt market. A second phase, added recently, will focus on developing the domestic bond market and includes devel- oping a long-term yield curve in government securities by encour- aging greater investment by wholesale investors in long-term government securities. It will address some of the key problems in the government securities market discussed earlier (see the appen- dix to this chapter for a commentary on the CMDP published in the SBP's 1997/98 Annual Report). Detailed plans for the second phase of the CMBP were to be released in early 2000. In addition, major market participants have begun a healthy dia- logue on policy matters affecting the development of the corporate debt market. This interaction has been given a helping hand by the SECP's new policy board, which includes five members from the private sector. Private sector participants are willingly sharing their knowledge with the SECP, MOF, and stock exchanges. Various market participants have suggested the SECP would benefit from technical support for the development of policy on particular debt instruments. Market Size and Structure: Term Finance Certificates Currently, major private sector Pakistani companies issue term fi- nance certificates when making public debt issues. This instrument arose from legislation enacted in 1984, which authorized the issue of redeemable capital securities. As a debt instrument, the TFC is slightly different from the traditional corporate bond because it was specifi- cally created to accord with the Sharia principles of Islamic law. The key difference is that the TFC substitutes the words "expected profit rate" for "interest rate." No TFCs were issued by private sector Pakistani companies in the 1984-1995 period. However, some were issued in this period by major government-owned corporations, most notably, the Water and Power Development Authority (WAPDA), the national power util- ity company. From 1988 to 1994, WAPDA issued Rs 22.25 billion of bonds to the public; the retail component appears to have been greater than that recorded in recent TFC issues. The total amount of these issues is more than six times the current value of TFCs raised by Pakistani companies in public issues since 1995. 180 PAKISTAN SURVEY The market's experience with WAPDA bonds proved unsatis- factory in two respects. The secondary market for WAPDA bonds was not sustained owing to the market maker's undercapitalization. As a result, liquidity gradually declined, and many retail investors became dissatisfied with their WAPDA investments. Second, WAPDA was at one point unable to repay some maturing bonds: investors had no option but to accept repayment at a later date. These experi- ences created lasting negative feelings about the bond market. Idsuance. Pakistani private sector companies have issued TFCs to the public only since 1995. To be an issuer, a private sector company must obtain the prior consent of SECP, which regulates Pakistan's securities markets. To gain the SECP's consent, the issuing com- pany must ensure that the TFC issue meets the following conditions: * The issue must have been given a credit rating by one of the two approved credit-rating agencies. These two rating agen- cies have associations with major international agencies.7 . The company must apply for listing on one of the three stock exchanges. • The issue must be underwritten. Underwriters to TFC issues have generally been investment banks and the two major broking firms dealing in debt securities. Of the eight recent TFC issues, two were undersubscribed (the shortfalls are not known). The other six issues were oversubscribed. Market- making arrangements are disclosed in the prospectuses. No market-making arrangements were made in the most recent issue (Dewan Salman Fibre Limited); according to the pro- spectus, "in view of the offer to the general public, there will be adequate liquidity in the market". _ A trustee must be appointed to look after the interests of in- vestors. To date, trustees have generally been investment banks (which have also participated in issues as underwriters or in- vestors). Trustees' annual fees are in the range of 0.7% to 0. 1% 7The Pakistan Credit Rating Agency (Pvt) Limited (PACRA) is associated with Fitch-IBCA Inc., and DCR-VIS Credit Rating Co. Limited is associated with Duff & Phelps Credit Rating Co. 1SI JOHN LEONARDO of the amount raised. The trustees do not appear to have the power to take independent action to look after the interests of investors; rather, they act only after the receipt of instructions from investors. * The proposed security for the issue must be specified. This is often a charge over specific property. * A minimum of 25% of the proposed issue will be raised from the public. (The rest is presold in pre-initial public offerings to institutional investors, as noted below.) * A redemption reserve must be established to cover the repay- ment of principal. The SECP's stance on this requirement appears to vary. The SECP agreed to waive the need for a redemption reserve in the Dewan Salman Fibre Limited pro- spectus, whereas there was no specific reference to this mat- ter in the Gatron (Industries) Limited prospectus. * The SECP's consent is not required if TFCs are issued as a private placement. Structure of TFC Isdued. TFC issues have both private and public components. Before the prospectus is approved by the SECP, the underwriters to a TFC issue must secure firm undertakings from wholesale investors to subscribe to the issue. The wholesale inves- tors are named in the prospectus and complete their subscriptions prior to the opening of the public component of the issue. This pro- cess is known as "Pre-IPO," meaning "pre-initial public offering." The profit (interest) rates set in the most recent TFC issue, that of Dewan Salman Fibre Limited in May 1999, are shown in table 9. The Dewan Salman Fibre Limited TFC issue carried a 19% ex- pected profit (coupon) rate, payable semiannually in arrears. It was issued in denominations of Rs 5,000. Nominal principal repayments are planned over the first three years. The residual principal will be repaid over the final two years of the issue compared with the tradi- tional practice of a bullet repayment at maturity. The zakat (wealth tax) is payable at the rate of 2. 5% of the value of every principal repay- ment. The tenor and repayment schedule is typical of the other issues. Actual TEC Idsueu. Because of the high interest rates paid on gov- ernment savings certif icates, TFC issuers have had to pay very high 182 PAKISTAN SURVEY Table 9. Dewan Salman Fibre Limited TFC Structure and Redemption Schedule, May 1999 Tlnor Ouludl4Q Pnneipal Frotit Gloss Withbolti& Redemplion Weath Rekmpfto Net (months) pxtncltiat raSeiptina re4.mtitin tax (tes wtitax) (Zwat) (ls zakt) tedmption 0 5000. 00 1.00 6 4999 00 1.00 475.00 476 O0 47 5 428.50 0. 03 475. 98 428 48 12 4998.00 1.00 474 91 475 91 47.49 428. 41 0. 03 475. 88 428. 39 18 4997 00 1.00 474 81 475 81 47 48 428.33 0.03 475 79 428 30 24 4996 00 1. 00 474. 72 475.72 47 47 428.24 0. 03 475.69 428 22 30 4995.03 1.00 474 62 475.62 47.46 428.16 0 03 475 60 428.13 36 4994.00 1.00 474. 53 475. 53 47 45 428.07 003 475 50 428.05 42 3745.50 1248 5 474.43 1722. 9 47 44 1675 49 31.21 1691.72 1644 27 48 2497.00 1248.5 355 82 1604. 32 35.58 1568 74 31.21 1573.11 1537.53 54 1248 50 1248.5 237 22 1485.72 23.72 1461 99 31.21 1454.50 1430.78 60 0 1248.5 118.61 1367.11 11 86 1355.25 31.21 1335.90 1324.03 Totals 5000.00 4034 67 9034.67 Note. Table s presented in the manner disclosed in the Dewan Salman Fitre Limited prospectus The "Net redemptiol' igure is denved by subtracting "withholding tax" and "wealth tax' trom the "gross redemption' tigure ''profit" rates in recent TFC issues, which has limited the number of new TFCs issued. Nevertheless, some companies have proceeded with TFC issues in order to develop a presence in the corporate debt market. The eight public TFC issues represent only two a year over the last four years (see table 10). Less than Rs 3.5 billion in such debt has been issued since 1995 (excluding oversubscriptions) with maturities of three or five years, and less than Rs 2.4 billion is now outstanding. These figures are relatively modest in relation to the estimated current value of outstanding government securities (Rs 270 billion, comprising treasury bills and federal investment bonds) and with respect to the overall size of Pakistan's economy. As table 10 demonstrates, all TFC issues to date have had fa- vorable credit ratings. Currently, it is not possible for companies without investment-grade credit ratings to gain SECP approval for issuing TFCs; the demand for such paper would be modest at present. Market participants, however, argue that SECP regula- tions should be modified to provide for such issues. Secondary Market Trading Volutmed for Corporate Debt. At present, the estimated monthly volumes of corporate debt securities traded on the secondary market are low (see table 11, which includes repo transactions). To date, only a small proportion of issued TFC stock 183 JOHN LEONARDO Table 10. TFC Issues in Pakistan, January 1995 to July 1999 1"er Piekaps SSGC NWlut to atoln biterf-bak SaVdl f3N Total TEK Pak SaUlm Dateof Feb. 1995 Oct.1995 Jan. 1996 Sep.1996 Jun.1998 Dec. 1998 Feb. 1999 May 1999 subscription Rating A+ AA A+ AA A+ A A+ A+ Coupon rate(%) 18 50 18.25 18.00 18 70 18.00 17 50 18.50 19 00 Total TFC 210 500 250 1,000 250 300 250 700 3.460 issue (Rs m) Pre lPOI 110 400 175 750 200 0 200 500 2,335 Public offering 100 100 75 250 50 300 50 200 1,125 Oversubscription 22 Under Under 70 20 26 35 176 (Rs rn) Subscribed Subscribed Outstanding 77 499 0 999 250 325 237 876 2,387 (Rs m) Tenor (years) 5 5 3 5 5 5 3 5 Grace period (years) 2 1 3 2 3 1 0 3 Redemption 3 anneal 4 annual Bu let 3 annual 4 half- 4 annual 6 half- 4 half- repayment yearly yearly yearly Effectivecost 18 90 18.65 18 93 19.30 18. 47 17.77 19.12 19 49 (% per annum)7 I IPO represents the amount of TFCs placed "firm" prior to the public issue proceeding. I Effective cost represents estimated internal rate of return based on the value of gross redemptions Source: First International Investment Bank Limited has been traded on the secondary market. Companies that have is- sued TFCs indicate that less than 15% of TFCs issued have been traded on the secondary market. Current market yields for TFCs do not bear a strong relation- ship to prevailing credit ratings. Changes in market yields are influ- Table 11. Turnover of Corporate Debt Securities in the Secondary Market (monthly averages) Year Turnover (millions of rupees) 1996 10-20 1997 20-30 1998 10-15 1999 (June half-year) 40-50 Source. Khadim Ali Shah Bokhari & Co. 0184 PAKISTAN SURVEY enced more by changes in market liquidity than by the associated credit risk. This indicates a thin and unsophisticated market. The yield curve in private sector corporate bonds is not well developed. Details of market yields for some leading TFC issues as at mid-Sep- tember 1999 are presented in figure 2; the market yields shown in figure 2 are less than those presented in table 10 owing to the gen- eral easing in market yields recorded during the 1999 September- half year. To date, the absence of market makers in corporate debt securi- ties has not significantly constrained secondary market activity as daily volumes remain relatively low and stock on offer is sold with- out undue difficulty. But from a longer-term perspective, increased market-making capability is essential. Until the SECP allows short selling, that capability will remain limited, and until a market maker is able to be in the market as a seller at all times as well as a buyer, there can be no real secondary market. It is not realistic to expect market makers to play the role of sellers if they are compelled to maintain an inventory of physical securities to meet any future de- mand for purchases. Of course, short selling carries potential finan- cial risks to the seller, so it may be helpful to ensure some technical assistance to would-be market makers before opening the market to short selling. The capital position of major market participants is comfortable and capital shortages do not appear to be materially limiting sec- ondary market development at this point. Figure 2. Pakistan FC Yield Curve, Mid-September 1999 18.2 18 - 17.8 - . 17.6 X 17.4 17.2 17- 16.8 16.6 - I 1 Yea 2 Years 3 Years 3.5 Years 4 5 Years (PA) (AA-) (AA) (AA-) (MA+) Maturity -185 JOHN LEONARDO Central Market Infrastructure Some corporate debt is traded on-line on the stock exchanges. How- ever, there is no publication of indicative bid/offer rates for corpo- rate debt securities on the Reuters' pages of market participants. Transactions in listed securities, done by members of the three stock exchanges, are settled through the settlement systems of the exchanges. Other corporate debt traded over the counter is settled privately, through physical delivery in exchange for checks. E'akistan's deposi- tory, the Central Depository Company of Pakistan Limited, was es- tablished in 1993 to provide efficient delivery, settlement, and transfer of securities transactions through a computerized book-entry sys- tem. The company has declared all listed equity securities eligible for its depository system. About 95% of the traded value of equity securities are registered in dematerialized ("demat," that is, electronic book-entry) form. In August 1999, the Central Depository Company began offering individual accounts. The depository company has set reduced fees for TFCs to encourage registration in demat form. The last two TFCs issued in Pakistan were eligible depository securities. Currently, 34% of Dewan Salman Fibre Limited TFCs are held in demat form. The Central Depository Company and the three Paki- stani stock exchanges are discussing the modalities of a centralized clearing and settlement system for listed debt and equity securities. Pakistan has two credit-rating agencies: the Pakistan Credit Rating Agency (PACRA) and DCR-VIS. PACRA was established in 1994 and is the largest operator; its major shareholder is Fitch lBCA, while IFC and the Lahore Stock Exchange are minor share- holders. DCR-VIS, is associated with Duff & Phelps. Market Participants Again, market participants are divided into issuers, investors, and intermediaries. Idsaer4. As noted earlier, since 1995 most of the TFC issuers have been private sector firms. Since 1984, public sector bond issues have amounted to more than Rs 30 billion, while private sector issues from January 1995 to July 1999 came to Rs 3.8 billion. 18 PAKISTAN SURVEY Several trends are pushing for more TFC issues. First, many major scheduled banks are close to, or at, their peak lending expo- sure to some of the largest and strongest public companies, which means companies with excellent credit ratings. These companies will most likely revert to TFCs for part of their term-funding require- ments, when their demand for term funding increases. Strong issu- ers such as these will be a welcome addition to the public debt market and will increase investor interest and participation in the market. Second, companies that previously relied heavily on develop- ment finance institutions will have to utilize some TFC funding in the future. In recent years, development finance institutions have provided a significant proportion of term funding for leading Paki- stani companies. These institutions in turn received considerable funding assistance from the multilateral development agencies. Lately, however, multilateral agencies have ceased funding the Pakistan development finance institutions, particularly for leasing companies. This has adversely affected the latter group's lending and invest- ment capability. Consequently, some major Pakistani corporations, particularly leasing companies, may be forced to rely on TFCs in the future for additional term debt funds. Any such increase in new TFC issues will assist in the development of the corporate bond market. Third, privatization of state-owned enterprises, which is likely to increase in the medium term, may force newly privatized SOEs to the corporate debt markets, to substitute for reduced government funding. Fourth, some SOEs may borrow in the public markets; the Pakistan government has already announced (on August 3, 1999) plans to issue TFCs to fund the prime minister's housing scheme. Fourth, using another avenue for new issues, government-owned corporations have made two securitizations (none yet by private firms). In addition, new rule changes made by the SECP policy board in November 1999, to approve the draft Companies (Asset-Backed Securitization) Rules, 1999, should encourage increased securitization in the near term. Under these rules, special-purpose vehicles regis- tered with the SECP will be able to mobilize new sources of funds by issuing bonds or TFCs against receivables. Finally, the stamp duty rate on newly issued securities was re- duced from 4.5% to 0.1% and set at 0.1% for subsequent transfers. 187 JOHN LEONARDO This reduces the cost of secondary market trading, which might in- crease trading and consequently make new issues more attractive. Investor4. Commercial and investment banks are the main investors in private sector TFCs, with the 46 scheduled banks being major players. According to information supplied by First International Investment Bank Limited, scheduled banks have invested Rs 1,128, 5 million in the IPOs of the TFC issues made since 1995. As of May 1999, the scheduled banks' investments in TFCs represented an es- timated 48.33% of IPOs subscribed (Rs 2,335 million) and 32.62% of total subscriptions to TFC issues (Rs 3,460 million). Life compa- nies and provident funds hold about 5% of total issues, while retail investment is negligible. Although no data were available, anecdotal evidence suggests that retail investors have shown interest in TFCs issued by public sector entities. In 1997 listed TFCs became "approved securities" for the pur- pose of meeting statutory liquidity requirements (SLRs) for non- bank financial institutions (NBFIs). Some market participants believe that TFCs should be considered eligible investments for meeting the SLR of scheduled banks as well. They believe this could promote additional demand for TFC paper. (Note that allowing corporate obligations for statutory liquidity purposes would be contrary to generally accepted central banking practice elsewhere in the world and seems unwise). Even if the latter position materialized, the rela- tive lower liquidity of TFCs compared with T-bills and FIBs is likely to remain a major influence on the investment decisions of sched- uled banks with respect to meeting their statutory ratio requirements. Recently established rules define approved investments for provi- dent funds and allow these investors to invest in TFCs. Since 1996, a limit has been placed on the portion of provident funds that can be invested in listed securities. At that time, the SECP introduced the "Employees' Provident Fund (Investment in Listed Securities) Rules, 1996," which apply to investments made by provident funds in listed securities. Under these rules, a provident fund's total investment in listed securities is not allowed to exceed a specified percentage of the total value of the fund. The ceiling for total investments in listed securities was increased from 10% to 20% in June 1997 and increased again to 30% in November 1998. Investment in debt securities of 188 PAKISTAN SURVEY listed companies must be restricted to securities that have been rated as an investment grade with a minimum rating of "BBB" by an ap- proved credit-rating agency. By contrast, life insurance companies are currently able to place up to 50% of their investments in non- government securities (shares, property, and debt). Demand from investors is likely to grow for several reasons. If feedback from institutional investors is any indication, the corpo- rate debt market could expand quite quickly in the immediate fu- ture, depending, of course, on the level of interest rates. First, there appears to be some unsatisfied demand for quality TFC paper by institutions such as insurance companies, mutual funds, and provi- dent funds, all of which manage long-term savings. The limited sup- ply of prime corporate debt paper reflects the modest total borrowing requirement by prime corporate borrowers in Pakistan at present, which is partly due to the present soft state of the economy. Second, the proposed restructuring of the insurance sector, con- tinued improvement in the financial position of scheduled banks, and privatization of the three remaining nationalized banks will indirectly assist the secondary market by helping to improve asset manage- ment in the banking and insurance sectors. That, in turn, should in- crease demand for debt instruments, both from IPO and in secondary market trading of corporate (and government) debt instruments. Third, since 1995, the government has sought to promote the corporate debt market by reducing stamp duties and taxation, in- cluding the withholding tax on profit (interest) payments. In the past, a withholding tax was charged on profits from TFCs. The 1999 federal budget extended an exemption on payment of the withhold- ing tax to all persons, including companies. However, erratic and unpredictable imposition of the withholding tax in the past has made some investors uneasy.8 A wealth tax (cczkazt) is payable by individu- als, gratuity funds, and pension funds on repayments of principal and profit from TFCs. However, no zakazt is payable by pension funds 'Withholding tax on TFC profits was removed in 1995 but reinstated in March 1998, owing to the government's urgent need for additional revenue. There was also some concern that private placements at that time were undermining the launch- ing of new public TFC issues, another reason for initially removing the withhold- ing tax in 1995. 189 JOHN LEONARDO on their investment in government savings certificates. This particular policy, together with the attractive returns available from these in- vestments, has made pensions funds less interested in participating in TFCs. Intermediaries. Fewer than 10 organizations may be described as "key intermediaries" in the Pakistani corporate bond market. Of the foreign banks, Bank of America has been very active as an adviser and arranger of TFC issues but does not act as an underwriter; Citicorp Investment Bank has also been active as an arranger/ad- viser and underwriter. Local investment banks active in TFC issues have included First International Investment Bank Limited and Orix Investment Bank. Of the local scheduled banks, Muslim Commer- cial Bank Limited has been the most active in TFC issues. Three securities firms have also been very active in TFC issues: Jahangir Siddiqui & Co., Khadim All Shah Bukhari & Company Limited, and UBS Securities (this firm was the lead arranger for a recent issue by a major government-owned utility). In addition to registering and supervising public issues of cor- porate debt securities, the SECP is endeavoring to upgrade the re- sponsibilities of trustees in TFC issues. Currently, the role of trustees is covered by the antiquated 1882 Trustee Act. The SECP must be proactive in monitoring the performance of trustees because one incident has already occurred in which the trustees in a TFC issue appear to have carried out their obligations to investors inefficiently. The SECP's role in such matters is of concern to members of the stock exchanges, especially in view of the past problems with WAPDA bond issues discussed earlier. IMPEDIMENTS TO BOND MARKET DEVELOPMENT Impediments to development of the government securities and cor- porate debt markets in Pakistan fall into three categories: those op- erating around the bond market, in the broader political, macro, and legal environment; those in other parts of the financial system, par- ticularly the banking sector and government securities market; and those "inside" the market itself, arising from regulatory matters, central 190 PAKISTAN SURVEY market infrastructure, and market participants. These are deemed to be influences outside the capital markets, but include forces within the country. Around the Bond Market Significant impediments to market development are posed by exter- nal influences, particularly political, macroeconomic, and legal fac- tors. Weaknesses in the broader financial system produce additional constraints. Political Impediments. Pakistan faces several problems on the politi- cal side. First, notwithstanding the financial sector reforms imple- mented progressively by the government since 1991, debt market participants still consider the level of political risk high. Material variations in past economic policy suggest that recent positive ini- tiatives may be difficult to sustain. The withholding tax on TFC profit payments reintroduced in 1998, for example, had to be removed in the following year. In addition, inconsistencies in the government's approach to eco- nomic management since 1991 frequently reflect difficulties experi- enced in meeting planned increased revenue requirements. Such inconsistencies create an impression of high political risk and make retail and wholesale investors hesitant about investing in long-term government securities and corporate bonds. In the past, inconsis- tencies in the taxation of investment income across different debt instruments have also made corporate bonds less attractive to inves- tors. Some government economic decisions reinforce concerns about political risk as they appear to compromise other major economic policy objectives. The 1999 budget decision to tax the undistributed profits of companies, for example, may reduce the reinvestment of tax paid profits by Pakistan-based multinationals and cause their interest in building strong local balance sheets to decline. It will also weaken the credit strength of local companies, making investment in their securities a riskier proposition for investors. Traditionally, policy implementation has been one of the government's weak suits, as can be seen in the privatization pro- 191 JOHN LEONARDO grams. Planned privatizations, including those of three banks (Al- lied Bank, Habib Bank, and United Bank) and two unit trusts (In- vestment Corporation of Pakistan and National Investment Trust) are behind schedule which means that recovery of the banking sec- tor and improvements in asset management are moving slowly. Prompt and successful privatizations are urgently required to demonstrate the associated benefits to the community. Macroeconomic Impediment4. Among the serious macroeconomic im- pediments to a corporate bond market is the crowding out caused by high government spending and low revenues. The Pakistan govern- ment has a significant fiscal deficit that creates a continuing need to borrow, and the large amount of outstanding government debt dis- courages private sector development by crowding out private bor- rowing. The excessive spending is due in part to the decision to carry and fund an extensive portfolio of poorly performing state-owned corporations and to rising government debt-servicing costs. The level of government revenue is relatively low and reflects real difficulties in generating additional revenue. Low savings rates in the country limit the resources available for investment in corporate debt securities. Capital formation has also been unimpressive, with only limited progress in the attempts to widen the government's revenue base. The large informal sector in Pakistan dilutes government revenue initiatives. Attempts to squeeze the private sector (for example, by taxing unrealized corporate profits) are not likely to be successful.9 BroaderLegalFramework. The low level of governance in Pakistan in both the public and the private sectors is reflected in the general inability to enforce commercial contracts. Without the rule of law, private sector development will proceed at a slow pace. Previous defaults by earlier bond issuers continue to dampen retail investor interest in corporate debt securities. 'Some of these problems are due to continued government involvement in the management of state-owned financial institutions, a legacy of Pakistan's now-dis- carded socialist approach to economic management, which prevailed from 1974 to 1991. ~192 PAKISTAN SURVEY Across the Financial System Looking across the financial system, one finds obstacles to market development both in the banking system and in the government se- curities market. The Banking System. Despite its recent marked recovery, the Paki- stan banking sector, remains vulnerable to unforeseen setbacks. The promised privatization of the three remaining state-owned banks is now scheduled for the first half of 2000. The state-owned develop- ment banks continue to have very weak balance sheets. The Government Securities Market. One major concern here is the lack of depth in the wholesale market for longer-term government securities. To develop the mnarket for long-term instruments, the SBP needs to introduce a new series of longer-term government securities, sold via auctions. These securities should carry cou- pons in line with prevailing trends in government securities mar- kets in the West. At the same time, steps must be taken to move wholesale investors from the high-yielding "unfunded" savings certificates into a more orderly and lower-priced market for FIBs. Savings Certificate yields must be reduced, so as to eliminate the tendency of wholesale investors to move into the retail markets. The government's decision in September 1999 to introduce a 10% withholding tax on income received from national savings schemes was a step in the right direction and has slightly reduced the rela- tive attractiveness of these investments (this decision was reversed recently). In addition, real-time infrastructure will be required in the fu- ture to support increased trading in government securities. Further- more, the SBP should encourage the MOF to place a ceiling on the maximum investment that insurance companies and provident funds can make in government savings certificates to ensure that such wholesale investors participate fully in the wholesale part of the market. Although the scheduled banks generally have a sound un- derstanding of the government securities market, the level of overall knowledge is lower in the other banks, insurance companies, and provident funds, and this needs to be improved. 193 JOHN LEONARDO Priorities in Addressing External Impediments Methods of removing external impediments can be assessed on a scale of high, moderate, and low depending on the degree to which they would help increase the volume of corporate bond trading. As tables 12 and 13 show, external impediments in Pakistan are all con- sidered to be significant, with none in the category of "low priority." Table 12. High-Priority External Impediments PolitiCal Inconsistent economic policies Inconsistent taxation of investment income across different debt instruments Slow progress in implementing planned financial sector privatizations Macroeconomic Crowding out effect on the private sector arising from the large government deficit Higher-cost retail borrowing used to fund the governments deficit Institutional access to retail debt instruments Large informal sector diluting overall effectiveness of economic policies and influence of financial institutions in the secondary debt markets 8roader financial system Wholesale government securities market provides limited market signals, especially for long-dated instruments Weak banking system constrains secondary market bond market trading Legalenvironment Ineffective commercial law Table 13. Moderate-Priority External Impediments Broader financial system Lack of awareness of the role of corporate debt securities in private sector development 194 PAKIS1'AN SURVEY Inside the Market Impediments Some important internal constraints to the development of Paki- stani debt markets are the impact of the supporting regulatory frame- works on market participants, the weaknesses in some sections of the supporting market infrastructure, and the overall level of knowl- edge of participants. Rule, Regulations, andRegulators. On occasion, the regulatory bodies (SBP and SECP) do not seem to have sufficient technical knowl- edge to understand or deal effectively with more complex issues and policies relating to debt market development. In the view of market participants, the SECP also needs to be more vigilant in monitoring all elements of the debt securities markets. Some recent short-term problems with proposed variations to the trust deed of an existing TFC bond issue were attributed to limited action by the SECP's predecessor. Another concern is that the SECP has made frequent changes to new policies and has thereby eroded market confidence in the commission. A shortage of resources has also hindered the SECP's work on some occasions. In regard to regulations, trustees of corporate debt issues need clear and detailed guidelines to help them avoid the kind of poor performance that occurred in a recent case involving a TFC issue. Rules should also be in place to ensure the operational independence of credit-rating agencies. The SECP in conjunction with the stock exchanges needs to reduce the time it takes to approve draft prospec- tuses for TFC issues. (This is a problem for the stock exchanges too.) According to market participants, several market practices need to be clarified or subjected to rules. In particular, they are calling for a standardized repo agreement and methodology for calculating accrued interest when determining debt security transaction settlement prices. Market Infrastructure. Live trading of corporate debt securities is limited to trading between members of the stock exchange. There is no central clearinghouse -no principal entity for the trading of cor- porate bonds and subsequent settlement of transactions. And mar- ket information is limited in its distribution. Details of non-broker secondary market transactions are not reported to the market. 195 JOHN LEONARDO Market Participantd. Issuers are deterred by four main problems. First, issuers and would-be issuers report that it takes far too long to secure SECP and stock exchange approval of prospectuses, and that this hinders planning. Second, issuing a TFC can be costly, particu- larly because of bank handling fees and stock exchange listing fees. Third, not enough real market makers are available to support an issue. Fourth, non-investment-grade companies have no access to the market, because it is still at a low level of development and the regulatory environment is not conducive to entry. Investors are also deterred by several factors. First, because of the weak governance system, they lack confidence in the market. Many are also concerned about whether corporate borrowers will repay loans according to agreed terms, especially after the WAPDA experience. Although not strictly related to the corporate bond mar- ket, the 1998 liquidity crisis experienced by the National Invest- ment Trust (a government-owned mutual fund invested heavily in equities), which culminated in the waiving of redemptions, also un- dermined overall investor confidence in non-equity securities, in- cluding corporate debt securities. Furthermore, the supply of issues is limited. Major institutional investors would welcome additional investment opportunities in TFC issues, subject to acceptable market profit (interest) rates, and could accommodate an increase in supply. But with the limited supply of TFCs, the relationship between credit quality and market pricing is often weak. Price is often influenced more by the short-term liquid- ity needs of investors than the financial risk of underlying securities. Long-term investors such as State Life and provident funds are not playing a major stabilizing influence in the secondary market be- cause trading volumes are so low. In addition, many investors, espe- cially retail investors, do not fully understand debt securities. The investing culture needs to change before the appetite of retail inves- tors is likely to expand beyond equities and government paper. Intermediaries, too, have difficulty participating in the market, first, because of the lack of short selling. Market makers are there- fore unable to provide two-way quotes to the secondary market or to develop a strong and deep secondary market. Approval of short selling for TFCs is unlikely for some time because of the limited volumes issued. Second, secondary markets are thin. To date, only 196 PAKISTAN SURVEY relatively small proportions of recent TFC issues have been traded on the secondary market, although brokers have had little difficulty moving the volumes that have been offered. This is the proverbial chicken-egg problem. Prioritied in Addressing Internal Impediments. As in the case of ex- ternal impediments, internal impediments were assessed on the ba- sis of which ones, if removed, would bring material increases in corporate bond trading volumes on a scale of high, moderate, and low levels of priority. The results of the assessment are presented in tables 14 and 15. No major identified internal impediments were considered to be of "low priority." RECOMMENDATIONS FOR REMOVING IMPEDIMENTS This section outlines specific recommendations for removing key impediments to the development of debt markets in Pakistan. External Impediments The recommendations for removing external impediments apply around and across the financial system. Around the Financial System. Given the recent change in govern- ment, political risk is clearly concern for investors. This risk will only be reduced through consistent and prudent economic manage- ment by the Pakistan government over the medium to long term. The government should make sure that it adheres to planned eco- nomic policies. Specific policy recommendations in this area have already been submitted by the multilateral agencies. A key macroeconomic issue is the size of Pakistan's fiscal deficit and overall government indebtedness. Multilateral agencies have put forth numerous recommendations for reducing the fiscal deficit and indebtedness. Briefly, they call on the government to reduce spend- ing, increase revenue, privatize many government-owned businesses, and reduce the size of the relatively large informal sector. The gov- ernment should also change its present approach to funding the fis- 197 JOHN LEONARDO Table 14. High-Priority Internal Impediments Rules, regulations, andregulators: Limited specialist knowledge at SBP and SPEC about certain debt instruments SEGP's limited financial resources, therefore limited ability to supervise the securities markets Deficiencies in the processes and policies employed by SBP and MOF for issuing wholesale and retail government securities Absence of any SECP guidelines for trustees and credit-rating agencies Lack of standard practices for issuing TFCsa Market participants: * Issuers Lack of a range of debt instruments Lack of knowledge of many debt instruments * Investors Ineffective management of some long term savings portfolios Limited supply Modest knowledge of debt markets (both retail and wholesale investors) Nonbank wholesale investors are relatively inactive in the secondary markets Limited educational opportunities for investors to learn about the debt markets * Intermediaries Delays in approval of draft TFC prospectuses issues by the stock exchanges and SECP High issue costs Absence of market makers Intermediaries and institutions are not allowed to short See the text for TFC issuer concerns. cal deficit by introducing new forms of long-term government secu- rities, preventing wholesale investors from investing in retail gov- ernment securities, and reducing the profit (interest) rates on retail government securities. In the broader legal, regulatory, and policy environment, a first step for the government would be to establish and maintain a neu- tral tax regime for investment income from debt instruments used in Pakistan. Second, the government should be encouraged to com- plete the planned privatizations of Allied Bank, Habib Bank, United Bank, Investment Corporation of Pakistan, and National Investment 198 fX PAKISTAN SURVEY Table 15. Medium-Priority Internal Impediments Rules, regulations, and regulators State-owned instilutions investment policies controlled by government Levying of stamp duty on atlotments and transfer of debt securities Central market infrastructure Limited capacity of SBP's manual clearing, sellement, and depository systems for wholesale govemment securities No real-time infrastructure in government and corporate debt securities Poor market disclosure of secondary market debt transactions Market participants Regulatory environment and present level of market development does not accommodating to borrowing by noninvestment-grade companies Potential longer-term capital constraints for intermediaries it volume picks up and interest rates rise Trust. Once completed, these planned privatizations, will increase competition in the banking sector and contribute to growth in trad- ing volumes in the secondary debt markets. Third, it is important to accelerate efforts already begun to improve governance in the pri- vate and public sectors. As part of this move, the commercial law should be reformed, and the commercial court system strengthened to ensure speedy and just resolutions of commercial disputes. A workable system of arbitration should be developed and promoted. Implementation of these recommendations would boost investor con- fidence in the medium to long term. And fourth, the Central Board of Revenue should be encouraged to remove the remaining stamp duty payable on transfers of debt securities. This would encourage additional trading in corporate debt securities. Acrosd the Financial System. In the area of banking, the SBP should be more proactive in promoting banking sector reform, especially since the Pakistani banking sector remains vulnerable in case there are any further unforeseen setbacks. In the government securities markets, new forms of long-term government securities should be introduced as the current instru- ments employed by SBP, which are issued at par and carry fixed rates, are unable to reflect market rates at the time of issue. Instead, "99 JOHN LEONARDO the SBP should issue long-term securities through auction and adopt either carry or zero-basis coupons. In general, the SBP should try to be more consistent in policy stances affecting the government securities market. In terms of market infrastructure, the SBP should move ahead with its planned computerization, to the extent practi- cable, to facilitate increased secondary market trading in govern- ment securities. The SBP should also improve the distribution of information on secondary market trading. Details of trading should be distributed as soon as information capability constraints are re- moved. The stock exchanges should enforce mandatory reporting of transactions in debt securities undertaken by its members. To enhance the investor base for government securities, the MOF must be encouraged to amend the present terms of issue for govern- ment retail securities to eliminate the distortions created by present policies. The amendments should include measures to prohibit whole- sale investors from investing in retail securities and further reduc- tions in the proft (interest) rates payable on these securities. The SBP must simultaneously introduce new long-term government se- curities as recommended earlier. In addition, the MOF should clarify the basis for establishing profit rates for government retail securi- ties and publicly disclose this information. Such action would pro- mote efficiency in the government securities markets. Internal Impediments The recommendations here pertain to rules, regulations, and regu- lators; market infrastructure; and market participants. Rules, Regulations, andRegulators. The knowledge and capabilities of debt market regulators should be strengthened. As noted earlier, the SBP, MOF, and SECP periodically require assistance to enhance their understanding of the technical issues associated with develop- ment of the debt market and to formulate associated policies. The government should increase its financial support to SBP, MOF, and SECP to allow them to take on greater roles in promoting the debt markets. Technical assistance inputs should be coordinated with those planned under the ADB's expanded CMDP initiative. The format of the proposed technical assistance could be accomrnodated by the 20flf- PAKISTAN SURVEY latter organizations within their debt market development activi- ties. All three organizations have adopted a staged approach to the development of the government securities and corporate debt mar- kets. They are also serious about developing the primary and secondary debt markets. In the case of market oversight, the SECP should be more vigi- lant in monitoring the debt securities markets; increased financial resources will be required to implement this recommendation. The SECP should also be encouraged to continue its search for ways to reduce the time it currently takes to approve draft prospec- tuses for TFC issues."0 For market infrastructure, the SECP should introduce clear and detailed guidelines for trustees in debt issues, be more proactive in monitoring their ongoing performance, and con- sider introducing rules to ensure the operational independence of agencies providing credit ratings on Pakistani debt securities. In market practices, priority should be given to evaluating the merits of allowing prominent financial institutions and intermediaries to short-sell debt securities. The SECP should promote standard mar- ket practices when necessary. The three Pakistani stock exchanges can contribute to these ends by continuing their recent efforts to reduce the approval time for TFC draft prospectuses and should be encouraged to reduce costs associated with TFC issues. More generally, the SBP, MOF, and SECP should be encour- aged to evaluate the merits of introducing new debt instruments to broaden Pakistan's debt markets. Priority should be given to evalu- ating the merits of establishing markets in short-term instruments, particularly bank accepted and discounted bills, bank certificates of deposit, and corporate/commercial paper. MarketInfradtructure. Pakistan needs to establish appropriate mar- ket infrastructure to support debt market trading clearing and settle- ment. This matter is being addressed by the ADB's CMDP. A consultative committee, sponsored by SECP and comprising repre- " Since this report was prepared in mid/late 1999, the Pakistani SEC has adoptd an abridged prospectus for TFC issues and is working on revised guide- lines for issuing TFCs. Both actions are intended to reduce the time and costs involved in issuing TFCs. 201 JOHN LEONARDO sentatives of the Central Depository Company, market participants, MOF, SBP, SECP, and stock exchanges should be established to assist in this process. Market Participantd. The regulatory improvements already mentioned for increasing issuance will also depend on the broader macropolitical environment and the ability to cultivate a broader investor base. Major public companies, the SECP, and the stock exchanges should there- fore work together to promote retail interest in TFCs over the longer term. Initially, the stock exchanges could form a committee to pro- mote an increased awareness among stock exchange members. To encourage more investors to enter the market, the SBP should re- move the MOF's restrictive investment policies on state-owned fi- nancial institutions and replace them with policies based on generally accepted prudential management policies. Financial management at state-owned financial institutions should be improved as well. In addition, the prospects for privatizing all other government-owned Financial institutions should be assessed and detailed plans prepared for the speedy privatization of those institutions where privatization is appropriate. Finally, the management of long-term savings port- folios is in urgent need of improvement. Considerable education will be required to improve the private sector's understanding of the corporate debt market and the role of different corporate debt instruments. That means long-term educa- tional services should be provided for the different groups partici- pating in the wholesale debt markets. Likely participants would include staff employed in the accounting, actuarial, banking, corpo- rate, financial, insurance, legal, and regulatory sectors. In the first stage of such an effort, interested Pakistani parties could be given technical assistance in the preparation of a discussion paper outlin- ing possible educational mechanisms and preliminary cost estimates, and taking into consideration international experience. Once agreed, a draft strategic plan would subsequently be prepared for the pre- ferred mechanism. Possible mechanisms include expanding the role of the present Funds Management Association, creating a new en- tity to service these needs, or adding this activity to an existing pri- vate sector organization that has a marked educational focus. There would be no difficulty obtaining a private sector sponsor for this 2202 PAKISTAN SURVEY exercise as several Pakistani organizations have expressed strong interest in this matter to the Mission. The provision of increased access to educational services pertaining to debt markets should help improve portfolio management skills in the private sector over the long term. Intermediaries must be able to engage in short selling in order to make markets and provide liquidity. The SBP and SECP should authorize scheduled banks, major long-term investors, and stock exchange firms who are major participants in the debt markets to carry out shorting transactions in debt securities. CONCLUSIONS Pakistan's debt markets are fragmented and not yet able to function properly. This state of affairs can be attributed to many factors dis- cussed in this paper, including lack of money markets and futures contracts. The absence of markets for CP and futures reflects the reaction of the SBP and SECP to past material governance prob- lems in Pakistan. Since short selling is not permitted, it is also diffi- cult to develop a secondary market. Today, the small corporate debt securities market is overshad- owed by the size of government borrowings. The processes employed in issuing retail and wholesale government securities must be changed in order to develop a more efficient yield curve for government se- curities. The necessary changes should be identified with the help of the MOF and SBP. The SBP should begin selling long-term tenors in government securities through auction. The MOF should discour- age investments in the retail government savings certificates by wholesale investors. The major constraints facing the corporate debt market at present are the limited supply of TFCs, the lengthy process for approving draft prospectuses, issue costs,'" competition from government sav- ings certificates, and minimal retail interest in corporate debt secu- rities. In view of recent progress and the likelihood of further TFC "See footnote 10. 203 JOHN LEONARDO issues, these constraints are expected to loosen over the next 12 months. The SECP and stock exchanges are taking steps to stream- line the TFC approval process and reduce TFC issue costs. There may be some further reduction in interest rates on government sav- ings certificates as well. Overall, the government wishes to see an easing trend in interest rates. It reduced profit (interest) rates on government savings certificates in April 1999, and further reduc- tions are likely. The MOF, however, is likely to insist on making this decision when it sees fit to do so. It may take a similar view on the recommended further privatizations and revised investment policies for state-owned organizations. Technical assistance should be supplied to SBP, MOF, and SECP on an as required" basis. This approach would alleviate the short- term constraints to debt market development initiatives. Technical assistance would also be required in establishing an organization to provide educational services for wholesale debt market participants. The SBP wishes to avoid any crisis in public confidence in the weaker scheduled banks. If the latter banks are unable to make sat- isfactory progress over the next two years, the SBP may well inter- vene to force some rationalization in the manner recommended above. The SBP has recently declined requests for banking licenses from three development finance institutions. However, it suggested that one of the development financial institutions should consider acquiring a scheduled bank (in view of its relatively strong cash position). The SBP and stock exchanges support the general objective of improv- ing transparency in the debt markets. Any objections to the above recommendations are likely to pertain to matters of detail. 204 PAKISTAN SURVEY APPENDIX: CAPITAL MARKET DEVELOPMENT PROGRAM The following text discussing the Capital Market Development Pro- gram was presented as Appendix VII in the SBP's 1997-98 Annual Report. To strengthen regulatory and institutional framework, eliminate market distortions, modernize and upgrade securities market infra- structure, make available more investment outlets to investors so as to inspire their confidence and improve efficiency of the market participants, the Government in collaboration with Asian Develop- ment Bank (ADB) formulated a Capital Market Development Pro- gram (CMDP). "The main objectives of the CMDP are to increase mobilization of long-term resources, bring about efficient resource allocation, making the capital market diversified and competitive and thus en- courage broader participation of issuers and investors. "The key components of CMDP include: * creation of an enabling policy environment, especially a level playing field to enhance competition; * strengthening governance, institutions, regulations, and su- pervision of the securities market; * improving and modernizing market infrastructure and its link- ages; * developing the corporate debt market; , reforming mutual fund industry; * developing leasing industry; and * promoting contractual savings through reforms of the insur- ance sector and pension and provident funds. "As part of reform of regulatory framework of capital market, the Corporate Law Authority (CLA), which is an attached Depart- ment of the Finance Division, is being restructured as an autono- mous Securities and Exchange Commission of Pakistan. The law governing boards of the three stock exchanges are being restruc- tured. Karachi and Lahore Stock Exchanges have already inducted seven outside directors each on their boards. Full-time Chief 205 ,JOHN LEONARDO Executives of the two stock exchanges have been appointed for the First time. "As regards infrastructure development of the market, consider- able progress has been made. All the three stock exchanges have developed and introduced automated system of trading. A Central Depository Company (CDC) has been established. It has become operational and is serving all the three stock exchanges. Fifty-six companies have already joined the CDC, and 135 more have been declared eligible for entry. All the companies would be joining the CDC by June 1999. The introduction of automate(d trading system in the stock exchanges and the operation of the CDC has [,stc] lent considerable degree of transparency to the market. It is also planned to establish a National Clearing and Settlement System and over the counter market for fixed-income securities with the technical assistance of ADB." REFERENCES Asian Development Bank. 1999. "Development of a Secondary Debt Market-Capacity Enhancement of the Securities Industry." Draft unpublished report, May 7. Asian Development Bank. 1999. "NSS & Competing Instruments of Fi- nancial Saving: The Need for a Level Playing Field, NSS Study." Unpublished report, July 22. Central Depository Company of Pakistan Limited. 1997/98. "Annual Re- port 1997-98." Jardine Fleming Research. 1999. "FY2000 Budget Review." June 18. Khadim Ali Shah Bukhari & Company Ltd. 1999. "Fixed Income Monthly." September. Merrill Lynch. 1999. "Federal Budget FY00: Review." June 14. State Bank of Pakistan. 1997/98. "1997-98 Annual Report." Lahore. State Bank of Pakistan. 1999. "Statistical Bulletin." Lahore, June. World Bank. 1996. "Pakistan Bond Market." report. September 27. World Bank. 1999. "Republic of Pakistan Insurance Sector: Review of Sector and Considerations for Privatization." May 12. 206 CHAPTER 13 Sri Lanka Survey: Issues In Local Bond Market Development LENNART KONIGSON & MALIN NYSTRAND ver the past few years, Sri Lanka has been working to develop its local currency bond markets, particularly for government securities, but for corporate securities as well. This study con- siders the prospects for developing corporate bond markets in Sri Lanka and the impediments to such efforts. It covers issues ranging from economic policy to the specifics of trading, clearing and settle- ment, and ways to facilitate bond market development. CURRENT STATE OF THE SRI LANKAN BOND MARKET The Sri Lankan bond market consists of government securities (T-bills and T-bonds), corporate and bank bonds listed at the stock exchange, and unlisted corporate bonds. There are no mortgage or infrastructure bonds, although some of the bonds issued by banks are for housing finance. As of mid-1999, the bond market (not in- cluding T-bills) totaled about LKR 77 billion (table 1). As figure 1 shows, Sri Lanka's bond market is small in compari- son with the markets of other Asian countries. Only China has a smaller bond market in relation to GDP. Government securities dominate the capital market in Sri Lanka, followed by bank depos- LENNART KONIGSON & MALIN NYSTRAND Table 1. Estimated Total Bond Market, mid-1999 (billions of Sri Lanka rupees) Bonsds Amont Listed nongovernment 6 Unlisted nongovernment 6 Treasury 65 Total 77 Source: Central Bank of Sri Lanka; Colornbo Stock Exchange, dealers' approximation for unlisted corporale bonds. its. The equity market is fairly small, while the corporate bond mar- ket is minimal (table 2). The Government Bond Market With the gradual increase in trading in government bonds in recent years, Sri Lanka now has an active primary and secondary market. Figure 1. Fixed Income Market as percentage of GDP' Sri Lanka Thailad Indonesia Philipines Hong Kong Malaysia China Singapore India South Korea 0% 10% 20% 30% 40% 50% 60% 70% 80% Source. Central Bank of Sri Lanka, Endo, Tadashi, 1998, The Indian Securities Market. ' Figures for Sri Lanka are from mid-1 999. Data on the fixed income market in other countries are from 1996, while GDP data are from 1995. SRI LANKA SURVEY Table 2. Size of the Markets, End ot 1998 (billions of Sri Lanka rupees) Government Domestic Debt per debt instrument LKR as per the end of 1998 billion percentage Listed corporate bonds 5 Rupee Securities 25 156 Unlisted corporate bonds approx. 5T-bills 120 27 Government domestic debt 447 T-bcnds 49 1 1 Equity market capitalization 11 7 Other 27 6 Bank deposits (Commercial banks and NSB) 372 Total 447 Source: Central Bank of Sri Lanka Annual Report 1998; Colombo Stock Exchange; dealer's approximation 'or unlisted corporate bonds; Central Bank ot Sri Lanka Annual Report 1998. The Primary Market. As of mid-1999, some LKR 65 billion of T- bonds were outstanding (figure 1). In recent years there has been a decisive shift from nonnegotiable debt securities (so-called rupee loans) placed directly with the main lending instituttions to market- able government securities (T-bills and T-bonds) sold through auc- tions. After a modest start with T-bond issuance in 1996, the government securities market now comprises bills and bonds with maturities ranging from 30 days to 5 years (figures 2 and 3). The primary market for T-bills and T-bonds consists of weekly sealed electronic bid auctions. Bids can be competitive or noncom- petitive.2 The central bank has the right to refuse up to 25% of the bids and to reduce the quantity of the offer accordingly. Thus, the primary market is more a call for bids than a real auction. Govern- ment securities are sold through primary dealers. As of mid-1999, Sri Lanka had 18 primary dealers in government securities, but this number is expected to drop to 8 or 9 in the near future.3 'From having previously been mainly noncompetitive a growing proportion of the bids have gradually become priced. Primary dealers in government securities are appointed by the Monetary Board, while the requirements and standards that apply to primary dealers are set by the central bank. The central bank may appoint secondary market dealers in government securities. There are no secondary dealers at present, however. Only appointed primary and secondary dealers can deal in government securities. 209 L.ENNART KONIGSON & MALIN NYSTRAND Figure 2. Outstanding Negotiable and Non-negotiable Bonds 300 250 200 150 50 - 100 1994 1995 1996 1997 1998 mid-1999 + Non-negotiable ---Negotiable (T bonds and T-bills) Source: Central Bank of Sri Lanka Annual Report 1998: Central Bank of Sri Lanka (as regards data to 1999). Figure 3. Outstanding Bills and Bonds 140 120 2 100 ° 80 z 60 - 40 20 1994 1995 1996 1997 1998 mid-1999 __- T-bills - T-bonds Source: Central Bank of Sri LankaAnnual Report 1998; Central Bankof Sri Lanka (as regards datafor 1999). 210 SRI LANKA SURVEY The Secondary Market. Trading in shorter maturities is quite active. This secondary market is dominated by bill repos and has gradually become the busiest segment of Sri Lanka's money and capital mar- ket. The secondary market in the shorter maturities has developed relatively fast while a short-term borrowing and lending market based on T-bills and bonds has grown even faster. This market, estimated to process LKR 4 billion to 5 billion daily, is commonly referred to as the repo market.4 The central bank offers repo and reverse repo rates but has otherwise not engaged in open market operations on a large scale. Of the LKR 447 billion outstanding government securities, only LKR49 billion, or 11%, consist of tradable treasury bonds. Non- negotiable, nonmarketable rupee securities are gradually being re- placed by T-bonds, which are growing at a steady pace. As noted below, trading is done through a central bank system, but provisions are being made for the trading of T-bills and T-bonds. At present, there is no screen-based bidding and dealers do not know the spreads and volumes of other dealers. The Yield Carve. The yield curve for government securities (from 30 days to 5 years) serves as a benchmark for nongovernment bonds. Figure 4 shows the benchmark yield curve as of June 1999. It also illustrates the interest anomaly that exists in the Sri Lankan capital market: bank deposits yield a lower return than the risk-free gov- ernment securities. The Corporate Bond Market Like the government market, the corporate bond market has a pri- mary and secondary level. The Primary Market. The nongovernment market consists of listed and nonlisted bonds. Listed bonds are normally referred to as corpo- rate bonds, although most of them are either issued or guaranteed by banks. The market is small in comparison with government borrow- 4 A large part of the so-called repo market is a money market using T-bills and bonds as collateral. 211 LENNART KONIGSON & MALIN NYSTRAND Figure 4. Benchmark Yield Curve, June 1999 15 14 - 13 - 12 - 11 A- 10 9 _ 3-3month 6-month 12-month 2-year 3-year 4-year 5-year T-bill and T-hond & NSB fixed deposit, effective yield p.a. --- Pan Asia Bank fixed deposit rate Source Adverftisements in Daily News, June and August 1999. ings, but it has grown rapidly from a single issue of LKR 1.3 billion in 1995 to 19 listed issues totaling LKR 6.7 billion in August 1999. As figure 5 shows, 11 issues were launched in 1998. The market rose from only 1 or 2 issues per year in the period 1995-97 to 11 issues worth LKR 2.6 billion in 1998. All but two of the issuers have been financial institutions (predominantly banks), driven by recently introduced capital adequacy requirements (most ol the bank bonds Figure 5. Listed Nongovernment Bonds Issues 1995-99 3000 12 2500 10 2000 7*1500 -J10004 500 2 0 0 1995 1996 1997 1998 1999 mths Amount of listed bonds + No of listed issues Source. Colombo Stock Exchange; Samarakoon Lalith P. *21 SRI LANKA SURVEY were issued in 1998), and by a need to correct maturity mismatches resulting from the banks' growing housing loan portfolio. Many bank bonds were placed with major depositors, which allowed banks to transfer liabilities from deposits to bonds, that is, into tier-2 capital, and thereby meet the new, more onerous, capital adequacy require- ments. Vanik, a licensed finance company, has been the most active bond issuer with five bonds listed as well as one unlisted issued in 1999. The total amount outstanding is LKR 1.75 billion (including LKR 200 million in unlisted bonds). The 19 bonds listed on the Colombo Stock Exchange (CSE) were issued by 11 companies, again dominated by banks and other financial institutions (table 3). Only two bonds listed on the CSE were issued by nonfinancial companies, namely, Ceylon Glass Com- pany and Overseas Reality (a property company that owns the World Trade Center in Colombo). Data on the volume and number of unlisted issues are limited. According to traders,unlisted outnumber listed issues and their esti- mated value is between LKR 6 billion and 9 billion. A large number of those issues have been privately placed with captive investors. The privatization of many plantations, for instance, was financed in part by bonds taken up by government entities. The privatization of tea plantations was funded in part by bonds (in lieu of equity) placed mainly with the two large provident funds. Secondary Market. In contrast to the primary market for nongov- ernment bonds, the secondary market has had limited trading. With a total of over LKR 6 billion of outstanding listed bonds, daily trad- ing has averaged only about LKR 3 million, and it has been confined to 9 of the 19 issues. Of those nine, only two have been traded on a regular basis. There is an over-the-counter (OTC) market, but it is restricted to securities that are not traded on Colombo Stock Exchange. These are mainly short-term corporate papers that are bought and sold by and between banks and other finance companies on a negoti- ated basis. There is no quotation system or any systematic dissemi- nation of market data, and clearing and settlement are handled on a gross basis, with a physical transfer of securities (see the next section). 213 LENNART KONIGSON & MALIN NYSTRAND Table 3. Bonds listed on the CSE Amount of Annual yield Maturity Other Date of issue Issuer issue (LKR) (percent) (years) inlormation Novemoer 1994 Overseas Realty 1,300 million 6 5 Secured by the propedies (semi-annual) o1 the company matures in November 19990 do ar-denominated. May 1996 Van k 150nmlion 20 3 Ursecured; matured May, 1999 August 1997 Vanik 150 m Ilion 18 3 Unsecured. (semi-annual) December 1997 Van k 1,000 million 15 10 Unsecured. (annual) June 1990 Ceyl nco 100 mil ion 17.5 4 Unsecured Securities (semi-annual) Augustl998 Haton 250 mi ion 13.5 5 Unsecured National Bank (quarterly) (commercial bank). August 1998 Hatten 750 million 14.2 5 Unsecured National Bank (annual) (commercial bank). August 1998 Cey on Glass 50 million 13 5 3 Capital and interest (semi-annua) guaranteed by DFCC. September 1998 Commercia Bank 250m lior 13.5 5 Ursecured. (quarterly) November 199B Commercia Bank 250 m Ilion T-bi -1% 5 Unsecured (commerc al bank). December 1998 Seylan Bank 208 million 13.5 5 Unsecured (monthly) (commercial bank) December 1998 Seylar Bank 368 mil on 14.37 5 Unsecured (annua) (Commercial bank). December 1998 Seylan Bank 24 mi lior T bill+1 5 Unsecured (annual) (commercial bank). December 1998 Vanik 363 million 150 5 Capilal (annual) guaranteed by USAID. December 1998 Vanik 37 million 142 5 Capital guaranteed (quarerly) by USAID. February 1999 Mercantile Leasing 220 mi lion 14 5 Capital and interest (semi-annual) guaranteed by IFC ard NDB. April1999 People's Merchant 150mi ion 13.5 4 Capital guaranteed Bank (quarterly) by FC and NDB. July 1999 Sampath Bank 500 million 14.2 annua 5 Unsecured 13.5% quarterly; (commercial). T-bill+1% 5 Unsecured (commercial bank) September 1999 Nalional 500 ml ion T-bill+1% 5 Unsecured Development Bank (annual) (development bank) (NDB) (12 5 guaranleed minimum return) 19 issues Total issued 6,620 million Source: Colombo Stock Exchange; Samarakoon Lalith P., 1999, Corporate Bond Market of Sri Lanka. 214 SRI LANKA SURVEY CURRENT STRUCTURES AND ISSUES The corporate bond market in Sri Lanka is shaped by both external and internal factors. Around the Bond Market On the external side, the primary influences are the political envi- ronment, the macroeconomic situation, and the broader regulatory environment. The Politica/Environment. Although Sri Lanka calls itself a "social- ist republic," it has a longer tradition of multiparty democracy than any other country in the region except India. The current coalition party in power, has its roots in Sri Lanka's socialist movement, yet has proven to be more market and reform oriented than its election campaign suggested. The opposition party, if elected in the next round, would not be likely to pursue substantially different economic poli- cies. The issues dividing the political parties in Sri Lanka are no longer mainly those of capitalism versus socialism. Despite the government's growing interest in economic reform and privatization, progress toward the latter has been slow. Union resistance in combination with increased electioneering is likely to preclude a more decisive reform of the financial sector in the near future. The government will therefore remain the regulator and the dominant actor on both sides of the capital market. At the same time, the move to give different government entities more distinct roles and to create more distance between the government and the enti- ties is likely to continue. Restrictions on capital flows between Sri Lanka and the international capital market are unlikely to undergo further liberalization as long as the government believes that those restrictions saved the country from the Asian crisis. Another problem for longer-term securities markets is that the civil war continues to consume financial resources and political en- ergy, although it has not posed a serious threat to Sri Lanka and its institutions. The war fuels the government's borrowing (both directly and indirectly through the rising interest on an ever-increasing na- tional debt). It also deters foreign investment. 215 LENNART KONIGSON & MALIN NYSTRAND The Macroeconomic Situation. Despite the ongoing civil war, Sri Lanka's macroeconomic environment has remained surprisingly stable. Budget deficits and inflation have been kept within manage- able bounds. In recent years inflation has hovered at around 10% a year (figure 6). The effect of inflation on exports has been managed by a crawling peg exchange rate regime. If allows the exchange rate to vary within a band that is pegged to a basket of currencies, which includes the U.S. dollar. The government runs small budget deficits in its borrowings, both in absolute terms and in relation to the region as a whole: the deficit has been between 8% and 11% over the past five years, (fig- ure 7). It has financed these deficits mainly by domestic market bor- rowings, with limited borrowings in the international market. The main lenders have been state pension funds, insurance companies, and banks. Although annual net domestic government borrowings increased from 3.4% of GDP in 1997 to 7% in 1998 (from LKR 30 billion to 71 billion), the margin of savings over government borrowings increased considerably (from 8% of GDP in 1994 to almost 11% in 1998; see figure 8). This suggests limited crowding out of private sector borrowings. Interest rate volatility is a problem, however. In recent years, short- Figure 6. Inflation (WPI) 1989-98 25% 20% 0 15% 10% 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 Source. Central Bank of Sri Lanka Annual Report 199i\ 216 SRI LANKA SURVEY Figure 7. Budget Deficit as Percentage of GDP 12.0% 10.0%= 80% 5.0% 4.0% 2.0% 0.0%- 1994 1995 1996 1997 1998 Source Central Bank of Sri Lanka Annual Report 1998 Figure 8. Savings and Deficit Financing 20 18- 16- 14 rZ w12- a 10- 8- w 6- 4- 2- 0- 1994 1995 1996 1997 1998 Savings / GDP -- Dnmestic deficit tinancinT / GDP Source Central Bank of Sri Lanka Annual Report 1998. term interest rates have ranged from a low of about 10% to a high of almost 20%. The main cause has been irregular government bor- rowings. The tendency has been for interest rates to drop during the first half of the year, then to increase during the second half along with borrowing requirements. 217 LENNART KONIGSON & MALIN NYSTRAND Broader Legal andRegulatory Environment. The taxes levied on the capital market, especially on listed securities, are by and large quite favorable for market development. Sri Lanka has no capital gains tax, for example, and companies that list their shares on the stock exchange are allowed several exemptions. The stamp duty and with- holding tax structure, which previously acted as a disincentive for capital market development, has gradually been overhauled and made more favorable. Stamp duty and withholding taxes on listed bonds were recently removed, as was the stamp duty on repos of T-bills. Moreover, bonds are favored over bank loans because of a 6.25% defense levy on interest charged by banks (which does not apply to bonds and debentures) along with a bank reserve requirement of 11% of deposits. Sri Lanka accounting standards (SLAS) are set by the Institute of Chartered Accountants of Sri Lanka and are in line with most currently valid international accounting standards. Under Sri Lanka's Companies Act, firms are required to maintain proper accounts and to keep a record on file at the Registrar of Companies. The act calls for certain disclosures, but does not insist that SLAS be followed. There is little monitoring of accounting standards, and penalties are very low. However, the CSE requires all listed companies to pre- pare financial statements, in compliance with SLAS. Furthermore, a new law requires all "specified business entities" and their audi- tors to prepare financial statements in accordance with SLAS for financial periods beginning January 1, 1999.5 With the exception of the banks' parate rights, Sri Lanka's legal system offers essentially the same adjudication possibilities as do other Anglo-Saxon legal systems. Bondholders' rights are contained in a trust agreement lodged with a trustee. The so-called parate powers Specified entities have been defined as entities engaged in the business of banking, insurance, leasing, and factoring; finance companies; quoted public com- panies; fund management companies; stock brokers and stock dealers; a company operating a stock exchange; unit trusts; public corporations; unlisted companies of a certain size (having either a turnover exceeding US$7 million, gross assets ex- ceeding US$4 million, equity or liabilities to banks and financial institutions exceeding US$1.4 million, or staff exceeding 1,000); a group of companies, any one of which falls within any of the above categories. ;218g SRI LANKA SURVEY allow trustees to foreclose on loan collaterals without the interven- tion of the courts. However, it takes a long time to settle default cases that reach the courts. Defaulters can obtain restraining orders against banks, which prolongs such cases even more. Recent legislation has made it easier for banks to enforce finan- cial contracts. Under their so-called parate powers, commercial banks are able to foreclose on collaterals without the intervention of the courts. These are unique powers and favorable for the banks, but they do not apply to financial institutions other than commercial banks. Since these institutions are dominated by the two state-owned banks, the privilege benefits mainly the state.6 Across the Financial System As the foregoing suggests, the banking system dominates the finan- cial sector in Sri Lanka. TheBanking SydtenL The interest rate anamaly in Sri Lanka, whereby bank deposits yield lower returns than government securities, can be traced to the banking system. The three large state-owned banks that dominate the banking sector have by far the largest branch network in the country. Through it they have traditionally offered what are perceived to be risk-free accounts. This is indeed the case for the largest deposit taker of those three, the National Savings Bank (NSB), since it invests mainly in government securities (75%) and in bank deposits or bank-guaranteed papers. In essence, the NSB, and to a somewhat lesser extent the two state-owned com- mercial banks, have served as retail savings collection agencies for the dominating borrower, the central bank. The NSB therefore has no option but to offer savings yields below those of T-bills and bonds. In practice, this is a way for the government to channel savings into government borrowings. This has also allowed the other banks to attract funds at below T-bill cost but to lend them to corporate cus- tomers with a relatively large spread of 5% to 7%. 6 Sri Lanka has also instituted specialized commercial courts. With those it should be possible to reduce the long processing time of the ordinary judicial sys- tem. 219 LENNART KONIGSON & MALIN NYSTRAND While state-owned banks maintain their grip on the market out- side Colombo (outstations), there are signs of increasing competi- tion for deposits in the Colombo area, both among the banks and between banks and other deposit-taking finance companies. A large part of the so-called repo market is one area in which banks and others compete actively for deposits. However, these are mainly short- term deposits and savings. Bank liquidity requirements were reduced in mid-99 from 12% to 11% of deposits. Previously, banks had to satisfy the liquidity requirements on a weekly as opposed to a daily basis, which created considerable volatility in the money market that also spilled over into the capital market. This has since changed, and daily fluctuations must now be within 10o of the weekly requirement. The Money Market. The money market comprises an interbank or call money daily market, T-bills, a repo market, commercial papers issued by corporations, and promissory notes (so-called pro-notes). The money market has been thin but the activity in the market has been increasing. The Government Securities Markets. As already mentioned, the gov- ernment securities market has been growing, and efforts are under way to make it more market oriented and active. The main concerns at this stage are market liquidity, usefulness of the securities as a benchmark, and dealer profitability. The lack of liquidity in the secondary T-bond market has two effects on the corporate bond market. First, it detracts from the yield curve and the value of T-bonds as benchmarks. Second, it does not allow corporate bondholders to manage the interest rate risk of their portfolio by hedging or diversification. Three factors detract somewhat from the usefulness of the yield curve as a benchmark. First, auctions in the primary market are more calls for bids than real auctions. Second, the yields for longer maturities are derived from quoted prices rather than actual trans- actions (the trade in longer maturities is not daily). Third, the gov- ernment influences decisionmaking among the dominating investors, which are state owned. 220 SRI LANKA SURVEY Another concern is the order-driven trading used for govern- ment bonds. In the past, trades were negotiated over the telephone and transactions settled through the transfer of physical paper and bank checks. But the central bank is currently developing a new screen-based trading system, to be used only by primary dealers and only for T-bills and bonds, which will be order driven.7 Moreover, by the end of 1999 T-bills and T-bonds were to be traded over the CSE, which is also order driven." In addition, trading through the CSE was expected to face high commissions. At the same time, trading through the CSE was expected to increase competition, since it is an alternative to the central bank system, in which trade has to be chan- neled through the primary dealers. The central bank is also developing a new system for clearing and settlement that will enable primary dealers to settle trades be- tween themselves and major investors through a "semi-scripless" system that employs book-entry and electronic registry, as well as paper. There will be a central depository and a book-entry system in which each primary dealer will have two accounts, one for the dealer's own trading and one for trading on behalf of customers. Trading on the primary dealer's accounts will be paperless. Trading on behalf of customers can be paperless but the option of providing the customer with a certificate will remain. If a certificate is provided, the bond will be transferred to another register. The name of the holder will not be registered in the central bank system, but by the dealer. Thus 7A local software Firm, SasiaNet, has developed a trading system in response to the central bank's request for proposal, but the system has not vet been ac- cepted by the bank or the dealers, purportedly because of the uncertainty con- cerning the number of primary dealers and the functioning of the related clearing and settlement system the Bank is working on. The SasiaNet system includes a sophisticated monitoring function for the central bank, in which the central bank cani follow every trade made by every dealer and thus have access to the position of each dealer on a real time basis. This would give the regulator access to consid- erably more information than is available in most large and developed bond mar- kets in other countries. 8 The physical security can then be deposited with the CDS at CSE, after which it can be traded through the CSE trading system, using the CSE clearing and settlement system. The same rules will apply to government securities as for other securities, so that, once deposited, the security has to be traded through the CSE system through CSE member stockbrokers. 221 LENNART KONIGSON & MNALIN NYSTRAND trading through primary dealers will be paperless, but not trading between other parties. Although it will be possible for large inves- tors to open a paperless investor account, trading based on bonds in such an account will have to go through a primary dealer. Settle- ments will be made on a delivery-versus-payment basis. All primary dealers have a current account with the central bank, which means that payments will be made by direct transfers between accounts. This new system should facilitate the development of the sec- ondary market in government securities. One of its important fea- tures is that all trades in the secondary market will have to be channeled through the primary dealers in order to be cleared and settled through the paperless system. The system will thus give pri- mary dealers extensive control over secondary market operations, but this could deter competition and further development of the sec- ondary market in government securities. At present, primary deal- ers also exercise control over the secondary market since all secondary market transfers have to be endorsed by a primary dealer or a bank. However, this is less of an issue for a body made up of 18 primary dealers. Since investor bond ownership will be registered with the primary dealers (who in turn will enter their consolidated register in the depository), the borrowing and lending of dematerialized secu- rities and, as a consequence, short selling, would be difficult for anyone but primary dealers. Of the 18 current primary dealers, only a few are active in the secondary market. The central bank's new rules, as already men- tioned, are expected to bring their number down to around 8. Some of the current nonbank primary dealers who will not remain pri- mary dealers under the new rules would like to be licensed as sec- ondary dealers, with access to central bank clearing and settlement, and would like to deal in corporate bonds as well. This would ben- efit the secondary market in both government and corporate bonds. Primary dealer profitability can also affect dealers' ability to support the market. Several of the large investors, such as Employee Provident Fund buy from primary dealers at no cost, that is, at no commission and no spread. From their position of dominance, the two provident funds have been able to squeeze the margins of pri- mary dealers to an absolute minimum. Since they and the NSB, also a primary dealer, account for a large proportion of the primary mar- 222, SRI LANKA SURVEY ket, the profit primary dealers can earn in the primary market is negligible. Moreover, primary dealers are not allowed to keep a spread on trades exceeding LKR 5 million. The reason for selling large quantities without profit is that the central bank has volume require- ments for the primary dealers, and the primary dealer's license can be withdrawn if these are not met. Several primary dealers claim they have to make their money in the secondary market. The difficulty they encounter in making profits may be one reason why the central bank has not tried to enforce the rule that requires primary dealers to post two-way prices, that is, to make a market. Perhaps that is also why the new scheme accords primary dealers privileges in the secondary market. Inside the Corporate Bond Market Within the corporate bond market, the principal factors that affect market development are regulations and regulators, central market infrastructure, market participants, and intermediaries. Regulations and Regulators. Whereas government securities come under the central bank's market rules and regulations, equities traded on a formal exchange must abide by SEC and CSE rules, although some changes have recently been made for corporate debt. One such regulation prohibits short selling for either listed securities or gov- ernment papers. This rule will not be easy to change, since short sell- ing of scripless securities does not appear to be possible under present laws. Second, there are no rules for unlisted fixed income securities. Bonds that are listed on the stock exchange and/or traded by licensed stock brokers fall under the indirect supervision of the SEC. The issuing of listed bank guaranteed bonds is also subject to central bank guidelines. Neither the central bank nor the SEC is responsible for unlisted nongovernment securities, except for commercial papers, for which there are central bank guidelines. Third, there are no regula- tions for the OTC market in which such securities are traded. Exist- ing regulations pertain to the financial intermediaries, which guarantee, market, and trade in these securities. Thus a listed bond cannot be traded in the OTC market. Only securities listed on the CSE can be settled through the CSE clearing and settlement system. 223 LENNART KONIGSON & MALIN NYSTRAND CentralMarketInfradtructure. The noteworthy features of the cen- tral market infrastructure are the channels of trading, the means of clearing and settlement, the OTC market, and credit-rating agencies. Trandiq. Corporate bonds are listed and traded on the Colombo Stock Exchange. All securities listed on the stock exchange must be traded through the brokers of the exchange and through its trading system. Trading at the CSE takes place through an automated screen-based system. Brokers trade through a wide area network that enables them to operate from their own offices. Trading is conducted on all market days between 9:30 a.m. and 12:30 p.m. As with most modern stock exchanges, the trading system is order driven. A number of new rules were recently introduced to facilitate the trading and issuance of corporate bonds on the CSE. To begin with, companies with unlisted shares are now able to list bonds. Listing rules for bonds are slightly less stringent than for shares. The SEC has also removed the stamp duty and withholding tax on listed bonds. In addition, clearing and settlement day has been brought down to T+1 for corporate bonds, while it remains T+6 for equity trade. Several features of the infrastructure at CSE are likely to hinder secondary trade in listed bonds. In particular, the system is order driven, there are no dealers or market makers, short selling is not possible, and stock brokers' trading privileges create high transac- tions costs and a "lock-in" effect (all listed securities have to trade through the stock exchange by its licensed brokers). The order-driven system is a hindrance because bond traders typically shun automated trading systems of this nature in favor of screen-based quotation systems and negotiations by phone. Indeed, attempts to introduce order-driven trading have failed in several developed bond markets elsewhere. The bond trade in official ex- changes is limited for a number of reasons. For one thing, bonds are traded on terms whereas stocks are sold and bought because of expectations related to the issuer. A stock in- vestor normally wants to buy a specific stock, while a bond investor looks for a combination of yield, maturity, liquidity, and credit risk. The bond investor seldom has a strong preference for a specific bond issue as long as his or her parameters are satisfied. Bond trading there- fore involves a larger proportion of negotiation than trading in stocks. 224 SRI LANKA SURVEY For another thing, bond trades typically involve much larger amounts than stock trades, and dealers' margins tend to be volume sensitive. In addition, transaction costs in the form of fees and com- missions are often higher in the stock exchange than in the OTC market. Furthermore, a stock exchange normally has no minimum trading volumes and shorter trading hours. Trading in bonds re- quires somewhat different skills than stock trading, and it is carried out largely by dealers, whereas brokers tend to dominate the trad- ing of stocks. Finally, the lower degree of transparency in the OTC- market might benefit the dealers, by way of reduced competition and higher spreads. Another important point to note is the broker cartel in the stock exchange and the noncompetitive manner in which fees and com- missions are established. Both have likely helped curtail stock and especially bond trading. In a highly competitive bond market, bonds are traded only on spread, which may be as low as 5 basis points. By comparison, the current CSE commission for bonds is 0.20% of the traded amount, to which may be added a considerable spread on account of the limited liquidity of the market. High brokerage fees and commissions as well as high spreads add to the "lock-in" effect, which dampens the secondary market. In the long run, brokers them- selves become the victims of a lack of market growth. Clearlnq anld settlement. The CSE clearing and settlement system in use in May/June 1999 allowed settlement to take place six days after trade at the earliest. This serves the needs of mainly equities and corporate bonds, but provisions are being made for T-bills as well. In September 1999 the system began allowing settlement one day after trade (T+1) for bond trades. Settlementtakes place through the Central Depository System (CDS), which also registers the transactions. This procedure is automated. Those who want to trade their securities will let their broker deposit the physical paper with CDS, and security ownership is then transferred by way of book- entries. Only transactions through the stock exchange can be reg- istered. The CDS is a separate company, which is fully owned by the CSE. Securities change hands by way of book-entries at CDS, and a schedule is presented for net settlement between brokers at T+6. 225 LENNART KONIGSON & MALIN NYSTRAND The settlement bank is ANZ Grindlays, and each broker and the custodian banks have an account there. The buyer has to pay the broker on T+5 in order for the check to clear by T+6. The buying broker is required to have sufficient funds on account at ANZ Grindlays. If a broker fails, which has not happened asyet, the Settle- ment Guarantee Fund operated by the SEC covers the loss. How- ever, any broker who causes a withdrawal from the fund will have their trading privileges revoked. Since securities are transferred at T-day, the buyer pays at T+5, and the seller is paid at T+6, there is no delivery versus payment. Because it is possible to reverse the transfer of the security on the trading day (T), the counterpart risk is limited to one day in the paperless system. The six-day delay from the day of the trade to settlement is in- tended to allow checks to clear. Check clearance takes one day within Colombo and the Western province, two to four days for many other parts of the country, and up to seven days for war-stricken areas. Since many brokers require buying customers to pay on T-day or even deposit an advance, the current practice creates a positive float on which brokers can earn interest. Allowing for settlement of trade in corporate bonds by T+ I would significantly reduce risks, but it might also cause difficulties for traders outside Colombo and a loss of a profitable float for Colombo brokers. Outstation buyers would be unable to meet the T+l limit because there are no facilities for wire transfer of funds and securities that are not lodged with the CDS have to be presented in kind. Some brokers have objected to the proposed T+1 rule. The loss of the float may not be the only reason. The time may be too short for brokers to notify clients of the trade, receive funds from clients, and allow for check clearance. The OTC market. Colombo has an OTC market, mainly in money market instruments such as commercial papers and so-called pro- notes, which are unlisted. There is reportedly very little trade in unlisted bonds, and the secondary trade in listed bonds must take place in the stock exchange. Under the proposed T-bond and primary dealer structure, pri- mary dealers will control all secondary market T-bond trades that are to be scripless, but their activities will be restricted to govern- 226S SRI LANKA SURVEY ment securities. A semi-scripless clearing and settlement system for government securities only will support a screen-based trading sys- tem. This will serve to keep dematerialized government securities out of the OTC market and will concentrate bond trading capital and competence on T-bills and bonds. The proposed new rules would, however, allow materialized T- bonds to be listed on the CSE and to be dematerialized and entered into its CDS. This would allow for secondary trading outside the purview of primary dealers in the case of materialized bonds. Credit-RatingAgencied. A credit rating agency (CRA) was set up in July 1999 and was to become operational before the end of 1999. It is to be owned by a number of institutional investors, including IFC, the central bank, NDB, and the Development Finance Corporation of Ceylon. The technical partner and major owner will be Duff & Phelps, the third largest credit-rating agency in the United States, which also operates a rating agency in India. The CRA is expected to affect the market in two ways: it should reduce the issuers' dependence on bank guarantees and thus the cost of issuing debt, and it should expand the market for corporate bonds. With a CRA, many institutional and contractual investors will not need to be as constrained as they are today by rules that prevent them from investing in corporate bonds other than those secured by bank guarantees. Credit rating would make it possible to reduce this restriction and allow portfolios to include different grades of bonds. Furthermore, credit rating is thought to improve the risk-reward structure in the bond market, by distinguishing between companies with different risk profiles. The result is likely to be a much wider range of interest spreads. Market Participants Market participants can be divided into issuers, investors, and in- termediaries. Idsuerd. As already noted, financial institutions make up most of the current investor base, with banks being predominant. There are sev- eral potential issuers in Sri Lanka, and these can be divided into _ 227 LENNART KONICSON & MALIN NYSTRAND three main groups: large, well-established domestic conglomerates, leasing companies, and telecom companies. Lnqe, vell-e,a/hJtbed domedticconiglomerates. This group includes firms such as Aitken Spence, John Keells, Hayleys, and Richard Peries. These companies have strong balance sheets and a good reputation, and they are well known. Many of them see the bond market as a means of reducing their dependence on short-term borrowings. Sev- eral are waiting to find out if CRA ratings would reduce the need for bank guarantees. Some consider the current risk-reward struc- ture to be inadequate in the sense that the price difference between them and higher-risk companies is too small. The CRA is expected to help address this problem. Lewsinq companies. With a growing number of banks starting their own leasing operations, leasing companies are not only experienc- ing increased competition but are having difficulty gaining access to funds and bank guarantees for bond issues. This group of compa- nies is extremely interested in bonds, and they place great hope in the CRA, which might enable them to raise funds without bank guarantees. Telecom companlesJ. Firms in this group include Suntel, Dialog, and Sri Lanka Telecom. The telecom industry is growing rapidly and thus has a large need for capital. Furthermore, it has good prospects for profitability. Bonds issued by those companies would probably stand a good chance of competing with bank loans in the area of cost. Invedtord. The principal investors in Sri Lanka are provident funds, insurance companies, banks, unit trusts, and individuals. The main categories of institutional bond market investors have deposits to- taling more than LKR 500 billion, almost half of which is invested in T-bonds and other government securities (see table 4). The bal- ance is made up of commercial bank loans. The annual increase in deposits is currently about LKR 100 billion, a growing portion of which is to be channeled through long-term investors such as pen- sion funds and insurance companies. With government borrowings 228 SRI LANKA SURVEY Table 4. Principal Investors Approximate figures Total Annual Government Corporate (LKR billion) capital inflow securities portfolio bond portfolio Provident funds 210 50 165 6 Insurance industry 22 4 13 0.5 Commercial and savings bank 300 50 40 <1 (in deposits) Unit trust 3 0.2 0.7 0.1 Corporations and individuals n.a n.a - around 4 Total 535 104 219 around 12 Sources EPF; ETF, approximations by private provident holders, Insurance companies; Central Bank ot Sri Lanka; National Savings Bank; unit trust management companies. currently running at LKR 40 billion to 50 billion annually but set to decline, there will be an increased demand for nongovernment long- term securities. The main concern about investors is that they are dominated by a few large state-owned institutions, mainly banks and two provi- dent funds. Generally speaking, this has limited competition among banks, pension and provident funds, housing finance institutions, and insurance companies in several ways. For instance, the govern- ment has used its regulatory powers to impose rules favoring gov- ernment entities. It has also excluded private sector competitors from government business in the case of both pension and provident fund management and insurance, and it has granted its own entities ex- clusive access to funds. Being both large and dominant, government entities have been able to set prices and extract margins on the basis of cost plus rather than market forces. This has allowed more effi- cient private companies to increase profit margins rather than com- pete on price and efficiency. Pensfion and prtivident/indf. The pension and provident fund sector is dominated by the two state-controlled provident funds, EPF and ETF, which together account for over 90% of the total market (see table 5).9 EPF is larger than ETF and has an investment portfolio of 9 There are no pension funds in Sri Lanka, except for a few small in-house pension funds serving only employees of the institution in question. Contributions 229 LENNART KONIGSON & MALIN NYSTI RAND Table 5. Pension and Provident Funds Provident fund Total capital Approximate Goverment Corporate (all amounts in LKR) market share securities holdings bonds holdings EPF 170 billion 81% 150 billion 3.1 billion, ETF 23 billion 11% 15 billion approx3 billion' Private provident funds approx. 17 billion 8% n.a n.a Total approx. 210 billion aAs of the end of 1998, EPF bond holdings consisted of one corporate bond (Vanik) worth LKR 10 mi lion, commercial papers worth LKR 55 mill on, plantation debentures wsrth LKR 1,643 million, and development bank debentures worth LKR 1,369 million. I Inc udes LKR 600 Wihion in corporate bonds. The remaining amount consists mainly of plantation debentures. Sources: EPF; ETF; approximations by private provident holders. approximately LKR 170 billion, of which 85% is placed in govern- ment-related investments. ETF's investment portfolio totals LKR 23 billion, of which approximately 60% is invested in government papers. Hence at least 65% of all provident savings are invested in government securities. EPF and ETF buy government securities mainly in the primary market. EPF and ETF face three problems. First, they are only allowed to buy corporate bonds guaranteed by banks. This restriction is likely to be lifted when the CRA comes into operation.10 to EPF and ETF are mandatory. They are deducted from the payroll along with income tax. Evasion is thought to be widespread, however. The funds are contri- bution schemes, and as such are fully funded. The accumulated benefits, including returns, are paid as a lump sum to the beneficiary on retirement or death. ETF, but not EPF, allows for withdrawal upon termination of employinent and also pro- vides a limited health insurance. A draft bill presented to Parliament at the begin- ning of the 1990s included proposals for the merger of the two provident funds into one provident and pension scheme that would allow for a combined lump- sum payment and annual payments. The bill was turned down after objections from the trade unions. ' Companies have a choice of contributing to EPF or to a closed company specific provident fund. There are about 200 such funds of different sizes but only two that serve the needs of more than one company. The Ceylon Chamber of Com- merce has a provident fund that manages the provident capital of employees of the Chamber's member companies, and there is another similar fund for plantation companies. With the exemption of privatized plantation companies, these funds have not been allowed to take in new members since 1996, when the Employees' Provident Fund (Special Provisions) Law No. 6 of 1975 came into effect. This law prohibits the establishment of new private pension and provident funds. 230 SRI LANKA SURVEY Second, the funds are very large in relation to the market. EPF is the single largest primary market investor, with an annual inflow of LKR 45 billion to 50 billion in funds. It would trade, on average, LKR 1 billion worth of securities per month in the secondary mar- ket. Now that its investment policy has been liberalized, EPF is gradu- ally moving into private sector papers and investments. However, it is looking into outsourcing of fund management for a smaller part of its portfolio and is considering proposals aimed at creating several separately managed funds. Third, the pension and provident system has been operating with- out any return on investment targets, so it can hardly be said to have the interests of the contributors at heart. Payroll deductions in favor of EPF and ETF have been mandatory, similar to those of taxes, and the funds have been used mainly for financing the government's budget deficit, as in the case of taxes. Furthermore, the government has used its regulatory powers to limit the scope of competing private provident funds. This has helped keep the national savings rate in Srri Lanka lower than in most other Asia countries. In.i,'rance conzpalte.s. Sri Lanka has seven insurance companies, two of which are owned by the state. The largest state-owned company accounts for over 50% of the market. The state insurers are per- forming poorly, while two of the private companies appear to be under financial strain. The capital of the entire insurance industry amounts to LKR 21 billion to 22 billion, of which approximately 20% is classified as free reserves and is available for investment in corporate bonds. The in- surance market is growing rapidly. During 1998 alone, the total life insurance market grew by 20%, and the figures for the first half of 1999 indicate an even higher growth rate. Thus the annual increase in investment capital from the insurance industry would amount to LKR 4.4 billion, of which approximately LKR 1 billion would be available for investment in corporate bonds and other nonspecified forms of investments."' "Total purchases as of April 1999 amounted to LKR 15.9 billion. 231 LENNART KONIGSON & MALIN NYSTRAND Efforts to modernize insurance regulation have been under way since 1995. A new law has been drafted that would allow for greater flexibility in choosing investment alternatives, so that only 30% of life and 20% of general insurance would have to be invested in gov- ernment securities, while the remaining capital would be open to all sorts of investment. When the law will go into effect remains unclear. For the time being, the insurance industry remains heavily regu- lated, although the resources devoted to supervision are very lim- ited. Under current regulations, the companies' capital less accumulated profits (the so-called reserve fund) has to be invested in the following manner: government securities must make up at least 50% of the reserve fund for life insurance (30% for general insurance), equity can make up a maximum of 5% of the reserve fund, and the remaining part can only be invested in a number of specifically permitted forms of investment excluding corporate bonds/ debentures and real estate. Accumulated profits (the so-called free reserve) can be invested as the company sees fit, and from this source investments are made in corporate bonds, debentures, and other such entities. At present, insurance companies have more than 50% of their total capital in government securities because few alternatives are open to them. Insurance companies buy mainly in the primary market. The government has protected the state-owned insurance com- pany from competition insofar as insurance by the public sector is concerned. It has also reduced the return on investment of the en- tire sector by onerous investment restrictions in favor of govern- ment debt. However, the private insurance companies seem to be able to take advantage of the inefficiencies in state institutions and to attract new customers and funds. Bank.s. The National Savings Bank is a wholly state-owned institu- tion. It collects savings through a large number of branches cover- ing the whole country. The NSB portfolio consists of 75% in government securities, and the bank is required to invest at least 60% of money deposited by customers in government securities. As of the end of 1998, the total deposits of NSB amounted to LKR 84 billion. Deposits increased 9% during 1998. Until recently, the NSB 232 SRI LANKA SURVEY was not allowed to invest in equity; in the area of corporate debt, the bank is allowed to invest only in listed bonds secured by bank guar- antee. Part of the portfolio is deposited with commercial banks and merchant banks. NSB holds approximately 25% of the outstanding government debt. In 1999, NSB planned to buy some LKR 20 bil- lion worth of government bonds. As to be expected, commercial banks use most of their funds for commercial lending activities. They invest mainly in government securities, partly because government bills can be included in their liquidity requirement. Most banks have some investments in deben- tures, equity, and listed bonds. During 1998, total deposits of com- mercial banks increased by LKR 37 billion (or 12%). The operating inefficiencies and dominating market position of the two state banks and the National Savings Bank are said to be responsible for Sri Lanka's interest anomaly, whereby bank deposit rates are considerably lower than the rate of the risk-free govern- ment securities. The losers are first and foremost the savers, and the winners are the private commercial banks, which enjoy a comfort- able spread. The housing finance sector has been dominated by two govern- ment institutions, but comrnercial banks and development banks have taken an interest in the sector. Competition is therefore likely to increase. Unit trittsi. Sri Lanka has five management companies at the mo- ment with 12 open-ended unit trusts, with a total value of LKR 3 billion. Approximately LKR 800 million is invested in fixed-income funds, of which LKR 1 10 million is in corporate bonds. Unit trusts are regulated by the SEC. The drastic decline in stock values over the past five years hit unit trusts rather hard, since their portfolios comprised mainly stocks. Yields have been low, at 3% to 4% a year. Many of the unit trusts entered the market in 1993/94, when the stock market reached a peak. Individuals who invested in the funds at that point have seen nothing but losses so far. The current inflow of funds is limited (LKR 2 million per month for the whole industry), and for some funds outflows are larger than inflows. Another drawback is that foreign investors have not been allowed to invest in unit trusts. During part 233 LENNART KONICSON & MALIN NYSTRAND of the last five years, foreign investments in the stock market ac- counted for as much as 50% of turnover. Indications are that this limitation might be removed in the case of equity funds. Some unit trusts have started fixed-income funds, of which ap- proximately 25% is invested in corporate instruments and the re- maining 75% in government securities. Unit trusts claim that the problem with corporate instruments is availability. Furthermore, government bonds are more liquid than corporate bonds, and thus preferable. IndividuaLo. Individuals invest in corporate bonds on a small scale, but they generally do not trade, preferring to hold to maturity. Some of the issuers of bonds-Vanik for instance-have advertised heavily in rural areas in order to persuade individuals to subscribe to their bond issues, and subsequently have gotten a number of individual subscribers (approximately 25% of the subscribers were individu- als). Other individuals buy their bonds through a broker. Since the unit trusts have been performing poorly and are at- tracting less and less funding from individuals, one of the most im- portant intermediaries between individuals and the capital market is missing in Sr; Lanka. Internediaries. The following intermediaries participate in Sri Lanka's securities market: 18 primary dealers, 15 stockbrokers, 14 custo- dian banks, and 5 moneybrokers. None of these entities act as mar- ket makers or so-called specialists, in the sense that they post firm buy and sell prices with maximum transaction volumes. The new central bank rules for primary dealers aim at creating a market maker situation. Dea/lerd. The only licensed dealers in Sri Lanka are the primary deal- ers in government securities. Although the CSE presents no formal obstacles to dealers, it has no rules for them as yet. Since there are no dealers, there are no market makers. The primary dealers in gov- ernment securities are the most experienced bond market interme- diates. Currently they deal in corporate securities as well, but their business volume in this area is limited by the small size of the corpo- rate bond market. Primary dealers would be in a good position to 24 SRI LANKA SURVEY develop the corporate bond market. The new central bank rules (re- quiring them to deal exclusively in government securities) will raise their cost for participating also in the corporate bond market since this would have to be done through another corporate entity. The CSE does not allow short selling, without which dealers and market makers must face high Finance costs for the inventory they have to maintain. Allowing short selling is likely to be compli- cated under the present laws, especially in the case of scripless secu- rities. The CDS at the CSE is a company of its own, but fully owned by the CSE. Clearing and settlement through this CDS are only available for securities listed on the CSE. This restriction is likely to inhibit the growth and development of the OTC market since it will place OTC-traded securities at a disadvantage. Money brokelrs. There are Five moneybrokers in the primary market. These brokers are important intermediaries for Primary dealers, but this is expected to change with the new trading system between the primary dealers. John Keells, Sri Lanka's largest conglomerate, re- cently closed down its moneybrokering firm, not because of losses, but because it did not see a future for this activity. Others can be expected to follow suit. Stockbrokeri.o Stocks and corporate bonds are listed and traded on the Colombo Stock Exchange, which is owned by the 15 stockbro- kers who are members of the exchange. Since they are brokers, these members are not allowed to act as dealers, that is, trade for their own account. There are provisions for dealers under the SEC rules, but so far no one has applied. Thus there are no dealers in Sri Lanka, except for the primary dealers in government securities. The CSE is not yet prepared to admit dealers, since there are no CSE rules for dealers. Brokers' commissions are fixed, determined by the CSE in consultancy with the SEC, and are fairly high by international stan- dards. All CSE trades have to be conducted through licensed bro- kers, and all securities listed on the CSE have to be traded on the CSE. These rules assure brokers of a certain level of remuneration, but they are also likely to inhibit the growth of the secondary mar- ket. Thus it affects the brokers negatively as well, since the volumes of their trades are less likely to grow. 235 LENNART KONIGSON & MALIN NYSTRAND CustodAn 6ank,s. Custodian banks execute trade on behalf of foreign investors, but they, too, have to trade through brokers. Since for- eigners are not yet allowed to invest in corporate bonds, they have little impact on the corporate bond market at present. This might well change in the near future. POSITIVE SIGNS FOR THE BOND MARKET Despite the impediments described earlier, Sri Lanka is interested in developing its corporate bond market. Potential users are show- ing interest in such a market, and several factors suggest that inves- tor demand for corporate bonds is likely to increase in the near future. Also important is the fact that corporate bonds are competitive with other financial instruments. The Need for a Better Bond Market Although they are growing rapidly, Sri Lanka's government and corporate bond markets are still very small in relation to GDP. In Assia, only China has a smaller bond market. An expanded bond market would improve the efficiency of Sri Lanka's capital market by lowering spreads, extending maturities, and raising the return on long-term investment portfolios. Lowering Spreads. A more developed corporate bond market would lower the spreads by increasing the competition for borrowers and among different financial intermediaries. As of mid- 1999, the prime rate for bank loans was approximately 17%, while the bank long- term deposit rate averaged about 11%. The spread of approximately 6% can be broken down as follows: 1% represented tax on inter- est,'2 1.5% was the bank cost of reserve requirement, and the re- maining 3.5% was the bank's lending margin.'3 This spread would be put under pressure if the bond market became larger and more 2 The tax on interest is 6%. 3 Sources for the data on bank borrowing are DFCC and borrowing corporates. 236 SRI LANKA SURVEY active, since the bond market would offer an alternative mainly for borrowers, but also for some investors. Furthermore, the bond market could increase the competition between banks and other financial intermediaries such as leasing companies, finance companies, and investment banks. As mentioned earlier, now that more and more banks are venturing into leasing, specialized leasing companies are finding it difficult to get bank loans. A well-functioning corporate bond market would allow them to partly offset the banks' advantage in both cost and the availabil- ity of funds. Extending Maturities. Sri Lankan commercial banks seldom offer loans for longer terms than two to three years. To date, only two long-term lending institutions (NDB and DFCC) have been able to meet demand by sourcing themselves in the international market or from multilateral development banks. Without further government guarantees, they are unlikely to be able to do so in the future. In Sri Lanka, as in many other developing countries, companies frequently operate with considerable asset liability mismatches. They have long-term assets financed by short-term bank borrowings, which are rolled over on a continuous basis. This exposes companies to unnecessary risks and stifles long-term investments in general. At present, NDB and DFCC loans are the only options for correcting companies' maturity mismatches. Bonds could help correct such mismatches, since they have longer maturities than bank loans, typically five years. Bonds can also in- crease the supply of housing finance and the competition for the mainly state-owned housing finance institutions. In the absence of sufficient demand for funds from the corporate sector, many banks have ventured into housing finance. Further growth of their hous- ing loan portfolios is constrained by a growing maturity mismatch, which can only be addressed by borrowings in the bond market. A large number of the listed corporate bond issues in recentyears have been driven by the need to correct this mismatch. Increas4ingReturns. With a larger and more developed corporate bond market, the main investors-the provident funds, insurance compa- nies, and banks-would be able to increase the return on their in- 237 LENNART KONIGSON & MALIN NYSTRAND vestment portfolios, since corporate bonds yield higher return than government bonds. The return on some of these portfolios at present hardly covers inflation. The primary market yields of bank-guaran- teed corporate bonds are currently 1% over the corresponding T- bond yield, but they trade at a discount in the secondary market. The secondary market is very thin and illiquid, however. If the avail- ability and the liquidity of corporate bonds were to increase, corpo- rate bonds would become even more attractive to investors. Moreover, a well-developed bond market would have a better diversity of yields in relation to risk and would allow investors to match their yield requirements with risk. A well-functioning sec- ondary corporate bond market would furthermore improve the risk-reward structure of the overall capital market and therefore stimulate thrift. The need for a better bond market is confirmed by the interest of potential issuers and investors' need for alterna- tive investments. Interest among Potential Issuers This review of the bond market in Sri Lanka included a survey of 15 potential corporate issuers in the financial sector, the telecom indus- try among the large conglomerates, and the plantation groups. The near-term (within one year) demand for funds for which bonds were considered was in the range of LKR 3 billion to 5 billion. In the medium term, the main demand for bond borrowings is likely to come from the banking sector for housing finance purposes (LKR 2 billion to 3 billion annually), the leasing industry (0.5 billion to 1 billion annually), the telecom sector (1.5 billion to 2 billion annu- ally), and industry (0.5 billion to I billion annually). This suggests an average of about LKR 5 billion to 7 billion worth of new issues per year in the medium term. If the privatization pro- cess picks up, the amount required by the industrial issuers could become considerably higher. Infrastructure and municipal bonds, on the other hand, are not likely to appear until later. The market currently requires nonbank corporate bonds to be secured by a bank guarantee. This increases the borrower's costs by 1 to 1.5%, which is seen as a considerable obstacle. Several sur- veyed companies expected the coming CRA to reduce the need for 238y SRI LANKA SURVEY guarantees and thus make bond issues more competitive with bank loans. This applies in particular to issuers such as leasing compa- nies, which compete with the banks' in-house leasing operations. The surveyed potential issuers were interested in promoting market making for their securities since this would help reduce costs (interest would be 0.25% to 0.5% lower, according to several large investors) and would facilitate exit for the issuer in case the market changed. Investors Will Demand More Corporate Bonds Several factors will combine to increase the demand for corporate bonds: the government intends to reduce its borrowing; financial sector reforms are likely to lead to increased pension and insurance savings, which are very low in Sri Lanka; and the liberalization of investment restrictions for provident funds, insurance companies, and unit trusts would benefit mainly corporate bonds. The three main categories of institutional bond market investors have total deposits of more than LKR 500 billion, almost half (LKR 220 billion) of which is invested in T-bonds and other government securities. The balance is mainly commercial bank loans. The annual increase in deposits is estimated to be LKR 100 bil- lion, and the long-term investors such as pension funds and insur- ance companies are likely to account for a growing portion of this. Table 6. Deposits of Main Investors (approximate figures in billions of Sri Lanka rupees) Holdings of Holdings of government corporate Investor Total capital Annual inflow securities bonds Provident funds 210 50 165 6 Insurance industry 22 4 13 0.5 Commercial and savings bank 300 50 40 <1 (in deposits) Unit trust 3 0.2 0.7 0.1 Corporations and individuals n a. n.a - around 4 Total 535 104 219 around 12 Sources. EPF, ETF, approximations by private provident holders, Insurance companies; Central Bank of Sri Lanka; National Savings Bank; unit trust management companies 239 LENNART KONIGSON & MALIN NYSTRAND With government borrowings currently at LKR 40 billion to 50 bil- lion a year but set to decline, there will be an increased demand for nongovernment long-term securities. Corporate Bonds Are Competitive Sri Lankan corporate bonds, typically secured by bank guarantees, have slightly higher yields than long-term government securities, but they cost the borrower less than the prime rate payable on bank loans. The annual cost of a bond to the issuer is made up of the coupon rate paid annually to the investor (14%), the one-time cost of issuance (0.4%), the annual bank guarantee fee (1.5%), and taxes. There are no taxes on listed securities, and most bonds that are listed are bank guaranteed. As of mid-1999, the approximate annual cost of a bank-guaranteed listed corporate bond with a normal 14% cou- pon was just under 16%.14 Bonds can compete with banks, both on the investment side and on the borrowing side. Most of the corporate bonds issued in Sri Lanka are bank guaranteed. They are therefore not corporate bonds in the normal sense, since they carry a bank risk and since the bank- ing sector is perceived as risk-free. Bank deposits yield approximately 3 percentage points lower interest than bank-guaranteed corporate bonds while carrying the same risk and costing approximately 1 percentage point less than bank loans for the issuer/borrower of funds. The 1 1% yield on bank deposit and the 14% yield on corporate bonds are not fully comparable, since the maturity period differs (1-2 years compared with 5 years). If the liquidity of the corporate bonds were to improve, however, bonds could be a competitive al- ternative. Currently there are little or no secondary market activi- ties in most bonds. Potential investors are therefore mainly those who can afford to be "locked in" for the term of the bond. Corporate 14 Sources for the data on cost of corporate bonds are Sri Lankan bond issu- ers. The cost of bond issuance, 0.4% per annum, is based on a five-year bond of LKR 500 million. The underwriting fee is 0.75% up front, making it 0.15% per year. Other costs of issuing, such as the listing fee, prospectus, and advertising, are approximately LKR 6 million, which equals 0.25% a year. 240 SRI LANKA SURVEY Table 7. Key Features of Longer-Term Bank Deposits, T-bonds, and Corporate Bonds Corporate bonds Feature Bank deposits T-bonds (bank guaranteed) Average maturity (years) > 2 15 3-5 Yield (percent) -11 -13 -14 Interest risk optional Fixed or market Market Market Credit risk None None Same as banks Liquidity risk None Low High bonds thus have a competitive yield, mainly from the perspective of the issuer but also from the perspective of the investor. The record also suggests that they have been competitive: all seven listed corpo- rate bonds that were issued in 1998 were oversubscribed. Corporate bonds also can compete with government bonds on the investor side. In the primary market, bank-guaranteed corpo- rate bonds currently have a 1% higher yield than the corresponding maturity T-bonds. Apparently this is a competitive yield since all issues have been oversubscribed. Almost half of all bonds listed on the CSE are bank bonds, and of those issued by other companies, some 50% are bank guaranteed. Vanik issued three unsecured bonds in 1996 and 1997, and Ceylinco Securities issued one in 1998. Recent issues of nonbank companies have been bank guaranteed, and the view of issuers is that investors demand bank guarantees until the CRA comes into operation. In effect, the credit risk of the corporate bonds issued up to now by and large corresponds to the bank risk. The key features and risk factors for longer-term bank deposits, T-bonds, and corporate bonds are compared in table 7. The three investment alternatives differ mainly with respect to their liquidity. The low liquidity risk on T-bonds'5 gives banks a long- term yield against shorter-term deposits. Interviews with large in- stitutional investors suggest that improved secondary market liquidity could reduce the "liquidity cost" by up to 0.5%. The current 1% spread between government and bank-guaranteed corporate bonds " Longer-term T-bonds are not traded on a regular basis but can be repoed through a central bank window. 241 LENNART KONIGSON & MALIN NYSTRAND compensates for the credit risk (that is, the bank risk) and the li- quidity risk. Investors estimate the liquidity risk to be about 50 per- centage points. In the future, when the Sri Lankan bond market has developed, there are likely to be fewer bank guarantees, and the spreads will be larger and more diverse. The credit risk for nonbank companies is likely to exceed the bank guarantee fee, but will depend on how and if the company is rated by the CRA. If market-making arrangements are introduced, it should be possible to reduce the liquidity risk. Considering the low bank deposit rates and the demand from issuers and investors for corporate bonds, bonds will probably be competitive even in the future. Figure 9. 120 - 100 3 80 Bank credit risk , 60- coo ,, Liqjid2ty risk 40- 20 0 Figure 10. 400 350- U Unrated credit risk 300 U Rated credit risk - 250 - Bank credit risk 200- l [iCuidity rise co 150 100- 50 Current Future SRI LANKA SURVEY RECOMMENDATIONS For the market to develop, attention needs to focus on four broad areas: * Improving the secondary T-bond market, which would allow better interest benchmarking-and thus enable primary deal- ers to trade in corporate bonds as well as government bonds and better manage the risks of their bond portfolio -and repo market supervision. * Helping the OTC market function better by intro- ducing scripless clearing system for OTC-traded securities; a screen-based quotation system; market making; short selling and borrowing and lending of securities; and regulation of OTC securities, issuers, and intermediaries. * Improving the regulatory structure and investment guidelines, by creating independent and specialized regulatory agencies, and modern and flexible investment policies for contractual investors. * Increasing the competition for banks, pension and provident funds, housing finance, insurance companies, and brokers and dealers. Improved Secondary T-bond Market To strengthen the secondary market, it will be necessary to improve interest rate benchmarks, allow primary dealers to deal in corporate securities, and rein in the repo market. Improve Interedt Rate Benchmarks. Although interest benchmarks have improved considerably, they need further attention. The secondary market in longer-term maturities is not daily and the yield curve for those securities is quotation rather than transaction based. Currently there are many small issues of government bonds. Fewer but larger issues (jumbo issues) would increase the free float of securities and enhance liquidity. Allow Primary Dealers to Deal in Corporate Securitied. Proposed new rules for primary dealers in government securities would require 243 LENNART KONIGSON & MALIN NYSTRAND them to deal only in government securities. This would create a seg- mented market, which could constrain the development of a corpo- rate bond market. It would earmark a large proportion of the capital available for financing bond intermediation for government securi- ties and would also prevent dealers from hedging corporate bond risks against T-bonds. The best way to reduce the market risk of corporate bonds is to hedge corporates against government bonds. To do this, dealers typi- cally maintain portfolios with a mix of treasury bonds, corporate bonds, housing bonds, and other entities. In fact, in most markets, corporate bonds are typically quite illiquid, but this is of little sig- nificance as long as the risk in mixed portfolios can be managed. Restricting the primary dealers to trade in government securities thus eliminates the most suitable trader in corporate bonds from the market. A bond dealer faces three risks. The first is credit or default risk, that is, the risk of default on the part of the issuers. There is no credit risk for government bonds. For corporate bonds, this risk is measured through credit ratings and is reduced by bank guarantees or other forms of security. The second is liquidity risk, that is, the risk of not being able to exit and recoup the capital at any given point in time. Liquidity risk can be reduced by market-making ar- rangements, since a market maker would be required to buy and sell at least once at quoted prices. In the government securities market, this obligation rests with primary dealers. The third is market or interest risk, which is the risk that movements in the interest rate will change the value of the dealer's inventory of securities. This is the main risk a dealer faces. Rein in the Repo Market. Over the course of 1999, Sri Lanka saw the rise of a short-term deposit market, using bank T-bills and T-bonds as collateral. The borrowing bank or financial institution typically issues a letter that entitles the lender to a proportion of the borrower's bond portfolio that corresponds to the value of the loan with accrued interest. Since bonds are bearer instruments, how- ever, this is likely to be an unenforceable security unless the lender has physical possession of the security. This so-called repo market is thus burdened with two problems: the same collateral can be 244 SRI LANKA SURVEY used for several transactions, and the security cannot be executed if it remains with the borrowing bank, since bonds are bearer in- struments. This practice, which seeks to reduce transaction costs by avoid- ing physical transfer of instruments, is defined as a repo transac- tion rather than a deposit. It allows the banks to avoid liquidity reserve requirements and in principle would allow a chain of bor- rowing and lending transactions collateralized by the same (often unidentified) securities. A default of one of the borrowers could have far-reaching consequences beyond its effects on the money supply. This particular segment of the repo market is not supervised and little detailed information is available on its volume. According to the banks, the entire repo market has a daily turnover of about LKR 3 billion to 5 billion. Create a Well-Functioning OTC Market Developments to date suggest that improved secondary market li- quidity would make corporate bonds more competitive by reducing the interest spread with respect to T-bonds. Even in large and well- developed capital markets, however, liquidity for corporate bonds is often limited. In order to create an active and functioning secondary market in corporate bonds, it is essential to open up the CDS to OTC traded securities; establish an automated quotation system for the OTC-market; allow short selling for dealers and market makers to come into the market; decrease risks by regulating OTC securi- ties, issuers, and intermediaries; and allow primary dealers to trade in corporate bonds. Information Systems. An OTC market cannot function well unless dealers have access to one or more quotation systems, where indica- tive prices are quoted together with who posted the price, the mini- mum volume for which it is valid, and other such information. Dealers typically use this information to negotiate deals by phone. Because of the large amounts and narrow margins in bond trading, counter- part risks are large. The clearing and settlement systems need to be fast, cost-effective, and risk-free. 245 LENNART KONIGSON & MALIN NYSTRAND Under current CSE rules, the OTC market cannot gain access to the existing scripless clearing and settlement system of the stock exchange. Meanwhile, the future central bank clearing and settle- ment system is intended only for government securities and only for dedicated primary dealers. This leaves the OTC market without ac- cess to a CDS. UnIlstedSecurities. There are no regulations for unlisted securities, except commercial papers, for which the central bank has issued some guidelines. The OTC market in which such securities are traded is not regulated either. The regulations that do exist pertain to some but not all of the financial intermediaries, which guarantee, market, and trade in these securities. A total of 10 merchant banks, 3 leasing companies, and 2 private development banks are currently outside the regulatory framework. Among the merchant banks, several are affiliated with commercial banks and/or active in the OTC market either as dealers/traders or as issuers. The regulatory framework should not allow regulated financial intermediaries to operate through unregulated affiliates. Because OTC securities and several OTC intermediaries remain unregulated, OTC securities tend to be riskier than regulated secu- rities. The credit risk on issuers would in addition be higher in the OTC market than at the CSE since there are no listing and disclo- sure requirements. Regulations aimed at OTC-traded securities and intermediar- ies should support self-regulatory mechanisms to the extent pos- sible. A national association of dealers is emerging, but to date it has focused on primary dealer issues. It will be important to ensure that others have a forum once the new primary dealer structure is in place. MarketMaking. Most well-developed bond markets rely extensively on market making in order to maintain liquidity. So far, this has not been the case in Sri Lanka. Market makers have to be dealers; that is, they must maintain inventories in the securities for which they are to make a market. The current rules of the stock exchange do not provide for this and the primary dealers in government securi- ties will not be allowed to deal in corporate securities. Unless and SRI LANKA SURVEY until this situation changes, the prospects for a secondary market for listed bonds will probably be very limited. The current prohibition against short selling is an additional impediment in the sense that it increases the market maker's costs and risks. Unless this restriction is removed, dealers are unlikely to want to offer market-making services. Market makers offer the market liquidity, but they do so at considerable risk in the case of illiquid markets. The value of li- quidity translates into lower cost and greater flexibility for issuers. A bond that at the time of issue is accompanied by the promise of a reasonably liquid secondary market is likely to sell at a lower yield than one without the corresponding feature. Liquidity for a paper also offers issuers the option of either buying back all or part of their bond or of adding to it (on tap) if and when market condi- tions change. Together, those factors give issuers an incentive to pay for a market-making service or to otherwise reduce costs and risks for market makers. Interviews with some large institutional investors in the Sri Lankan market suggest that they might be willing to accept a reduction in yield of up to 50 basis points for a corporate bond that has a market-making feature. More Competition for Banks, Pension and Provident Funds, and Housing Finance and Insurance Companies Government dominance has served to limit competition among banks, pension and provident funds, housing finance institutions, and insurance companies in several ways. The government has used its regulatory powers to impose rules favoring government enti- ties. It has also excluded private sector competitors from govern- ment business in the case of both pension and provident fund management and insurance, and it has granted its own entities ex- clusive access to funds. As a result, government entities have been able to set prices and extract margins on the basis of cost plus rather than market forces. This has allowed more efficient private companies to increase profit margins rather than compete on price and efficiency. Table 8 shows the market share of state-owned com- panies. 247 LENNART KONIGSON & MALIN NYSTRAND Table 8. Market Share of State-Owned Companies and Institutions (percent) Institution Share Commercial banks 50-60 Pension and provident funds 80 Housing finance 75 Insurance industry 60 BANKS Contractual Savings Broker and Dealer Privileges. Under the CSE rules, brokers have exclusive privileges and engage in noncompetitive pricing in a car- tel-like manner. That is to say, all trades in listed securities have to go through licensed brokers that are members of the stock exchange and through its trading system, and only securities listed on the stock exchange can be settled through the stock exchange clearing and settlement system. The secondary market in government securities may also evolve into a cartel, since all trades in the secondary market will have to be channeled through the primary dealers in order to be cleared and settled through the paperless clearing and settlement system. Fur- thermore, primary dealers function as trustees for their clients and register their clients' bond holdings in bulk in the central bank de- pository, and a large investor, who can maintain a paperless inves- tor account with the central bank depository, has to go through a primary dealer. Primary dealers were accorded privileges in the secondary mar- ket probably because it is difficult for them to earn good returns in the primary market. Primary dealers are not allowed to keep a spread on trades exceeding LKR 5 million, and the large investors do not pay any commission. These rules ought to be changed so as not to favor primary dealers in the secondary market. Otherwise, these cartels will segment the market and thus stifle market growth and product development. In particular, they may hinder the development of a market-making function, which is es- sential for the development of a corporate bond market. 2A8 SRI LANKA SURVEY Needfor New Bond MarketActord. Several new services are needed to facilitate bond market development. They, in turn, will require new skills and financial strength, especially in the area of market making and underwriting. Some of the major local financial institu- tions are considering establishing a bond house with a view to meet- ing these and other needs. The bond house would develop a corporate and retail investor network for market making and active trading in corporate and government debt, would invest and take positions for its own account (that is, act as a dealer), and would structure debt instruments and manage issues. The bond house would become a CSE dedicated debt broker and member. The CSE is expected to create rules for dealers for the bond house to function in this capac- ity. However, short selling remains an issue. Better Regulatory Structure and Investment Guidelines The institutional structure of Sri Lanka's regulatory framework was created during an era when the state rather than the market domi- nated the economy. This structure needs to be modernized and im- proved to ensure that it will have the capacity and competence to deal with the realities of a free market economy. The central bank of Sri Lanka combines a large number of functions, which in most market economies have been separated into different institutions. Most modern economies have established separate banking sector regulators and supervisors as well as sepa- rate national debt offices in order to allow their central banks to focus only on the critical issues of price stability and exchange rate policy. The need for specialized competence and administrative and political independence will eventually force Sri Lanka to move in a similar direction. This would probably mean setting up a separate National Debt Office, which would be the issuer of all government securities and the manager of the entire national debt. It could also comprise a single financial sector regulatory body (as in the United Kingdom or Sweden) covering all financial intermediaries, that is, banks, merchant banks, finance companies, insurance companies, pension and provident funds, and unit trusts. An alternative could be a small family of specialized regulators such as those set up in 249 LENNART KONIGSON & MALIN NYSTRAND several other countries. Whatever regulatory structure is chosen, it must address three important issues: regulatory gaps and overlaps, professional competence, and supervisory capacity. In the area of guidelines, the government has hitherto exercised full control over the investment policies of the provident funds so as to meet government financing needs. Under current rules, the two provident funds (EPF and ETF), private fund managers, and the insurance industry can invest a small portion of their funds, but only in bank-guaranteed corporate bonds. This has hitherto not acted as a constraint to market development since the volume of available corporate bonds has been within those limits. However, it has con- tributed to the very low returns on provident and insurance com- pany funds, among other factors. Pension savings in Sri Lanka are very low by international standards. Although government influence is gradually being reduced, poli- cies have yet to relate portfolio risk to return objectives. One conse- quence of the liberalization is that EPF and ETF are investing in both equity and corporate bonds. The small size of those markets, in combination with the large concentration of resources in these two organizations, has nevertheless allowed them to become extremely large market actors. ETF is the largest single buyer on the stock exchange and together with EPF it effectively determines the spread of corporate bonds over T-bonds. It has been proposed that the investment portfolio of provi- dent funds be split into smaller portfolios, each under separate man- agement, so as to increase market diversification and focus more on return on investment. As a first step in that direction, EPF is proposing to outsource fund management for a smaller part of its portfolio. Another way to increase market activity would be to al- low more private pension and provident funds to compete for cor- porate clients. At present, Sri Lanka has only two private open provident funds, and they have not been allowed to accept new members since 1996. EPF has argued that a competent pension regulation authority needs to be put in place before more funds are opened up. Insurance companies face even more onerous restrictions, since insurance supervision is limited, those rules tend to be ignored. A proposal for a new law that would allow greater flexibility in invest- 250 SRI LANKA SURVEY ment alternatives as well better supervision was drafted in 1995 but has yet to be promulgated. In order to improve the regulation of the provident and pension fund sector and the insurance sector, separate regulators with in- creased supervisory capacity should be created for these two sec- tors. The provident fund regulatory role of the EPF Statutory Board and the Department of Labor would then be repealed. Requiring provident funds, unit trusts and insurance companies to mark to market their portfolio values would create better transparency and improve the incentive to trade for better returns. In sum, the regulatory environment can be improved by sepa- rating banking sector supervision from the central bank and strength- ening the supervisory capacity; creating a National Debt Office separate from the central bank; creating a provident and pension fund and insurance sector regulator, with increased supervisory ca- pacity, and repealing the provident fund regulatory role of the EPF Statutory Board and the Department of Labor; and promulgating the proposed new insurance law. CONCLUSIONS Sri Lanka has taken numerous steps to build its local currency bond markets. It is suggested that the next steps should focus on: First, improve T-bond trading. Suggested actions are: have fewer and larger T-bond issues (increased free float) and hence more li- quidity; allow primary dealers to deal in all kind of bonds and stocks so that they can manage corporate bond portfolio risks better; and strengthen repo market supervision to reduce default risk due to the current practice of using nonenforceable collateral Second, strengthen the OTC market. A well-functioning O'I'C market would require a scripless clearing system, a quotation sys- tem, and market makers who can sell short, borrow and lend secu- rities. Thus, Sri Lanka should consider allowing regulated OTC-traded securities to use the CDS; establishing an automated quotation system for the OTC market; allowing dealers to sell short and thereby promote market makers; allowing primary dealers to 251 LENNART KONICSON trade corporate bonds; and reducing investor risk by regulating OTC securities, issuers, and intermediaries. Third, improve competition in the market. Several broad-based, well-known measures would contribute to this end such as restruc- turing state-owned banks, provident funds, and insurance compa- nies into a larger number of smaller units; privatizing the majority of the resulting units, as well as the state housing finance units; and allowing banks, housing Finance, broker/dealers, and contractural savings institutions to compete with another. More specific to the bond market, it is suggested that Sri Lanka liberalize protective CSE rules so that listed securities can be traded outside the exchange (OTC); limit primary dealers privileges to avoid cartelization of government securities trading; amend the proposed government securities trading system so that primary dealers can earn a reasonable return in the primary market and all dealers can compete on equal terms in the secondary market; and require mark- to-market accounting by banks and contractual investors. CHAPTER 14 Bangladesh Survey: Issues in Local Bond Market Development MIKAEL KVIBACK L ike emerging-market countries around the world, Bangladesh could benefit from having a local-currency, fixed-income secu- rities market. At present, its main fixed-income financial prod- ucts are bank deposits, bank loans, government savings certificates, term loans, treasury bills, and government bonds and corporate debt (syndicated loans, private placement, and debentures). But in gen- eral the corporate debt market is still very small compared with the equity market (see table 1). Table 1. Instruments Available in Bangladesh Ntinel amount (billions of takas felatt e size (%) Instruments 474 31.9 Bank Loans 580 39.1 Deposits 184 12.4 Term loans (as of June 1998) 123 8.2 Government saving certiticates 55 3.7 Government bonds 40 2.6 Treasury bills 27 1.8 Equity (issued value) (Not puolicly available) Private placement 1.3 0.1 Debentures Sources Bangladesh Bank, National Savings Bureau. Dhaka Stock Exchange MIKAEL KVIBACK Numerous factors in Bangladesh today suggest that Bangladesh will not be able to develop an active, local-currency fixed-income market. Economic and financial transactions are highly regulated, and the economy does not provide a sufficient number of appropri- ately structured and skilled issuers and investors. Although the gov- ernment recently began privatizing selected state-owned companies and deregulating the financial market, progress has been slow, leav- ing financial market participants skeptical about whether the gov- ernment can succeed in this endeavor. Bangladesh Finds it difficult to move forward for several rea- sons: weak governance at the institutional and market levels; high nonperforming assets among the nationalized commercial banks (NCBs); poorly defined and overlapping responsibilities of the Bangladesh Bank, Securities and Exchange Commission, and Min- istry of Finance; and the lack of incentives and private initiative to drive market developments. These four problems are the principal obstacles to the develop- ment of bond markets in Bangladesh. The government is aware of them, and the World Bank and other organizations have been push- ing for solutions. However, change is slow. Although there is no meaningful base for a secondary market in corporate bonds today, the Bangladesh economy may well grow at an attractive rate in the future, and if it does, capital-intensive in- dustries such as gas and telecom will invest heavily. Thus Bangladesh will eventually need an efficient capital market that can mobilize domestic and foreign resources for investment. For the time being, however, Bangladesh should focus on creating a well-organized, regulated, and attractive primary market in both public and private placements. This discussion is about some of the impediments to the development of fixed-income market in Bangladesh and some ways to remove them. MAJOR IMPEDIMENTS TO BOND MARKET DEVELOPMENT The obstacles to bond market development can be divided into three broad categories: those around and across the market, and those inside the fixed-income markets. 254 BANGLADESH SURVEY Around and Across the Market The obstacles in this group stem from the political situation, the macroeconomic situation, and the broader financial system. The Political Situation. The People's Republic of Bangladesh has been a parliamentary democracy since September 1991. The present government is headed by the Awami League which has an absolute majority, but the opposition party has stepped up its nationwide program of strikes, processions, and mass meetings. These activities have weakened the government's intentions to foster changes such as the development of the financial market. In addition, certain commercial and financial regulations are outdated in that they tend to focus on institutions rather than func- tions. Governance and accountability are lacking in certain areas, and there are elements of inefficiency in the financial system, mainly concerning the state-owned banking sector. Although the govern- ment is aware of these problems, it has been slow to improve gover- nance and develop strong institutional capacity. The problems created by these weak institutions are compounded by an increasingly con- frontational political environment. At the same time, the government has committed itself to launch- ing financial reforms that could help accelerate the country's rate of growth. The main goal of these reforms is to reduce the direct con- trols on the financial system, and to deregulate and introduce a new set of market-oriented approaches to financial sector activity. The Bangladesh National Budget for 1999-2000, for example, earmarks funds for the creation of a central depository system (CDS) to help streamline trading at the stock exchanges and improve authentica- tion. Furthermore, a proposal is under scrutiny that would amend the Trust Act to allow provident and pensions funds to invest in the capital market. To achieve that goal, it will be essential to ease the bad-loan situation, which is draining the country of its monetary resources. But certain factions in Bangladesh oppose those aims and commitments. Since no one has stepped forward to "champion re- form," the government appears unwilling and unable to undertake the requisite changes in due time. Because the political environment is so fragile, laws and regulations are not being fully enforced. 255 MIKAEL KVIBACK Macroeconomic Situation. Bangladesh's macroeconomy was fairly strong throughout the 1990s, with growth rates averaging a respect- able 5%, and inflation averaging a modest 9%-10%. The primary fiscal deficit during the past five years has averaged about 5.5% of GDP, which has generally been within sustainable limits. (However, the consolidated public sector deficit, taking into account losses in- curred by state-owned enterprises, is much higher and underscores the need for improved fiscal management, although foreign exchange reserves have become more stable recently owing to impressive ex- port performance and reduced imports.) Heightened foreign inves- tor interest in the country's natural gas sector has opened up tremendous possibilities. But despite these positive elements there are some serious con- straints on the development of active corporate bond markets in Bangladesh. First, Bangladesh is one of the poorest countries in the world, with approximately 125 million inhabitants, of which about 60 million live below the poverty line. Although its GNP growth rates-in the range of 4%-5%year-are attractive, they suggest that it will take Bangladesh 25 years to double its per capita income. In order to reduce the incidence of poverty to about 11%, as it hopes to do, Bangladesh will have to achieve economic growth rates of 7.5% or more ayear. According to several studies (see, for example, World Bank, "Bangladesh, Key Challenges for the Next Millennium," April 1999), economy has the capacity to move out of poverty with in- creasing speed, but that will require decisive policy actions in sev- eral areas, not least of which is the financial market. However, a sense of urgency is missing in policymaking, despite the growing imbalances in the economy and crowding out as Bangladesh continues to channel vast monetary resources into servic- ing bad loans. Given that macroeconomic changes can happen in short periods of time and that nonperforming loans, which account for a third of the loan portfolio, can create financial sector vulnerability, the bad-loan situation could trigger a severe liquidity crisis nationwide. It can take decades to build a fixed-income market in the wake of such crises. This issue clearly needs immediate and focused attention. If the country's positive macroeconomic trends continue into the future, the fiscal deficit and bad-loan situation will ease up and these factors would pose less threat to the financial market. .256- BANGLADESH SURVEY Broader Laws and Regulations. Certain omissions or drawbacks of the broader laws and regulations directly affect development of the fixed-income market. First, with regard to the ownership of land, the law provides for the registration of deeds rather than of owner- ship, which makes it impossible to take land as collateral for bond issuance. Second, the law makes arbitration a cumbersome and slow process; moreover, foreign arbitration awards are not enforceable in Bangladesh. Third, in terms of obtaining issuers, there is no privatization law to lend transparency and authority to the privatization process, although one is at present being drafted. Fourth, Bangladesh's laws represent a mixture of codified British common law and legal principles from various religious heritages. Although the court system derives from a common law tradition, Bangladesh courts are limited in their ability to function effectively. In view of these constraints, the legal system can move only so fast in amending the laws and enacting new ones, even though the government acknowledges the need for such changes. Contract laws and commercial codes seem to be fair, but ensuring that they are observed is difficult because of a weak adjudication system. Broader Financial Sys4tem. The broader financial system includes the banking sector, nonbanking sector, government securities mar- ket, and short-term money markets. Bankinq sector Bangladesh's banking system, which is dominated by state-owned NCBs, creates two serious problems for a local corpo- rate bond market.' First, the system provides low-cost loans to state- owned enterprises, which account for a large part of the corporate sector. This undermines development of the corporate bond market because other financial institutions are unable to compete with these "underpriced loans." Indeed, the state-owned enterprises constitute a large part of the NCBs' business. To complicate matters, develop- ment financial institutions (DFIs) also provide low-cost loans, priced at a small percentage over bank deposits for similar maturities. 'The Bangladesh banking sector is composed of the central bank (Bangladesh Bank), 4 state-owned commercial banks, 4 specialized banks, 13 local private banks, and 12 local foreign banks. 257 MIKAEL KVIBACK Second, the banking sector is faced with a substantial number of bad loans; nonperforming assets account for about 30% of total as- sets. Although these nonperforming assets can be said to create a need for an active bond market, to the extent that banks are con- strained in new lending and thereby cannot meet the funding needs of corporate borrowers, they also rob the bond market of needed investors. Yet the state-owned banks just keep on making bad loans. Nonbankincqsector. The nonbanking portion of the financial sector con- sists of two small stock exchanges (Dhaka and Chittagong),2 both of which have still not recovered from the bull market problems of 1996, which left the public suspicious of corporate institutions because it is hard to get them to disclose their Figures. At that time, the stock ex- change experienced a hefty run-up in prices owing to a large inflow of funds from retail investors. This inflow, drawn by the prospect of easy money, was a new experience for the Bangladesh people, but it lasted only the second half of 1996. In those six months the index soared from 500 to 3500 and the market came crashing down to about 600. The stock market has not recovered yet: in May 1999 the index hit a 63-month low, at about 465. The average daily turnover in the spring of 1999 was about US$1 million to US$2 million. The weak operating performance by listed companies and low confidence in the market overall has made it difficult for the market to recover. In sum, the nonbanking sector has not evolved in a way that would allow it to play an active role in the financial system. Nor, as discussed in the section on intermediaries, is it prepared to play an 2 The nonbank section is made up of 2 stock exchanges (Dhaka and Chittagong), 170 active brokerage firms, 19 nonbanking financial institutions, and 17 merchant banks. The Dhaka Stock Exchange (as of April 1999) has 208 listed companies on the equity side, 9 mutual funds (of which 8 are issued by the Investment Corpora- tion of Bangladesh, ICB), a state-owned mutual fund company, 11 debentures, and a total market capitalization of approximately US$ 1.1 billion, of which equity stands for approximately US$1 billion. The Chittagong Stock Exchange has 136 equity shares listed, 9 mutual funds, 5 debentures, and a market capitalization of about US$825 million. Membership is open to foreigners at the stock exchanges. Trading is done through an automated real time system and settlement occurs on T + 5. BANGLADESH SURVEY active and skilled leadership role in developing and participating in an active fixed income market. Government securities m,arket. The government securities market in Bangladesh is small, does not provide much of a yield curve to sup- port a corporate bond market, and does not provide intermediaries with skills and a profit base to support the corporate bond market. At present, the government issues long-term savings certificates at high interest rates and government bonds, and it only has market- oriented rates for T-bills. At the shorter end of the market, T-bills are auctioned weekly for 91 days and the Bangladesh Bank (BB) occasionally issues pa- per for 180 days, 365 days, and 720 days. Commercial banks partici- pate in auctions weekly for 91-day T-bills, whereas the others are issued occasionally. Accepted bids are noted in the newspapers. The market is small, with outstandings of about US$800 million. There is no secondary market and no market for repurchase agreements ("repos"). T-bills are transferable, but settlement is manual and very slow, done through BB. On the whole, T-bills are mainly used to satisfy statutory liquidity requirements (SLRs). The past few years have seen a clear bias for short-term borrowing. Government bonds, with maturities ranging from 3 to 25 years, are issued when needed; they do not create a yield curve as T-bonds are nontransferable, mostly because they are issued to recapitalize state-owned banks. Their notable feature is that they are guaran- teed by the government and are eligible for SLRs. Government savings certificates (GSCs) range in maturity from three to eight years. GSCs are offered to different types of investors in the retail sector (but small corporates are allowed to invest). The types of investors are mostly individuals and families but also in- clude charity and provident funds. GSCs are issued in series through the year. The holder may redeem them at par at any time. Finally, GSC issuances offer significantly higher rates than local bank deposits, which create a relatively high rate for risk-free and tax-free government securities. This establishes a high benchmark rate for corporate fixed-income securities, creating a disincentive to invest in corporate securities. GSC rates are 2%-3% higher after tax compared with rates on other government paper. GSCs create a 259 MlKAEL KVIBACK high benchmark interest rate foundation for corporate securities. That matters because it is very hard to compete with risk-free gov- ernment debt. At present, Bangladesh law and the government's fiscal and mon- etary policy combine to create a financial market monopoly for GSCs and NCBs, which in turn keeps alternate financial intermediation from emerging. Bangladesh needs a healthy nonbank financial insti- tution (NBFI) sector to increase mobilization and make competitive financing available in a fixed-income market. To achieve that end, it must break the NCBs' monopoly. Although the government is aware of this problem and has put forward some relevant reforms, there are no real incentives to speed up the process, maybe because of political considerations. Short-term nioney markets. Money markets provide another founda- tion for bond markets. The money markets in Bangladesh are quite small. There s' an interbank market, in which commercial banks bor- row and lend to adjust their short positions (the size of this market is not publicly known). Normal maturities range from overnight to 30 days. Bangladesh also has a forward market for U.S. dollars against the taka, but only for short maturities. There is no commercial paper market. Inside the Fixed-Income Markets The important factors to consider inside the fixed-income markets are regulators and regulations, central market infrastructure, and intermediaries. Regulators and Regulations. One impediment at the regulator and regulation level is the overlapping authority between the two finan- cial market regulators, Bangladesh Bank and the Securities and Ex- change Commission (SEC), and no clear jurisdiction over the fixed-income market. In general, BB regulates the commercial banks and their activities, while the SEC regulates the NBFIs, the two stock exchanges, and the capital market. A second problem is that the SEC has no authority to issue rules and regulations, and the procedure as a whole is long and drawn 260 BANGLADESH SURVEY out. As a result, the SEC has not proposed any regulations for the issuance of bonds or debentures. All rule proposals must first be submitted to the Minister of Finance for approval and then passed on for approval from Ministry of Law. Furthermore, potential issu- ers have to look at various sets of regulations and follow a long and cumbersome procedure. Third, although the SEC requires listed companies to meet in- ternational standards on accounting and auditing, accounting infor- mation appears to be of doubtful quality and reliability. Fourth, the Securities and Exchange Act of 1993 confers vast regulatory authority on the state, and is regarded as a constraint on capital market development. There is a board of policymakers. Three of its members are appointed by the state, another is from the Min- istry of Finance and one from the central bank, and the chairman is appointed by the government. Fifth, in the present system, a company can float debentures up to a maximum amount of its current asset value and has to register its assets in the name of the Trustee as Security. Hence there is no provision for floating unsecured debentures. Central Market Infrastructure. In the absence of a secondary mar- ket in fixed-income securities, no effort has been made to build up a central market infrastructure to support it. Bangladesh only has a telephone market for T-bill trading and central market infrastruc- ture at the stock exchange for trading equities and debentures. In the T-bill market, the counterparts call each other and settle trans- actions without any transparency in real time for other participants in the market. At the stock exchange, the debenture market is fully automated. The debenture market has a somewhat more transpar- ent order-matching system in that bids and offers are entered in the computer and then matched automatically. Bangladesh has no central depository system, though one is ex- pected to start operating in 2000. Today, clearing and settlement are done manually, which creates various risks to completing a transac- tion. Also lacking are a credit rating agency, research and informa- tion companies, and market information on screens; market participants are referred to other media, such as the daily financial newspaper, and thus experience a delay in obtaining essential eco- 261 MIKAEL KVIBACK nomic information. According to some participants, even that infor- mation is often unreliable. MarketParticipantd. Market participants can be divided into issu- ers, investors, and intermediaries. Issue,lv. The foremost impediment here is that Bangladesh lacks a significant number of potential, good-quality issuers. Its economy continues to be agriculturally based; agriculture accounts for nearly 30% of the country's GNP, and more than 70% of the labor force is engaged in agricultural activities. The industry and service sectors contribute 20% and 50%, respectively, but compared with landhold- ings, the average size of industrial and commercial enterprises is rather modest. Most private sector enterprises are small and owner-run, many are of "cottage size" and most are in the garment industry, which to date depends largely on short-term bank loans for financing. These enterprises could benefit from longer-term funding but are neither large enough nor well known enough to issue bonds. Most of the large-scale industrial units and commercial enterprises are state owned. Their shares are not listed, and they do not offer debentures since their financing needs are met by the government or by the state-owned NCBs. These state-owned firms generally stay outside the capital market. The privatization program for state-owned com- panies works too slow to influence the market. Second, although Bangladesh has a debenture market, to date only a small number of well-known issuers have used the market (see table 2). The liquidity in those debentures at the stock exchange is insignificant because of the small number of investors and their buy-and-hold mentality. The investor community does not seem to find this market too attractive owing to weak disclosure by the issu- ers, which in turn reduces credibility and investor confidence. Third, companies find that issuing debt is costly, both in mon- etary and nonmonetary terms. The interest rate distortion due to the GSCs mentioned earlier raises the ongoing cost of borrowing, while various up-front costs amount to about 7% of the value of the issue (these include registration costs -that is, stamp duties -totaling about 2.5% of the issue value). BANGLADESH SURVEY Table 2. Prominent Issuers in the Debenture Market Issued debenture Debenture Coupon Year of flotation (in millions of laka) Beximco Infusion Ltd 17 1992 14.5 Beximco Synthe'ics Ltd. 14 1993 240.8 Bangladesh Chemical Industries Ltd. 17 1993 3.2 Eastern Housing Ltd. 15 1994 202.5 Beximco Knitting Ltd, 14 1994 188.4 Beximco Fisheries Ltd. 14 1994 94.3 Beximco Textile Ltd. 14 1995 222.8 B.D. Zipper Ind. Ltd. 14 1995 22.4 Beximco Denim Ltd. 14 1995 278.5 Bangladesh Luggage Ind. 14 1996 135.0 Arami Cement LUd. 14 1998a 112 5 a No debentures were issued in 1997 Fourth, it is difficult to persuade issuers to disclose sufficient information about their companies (although prospectus require- ments for listed debentures do seem fair). Yet another problem is that most potential issuers are unwilling to take the opportunity cost involved in issuing a long-term bond. In addition, the absence of a yield curve makes pricing difficult. Investors. On the investor side, few investors are sophisticated enough to think about investing in bonds. About 80% of the base here is made up of retail investors, whose primary concerns include the equity at the stock exchange or the government savings certificate. Of the few institutional investors that could support a bond market, most are either prevented from investing in corporate bonds by restrictive guidelines or are not professionally managed. The major institutional investors are the Investment Corporation of Bangladesh-a government-owned financial institution -and the insurance companies. The mutual fund industry in Bangladesh is the exclusive domain of ICB. There are no private mutual funds to mobilize savings toward the debt market, and the IOB's monopoly has prevented new investor companies, that is, mutual funds, from developing in Bangladesh. There are provident and pension funds (total assets managed amount to Tk 6.7 billion; see The Financial Express), self-managed by public and private corporate entities, but none are professionally managed. The pension obligations of the gov- MIKAEL KVIBACK ernment are not funded. The Trust Act of 1882 prohibits those funds from being invested in equities, corporate debentures, and private money market instruments. In addition, no protective laws are in effect to ensure that inves- tors will get their dividend and capital back. Missing are higher audit standards together with SEC regulations on disclosure standards in prospectus along with arbitrary institutions. Furthermore, most investors lack a trading mentality and just buy and hold because of SLR requirements or because they do not know how to trade. Few foreign investors are attracted to this, mainly because of the weak disclosure by the borrowers. As for the general public, it has little understanding of debt products, and the intermediaries are not much help because few engage in research on markets, compa- nies, and industries to encourage investment. Intermediaries. Intermediaries in Bangladesh lack many of the skills needed to foster an active local corporate bond market. As men- tioned earlier, commercial banks dominate the financial sector and not enough intermediaries are skilled in securities. Few are able to identify issuers and investors and bring them to the market. They provide little or no research analysis on industries or companies to encourage investment in the local debt market. Too few private merchant banks are able to conduct financial advisory and trust ser- vices. Nor do any feel motivated to become a market maker for an issue. Hence the market is illiquid, with large spreads. At the same time, the fee structure and pricing are high enough to allow interme- diaries to make money, but because transactions are so limited, the intermediaries seldom make money. Even if they are able to partici- pate, intermediaries are reluctant to take any risk in dealing. RECOMMENDATIONS ON HOW TO REMOVE THE IMPEDIMENTS The various impediments to bond markets in Bangladesh pose a large challenge for policymakers. Nevertheless, some suggestions can be made for dealing with them. 264 BANGLADESH SURVEY Overall, given the current situation in Bangladesh, the country should focus on developing a well-run primary market, both for pri- vate placements and for public offerings. That means several steps need to be taken to fix the inside elements of the market, in addition to some changes around and across the financial system. Inside the Market Rules, Regulations, and Regulators. The role of the BB and SEC in regulating the fixed-income market needs to be clearly defined in detail so that appropriate regulations can be written for the public, private, and secondary markets. These regulations should ensure that each market is encouraged and protected. In view of situations around the world, it likely is best to have the SEC regulate the fixed- income market. But whichever agency is chosen, the regulator must be educated appropriately to ensure that it fully understands the product and is able to supervise the markets, monitor the risks in the markets and the intermediaries, and enforce its power where necessary to ensure a quality market. CentralMarketInfrastructure. Bangladesh should consider whether to develop a central clearing, settlement, and depository institution. Such an entity would support both the equity and debt markets. MarketParticipants. There are too few professional participants in the Bangladesh market to create an effective secondary market in fixed-income instruments. Activities in the market are as yet too limited because the government is unable to create an effectiveyield curve. When such a base is established, market participants will know their relative value for issuing and investing, which in turn will attract new participants to the market. The government also needs to support private initiatives to bring intermediaries to the marketplace. They, in turn, bring the trading mentality to the market that is essential for a secondary market. The best way to do this is to create incentives for professional people to establish their own profitable business. With such a base, the mar- ket will drive itself and private initiatives will ensure diversity in fixed-income instruments. To build investor confidence, the market 265 MIKAEL KVIBACK needs strong accounting rules and regulations comparable to inter- national standards. To that end, the government should strengthen and supervise the accounting rules and controlling body. Factors Around the Market Macroeconomic. At a more general level, to foster market develop- ment, Bangladesh needs to bring more competition into the finan- cial sector through deregulation and privatization. The cotrntly appears to be moving in that direction, but the speed is slow. The govern- ment needs to accelerate its efforts in this area. Broader Financial Sydtem. Ideally, there are several ways that Bangladesh might work to improve operations in its government securities market, to create a market that provides an interest rate structure that supports the entire financial system and a benchmark for corporate bond offerings. However, the analysis performed for this study was not sufficient to determine whether and if so when certain changes might best be suited for Bangladesh. But some sug- gestions can be made which Bangladesh might consider over time, as it seeks to improve operation of its government securities market. More specifically, instead of issuing tax-free and nontransfer- able government bonds to the retail market, the government should consider issuing its bonds in the marketplace. It might issue T-bills and T-bonds with a broader maturity base, transparent pricing, which are tax-neutral and transferable. To start with, efforts might focus on building an effective money market (O/N-365 days), and from that base it may be possible over time to create new short-term in- struments such as repos, futures, short-term interest rate swaps, commercial paper, and a USD/Taka forward market. It is important to build a more sophisticated interbank deposit market with differ- ent maturities. This will help create the everyday price fixing needed to price other financial products (for example, leasing agreements), and it can help create a forward rate agreement (FRA) market. Cre- ating an effective yield curve will help provide a foundation for ulti- mately creating a diverse secondary market. i66 CHAPTER 15 Nepal Survey: Issues in Local Bond Market Development M1KAEL KVIBACK The financial market in Nepal is relatively undeveloped. A tiny corporate bond market is in operation; that is to say, a few pri- vate placements have been made. The government market is more developed, but prices are not market oriented. Generally speak- ing, activity in the financial market is hampered by four problems: the issuer and investor base is insufficient, infrastructure of laws, regulations, and institutions is weak; Nepal Rastra Bank (central bank) and the Securities Board of Nepal (Securities Board) have overlapping roles; and incentives and private initiative are not strong enough to drive market developments. A number of impediments discussed later in the report can be traced to these overarching problems. More specifically, the ability to develop the local corporate bond market is seriously constrained by a weak supply of and demand for the product. The number of potential blue chip issuers and the size of the collective investor base are insufficient to create an institu- tionalized market, and too few financial instruments are available in which to invest. An effective interest rate structure, a fundamental ingredient of an efficient and deregulated business environment, is absent, as are the credibility, accountability, and trust that come from solid corporate governance. In many countries that are moving to a more market-oriented economy, the required major changes in eco- MIKAEL KVIBACK nomic strategy have been achieved, and the principal challenges now lie in building the infrastructure of laws, regulations, and institu- tions needed to implement that strategy. Much of this effort involves transforming the civil service and judiciary and improving the qual- ity of governance in the corporate and public sectors. Nepal has taken a few steps toward these ends, backed by international sup- port. But a more systematic approach is needed. Nepal could benefit from an efficient corporate bond market, as such a market would provide an outlet for resolving the country's bad-loan situation. It would also bring necessary competition into the Financial sector, particularly among financial intermediaries, which at present are almost wholly dominated by the local commercial banks. This review deals with the situation in Nepal today, the major impediments to developing the corporate bond market there, and some ways it might move forward. The overall conclusion is that Nepal should focus on developing a well-regulated private place- ment market, one that gives borrowers access to long-term, local currency funding by matching them up with investors who are look- ing for long-term local currency products. CURRENT STATE OF THE NEPALESE BOND MARKETS Nepal has two types of securities markets: government and corpo- rate markets. The Government Securities Market The government uses three instruments to meet its financing require- ments: short-term treasury bills, national savings certificates (NSC), and overdraft financing from the Nepal Rastra Bank (NRB). A glo- bal limit has been set on net domestic borrowing, which includes net new issuance of treasury bills and NSCs. The 91-day T-bills constitute the large proportion of these in- struments (and are issued through weekly auctions at which the primary dealers are the commercial banks). The NSCs are encashable on demand by the holder at par and have maximum outstanding maturities of seven years. This means that the effec- NEPAL SURVEY tive term to maturity is considerably shorter as they are encashable at any time. In theory, overdraft lending is limited by a "gentleman's agree- ment" between the Ministry of Finance and the NRB, but in prac- tice the limit is often exceeded to finance budgetary shortfalls. On a periodic basis, the overdraft is then converted into nonmarketable bonds, which are held on the NRB's balance sheet. As a result of these arrangements, the Ministry of Finance has little interest in debt management and market development. A well-functioning primary market is fundamental to the creation of an active secondary market in government treasury bills and bonds. The T-bill primary market appears to be functioning effectively with transparency ensured and procedures for announcing upcoming auc- tions (on a weekly basis). Successful bidders are informed of their success in a simple and not unduly bureaucratic way. The retail mar- ket dominates the primary market in NSCs, while the NRB addresses their issues. NSCs are issued at a fixed price. If the issuance is undersubscribed by the retail market, the balance can be subscribed by institutional investors. The NRB and finance companies distribute the issues. Currently, there is a withholding tax of 5% on the final holder of T-bills at maturity. There is a repo market for T-bills with a maturity of up to seven days. Settlement is done manually. Secondary market activities in T-bills are low at present. The repo facility is not used by the NRB in its open market operations. In the absence of a secondary market in government bonds, there is no effective yield curve. The Corporate Securities Market Nepal has a number of privately placed corporate bonds. The Na- tional Provident Fund provides term loans, which are quite similar to privately placed bonds. Because of the present undeveloped state of the capital market, both individuals and institutions place far more of their savings in bank deposits and fixed-interest government se- curities than they would if the capital market were more developed. Long-term savings that should be invested in the capital market are going into short-term instruments. The fact that the central bank is able to raise money at negative real interest rates shows that there is no lack of liquidity in the economy at present. 269 MIKAEL KVIBACK IMPEDIMENTS TO BOND MARKET DEVELOPMENT Impediments to bond market development can be divided into three main groups: those around the bond market, those across the finan- cial system, and those inside the bond market. Around the Bond Market Impediments in this group relate to the political situation, the mac- roeconomic situation and broader laws and regulations. The Political Situation. To accelerate the rate of economic growth, the government has committed itself to deregulating the economy, introducing more market-oriented approaches to economic activity, and reducing its controls over economic affairs. But to succeed at this plan, it needs to intensify its effort to dispel policy uncertainty, instill business confidence, and build on recent progress by further strengthening governance institutions. Stronger and mutually sup- portive regulatory policies are needed to consolidate stability. Since no substantial political changes can be anticipated at the present time, the impediment just described will not go away that easily, although a growing economy might help to shorten the transition period. The Macroeconomic Situation. Three areas of macroeconomic con- cern are the relatively poor state of the economy, the fiscal deficit, and debt management. Nepal's economy has achieved modest growth in recent years. Since fiscal 1994/5, GDP growth has ranged be- tween 2.8% and 5.4%. Economic growth runs at about 3%. The two largest export industries are carpets and garments, which have grown rapidly in recent years. Despite the transition to democracy, which has caused a certain amount of political instability (five governments since 1994), the macroeconomy has remained relatii'ely stable, with low rates of inflation and a reasonably stable currency.' ' The principal unit of currency is the taka (50 taka equal 1 U.S. dollar), which has frequently depreciated to boost exports and to maintain a competitive edge. The currency is convertible. Exchange control policy rests within the Min- istry of Finance and is administered by Bangladesh Bank. 270 NEPAL SURVEY The establishment of parliamentary democracy in 1990 opened the door to a program of,sweepifig economic reforms, carried out be- tween 1991 and 1994. This embraced policies on finance, foreign and domestic trade and investment, industry and foreign exchange, and a start to privatization. A change of government then led to a hiatus in privatization and set back macroeconomic stabilization, but the governments in power since September 1995 have relaunched both programs. Major state-owned enterprises are active in the ser- vice sector, banking, insurance, telecommunications, the international airline, and oil and fertilizer importation and distribution. Most of these enterprises have been designated for privatization, which in some cases has already begun. Foreign interest in the hydropower sector has opened tremendous possibilities. However, Nepal is still a small and very poor country, bordered by two of the largest countries - and economies-in the world, In- dia and China. It is at an early stage of economic development: 40 percent of its 22 million inhabitants are classified below the poverty line, agriculture accounts for some 42% of GDP, and more than 80% of the population live in the countryside. With a per capita income of about US$204, it is poor even by South Asia's relatively low stan- dards of per capita national income (around US$336). Although inflation has declined sharply from the 20% level reached in the early l990s, inflation has hovered near the double- digit level over the past five years. Moreo(iet; the country has a large current account deficit relative to GDP. The fiscal budget for 1999/ 2000 amounts to approximately US$1.1 billion, with a budget defi- cit of approximately 30%. About 75% of the deficit is financed through foreign loans and grants, the balance through domestic borrowings. In addition, development activities remain slow, and the country is facing an increasing incidence of poverty, human suffering, and in- efficiencies of government institutions. The agricultural sector has been hampered by low productivity while growth in the manufac- turing and service sectors has been hindered by a lack of infrastruc- ture, power shortages, and a relatively high cost for capital due to inefficiencies in the financial sector. Moreover, the economic environment remains uncertain. Mac- roeconomic changes can happen in short periods of time, and be- cause the financial sector remains vulnerable, the fiscal deficit can 271 MIKAEL KVIBACK easily contribute to a severe liquidity crisis. Fiscal debt management should structure the borrowing needs to create an optimal balance between short- and long-term borrowing. Up to now, there has been a clear bias for short-term borrowing. Increasing long-term debt is- suance to shape an efficient yield curve would also help reduce the liquidity risk. Broader Laws andReguZltions. In general, laws and regulations are relatively undeveloped in Nepal. Broader rules would aid in devel- oping a corporate bond market. Indeed, a key challenge for Nepal today is to build a proper infrastructure of laws, regulations, and institutions. Civil servants, judges, and lawyers have limited train- ing in commercial subjects. In addition, the judiciary system is re- puted to be slow, and judges are said to lack training in financial matters. There is no efficient arbitration system and no bankruptcy law. Another crucial matter is tax collection. Tax policy appears to have hindered capital market development by increasing transac- tion costs and creating disincentives for investment in the equity and private bond market. Thus NSCs, for example, are free of tax, while investment in bonds and debentures issued by private issuers are taxed. Dividends are subject to taxation at the corporate level; however, the individual recipient is not subject to taxation. Capital gains on securities transactions are taxed as ordinary income to cor- porations and individual investors, whereas in most emerging mar- kets capital gains on investments in stocks and bonds are not taxed. With regard to individuals, there is no specific reference to capital gains in the tax code. It is treated as "other sources" of income. If Nepal is to mobilize domestic and foreign capital, it must improve the integrity and transparency of tax collection. Across the Financial System The overall philosophy governing the financial markets, banking sector, and government securities market tends to make it difficult to develop the corporate bond market in Nepal. Nepal has a by and large government-led financial system dominated by a state-owned bank, with limited competition and a limited capital market. The 272 NEPAL SURVEY government plays a key and pro-active role in the financial sector, owning key financial institutions, and retaining significant influence over the activities of the joint venture banks through direct partici- pation in the share capital and consequent membership on the boards. The two state-owned banks have a market share of approximately 70%. There is little competition in the financial market, owing partly to the absence of an effective capital market, which makes it pos- sible for privately owned banks and joint venture banks to enjoy very high profits. The central bank is also involved in several activi- ties, and exercises important political influence on banking and fi- nancial activities. The Banking Sector. The state of the banking sector-its ability to provide sufficient funds to corporate borrowers and to be an issuer, investor, and intermediary in the local fixed-income markets-in- fluences the ability of those markets to develop. Nepal's financial sector is dominated by two state-owned commercial banks and sev- eral smaller state-owned development banks (see table 1), which together control approximately 60% of the total assets held by bank and nonbank financial institutions (NBFIs). At the present time, Rastriya Banijya Bank (RBB) is 100% government owned. The government has divested its ownership of Nepal Bank Limited (NBL) to 41%. In addition, there are nine privately owned joint venture commercial banks with total assets accounting for about one-third of commercial banking assets at the end of fiscal 1996/97. The pri- vate commercial banks are believed to be well run and profitable; however, the government-owned banks are in poor financial condi- tion. The two state-owned commercial banks are suffering from bad Table 1. The Banking Sector in Nepal Type of bankk Niber Central bank, Nepal Rastra Bank 1 Commercial banks 14 Development Banks 2 Finance Companies 44 Rural development banks 5 Cooperative societies 29 Nongovernmental organizations 30 Postal saving bank 1 273 MIKAEL KVIBACK loans. These run at approximately 30%, but if measured by interna- tional standards, the level is as high as 50%. Reserve requirements are 12%. Interest rates in Nepal are deregulated, and the banks are free to set their own deposit and lending rates. Spreads between lending and borrowing rates continued to widen in recent years, while the overall performance of the banking system showed little improve- ment despite the growth of private banks in the 1990s. However, the Ministry of Finance has imposed a rule on banks limiting the maximum spread between borrowing and lending rates to 5% at the most. (Apparently banks use different maneuvers to circumvent this rule, but it is an example of the kinds of directed regulations that are used instead of market-oriented approaches.) Notable is the interest on treasury bills (91-days), which had fallen to less than 1% by the end of 1998. The lack of investment opportunities and an inefficient financial market are two reasons for that decline. At present, Nepalese law, the government's fiscal and monetary policy, and regulatory indecision combine to create a financial mar- ket monopoly for the state-owned banks. This in turn restricts the development of alternative financial markets and intermediating in- stitutions. The government is aware of the situation and the need to change it and has started reducing bank dominance with release of 60% of NBL. But there are no real incentives to speed up the pro- cess, perhaps because of political considerations. Even if political change seems unlikely in the near future, Nepal can improve the market in certain respects and remove some of the impediments, most notably the lack of clear and concise bankruptcy laws. This has not only hampered loan recovery by banks and other lenders but has discouraged lending to the private sector, especially to small businesses, and has created a bias in favor of public sector lending. The banking sectors' efficiency has been weakened by a priority lend- ing policy and the forced expansion of rural branches. The government securities market lacks the yield curves and active dealer communities needed to develop a sound corporate market. No long-term bonds are issued. One issuer that needs an efficient bond market is the government. At present, it issues NSCs to the general public at predetermined interest rates and offers only short-term securities in the market at market rates. The govern- 274;- NEPAL SURVEY ment also needs a bond market to push state-owned companies to borrow from the market instead of require financial needs from the fiscal budget. A bond market would eventually help resolve the bad- loan situation. Nonbank Financial nstitutions. In addition to the commercial banks, there are 44 deposit-taking finance companies. These companies were the fastest-growing segment of the financial sector during the 1990s. They are quite small in comparison with the state-owned banks and other NBFIs. The other major institutions in the financial sector are the Agricultural Development Bank and the Employee Provident Fund. Each of these institutions is about three times as large as the combined assets of the finance companies. The iMoney Marketd. Aside from treasury bills, Nepal does not have a money market. Development of a market in commercial paper and other money market instruments would provide an alternative source of funding for the private sector and would foster greater comnpeti- tion in the financial sector. An effective short-term money market is also a prerequisite for an effective long-term market. Equity Markets. The state of development of equity markets indi- cates how well versed issuers, investors, and intermediaries are in dealing with securities at the primary and secondary market levels. Nepal's overall market is still in its infancy, however. The Nepal Stock Exchange (NEPSE) is a late development, founded only in 1993. It is owned by the government and is a not-for-profit institu- tion. At present, 107 companies are listed. Very few companies listed on the exchange make a book profit on which a dividend can be paid. Most trading takes place in shares of financial institutions and the few multinational corporations operating in Nepal (such as Coca- Cola and Unilever) Those companies pay dividends. The reporting standards and dissemination of information are in need of consider- able improvement. Manual procedures are used for dealing and settle- ment. Foreigners are not permitted to invest at the exchange. Insider trading, lack of credibility, weak accountability, lack of knowledge about capital markets, low investor confidence, and high interest rates are the main constraints to further development. 275 MIKAEL KVIBACK The high cost of capital is one important obstacle to private sec- tor capital formation and economic development in Nepal, and that high cost is associated with the lack of competition in the financial sector and the poor performance of the government-owned banks. The (real) bank deposit rate is negative to neutral. NSCs offer higher rates than bank deposits and are tax-free. The lending rates of gov- ernment financial institutions are lower than those of other institu- tions, and this distortion affects the capital market. The high rate on risk-free and tax-free government securities establishes a high bench- mark rate for corporate fixed-income securities. Underpriced loans from government institutions further undermine the healthy devel- opment of the corporate debt market. Inside the Bond Market Factors of concern within the bond market relate to rules and regu- lations, central market infrastructure, and market participants. Rules and Regulations. The main concern here is that Nepal's finan- cial market regulators -the NRB and Securities Board -have over- lapping roles and that few regulations have been established for the market. Under the current regulatory scheme, finance companies and commercial banks acting as underwriters, issue managers, mar- ket makers, or portfolio managers are supervised by the NRB. Its job consists of administering prudential standards, financial report- ing, off-site monitoring, and on-site inspections for compliance with applicable rules and regulations. The NRB is also responsible for the overall supervision of commercial banks and finance companies, including their activities in the markets for government securities. Since finance companies and banks are permitted to underwrite the securities of private issuers, the central bank plays a significant role in establishing prudential standards and supervising intermediaries engaged in investment banking (issue managers and underwriters) and other capital market activities. The Securities Board regulates capital market activities. It was established in 1993. Among other things, the Securities Board is re- sponsible for the supervision of NEPSE, its members, disclosure requirements for listed companies, and the licensing of issue manag- NEPAL SURVEY ers, underwriters, portfolio managers, market makers, and stock- brokers. It is also responsible for the registration and supervision of investment funds. At present, the Securities Board is drafting regu- lations under the Securities Investment Trust Act for consideration by the Minister of Finance. The duplication in the functions of these regulating bodies makes it difficult to regulate the securities market. In fact, some laws and supporting regulations ignore important areas of the market alto- gether, many key provisions are inadequate, and others create ma- jor legal problems. Many laws and regulations relating to capital market activity require a thorough overhaul. There is at present no set of laws and regulations for the fixed-income market. Moreover, the Securities Board has limited staff (with 12 pro- fessional and 6 support workers) and needs stronger numbers and training to perform the regulatory, supervisory, enforcement, and market development tasks assigned to it. The staff has had no for- mal training, or internships in securities market regulation and su- pervision in the past five years. The adequacy of funding and staffing of the Securities Board and self-regulatory organizations (that is, the NEPSE) is an important issue in Nepal. The functions of broker and dealer have traditionally been separated in order to avoid po- tential conflicts of interest. As a result, the financial services indus- try is highly segmented. Yet another concern is that the disclosure of financial statements and accounting reports is unreliable. Companies are reluctant to publish objective audit accounts because the tax inspector has wide discretionary powers. Central Market Infrastructure. Nepal has virtually no central mar- ket infrastructure to support a bond market. It has no central de- pository system (CDS), no market information system or research companies, and no credit rating agency. Nor does it have efficient economic journalism. The stock exchange could conceivably be used to trade bonds, if they were developed, but there is no credit rating agency, no central depository, and no financial research or information. Without a sec- ondary market for fixed-income securities, there has simply been no incentive to develop central market infrastructure. The only market 277 MIKAEL KVIBACK information system of any importance is found in the primary T-bill sector, where counterparts telephone one another to do a trade and settle with fairly effective transparency in real time on behalf of other participants in the market. Clearing and settlement are all done manually. Lacking market information on screens, market partici- pants must rely on other media, such as the daily financial newspa- per. In sum, the central market structure for a fixed-income market would have to be built from the ground up. Market Participants. Market participants consist of issuers, inves- tors, and intermediaries. IdtUerls. For issuers, as already noted, the corporate bond market is extremely small, with only a few privately placed issues. State-owned blue chip corporations and banks, which might be the most relevant candidates to use the market, do not use it at all. There are several impediments to issuance in the market today. The main problems are that the privatization program for state-owned companies is too slow to influence the market, the country has a large number of small- scale, garment and carpet-oriented companies that do not need and could not issue bonds, and regulations and guidelines hinder insur- ance companies and pension funds from investing freely in the bond market. And given that there is no apparent benefit of using the market, companies are reluctant to publish objectively audited ac- counts. This deprives investors and lenders of the essential basis for judging the investee company. At the same time, Nepal has some prospects for economic growth, and which would create a need for more long-term local capital. In a growing economy a number of corporations and financial institu- tions are likely to emerge as profitable and attractive issuers that need the bond market-such as hydropower, tourism, and telecom companies. For those companies, bank term lending is constrained by credit-risk concerns. Nonbanking financial institutions are re- stricted in their ability to compete with banks and in all likelihood could not come up with appropriate amounts of funds. Thus a bond market would be useful. The proposed privatization program could push development of a bond market along. For various reasons, the privatization pro- 278 NEPAL SURVEY gram launched in 1992 has been delayed, although 10 enterprises were privatized between 1992 and 1994, and 6 SOEs between 1996 and 1997. Of Nepal's 40 SOEs, approximately 20 are possible can- didates for privatization in the next five years. For this to happen, it would be necessary to develop the capital markets and to strengthen public confidence. The government is also promoting development of the private sector, and has encouraged private investments in power generation and domestic airlines. Official policy is to stimulate private finance in sectors such as telecommunications. The entrepreneurial attitude is evident in a wide range of fields. In the travel and tourism sector, approximately 40 new companies, all large by Nepalese standards, have recently been formed to establish hotels. When the transport sector was deregulated in 1993, nine new Nepalese domestic air- lines were created. Nepal could accelerate economic growth also by exploiting the potential in hydropower, tourism, telecommunications, and agricul- ture. The first three of these could create great opportunities for the private sector to develop in Nepal. The government seems deter- mined to attract private initiatives, as no budgetary funds are avail- able to undertake these investments. Investo,rs. Nepal's institutional investor base is not sufficient to sup- port an active debt securities market (see table 2). Nepal has 12 insurance companies with total assets of approxirnately US$60 mil- lion. Of these, 10 are local and 2 are joint ventures with Indian com- panies. There are two small investment funds that help mobilize the savings of small investors: Citizen Investment Trust and NIDC Table 2. Institutional Investors in Nepal Type Number Insurance companies 12a Provident fund 1l Investment fund 2 Private pension funds n.a. n.a. Not available. One of these is state owned, its hold assets amount to about US$60 mi lion, or 75% of total assets held by insurance companies. o Its total assets are In the neighborhood of US$190 million. 279 MIKAEL KVIBACK Capital Market (Nepal Industrial Development Corporation). There is one large and growing provident fund, Employees Provident Fund (EPF). However, it does not invest to any significant extent in the capital market. For example, 60% of EPF's assets are invested in bank deposits, 30% in government securities, and the balance is in syndicated bank loans, among others. There are also some private pension funds. Since these are not subject to any registration or re- porting requirement, the exact number is not known. The institutional investor base is very small partly because banks have tended to dominate the financial sector. Moreover, regulations and guidelines constrain the investments of some institutional in- vestors in the bond market. Others are willing to risk investing in long-term securities but not to trade because they know little about how to do so. There are no foreign investors in the market. Indi- vidual investors lack the sophistication or the information they need to make certain investment decisions. Since the retail sector domi- nates the market and there is no private mutual fund industry to mobilize savings toward the debt market, savings go into the bank- ing system and the NSCs. Stock brokerage firms are too small to provide ancillary services such as research. Also, resource misallo- cation has been encouraged by practices in the insurance industry and in the administration of provident funds, which has contributed to excessive liquidity in the banking sector and a shortage of long- term funds to finance private business expansion through the capi- tal markets. Given the lack of investment opportunities in the Nepalese capi- tal market, insurance companies are not active players in this mar- ket, either as intermediaries or as institutional investors. Neither they nor their subsidiaries are registered with the Securities Board as brokers, dealers, market makers, issue manager, underwriter, or portfolio manager. Investors could benefit significantly from having a market: it would enable them to diversify their investments, improve their risk management, and obtain betteryields/returns from investing in fixed- income instruments compared with bank deposits. Thus the demand side of the capital market needs to be developed, in the form of insti- tutional investment. With increased institutional investment, the professional fund management industry would be able to gain a foot- 28 NEPAL SURVEY hold in Nepal. Sophisticated institutional investors are in a position to demand quality information from issuers of securities; they can help improve the flow of information to the market. Various problems also hamper the activities of intermediaries who might help promote the market. First, with broker and dealer functions completely separated, it has been difficult to develop di- versified securities firms that can offer a full range of broker, dealer, and investment banking services to meet the needs of issuers and investors. The securities law defines the five types of services -bro- ker, market maker, dealer, issue manager, and underwriting services - and the Securities Board requires a separate license for each of the five types. Under the current regulatory framework, a broker may only perform the function of a broker, while firms engaged in the other four categories of securities activities may not act as a broker on the stock exchange. As a result, Nepalese finance companies are able to perform the relatively risky functions of market maker and underwriter, but they are precluded from taking on the less risky function of acting purely as a broker, buying and selling stock on behalf of customers. Consequently, they have chosen to perform the function of underwriter rather than act as broker on the floor of the stock exchange. Brokers may not serve as underwriters. Furthermore, the brokerage business is not as economically vi- able as a stand-alone business at the present time, owing to the low transaction volume of recent years. Since a stock brokerage firm may not perform the economic function of issue manager or under- writer, brokers have limited ability to participate in the distribution of securities in the new issues market. As a result, the primary (new issues) market has become artificially separated from secondary markets in securities, and the stock market deprived of additional potential sources of liquidity. While full service securities compa- nies should be encouraged to develop in Nepal, individual brokers should not be permitted to perform the function of both broker and dealer in the same security. This move is necessary to avoid poten- tial conflicts and possible attempts at market manipulation. In the absence of secondary markets in fixed-income instruments and greater stock market activity, competition in the nonbanking financial industry is hard to find. Although small intermediaries are in abundance, the majority of them do not have the professional skills 281 MIKAEL KVIBACK to identify issuers and investors and bring them to the fixed-income market. Although the market pricing and fees structure seems to be high enough to allow the intermediaries to make money, the volume of business transactions activities is so low that they do not earn money being a dealer. There are no market-maker functions and intermedi- aries will not take any risk in dealing even if they are able to partici- pate. To build a secondary market, intermediaries need incentives to act as market makers. With state-owned commercial banks domi- nating the intermediating sector, there is little competition from nonbanking financial institutions. And with little or no research analysis on industries and companies, few investors are encouraged to consider the local debt market. The retail sector dominates the market, and because there are not enough private mutual funds to mobilize their savings toward the debt market, those savings go into the banking system. SEVERITY OF IMPEDIMENTS As table 3 shows, there is no meaningful base for a secondary mar- ket in fixed-income instruments in Nepal at present. The govern- ment has been attempting to reform the financial sector with assistance from several donors, but progress has been slow. Enabling legisla- tion needs to be passed. Regulatory bodies remain too weak to foster market develop- ment, and the overlapping structure between NRB and Securities Board makes them even weaker. There has to be an educating pro- cess for all the market participants who will require patience and diligence. If those impediments can be overcome in a constructive way, the marketplace develop private initiative and responsibility to remove many of the other impediments mentioned in this report. RECOMMENDATIONS AND CONCLUSIONS Nepal has taken a few steps toward becoming a more efficient mar- ketplace, but it still has a long way to go. Several critical founda- NEPAI SURVEY Table 3. Summary of Key Impediments External Situation Political Weak governance institutions Incentives and private initiative to drive market developments Macroeconomic Fiscal deficit Government debt management Broader financial system Monopolized banking sector Insignificant nonbanking sector Inferior interest rate structure No Securities Board guidelines on fixed-income issuance No yield curve on govemrnment securities No secondary fixed-income market No efficient money market No efficient primary market in treasury bonds hntemal Situation Regulations and regulators A weak infrastructure of laws, regulations, and institutions Overlapping roles of NRB and Securities Board Weak auditing and accounting standards Central market infrastructure: Trading, Clearing, Settlement, Depository No critical mass of trained ad experienced professionals No credit-rating agency No central depository system (CDS) No research or information institutions Market Participants Jssuers No long-term marketable bond issuance from the govemment State-owned blue chip corporates and banks do not use the bond market Privatization program for state-owned companies too slow to influence the market Enterprise structure of small companies does not need bond issuance Interior interest rate structure sets high benchmark rates for corporate debt Regulations and guidelines hinder insurance companies and pension funds from freely investing in the bond market Investors Undeveloped mutual fund sector No experience in trading culture Guidelines hinder investment in corporate bond market Intermediaries Dominance by state-owned bank restricts competitiveness Low activity, low earnings Tiny information system No market-maker function MIKAEL KVIBACK tions for active financial markets do not exist and there are con- straints on the attributes and abilities of key market participants that would make trying to build a corporate bond market extremely difficult at this time. Given the current state, Nepal might best focus on making sure that its private placement market for corporate bonds works as ef- fectively as possible so that local corporations can reduce their risk by having access to longer term, local currency funding to the great- est extent possible. Alternative sources of finance likely will be needed by local corporations because of constraints on bank lending. To- wards that end, issuers should be educated about the benefits of relying on bond financing so they can reduce their risks; investors that need long term assets should also be educated about and en- couraged to place their funds in bonds. Regulations governing the private placement market might be adopted and a small OTC mar- ket might be established so that a mechanism exists to trade bonds. At the regulatory level, clear boundaries are needed between the NRB and the Securities Board. The NRB could act as a central bank only, with responsibility for areas such as monetary policy, payment systems, and bank supervision. The Securities Board could be the sole regulator of the corporate fixed income market, which will help reduce confusion and help increase its credibility in the marketplace. On a more general level, Nepal should continue working on its privatization program and on fundamentals such as corporate gov- ernance and accounting that are needed before public markets can be used effectively and which also support the quality of private markets. It might also consider ways to improve its government bond market; for instance, ideally speaking, any country that wishes to create a government bond market which supports the overall finan- cial system and corporate bond market should issue frequently along the yield curve. More extensive analysis is needed, however, to de- termine if and when Nepal should take such actions. Contributors JOHN BROADBENT Chief Manager Domestic Markets Reserve Bank of Australia ALISON HARWOOD Principal Securities Markets Specialist Financial Markets Advisory Department International Finance Corporation (IFC) ASSAAD JABRE Vice President Portfolio Management and Advisory Operations International Finance Corporation (IFC) A.S. JAYAWARDENA Governor Central Bank of Sri Lanka YONGBEOM KIM Associate Director Securities Policy Division Ministry of Finance and Economy Republic of Korea LENNART KONlGSON Managing Director Swedish Development Advisers Sweden CONTRIBUTORS MIKAEL KVIBACK MK Financial Consulting Sweden JOHN LEONARDO Lead Consultant Admiralty Group Partnership New Zealand RAKESH MOHAN Director National Council of Applied Economic Resources India 1MALIN NYSTRAND Swedish Development Advisers Sweden MICHAEL PETTIS Managing Director Capital Market Strategies Bear, Stearns & Co. New York MAMTA SHAH Chief Investment Officer Treasury International Finance Corporation (IFC) RANJIT SINGH General Manager Economic Analysis and Financial Policy Securities Commission of Malaysia USHA THORAT Chief General Manager Reserve Bank of India 28 THE WORLD BANK 1818 H Street, N.)A Washington, D.C. 20433, U.S.A. TELEPHONE: (202) 477-1234 FACSIMILE: (202) 477-6391 TELEX: MCI 64145 NVORLDBANK Mci 248423 WORLDBANK INTERNET: www.worldbank.org E-MAIL: books@worldbank.org Building local currency bond markets has become an increasingly important topic for emerging market countries, particularly since the Asian financial crisis. This book reviews why countries should consid- er building local currency corporate bond markets and how to evalu- ate what is needed and what might be developed. It includes eight country studies: Republic of Korea, Australia, Malaysia, India, Paki- stan, Sri Lanka, Bangladesh, and Nepal. It is based on the IFC-spon- sored South Asia Debt Market Symposium held in October 1999 in Sri Lanka, which brought together regulators and market participants from around Asia and other parts of the globe to discuss issues in de- veloping local markets. The book should provide valuable insights for countries that are addressing these issues today. *IFC I INTERNJTIuNAI F]xNcE CoPorroNIB 0-8213 -481o9-1 A- M-hil, 4k Wov It B. CrI 0 24p