67120 v1 Developing a Regulatory Framework for Municipal Borrowing in India Volume 1 September 2011 Developing a Regulatory Framework for Municipal Borrowing in India Volume 1 (Change background colour as needed) September, 2011 Acronyms and abbreviations ADB : Asian Development Bank CBI : Central Bank of India CVC : Central Vigilance Commission DEA : Department of Economic Affairs DRT Act : Debt Resolution Tribunal EMMA : Electronic Municipal Market Access GoI : Government of India HUDCO : Housing & Urban Development Corporation Ltd. IDFC : Infrastructure Development Finance Company Limited IIFC : India Infrastructure Finance Company Ltd. IL & FS : Infrastructure Leasing & Financial Services Limited JNNURM : Jawaharlal Nehru National Urban Renewal Mission LIC : Life Insurance Corporation of India MDT : Municipal Debt Tribunal MMRDA : Mumbai Metropolitan Region Development Authority MoUD : Ministry of Urban Development NHB : National Housing Board PFRDA : Pension Fund Regulator PMDO : Pooled Municipal Debt Obligation PPP : Public Private Partnership RBI : Reserve Bank of India SARFAESI Act : Securitisation and Reconstruction of Financial Assets and Enforcement of Security Interests Act (2002) SCB : Scheduled Commercial Bank SEBI : Securities and Exchange Board of India SEC : Securities and Exchange Commission SLR : Sovereign Lending Requirement DEVELOPING A REGULATORY FRAMEWORK FOR MUNICIPAL BORROWINGS IN INDIA SPV : Special Purpose Vehicle TNUDF : Tamil Nadu Urban Development Fund ULB : Urban Local Bodies UTI : Unit Trust of India Exchange rate on July 30, 2010 US$ 1 = INR 47 Study team This Report was developed by a World Bank team led by Roland White and Raghu Kesavan (Co-TTLs), which included Vasudha Sarda Thawakar (Research Analyst), Charles Jokay (Lead Technical Consultant), Sujatha Srikumar (Consultant, Powertec), Piyush Joshi (Consultant, Clarus Law Associates), and Om Mathur (Consultant). As part of the study Ms. Srikumar and Mr. Joshi authored separate detailed technical reports on the Indian Municipal Market and Regulatory issues which may be found at Annexure 1 and 2 respectively. Table of Contents Composition of Steering Committee 7 1. Situation and Problem Statement 9 1.1 Why Municipal Borrowing in India? 9 I.1.1 Objectives 9 I.1.2 The Need for Municipal Borrowing in India 9 1.2 Factors Affecting Municipal Borrowing 12 1.3 Methodology 13 2. Overview of the Municipal Debt Market 15 2.1 General Borrowing Trends 15 2.2 Types of Lending and the Supply Side 17 2.2.1 Government Institutions 19 2.2.2 Banking Institutions 19 2.2.3 Specialized Infrastructure Finance Entities 19 2.2.4 Capital Markets 20 2.3 Observations on Trends 21 3. Analysis of Indian Municipal Borrowing Regulatory Environment 24 3.1 International Context: Models for Regulation of Municipal Borrowing 24 3.2 The Regulation of Municipal Borrowing in India: An Overview 27 3.2.1 Constitutional Structure 27 3.2.2 State Regulation of Municipal Borrowers 28 3.2.3 Central Regulation of Municipal Lenders/Market 29 3.3 The Regulation of Municipal Borrowing in India: Points of Analysis 31 3.4 The Regulation of Municipal Borrowing in India: The Overall Way Forward 33 4. Policy Recommendations on Local Borrowing Regulatory Framework 36 4.1. Recommendations on Central Regulation of Lenders 36 4.1.1 RBI Master Circular on Loans and Advances 36 4.1.2 Tax Treatment of Municipal Bonds 37 DEVELOPING A REGULATORY FRAMEWORK FOR MUNICIPAL BORROWINGS IN INDIA 4.1.3 SEBI Bond Listing Requirements 37 4.1.4 Insurance and Provident Fund Regulations 38 4.1.5 Central Vigilance Commission 38 4.2 Recommendations on State Regulation of Borrowers: Ex Ante Rules 38 4.2.1 Definition of Debt 39 4.2.2 Purpose of Borrowing 39 4.2.3 Borrowing Approval 40 4.2.4 Limits on Long Term Borrowing 41 4.2.5 Security and Collateral 41 4.2.6 Guarantees 41 4.2.7 Loan Tenor Limits 42 4.2.8 Monitoring, Reporting and Database 42 4.3 Recommendations on State regulation of Borrowers: Ex Post Procedures 42 4.3.1 Short to Medium Term Recommendations 43 4.3.2 Long Term Recommendations 43 4.4 Phasing in Reforms 45 4.5 Impact on Municipal Lending 45 Annexures: (in Second Volume of the Report) Annexure 1: Developing a Regulatory Framework for Municipal Borrowing in India Annexure 2: Indian Regulatory Framework Governing Municipal Borrowing Annexure 3: List of individuals consulted Annexure 4: Data Problems Annexure 5: Bond Transactions (World Bank, WSS Study, 2007) Annexure 6: Suggestions for State-Supervised Insolvency Procedural Steps Annexure 7: Suggested General Procedures for Municipal Borrowing COMPOSITION OF STEERING COMMITTEE Composition of Steering Committee Name Organization/Designation 1. Chair: Mr. Arun K. Mehta Ministry of Urban Development/Joint Secretary 2. Co-Chair: Dr. KP Krishnan Ministry of Finance/Joint Secretary 3. Ms. Meena Aggarwal Ministry of Expenditure/ Joint Secretary 4. Dr. B. M Misra Reserve Bank of India/Advisor, Department of Economic Policy & Analysis 5. Dr. K. V. Rajan Reserve Bank of India/CGM, Internal Debt Management 6. Mr. Anjan Barua State Bank of India/CGM 7. Dr. R. K. Nair Security and Exchange Board of India/Executive Director 8. Mr. S.R Ramanujam CRISIL/ Head, Urban Infrastructure Advisory 9. Mr. Raghav Chandra Madhya Pradesh /Principal Secretary, Urban Administration & Development 10. Ms. Gauri Kumar Gujarat/Principal Secretary, Urban Development & Housing 11. Mr. Manu Kumar Srivastava Maharashtra / Secretary, Urban Development 12. Mr. Niranjan Mardi Tamil Nadu/Secretary, Housing & Urban Development 13. Mr. V. Sivasubramaniam Department of Economic Affairs/Director, Infrastructure Finance 7 DEVELOPING A REGULATORY FRAMEWORK FOR MUNICIPAL BORROWINGS IN INDIA 8 CHAPTER 1 • SITUATION AND PROBLEM STATEMENT Chapter 1 Situation and Problem Statement 1.1 Why Municipal Borrowing in India? 1.1.1 Objectives A t the request of the Ministry of Urban Development (MoUD) and Department of Economic Affairs (DEA), the World Bank has undertaken a non-lending technical assistance exercise on the regulation of municipal borrowing in India. Most of the analytical studies till date have focused on the fiscal aspects of municipal borrowing rather than regulatory hurdles. The chief focus of this report is on the regulatory and legal conditions that currently hinder, but if altered could encourage, the appropriate expansion of this municipal borrowing. The central objective of the exercise has been to develop proposals for improving the regulatory environment pertaining to municipal borrowing, most of which – given India’s constitutional structure – would need to be specifically crafted and enacted at the state level. More particularly, this report: • Outlines the need and rationale for expanding access to credit finance on part of municipalities in India (rest of Chapter 1); • Provides an overview of the existing municipal debt market (Chapter 2); • Provides an overview of the chief characteristics of the regulatory environment pertaining to municipal borrowing in India, places the existing regulatory system in international context, and outlines a suggested overall direction for reform (Chapter 3); • Provides specific recommendations to improve the regulatory regimes over which the state and Union governments have respective control (Chapter 4). 1.1.2 The Need for Municipal Borrowing in India Expectations related to municipal borrowing in India are high, whereas the actual level of borrowing activity is currently very low. As is widely known, there is a massive need for capital investment in municipal infrastructure, only part of which can be met with flows of grant funds from JNNURM and other programs. While all Indian Urban Local Bodies (ULBs), including JNNURM participants, spent some Rs 180 billion on capital expenditures in the 2007-2008 budget year1, a mere 3% of those capital expenditures were funded by estimated borrowing of Rs 5 billion during that budget cycle (Chart 1). This low level of borrowing is consistent with international experiences elsewhere when local governments are in the early stages of embarking on borrowings of their own without higher 9 DEVELOPING A REGULATORY FRAMEWORK FOR MUNICIPAL BORROWINGS IN INDIA Chart 1: Total ULB Revenues and Expenditures, Rs billion, 2002-2008 (Thirteenth Finance Commission Data) 500 450 400 350 Rs. Billion 300 Revenue Expenditure 250 Capital Expenditure 200 Total Revenues 150 100 50 0 2002-03 2003-04 2004-05 2005-06 2006-07 2007-08 government support and guarantees. It is also apparent that very substantial infrastructure spending on urban development is being undertaken by non-ULB bodies, such as Development Authorities and state government departments. In other words, the relatively low level of investment by ULBs is deceptive – much of the urban infrastructure investment is not visible in ULB borrowing and budgetary data is taking place through other entities. Notwithstanding this, the very modest level of ULB borrowing has taken place in the context of robust growth in ULB expenditures, revenues and capital investment since 2002. Both total ULB revenues and expenditures have doubled between the 2002 and 2008 budget years. Own revenues increased by a factor of 1.5, while other revenues that include transfers from States and from the Union level, nearly trebled. Data on direct 2 ULB borrowing in India is not available on an aggregate, primary-source basis3. However, the amount of total borrowing (to be detailed in later sections) during 2002-2008 can be very conservatively estimated at about Rs 23 billion, based on secondary and calculated data provided by rating agencies, major lenders, the Reserve Bank of India and other sources. This amount – about Rs 4-5 billion per year on an average between 2002 and 2008 – is almost equal to the negative net funds balance (the difference between total expenditures and total revenues) of all Indian ULBs in the period. The remaining negative balance was made up using accumulated surpluses. As seen in Chart 2, total ULB expenditures have exceeded total revenues in the period 2002-2008 by Rs 24.59 billion. This net funds gap had to be filled with borrowing, apparently not shown as a source of revenue in available budget execution data, and by using accumulated surpluses that are quite visible in the 2005-2008 period. While both the budgetary and debt data reflect all ULBs, the JNNURM cities are dominant in terms of both population and size of budget. 10 CHAPTER 1 • SITUATION AND PROBLEM STATEMENT Chart 2: Total Revenues vs. Expenditures of all Indian ULBs (Rs billion, Finance Commission Data) 500 400 300 Rs. Billion Total Revenues 200 Total Expenditure 100 Net Funds Balance 0 –100 2002-03 2003-04 2004-05 2005-06 2006-07 2007-08 The massive urban capital investment needs in India, as calculated by various Finance Commissions and expert bodies (to be detailed later), cannot be met by the historically low levels of borrowing as well as accumulated reserves and anticipated grants and transfers. For instance the Report on Indian Urban Infrastructure and Services projects an investment requirement of Rs 39,200 billion (about US$ 870 billion) over the next 20 year period. Similarly a McKinsey study on Indian urbanization projects an investment need of US$ 1.2 trillion. If even a fraction of these needs are to be met by ULBs, their access to credit finance will need to expand greatly. One way to view the municipal borrowing issue is through the prism of JNNURM, GoI’s flagship program in the urban sector. Estimated total capital expenditure needs under JNNURM are Rs 1,292.79 billion during 2007-2012. Even with the promised Rs 1,000 billion in combined Union and state grants, less than the estimated capital needs, ULBs will have to provide at least Rs 200 billion in matching funds during the same period. The anticipated share that a ULB receiving JNNURM funds and state supplements could be responsible for is up to 20% of funded project cost in this scenario. This still leaves Rs 290 billion in unmet needs using the JNNURM calculation, i.e. an annual average of at least Rs 40 billion in new borrowing per year between 2007 and 2012. As total capital investment is still a fraction of what JNNURM foreshadows, potential non-grant funding needs are obviously many orders of magnitude higher than what has been mobilized to date. Ultimately, such needs will have to be met by the municipal tax base. To a significant extent, Indian cities should be able, in principle, to rise to this challenge: They are growing rapidly in economic terms, the incomes of city residents are rising, and asset values are increasing. 11 DEVELOPING A REGULATORY FRAMEWORK FOR MUNICIPAL BORROWINGS IN INDIA Two financing approaches4 are possible: the accumulation of operating surpluses (“pay-as-you-go�) and borrowing. The former has been relied on extensively to date and, to the extent that accumulated surpluses still exist, they are likely to be quickly exhausted as JNNURM counterpart funding requirements are implemented5. Apart from such practical considerations, it is an inefficient mechanism that is unlikely to prove capable of generating funds at the scale needed for bulky infrastructure investments, and is inter-generationally inequitable. If structured appropriately, the development of a municipal debt market in India has the potential to begin to deal with these problems – as it has, to a greater or lesser degree, in countries as diverse as the USA, Poland, South Africa and Hungary. Hence the increasing interest on the part of policymakers on what can be done to unleash this potential. 1.2 Factors Affecting Municipal Borrowing In any country, four basic factors affect the size and character of municipal borrowing: a. the intergovernmental fiscal framework; b. municipal creditworthiness; c. the nature of the domestic debt market in general; and d. the regulatory framework that relates specifically to municipal borrowing. The intergovernmental fiscal framework, including shared and assigned taxes, as well as fiscal transfers, directly influences the ability of municipal entities to generate operational surpluses that could be used for debt service or investment. Changes in the intergovernmental fiscal system (e.g. in India’s case, the recent elimination of the Octroi6) directly affect the extent of operational surpluses at ULB level. Furthermore, the fiscal framework includes not just revenue, but also expenditure assignment. The scope of institutional responsibility for the provision of infrastructure obviously directly influences demand for investment funds, a part of which may be expressed in terms of demand for loans and bond issues. In addition to revenue and expenditure issues, municipal creditworthiness also depends on the quality of local accounting and financial management systems, on the availability of reliable financial data, on the human resources responsible for running the local governments, and on the political stability and leadership of the local government system. The general domestic debt market for both term loans and bonds is an exogenous factor. The availability of capital for investment in municipal projects is partly determined by relative considerations such as the business plans of SCBs, both private and public, as well as market opportunities in other sectors that distract attention and potential funding from investment in ULB loans and securities. In India’s case, domestic debt market players may find the municipal sector relatively uninteresting, given a. the relative magnitude of India’s state and Union government borrowing; b. the existence of public and quasi-public lending institutions giving soft loans; 12 CHAPTER 1 • SITUATION AND PROBLEM STATEMENT c. strong regional banking institutions with other financial interests7. In 2008, “local and other quasi-governments,� as they are called by RBI, accounted for only 5% of lending flows from SCBs to the public sector in India. Other public and quasi-public entities, such as development authorities and such may also borrow, but their statistics may appear in State balance sheets, especially if loans from multilateral institutions such as ABD are lent to entities with State guarantees, and these funds are then on-lent for infrastructure projects. In these cases, neither the borrowing nor the assets being “created� appear on ULB balance sheets or budgets. The regulatory framework relating specifically to municipal borrowing comprises four interrelated factors: i. the ex ante rules and procedures governing ULB access to credit finance and the origination of debt; ii. rules and regulations pertaining to investors in municipal risk; iii. information on, and monitoring of, municipal debt; and iv. events in the instance of default on debt service obligations and municipal fiscal distress (often referred to as “municipal bankruptcy� in the most extreme cases). Factors (a), (b) and (c) have all received significant analytical attention in India in recent years. This study focuses specifically on factor (d) – the regulatory framework. The fact that it does so should not be taken to imply that it is more important than any of the other three. If municipal borrowing in India is to expand substantially over time, improvements will need to be made at all levels. However, the scope of the current report is quite specific – after providing some general information on the nature of the municipal debt market in India, it focuses in particular on the issues relating to the regulation of municipal borrowing and suggests both a general framework and specific measures for introducing concrete and tangible improvements in this area. 1.3 Methodology The current study used a blend of methodologies. The analysis is based on local research on borrowing patterns and the legal framework as well as fieldwork in four key states, Madhya Pradesh, Gujarat, Maharashtra and Tamil Nadu. Interviews with stakeholders at the Union, state and ULB level, local research on the borrowing environment, as well as an exhaustive review of laws in multiple states, is included in two detailed reports (Annexure 1 and Annexure 2). Dialogue with the Steering Committee overseeing the work and interviewees (see Annexure 3) served as the basis for many of the conclusions and recommendations presented in the report. The legal and financial reports, as well as interviews and extensive interaction among an Indian and international team led to a set of policy prescriptions that form Chapter 4 of this document, while other detailed findings and data are available in the Annexure 4 – 7. The general conclusions regarding the market situation and regulatory framework, combined with the recommendations to be considered at both State and Union levels, form the core of the report. 13 DEVELOPING A REGULATORY FRAMEWORK FOR MUNICIPAL BORROWINGS IN INDIA End Notes 1 Calculations based on Bank staff calculations using published Finance Commission data. 2 This does not include borrowing by other bodies such as port and road authorities or development entities that could indirectly support ULB services. 3 RBI does not track direct lending to ULBs, only to what it calls, since 2008, �local and quasi-government�. As of March 2008, only 17% of SCB lending went to the public sector, and less than 1% to the local and quasi-local governments that cover a lot more than just ULBs. Banks do not have to report direct lending to ULBs as RBI does not have such a category. States do not have consolidated balance sheets for ULBs and ULBs do not report borrowing as a source of revenue on a consistent basis. Hence any estimation of total ULB lending can only be approximated by combining sources of information such as rating reports, and extrapolating from bank disclosure statements. An alternative would be of course a detailed analysis of major ULB annual financial reports, but that may be misleading, as there are inconsistencies in the accounting practices used. See: http://www.rbi.org.in/scripts/PublicationsView.aspx?id=11422 4 Besides fiscal support from higher levels of government and Public Private Partnerships (PPPs) 5 The anticipated effect of JNNURM is that it will lead to massive new borrowing as estimated by numerous studies such as the FIRE-D project, the ADB and The World Bank�Market Finance for Local Service Delivery�, (known as WSS study). For example, the WSS study calculates a borrowing need of Rs 140 – 190 billion for 23 JNNURM cities. Of 240 Detailed Project Reports (DPRs), only 25 indicated any probability of borrowing. FIRE-D (2008) calculated that JNNURM is to sponsor projects with a total cost of Rs 1,205.36 billion, of which Rs 530 billion comes from central grants Rs 200 billion from state grants, leaving cost sharing at the ULB level of Rs 460 billion over 15 years. The ADB (Crisil, 2006 and 2007) estimated that with a state cost sharing of 10% – 20%, in the 2007-2012 period, an additional Rs 56 billion would need to be borrowed, with Rs 350 billion to be borrowed overall by ULBs during full implementation of JNNURM. 6 The reference to Octroi is only to illustrate the impact on local revenue sources, though otherwise Octroi is considered an inappropriate local tax due to the various distortions it introduces. 7 For example, RBI data on sectoral deployment of gross bank credit as of March 2009 indicates that banks have a gross non-food lending stock of Rs 26,023 billion. Estimates of total municipal borrowing not accounted for separately in RBI data, including bonds issued, amount to about Rs 30 billion, or less than one tenth of a percent of bank lending by stock. In other words, ULBs are a barely detectable client for overall bank lending. See http://rbi.org.in/scripts/AnnualReportPublications.aspx?Id=927. 14 CHAPTER 2 • OVERVIEW OF THE MUNICIPAL DEBT MARKET Chapter 2 Overview of the Municipal Debt Market 2.1 General Borrowing Trends T erm lending and municipal bonds are the visible, direct forms of lending to ULBs in India. Loans to non-ULB entities that end up financing “municipal� infrastructure projects in the roads, transport, water, etc. sectors are not counted as direct lending to ULBs, since both the asset and the liability show up on the balance sheet of an entity outside that of the ULB upon whose territory the investment is being made. Term lending consists of loans from Government sponsored institutions or funds (part publically owned, part privately owned) such as HUDCO, PMDO or TNUDF, and commercial loans from Scheduled Commercial Banks (nationalized, state-owned, private). Municipal bonds encompass the purchase of municipal securities by many of the above institutions, including SCBs, as well as by other capital market players such as institutions with long-term investment horizons. The aggregate level of ULB borrowing for all of India is low. Local government borrowing patterns are presented below, with the caveat that SCB flows had to be extrapolated from RBI data, as ULBs are not shown as a separate class of borrowers. The borrowers constitute a range of larger Chart 3: Summary of Annual Lending Flows to ULBs 1997-2009 (Rs billion) 6 5 4 HUDCO Rs. Billion 3 Bonds RBI SCB flows (est) 2 Estimated annual flow 1 0 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 Source: Compiled by authors, HUDCO and RBI data as reported by July, 2009. Excludes Mumbai’s MMRDA loan of Rs 28.02 billion 15 DEVELOPING A REGULATORY FRAMEWORK FOR MUNICIPAL BORROWINGS IN INDIA and smaller ULBs: for example, less than half of recent HUDCO borrowers are JNNURM cities. Borrowers are few and far between, as only 30 cities in India have ever accessed the bond market or commercial borrowing by some estimates, (as indicated by the bond issue table in the Annexure 5). All data for 2009 are estimated, except for HUDCO, which disclosed its ULB lending for that year. Chart 3 is based on municipal-level data from four states (Gujarat, Maharashtra, Tamil Nadu and Madhya Pradesh) and aggregate data for all of India compiled to the best extent possible, given data availability constraints. Chart 3 shows several trends that may not necessarily be related to each other in a causal manner: • Total estimated lending flows increased steadily from 2000 to 2006, with a flat tendency at about Rs three billion per year afterwards. • Strong growth in bond issues was evident between 2000 and 2005, with no new issues in 2006, 2008 and 2009, and only one small issue in 2007. Bond issuances have virtually come to a standstill except for sporadic issues by Ahmedabad Municipal Corporation. • Unfavorable changes in prevailing interest rates and the regulatory cap on tax-free yields reduced the attractiveness of tax-free municipal bonds after 2006. Despite some 35 rated municipalities, the existence of “investment grade� potential issuers does not seem to inspire any activity in the bond market starting in 2006 8 . • Lending by SCBs shows healthy and consistent growth from 1997 to 20089, with an estimated flat line from 2008 onwards. • HUDCO’s importance as a lender seems to be declining, although it appears to be the most persistent lender to the ULB sector. In FY 2008, HUDCO lent Rs three billion to Bangalore, distorting the otherwise strong downward HUDCO trend line. • State-owned banks are still lending to ULBs, but their ULB-specific data could not be separated in reports they file on an aggregate basis with RBI. As part of the study, an assessment of the potential borrowing capacity of nine ULBs was undertaken, based on financial parameters such as revenue profile, expenditure profile, operating surplus/deficit, current debt levels and finance charges (see Annexure 1 for details). The cities were selected among the JNNURM cities and represent a mix of larger cities that are likely to have a potential for borrowings. It is based purely on past financial performance and does not factor in any potential improvements in revenue performance on account of reforms currently underway, and is meant to provide a sense of which municipal governments are in a position to raise additional debt to support on-going capital expenditures programs in the four states analyzed in this study. It should be noted that the assessment is not a reflection of creditworthiness10, but merely measures borrowing capacity on a conservative basis. In addition, the regulatory limits set out for municipal government borrowing in accordance with prevailing legal jurisdictions are compared with estimated borrowing capacity to understand whether the current regulatory regime acts as a deterrent for additional municipal borrowing. The assessment concluded that most of these ULBs are borrowing well below their projected capacity (as demonstrated in Table 1), and that statutory limits, if they are 16 CHAPTER 2 • OVERVIEW OF THE MUNICIPAL DEBT MARKET operationalized at all, are mostly irrelevant to borrowing decisions. Table 1 shows that the cities of Rajkot and Bhopal have already exhausted their borrowing limits assessed on the basis of past financial performance. Therefore, these cities would only be able to borrow additional amounts based on substantial improvements in their financial profile linked to implementation of reforms. Otherwise they would be able to borrow limited amounts only after existing debts have been extinguished. Table 1: Estimated Borrowing Capacity Estimated Actual Debt Additional Statutory Actual Estimated Rating Capacity Capacity Limit Debt/ Capacity/ Estimated Statutory Capacity Limit Figures in Rs million % % Ahmedabad 7,000 4,940 206 70 CCR A+ Surat 10,500 240 1,026 2.28 CCR AA– Rajkot Nil 240 –24 NA CCR A– Nagpur 3,000 1,420 158 11,000 53 27 CCR A M.C. 43,000 28,020 1,498 111,000 65.1 39 AA Mumbai Bhopal Nil 1,790 –179 5,760 NA 0 BBB– Indore 900 1,070 –17 8,250 119 11 BBB Chennai 3,000 955 204.5 32 Ir BBB+ Coimbatore 1,000 660 34 66 Ir BBB+ 2.2 Types of Lending and the Supply Side Short-term balance sheet lending dominates the Indian banking sector. For this reason, SCB loans to ULBs tend to have tenors of 3 – 7 years, with longer loans being offered by HUDCO (up to 15 years), or LIC and IIFC (5 – 10, up to 15 years). State guarantees, though less common, are still required by HUDCO and by the dominant state-owned banks within a region. They insist upon a variety of security mechanisms, depending on their risk profile and their relationship with both the state organ that grants permission, and the borrower whose accounts they often manage. HUDCO appears to be slowly withdrawing from the municipal lending market. Others such as IDFC and IL & FS tend to rely on “private� transactions, that is, lending to special purpose vehicles and so on, but not directly to ULBs unless it is through special intermediation mechanisms such as the pooled municipal debt obligation (PMDO) credit facility. One of the characteristics of municipal borrowing in India is that true project finance, in which lenders can access only the revenue flows of the project itself and 17 DEVELOPING A REGULATORY FRAMEWORK FOR MUNICIPAL BORROWINGS IN INDIA cannot reach into the municipal balance sheet, does not exist at the municipal level (in other words, while security mechanisms such as escrow accounts are commonly found, in the event of their failing the full balance sheet of the municipality is technically available for the satisfaction of debt claims). As a consequence of the well-known inability of India’s courts to enforce commercial contracts timeously, and the aversion of courts to attach municipal revenues and/or property, market players have developed several forms of security to avoid reliance on the legal system. In some cases, lenders such as HUDCO insist on State guarantees for their municipal loans, despite difficulties in enforcing them11. In others (and increasingly frequently), lenders rely on escrow accounts, enabled by legislation throughout India. These accounts are closely monitored by all stakeholders, and allow the lender to seize funds from the account if payments are not made. Certain revenues flow to the escrow account, and borrowers are committed to keeping the accounts “full� up to the agreed amount12 . However, the enforceability of these escrow arrangements in the event of municipal fiscal stress is yet to be seriously tested. For example, these escrow accounts capture flows from user fees. Given these factors, banks that have an existing banking relationship tend also to provide short- term loans to urban local bodies (e.g. Bank of Maharashtra). The existing banking relationship primarily comprises the provision of banking accounts to the municipal government, banking accounts to the employees of the municipality and ancillary services to the employees (retail loans – personal, home, car, etc.). This enables the banker to have constant oversight on the daily volume of cash flows in the banking accounts of the municipal governments, the key vendors and the customers of the municipal government. This provides confidence to the banker and mitigates risk perception arising out of any short-term credit exposure. Banks that do not have this relationship with the municipal government are not in a position to assess the municipal cash flows accurately, given the weak accounting systems. Approximately 30 bonds have been issued by 14 ULBs since 1999. Their tenors vary between 5 and 15 years, with fixed interest rates, and about half of the issues are tax-free. The average tenor of the 25 issues, examined by the World Bank Water and Sanitation Program South Asia report in 2007, is 8.2 years, with the average fixed interest rate of 9.66% (declining from issues at 14.75% in 1999 to some at 5.9% in 2005), and an average issue size of only Rs 700 million – refer to Table in Annexure 513. Various types of institutions are active in financing municipal infrastructure projects in India: • Government institutions, established, owned and overseen by the public sector (HUDCO, LIC); • Scheduled commercial banks (public sector banks, either nationalized or owned by a government agency, Indian privately-owned banks, foreign banks licensed in India); • Specialized infrastructure finance entities that are privately owned or have only partial or indirect public ownership (IIFCL, IDFC, IL & FS); • Sector-specific municipal development funds (PMDO, TNUDF); • Capital markets. 18 CHAPTER 2 • OVERVIEW OF THE MUNICIPAL DEBT MARKET 2.2.1 Government Institutions Housing & Urban Development Corporation Ltd. (HUDCO) was incorporated as a fully-owned Government Company with the main objectives of (i) financing housing and urban development projects, (ii) financing building material industries, and (iii) setting up of new townships. In urban infrastructure financing, HUDCO’s lending is skewed toward the energy and commercial sectors since 2002. This is due to diminishing ability to extend state government guarantee as a security of the borrowing agencies like ULBs, Water Supply and Sewerage Boards and non-competitive interest rates. HUDCO has to borrow from banks and financial institutions at the prevailing market rates. The Life Insurance Corporation of India (LIC) is the largest life insurance company in India and also the country’s largest investor. LIC exposure as loans and advances to various entities for infrastructure and social purpose engaged in the water and sewerage sub-sector has been less than 0.50%. Prior to the economic liberalization of 1991, LIC was providing loans to urban local bodies and statutory boards for water supply and sewerage against state government guarantees. Over time, LIC’s incremental lending to this sector has come down. LIC once faced significant recovery problems on its exposure to the urban sector. 2.2.2 Banking Institutions Currently, India has 88 Scheduled Commercial Banks (SCBs) – 27 public sector banks (that is, with the Government of India holding a stake), 31 private banks (these do not have government stake; they may be publicly listed and traded on stock exchanges) and 38 foreign banks. SCBs with state or GoI ownership are the most active. Especially dominant are the relationship banks in each state, such as Bank of Maharashtra, who go beyond normal commercial services and provide advisory services due to the strong relationship between their owner and the state administration. Term lending and short-term cash flow lending seems to be based on local banking relationships, where risks are addressed by an intimate knowledge of the borrower’s cash flow and its ties to suppliers and major taxpayers. Those without intimate local knowledge have to resort to pricing and security mechanisms that may be unacceptable in comparison with these relationship banks. 2.2.3 Specialized Infrastructure Finance Entities Several specialized development funds and facilities finance local infrastructure projects and have some exposure to ULBs, although the widely-held preference is to lend to SPVs and private infrastructure providers, and not directly to ULBs: • India Infrastructure Finance Company Ltd. IIFCL was established in January 2006 as a wholly-owned Government of India company and commenced its operations from April, 2006. As of April 30th, 2009, Rs 183.4 billion has been sanctioned to 86 projects in the transport, power, and urban infrastructure sectors. • Infrastructure Development Finance Company Limited (IDFC) was set-up as a company focused on development and financing of private infrastructure. Government of India 19 DEVELOPING A REGULATORY FRAMEWORK FOR MUNICIPAL BORROWINGS IN INDIA earmarked an amount of Rs 10 billion as its contribution to this company. IDFC was conceived as a Public Private Partnership, with GoI as a 40% equity shareholder. While in the initial years telecom was the mainstay, the portfolio gradually shifted to a higher quantum of assets in energy and transport sectors. • Infrastructure Leasing & Financial Services Limited (IL&FS) was established by the Central Bank of India (CBI), Housing Development Finance Corporation Limited (HDFC) and Unit Trust of India (UTI). Over the years, IL&FS has inducted many institutional shareholders. IL&FS has a distinct mandate – catalyzing the development of infrastructure. IL&FS has conceived and promoted a pan-India facility for financing urban infrastructure, the Pooled Municipal Debt Obligation Facility. • Pan-India Pooled Municipal Debt Obligation (PMDO) Facility was established to address the gap in funding requirements by providing financial assistance to the urban sector for developing infrastructure. IL&FS, in partnership with IDBI, IIFCL, Canara Bank and eleven leading domestic banks, launched a Rs 30 billion Pooled Municipal Debt Obligation Facility. • State specific Tamil Nadu Urban Development Fund (TNUDF): TNUDF is the first Public Private Partnership between the Government of Tamil Nadu and three financial institutions (ICICI, HDFC, and IL&FS) for providing long-term debt for civic infrastructure in a non-guarantee mode. TNUDF was set up with the following objectives, namely to fund urban infrastructure projects which improve the living standards of the urban population, facilitate private sector participation in infrastructure through joint venture and Public Private Partnership, and improve the financial management of urban local bodies. TNUDF is currently implementing Phase III with World Bank assistance of US$ 300 million (Rs 14.1 billion). This assistance comprises capital grants and loans to ULBs. 2.2.4 Capital Markets The overall capital market in India has two main segments: government and corporate securities. The government securities market consists of central and state government securities, while the corporate securities market consists of FI bonds, PSU bonds and Debentures/Corporate bonds. Government securities form the major part of the market in terms of outstanding issues, market capitalization and trading value. During 2007-08, the government and corporate sector collectively mobilized Rs 3722 billion from the primary debt market – a rise of 27% as compared to the preceding year. About 69% of the resources were raised by the government (central and state governments), while the balance was raised by the corporate sector through public issues as well as private placement. Municipal bonds are visibly in only a very small portion of the capital market in India (Chart 4). The demand for municipal bonds has become muted since 2005 as indicated above. There have been some unsuccessful attempts at bond issuances. For example, Nagpur attempted a bond issuance in 2007 worth Rs 1.28 billion but was able to obtain commitments only for Rs 210 million, primarily 20 CHAPTER 2 • OVERVIEW OF THE MUNICIPAL DEBT MARKET Chart 4: Bonds Mobilized (Rs million) by Municipal Governments 3,500 3,000 2,500 Rs million 2,000 1,500 1,000 500 0 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 on account of poor market timing, due to which the potential investors were unable to utilize the tax-free benefits. Similarly, Indore attempted bond issuance in 2002-03 for Rs 500 million, but could receive commitments only for Rs 37.2 million due to concerns on credit quality, despite the rating. 2.3 Observations on Trends As indicated earlier, this study does not aim to provide a thorough analysis of overall municipal borrowing trends in India. However, a few brief observations on the patterns identified above may be helpful: • While it would be speculative to assert a causal relationship between the introduction of JNNURM and the diminution of municipal borrowing activity in India since around 2006, the co-incidence of the two is striking. Field observations, as well as some of the early results of the various assessments being conducted on JNNURM, indicate that the program is largely being implemented without de facto insistence on ULB funding requirements. Regardless of whether these counterpart funding requirements and levels constitute good policy, it is arguable that the rapid increase in grant funding available to JNNURM cities, which are also the biggest potential municipal borrowers, has crowded out any effective demand for credit finance, particularly in an environment where technical capacity to prepare and implement bankable projects is limited; • Changes in the intergovernmental fiscal structure – such as the elimination of a local tax, have a negative impact on the own source revenue position of LGUs, and hence of their debt service capacity. It is not yet clear what the implications of the potential 21 DEVELOPING A REGULATORY FRAMEWORK FOR MUNICIPAL BORROWINGS IN INDIA introduction of a Goods and Services tax (GST) may have on this situation, but it could be significant; • Underpinned by attenuated service-delivery accountabilities deriving from the fragmented nature of the local government institutional structure in India, there seems to be an unstated reluctance on the part of ULBs to engage with the lending process in order to fund project expenditures. This is compounded by a lack of capacity within the treasury function in ULBs to engage effectively and confidently with lenders, or operate in a decision-making modus operandi that is compatible with the private sector. On the lender side, this is matched by a perception that the municipal sector is small, opaque and politically risky relative to other sectors where loan origination effort does not need to be as great; • Some lenders enjoy unique advantages in providing capital to the municipal sector, while other investors encounter explicit regulatory requirements and other obstacles that make municipal assets less desirable. Examples include government or state- owned lenders and banks that have considerable experience with municipalities, and enjoy the benefit of being considered exempt from public tendering. • Despite the low levels of municipal borrowings, all the initiatives in commercial financing of urban infrastructure in the country, such as Municipal Bonds, Pooled Finance Development Scheme and the Pooled Municipal Debt Obligation Facility have used innovative credit enhancement mechanisms to address specific investor concerns about liquidity and difficulties in enforcing security against ULBs. While all the municipal bonds structured so far are basically general obligation bonds, they have been supported by escrow of specific revenue streams, either from a project or from a tax source, in order to provide liquidity and ensure timely payments against the bonds. A charge/intercept on cash flows  in addition to physical asset has been used to provide comfort to lenders in the Pooled Municipal Debt Obligation Facility. Project finance transactions involving SPVs promoted by urban local bodies in PPP arrangements have also been financed by the Pooled Municipal Debt Obligation Facility. The Pooled Finance Development Scheme of the Government of India is based on the State Revolving Funds concept in the US.  Insuchcases,boththecentralandthestategovernmentprovidecashcollateraltocreditenhance a bond offering backed by cash flows from projects of many small municipalities. End Notes 8 Creditworthiness as demonstrated empirically or expressed in terms of an investment-grade rating does not seem to be directly related to a demonstrated willingness to borrow. This was pointed out by several studies, including one by the ADB in 2007, as well as by FIRE-D on several occasions 2008-09. 9 RBI has indicated that they are neither obligated nor really motivated to keep track of the ULB sector as a separate category within the general lending to “public bodies� aggregate (that of course includes water boards, state governments, etc.) and for this reason, ULB lending by SCBs is difficult to plot. 10 It is assumed that the average of current surpluses calculated for the most recent three years for which the data is available 22 CHAPTER 2 • OVERVIEW OF THE MUNICIPAL DEBT MARKET will continue, and that half of these surpluses are available for servicing future debt obligations, apart from meeting existing debt charges (computed on the basis of three year averages). Half of the operating surpluses are assumed to be available for debt servicing since this implies a Debt Service Coverage Ratio (DSCR) of 2. 11 In fact, the security required by HUDCO generally comprises of: (i) State guarantee, (ii) escrow account, (iii) charge over municipal revenues and identified municipal bank accounts; (iv) mortgage of specific municipal immovable property. See the legal report in Annexure 2. 12 Some of the State municipal laws have been amended to specifically allow for the creation of special purpose escrow accounts to trap identified revenues only for the purposes of providing security to the identified loan(s) for which they are created. Escrow accounts differ from sinking funds in the sense that, unlike a sinking fund, an escrow account is not dependant on a separate payment being made by the municipality into a demarcated account. An escrow account captures the identified revenue stream directly (i.e. all identified receivables are directly deposited into the escrow account) and after maintaining a specific cash reserve, the revenue can be transferred to the municipal account. However if there is a default by the municipality in the repayment of the loan, the escrow account starts to trap the identified municipal revenue (without releasing any monies into the general municipal fund) and makes direct payment from it to the lenders, till such time as the default is cured. The flow of revenue from the escrow account into the general municipal fund commences only after the relevant default has been cured. Thus, unlike a sinking fund, an escrow mechanism is not dependent on the municipality and is not under its control. A separate bank appointed as the escrow agent manages the escrow account. 13 For details on a generic lending transaction, please see Attachment A of the Credit Market Report (Srikumar, December 2009) and Attachment D of the same report for details on the Visakhapatnam Municipal Corporation bond issue of 2004. Both are to be found in Annexure 1 to this Report. 23 DEVELOPING A REGULATORY FRAMEWORK FOR MUNICIPAL BORROWINGS IN INDIA Chapter 3 Analysis of Indian Municipal Borrowing Regulatory Environment 3.1 International Context: Models for Regulation of Municipal Borrowing T he regulatory environment pertaining to municipal borrowing in India can be assessed within a conceptual framework built on international practices. This sub-section of the report briefly summarizes the major characteristics of the different approaches that can be found internationally and their relative dis/advantages. The following sub-section provides an overview of the key features of the Indian system, places India in the international context, and outlines a suggested approach to the overall direction of policy reform. This provides the basis for the specific proposals made in Chapter 4. Three main approaches to the regulation and control of municipal borrowing can be identified internationally14: • “Market based�, where decisions about municipal borrowing are made by the borrowers and lenders within an overall legal framework and some level of administrative oversight, but without transaction specific higher-level authorization or detailed rules regarding the amounts and terms of borrowing transactions; • “Rules based�, where decisions about borrowing are made within a more tightly circumscribed set of parameters outlined in a detailed set of rules that are constant. Higher-level approval of specific transactions may be required, but this is largely limited to compliance with the rules themselves, rather than the underlying merits of the transaction, or the investment that it is funding15. This is based on a comprehensive definition of debt (lacking in India) and consistent accounting practices; • “Direct control� systems, where the emphasis is on the ad hoc approval of specific municipal transactions by higher levels of government, which have extensive discretionary powers in respect of the approval process. These approaches16 are elaborated below. First, a few qualifiers may be helpful: • To clarify the use of the term “rules based�, all systems of regulating municipal borrowing naturally have “rules� of some kind or another. The differences between them lie in what the rules are, how they are used, the processes through which the rules are exercised, the respective powers of different levels of government in decision- 24 CHAPTER 3 • ANALYSIS OF INDIAN MUNICIPAL BORROWING REGULATORY ENVIRONMENT making about borrowing, and the implications for risk allocation. The key question is whether the rule can be objectively operationalized and used in practice; • No country in the world represents any one of these systems in its purest form. Rather, the three systems are best understood as tendencies or points on a triangular spectrum. The actual systems that can be found in specific countries comprise a concrete mix of these approaches to a greater or lesser degree (see Chart 5); • Approaches to municipal borrowing evolve in particular contexts and are not divorced from the broader intergovernmental structure, financial system, and legal- administrative traditions of the countries within which they emerge. The borrowing powers of local governments are likely to reflect broader constitutional and legal arrangements regarding the level of general autonomy of local governments, for example, and the evolution of municipal borrowing frameworks will be circumscribed by these legal and institutional realities at any given time. Market-based systems: The defining characteristic of a market-based system is that, while there may be some general rules relating to the purposes of borrowing and the process through which debt is issued, the quantum of municipal borrowing, the transactional decision and the details of the loan structure are left to the lender and borrower, and higher-level government authorization is not required. To the furthest extent possible, risks are thus fully allocated to the lender, borrower and other market participants (bond insurers, for example) and do not fall on higher levels of government. While there may be some limits on the purposes of borrowing, a market system by definition means that municipalities have full autonomy to borrow. This does not prohibit market systems from requiring competitive bids, perhaps popular referenda for bond issues or the imposition of special, dedicated local taxes. The opinion of outside experts, such as bond counsel in the United States or rating agencies, may also be required as a part of the process. So far as default situations are concerned, market systems, or systems with mostly market features, must rely either on administrative or court-supervised default and bankruptcy procedures (examples include the US and South Africa). The chief advantages of a market-oriented approach are that the system allows for flexibility (borrowers and lenders are in the best position to decide on details such as the appropriate quantum of borrowing, the most advantageous loan structures, and so on) and for transactional efficiency. Moreover, because risk is borne by the lenders and borrowers and neither ex ante nor ex post approval is required, moral hazard on the part of higher level governments is minimized. However, for such a system to work effectively, important conditions need to be present: markets should be free and open, information on the finances of borrowers must be freely available (i.e., accounting and financial management standards should be high), municipalities need to have sufficient capacity to engage with lenders on equal terms, and there should be no possibility of bailouts (if there are, the moral hazard advantage collapses). Thus there also needs to be a clear institutional procedure in the case of default, which allows a certain minimum level of public service delivery to be maintained, while observing and protecting the rights of creditors. 25 DEVELOPING A REGULATORY FRAMEWORK FOR MUNICIPAL BORROWINGS IN INDIA Rules-based systems: In rules-based systems detailed standards and limits are specified in laws and derivative regulations regarding the quantum and character of the borrowing, the usage of debt, debt authorization procedures, counterpart funding requirements (in some cases), and so on. Often these rules attempt to mimic market discipline through placing limits on the indebtedness of municipal governments by linking them to some or other measure of their debt-service capacity. Higher level authorization may or may not be required. However, where it is (e.g. in Serbia), the authorization process is confined to a certification of whether the particular transaction being scrutinized complies objectively with the rules as stated, and does not extend to an assessment of the merits of the proposed borrowing or the specific investment that it may be funding. Deadlines are clear and observed. Where authorization is not required (e.g. Poland, Hungary), the investor runs the risk of ensuring that the loan is in full compliance with the rules as stipulated: where a prudency rule or other legal requirement is ignored, investors stands to lose their claims if the transaction is later challenged in a court or as part of a default procedure. The chief advantages of rules-based approaches are that they are transparent and even-handed, and provide an environment in which both investors and borrowers can relatively easily gauge transaction risk (e.g. the likelihood of a loan approval and the time it will take). The main disadvantages are that they lack flexibility (a rule that is suitable for a small, poor jurisdiction is unlikely to be appropriate for a large, rich one) and they often foster the development of behaviors and practices aimed at circumventing the rules that have been established (e.g. the use of debt instruments such as sale and leaseback arrangements that are not included in debt limits). There is also often a residual murkiness regarding the true meaning of higher level approval and the extent to which it implies some sort of underwriting or guarantee. In order to avoid these problems, rule- based approaches need to stipulate meaningful criteria regarding the things they wish to regulate (for example, the level of municipal borrowing relative to debt service capabilities), comprehensive definitions of debt and provisions designed to limit the scope for off-budget operations, and provide clarity on the exposure of higher level governments in situations of municipal default (preferably to limit it altogether by eliminating implicit guarantees). As with market-based approaches, rules- based approaches usually have in place administrative and/or court-supervised procedures to deal with non-payment, debt restructuring, and the maintenance of vital services. Direct control systems: Direct control approaches involve the exercise of direct discretion by higher level governments (central governments in unitary systems, often state/provincial governments in federal systems) over the borrowing activities of municipal jurisdictions. This may take a number of different forms but is most often characterized by the review and authorization of individual borrowing operations, including the specific terms and merits of the transaction, often on grounds that are ad hoc, variable or unclear. In a number of countries, direct control systems exist de facto although the system, on the face of it, is supposed to be rules-oriented. Such cases arise when the rules supposedly governing municipal borrowing are routinely ignored by higher level governments during the approval process (often because the rules themselves are poor and have little effective relationship to the risks they are supposed to be mitigating), and in reality, a process of ad hoc decision-making that lacks transparency and procedural and/or substantive clarity arises. 26 CHAPTER 3 • ANALYSIS OF INDIAN MUNICIPAL BORROWING REGULATORY ENVIRONMENT Direct control systems also tend not to have clear or institutionalized work-out procedures regarding municipal default. Typically, ad hoc processes, overseen by a higher level of government, emerge when such situations arise and deals are struck (usually behind closed doors) on a case-by-case basis. As is discussed later, this is predominantly the situation that has arisen in India. For the moment, it is important to note that the only real advantages of a direct control system relate to the management of macro-economic risks or sovereign borrowing strategies that typically arise in the context of the borrowing activities of states in federal systems. In a municipal context, where such concerns are not really germane, direct control systems have little up-side: they can be subject to intense political lobbying, they greatly magnify transaction risk (as neither lenders nor borrowers can be sure of how loan applications will be evaluated or the timing of the process), they involve higher level governments in micro-decisions that are best left to local governments, and they are conducive to moral hazard (as higher level substantive approval is invariably taken to imply some sort of underwriting of the loan transaction). Systems in specific countries: As indicated earlier, most countries where municipal borrowing is permitted, involve some sort mix of the three “ideal type� systems outlined above. Moreover, systems evolve over time. The United Kingdom, for example, went from a rigid central government controlled and dictated municipal borrowing system to a rules-based system in 2003. Previously, each municipal borrowing was approved at the cabinet level by a Deputy Prime Minister, with nominal borrowing limits determined by a central government department. After 2003, this changed to a “prudential� system where a departmental secretary of state (Communities and Local Government) is to approve each borrowing, as long as it is considered prudent. The regulatory framework of local government borrowing has also been developed over a decade-long-period in the European transition countries and continues to evolve. Chart 5 provides a diagrammatic overview of the location of different countries with respect to the three systems, and places the Indian system in this context. For this to be meaningful, it is first necessary to briefly outline the chief characteristics of the regulation of municipal borrowing in India, as it currently exists. 3.2 The Regulation of Municipal Borrowing in India: An overview 3.2.1 Constitutional Structure The constitutional structure in India, and particularly the 74th Amendment, provides the basis for the way in which borrowing by municipalities, and the activities of lenders in the municipal market, is regulated17. In essence: • States governments have full powers to regulate the activities, roles, powers, duties and financial activities of municipalities in their territory, including all municipal borrowing activity, even though pre-independence legislation such as the Local Authorities Loan Act of 1914 is still in force in those areas (not many) where relevant State legislation has not superseded it. In other words, all aspects of ex ante 27 DEVELOPING A REGULATORY FRAMEWORK FOR MUNICIPAL BORROWINGS IN INDIA municipal borrowing regulation – such as the purpose, authorization, limits, financial characteristics, tenor, securities offered, sources of repayment, accounting, reporting and local government disclosure related to borrowing or bond issues – depends on State-level regulation and procedures. These powers also include potential ex post State administrative intervention in case of municipal default, as well as reorganization and debt work-out arrangements. • The Union government is granted the power to regulate lenders (banks) and lending instruments through the Reserve Bank, the financial markets through SEBI, the taxes regime surrounding municipal investment through the Ministry of Finance (e.g. guidelines for tax free bonds), and the issuance of government securities18 . • In addition, a number of Union-level laws and regulations, some of which have state- specific regulations, have (or could have) an impact on municipal borrowing. The most important of these pertain to procurement and anti-corruption (CVC) and debt recovery (SARFAESI and DRT)19. 3.2.2 State Regulation of Municipal Borrowers Ex ante regulation: Annexure 2 provides a detailed outline of the State regulations pertaining to the borrowing activities of municipalities in each of the four states that were the focus of this study. As is indicated there, the nature of the regulations pertaining to municipal borrowing varies somewhat from state to state. In broad terms however, a number of general features can be identified: • Borrowing approval: In all states, ULBs require the approval of the state government (normally the Department of Municipal Affairs) to borrow. In most instances, this is on an individual case-by-case basis, though in others (e.g. Tamil Nadu) a financial plan is approved whereafter ULBs are free to borrow within this limit without further approval. The research conducted for this study showed that in all cases where some sort of substantive, merit-based scrutiny of the proposed loan transaction is conducted, there are no clear criteria governing the assessment (or at least none which are clearly documented and understood by the applicant ULB or investors). The process through which the assessment is conducted is highly opaque, and there is no mandated timing. See Annexure 7 for description of both the current procedure as well as a suggested set of steps for future consideration; • Loan limits and structures: Most states stipulate borrowing limits, loan tenors, and so on in law (Gujarat being an exception) using a variety of standards such as the volume of total borrowing to the value of ULB-owned property, or the size of the potential tax base. Invariably, these limits have little relationship to the actual creditworthiness or borrowing capacity of ULBs, and the data available to develop these indicators is poor. In sum, they do not provide a useful or effective way of managing the borrowing risks of municipalities. Moreover, in practice these limits appear, de facto, to be ignored in the loan authorization process; 28 CHAPTER 3 • ANALYSIS OF INDIAN MUNICIPAL BORROWING REGULATORY ENVIRONMENT • Security: All states allow for the pledging and escrow of local revenues as a form of security, and, possibly because of the weaknesses of the regulatory system as a whole, most loan transactions are characterized by structures of this kind. While some states (such as Madhya Pradesh and Maharashtra – in the case of Nagpur Municipal Corporation) also allow for the mortgaging of municipal property to provide collateral, others do not. In general, it appears that the limits on mortgaging are unnecessarily restrictive; • State guarantees: As indicated earlier, the issuance of state guarantees for municipal loans has become less common over time, perhaps affected by the introduction of the Fiscal Responsibility legislation pertaining to the states. In all cases, the issuance of such guarantees requires approval of the relevant state Department of Finance. However, as is the case with the issuance of borrowing approval, the criteria for the issuance of guarantees, the authorization process and its timing are unclear and opaque20. Ex post procedures: In general terms, the approach to what is often loosely termed “municipal bankruptcy� (i.e. resolution of the financial affairs of municipal loan default or municipal financial crisis) in India is marked by a paradox: On the one hand, state governments have considerable powers to intervene in such cases; on the other, there are basically no institutionalized frameworks or legislated procedures in terms of which such an intervention can be organized or effected. Moreover, because ULBs are regarded as “public bodies� under Indian law, the generally applicable debt recovery statutes that have emerged in the corporate sector to supplement the regular Indian court system (which, for reasons beyond the scope of the current report, generally fails to settle tort claims effectively within a reasonable time period), such as the Debt Recovery Tribunal Act and SARFAESI, do not apply and thus cannot provide an effective channel to settle creditor claims. Moreover, in no state is there a law that provides for the attachment of property vested with the municipality for the purpose of enforcing lenders’ rights21. The process whereby municipal default, or the threat of default, is handled in India is extremely opaque and there is almost no reliable or objective data concerning it. The research undertaken for this study (including extensive discussions with investor organizations which, on this issue, were necessarily anecdotal) indicates that while default or potential default is not uncommon, in the absence of any municipal bankruptcy process it is generally handled through a three-way negotiating process involving the borrowing municipality, the relevant state government, and the investor, with the state government playing a very significant role. By and large, the “work out� process is ad hoc, often strongly relationship-driven, and is usually affected by a range of considerations – political and commercial – which are outside the merits of the immediate financial issues at stake. 3.2.3 Central Regulation of Municipal Lenders/Market Four forms of central regulation have a direct impact on the willingness and the ability of lenders and investors to engage the municipal market: (i) RBI regulations; (ii) the tax treatment of municipal 29 DEVELOPING A REGULATORY FRAMEWORK FOR MUNICIPAL BORROWINGS IN INDIA bonds; (iii) insurance and provident fund regulations; and (iv) SEBI bond listing requirements. Unlike state regulations pertaining to borrowers, most of these regulations have attracted attention in previous studies of the municipal market (most notably in the USAID FIRE-D exercise), and they are subject to detailed assessment in Chapter 4. Broadly, however, their impacts are evident in four main areas: • In the way it is interpreted, the RBI master circular (Master Circular on Loans and Advances, June 2009) is often taken to mean that direct lending to ULBs by SCBs is not permitted. As evidenced from the SCB lending in some states (e.g. Maharashtra) this is not correct, but in some states (e.g. Kerala) a number of potential transactions have been terminated as a result of this interpretation. In general, the existing interpretation appears to have a dampening effect on the willingness of parties to pursue transactions on both the borrower and lender sides; • The interest-rate limitations of GoI regulations pertaining to the tax exemption of municipal bonds creates no incentive for investment in this sector in the current interest rate environment. Other “public purpose� tax free instruments are handled on a case-by-case basis without interest rate limitations, and municipal paper is placed at a competitive disadvantage relative to these. The approval process is also opaque and time consuming and is only granted for the fiscal year of application, so potential investors face substantial expiry risk; • IRDA guidelines and the removal of municipal bonds from the sovereign lending requirement (SLR) list in 2002/03 similarly places municipal bond issues at a competitive disadvantage relative to other forms of government paper. Redressing this would most likely imply some measure of relaxation of prudential stringency – a policy judgment that would need to be carefully considered – but it is worth noting that in these areas, municipalities are not on a level regulatory playing field even with private sector players in the infrastructure sector. • SEBI rules regarding the listing of municipal bonds and public disclosure requirements are derived mainly from the corporate sector and are inappropriate in some respects for municipal instruments22 . In addition, CVC guidelines relating to tendering requirements in respect of banking services such as underwriting loans and bond for municipalities are ambiguous and unclear. Since tendering only applies to private suppliers of goods and services23, some municipal corporations favor public sector banks and institutions such as HUDCO, simply to avoid tendering and to be on the “safe side� in a future Vigilance Commission inquiry. This has had constraining effects on the activities of SCBs and other market players, and has also placed a few lenders – such as HUDCO – at an unfair competitive advantage in the sector (with possible negative cost consequences for ULB borrowers). 30 CHAPTER 3 • ANALYSIS OF INDIAN MUNICIPAL BORROWING REGULATORY ENVIRONMENT 3.3 The Regulation of Municipal Borrowing in India: Points of Analysis With regard to the three basic approaches to the regulation of municipal borrowing discussed earlier – market, rules-based, and direct control – the system in India is most accurately understood as a de facto direct control system (see Chart 5 below). Some rules that broadly resemble those seen in rules-based systems can be found, but in India these tend to have little rational relationship to the things they are supposed to control (risk of excessive borrowing, for example), and they seem to play almost no role in loan authorization decisions. Rather, the crux of the system lies in the application of state level discretion to the authorization of borrowing on an ad hoc basis. Chart 5: Regulatory System of India in International Context Indian System in International Context: Current and Proposed MARKET BASED SYSTEMS Well defined systems USA Semi-defined systems Poorly-defined systems South Africa Poland Proposed policy orientation Hungary United Kingdom Philippines Mexico INDIA Austria RULES BASED SYSTEMS DIRECT CONTROL SYSTEMS Moreover, so far as direct control systems are concerned, India has a particularly poor one: The criteria which actually determine borrowing authorization decisions are unknown, and the process through which these decisions are made is highly opaque and lacks both a clear timeline and explicit structure. These features also characterize events surrounding default or potential default: while state governments have strong latent powers to intervene in the financial affairs of ULBs in such instances, there is no structured municipal bankruptcy process through which competing imperatives – specifically the need to keep public services running while respecting creditor claims – can be reconciled. The process through which such situations are resolved appears to be entirely ad hoc and discretionary. 31 DEVELOPING A REGULATORY FRAMEWORK FOR MUNICIPAL BORROWINGS IN INDIA As in other countries, the approach to the control and regulation of municipal borrowing in India is reflective of the overall legal and institutional environment, particularly of the intergovernmental structure. The chief weaknesses of this structure – which have received extensive analytic attention elsewhere24 – also have an impact on municipal borrowing activity. This ambivalence is reflected in the way in which municipalities are treated in the Constitution of India and the regulations governing financing in India, which fall under the jurisdiction of the central government: on the one hand, municipal bodies are regarded as a constitutional tier of governance discharging mandatory functions, whose assets and manner of their disposal are regulated and limited by the statutes under which the municipalities are created. This makes municipal assets/properties/revenue fall outside the scope of generally applicable debt enforcement legislations (such as SARFAESI and DRT Acts) and leads to a limitation on the extent to which remedies would be available to lenders in the event of default. On the other, the IRDA, for example, categorizes municipal securities as “non- governmental�, which places them at a competitive disadvantage relative to government securities. In other words, so far as central regulation is concerned, municipalities tend to fall between two stools, and do not attract the advantage of being categorized as either purely public or purely non- public entities. The net impact of all of the above is that the prevailing regulatory regime in respect of municipal borrowing dampens potential borrowing activity and makes the sector unattractive for investors relative to others. Both borrowers and lenders run substantial transaction risk – and while these risks are obviously high, there is little way in which they can be sensibly quantified or managed. In this environment, short of substantial reform, it is difficult to see how the conditions for the expansion of municipal borrowing activity are likely to emerge from the very low existing base. It should also not be forgotten that the present regulatory approaches have negative consequences beyond just the direct constraints they impose on borrowing activity. For example, the ad hoc and largely hidden “interventions� of state governments in cases of default, or potential default, which inevitably involve some sort of municipal bailout, tend to undermine hard budget constraints and entrench the lack of predictable accountability or the moral hazard of random bailouts. The relative impact of the regulatory regime as a factor in determining the overall trajectory of municipal borrowing in India needs to be kept in perspective. Of the four factors outlined in the Introductory Chapter 1 – the intergovernmental fiscal framework, creditworthiness, the nature of the domestic capital market, and regulation – the regulatory issues may not be the most important. For example, there is evidence to suggest that one important variable in determining the decline of aggregate municipal borrowing since around 2007 might be the rapid increase in the quantum of the grant funds that have been provided to larger Indian cities through JNNURM (which, to date, has not been stringent in insisting on up-front counterpart contributions as originally intended). Nonetheless, the regulatory system is clearly one important determinant; moreover, it is one that appears to be relatively tractable from a reform perspective and is thus worth doing something about. Chapter 4 provides detailed recommendations for the reform of this environment; the rest of the current section provides the rationale for the overall direction this reform should take. 32 CHAPTER 3 • ANALYSIS OF INDIAN MUNICIPAL BORROWING REGULATORY ENVIRONMENT 3.4 The Regulation of Municipal Borrowing in India: The Overall Way Forward The broad thrust and character of regulatory reform needs to be determined by a consideration of two sets of factors. First are normative goals. Put most simply, India needs (i) to expand the market to finance the extension and improvement of urban infrastructure by ULBs, while (ii) minimizing the risk of moral hazard through the proper identification, allocation and pricing of risk. The second factor is, existing institutional and structural constraints. Four of these require particular mention: • the hierarchal relationship between states and ULBs is ingrained in the constitutional and institutional structure of the Indian state and it is not likely that states will give up many of their current powers to regulate local governments in the short to medium term. Moreover, intergovernmental fiscal flows are structured and managed such that the emergence of realistic and consistent hard budget constraints is unlikely in the short-medium term; • accounting and financial management systems in ULBs largely remain weak. Improvements are being made in this area: The introduction of accrual accounting is one of the key items on the JNNURM reform agenda and such efforts are underway or completed in many states/ULBs such as Tamil Nadu, Karnataka, etc.; but getting accurate information on the finances of potential municipal borrowers is likely to remain a challenge for investors in the foreseeable future; • the court system in India cannot be relied on to serve as an effective mechanism for curing defaults and non-payments. Processes relating to (potential) municipal default will thus have to rely on some sort of extension and modification of the systems that have emerged to supplement court processes for settling debt claims in the corporate sector; • relative to other funding flows into the urban sector, municipal borrowing remains extremely modest and it is a very small fraction of overall lending activity in the country. While national policy in recent years has demonstrated a very significant change – in that the urban sector is now a primary concern of both national and state governments and is attracting much greater attention and resources – ULBs remain a fairly weak political constituency. In other words, while pressures for reform of municipal borrowing regulation are building, these are not yet particularly powerful and a number of other issues and initiatives in the urban sector tend to attract most policy attention (JNNURM being a case in point). In this context a feasible reform goal is an incremental set of improvements that shift the overall regulatory approach in India from a poor de facto “direct control� system to an explicit and de facto “rules-based� system. On the ex ante side, states would retain authorization and oversight powers, 33 DEVELOPING A REGULATORY FRAMEWORK FOR MUNICIPAL BORROWINGS IN INDIA but these would shift substantially in character such that: i. logical rules rationally related to the actual risks they are intended to manage would replace the current ones that are largely irrelevant to things like actual borrowing capacity (and are anyway ignored); ii the authorization decision would be limited to compliance with these rules rather than the substantive merits of the transaction or the project it is funding; iii the authorization process would be explicit, transparent and fair, would involve appropriate parties, and would be time-bound. On the ex post (default and municipal bankruptcy) side, an institutional system pertaining to the exercise of security and attachment of assets, as well as the permanent settlement of creditor claims while protecting the on-going delivery of public services and ensuring that risk is properly allocated, needs to be developed. Given the inevitable complexity and sophistication of such a system, an initial step would be to adapt one of the existing mechanisms, which currently do not cover the public sector, to improve the exercise of security options. So far as central regulation (of borrowers and instruments) is concerned, the overall goal is to rationalize current approaches and place municipal instruments on a level playing field with other public sector borrowing. To the extent that – as for various other sectors – overall policy allows for some sort of subsidization of municipal borrowing (through the tax system, for example), rules should be adjusted to create incentives sufficient to attract lenders into this market. The overall thrust of these reforms, in conceptual terms and relative to international experience, is represented by the red arrow in Chart 5. The detailed recommendations for the specific measures required to effect this change are outlined in the following section. End Notes 14 This framework is loosely based on the work of Teresa Ter-Minassian and others regarding sub-national borrowing in general (see, for example, her IMF Paper on Policy Analysis 96/4, “Borrowing by Subnational Governments: Issues and Selected International Experiences� at http://www.imf.org/external/pubs/ft/survey/pdf/112596b.pdf. Note that the framework has been adapted for usage with specific regard to municipal borrowing. This has material implications for the framework and particularly for the categorization of countries in terms of it. 15 An exception would be Slovenia and Latvia, for example, where national commissions decide upon the technical merits as well as the financial feasibility of municipal borrowing projects, based of course on published and operationalized rules. 16 Ter-Minassian discusses a fourth, “cooperative� model that involves sophisticated negotiations between several layers of government on issues such as limits, budget priorities, as well as fiscal equalization. This approach exists in countries with sophisticated, consensus- based budgeting such as that in New Zealand and Scandinavia. 17 For details see the Legal and Regulatory report in Annexure 2. 18 The Public Debt Act, 1944 governs only “government securities�, which are defined in a limited manner to cover only securities issued by either the Government of India or a state government. Thus, securities issued by a municipality are not “government securities� covered under the Public Debt Act, 1944. The Government Securities Act, 2006 reflects the same set of definitions as the Public Debt Act, 1944 and does not cover municipal borrowing. The main central legislation that is relevant to municipal borrowing is The Local Authorities Loan Act, 1914. This law was however, not applicable to a substantial portion of India. 34 CHAPTER 3 • ANALYSIS OF INDIAN MUNICIPAL BORROWING REGULATORY ENVIRONMENT 19 A number of areas of regulation involve some level of mixed or shared authority and responsibility between the central and state governments – particularly that relating to the (a) CVC, (b) DRT and SARFAESI. These are most accurately described as “Mixed Agency� areas. However, for simplicity’s sake they are dealt with under central and state regulation respectively. 20 In principle, of course, such guarantees create moral hazard and, in line with what seems to be the broader policy orientation of GoI, it would be best if they became an increasingly rare feature of municipal borrowing in India. 21 The Union acts are The Presidency Towns Insolvency Act, 1909, The Provisional Insolvency Act, 1920, The Civil Procedure Code, 1908, The Recovery of Debt Due to Banks and Financial Institutions Act, 1993, The Securitisation and Reconstruction of Financial Assets and Enforcement of Security Interest Act, 2002. The Presidency Towns Insolvency Act, 1909, and the Provisional Insolvency Act, 1920. They deal with insolvency of individuals or corporate entities. Thus they are not applicable for insolvency of statutory authorities such as municipalities. 22 Special municipal disclosure standards have been implemented in Hungary in 1996, and suggested for Serbia by LGID Ltd. under the World Bank project on “Serbia Capital Markets Technical Assistance – Municipal Bond Development� in June, 2009. In the United States, special boards have developed voluntary industry standards. Naturally, rating agencies may impose their own disclosure standards before SEBI develops its own. 23 One of the guidelines issued by the CVC is on “Transparency in Works/Purchase/Consultancy contracts awarded on nomination basis� which says that “it is needless to state that tendering process or public auction is a basic requirement for the award of contract by any government agency, as any other method, especially award of contract on nomination basis, would amount to a breach of Article 14 of the Constitution guaranteeing right to equality, which implies right to equality to all interested parties.� 24 The recent McKinsey report “India’s Urban Awakening: Building Inclusive Cities, Sustaining Economic Growth� is a prominent example. 35 DEVELOPING A REGULATORY FRAMEWORK FOR MUNICIPAL BORROWINGS IN INDIA Chapter 4 Policy Recommendations on Local Borrowing Regulatory Framework T he structure of the recommendations made here follows the structure of regulatory responsibility deriving from the Indian constitution as outlined in par. 46. First, regulations relating to lenders and lending instruments, which fall under the responsibility of the Union government, are dealt with. Thereafter, regulations relating to rules dealing with the ex ante borrowing activities of municipalities and ex post procedures on municipal default and insolvency are discussed. A number of areas of regulation involve some level of mixed or shared authority and responsibility between the central and state governments – particularly that relating to (a) the CVC, (b) DRT and SARFAESI. These are most accurately described as “Mixed Agency� areas. However, for simplicity’s sake they are dealt with under central and state regulation respectively. 4.1 Recommendations on Central Regulations 4.1.1 RBI Master Circular on Loans and Advances In relation to loans for infrastructure projects being implemented by ULBs, RBI regulations classify only loans given to a special purpose vehicle created to implement such projects as “infrastructure loans� and not loans given directly to a ULB, even if it is for the specific purpose of being used in one or more infrastructure projects. In the case of lending to an identical project where the borrower is a special purpose vehicle, it qualifies as “infrastructure lending�. This definition potentially disadvantages those lenders who want to exceed their 15% exposure limit to any one borrower or project and avail of the additional 5% limit permissible for infrastructure; in addition, this distinction has been adopted by other agencies (such as the Planning Commission), and ULBs face potential unequal treatment in respect of future policies. Other bodies that use RBI’s definition then do not examine the content of the project and instead categorize on the basis of legal personage of the formal borrower. In some cases, the RBI Master Circular has been interpreted by some lenders as well as some municipalities to mean that scheduled banks are prohibited from lending directly to municipalities, and can only provide loans to SPVs created by them for specific projects. Although currently no municipal borrower threatens to exceed an SCB’s one borrower exposure limit, eliminating this disadvantage is a no cost method of at least attracting the attention of lenders to larger municipal exposures in the future. Recommendation 1: Stakeholders should be informed at state level that direct lending to ULBs is allowed under the circular. In addition, state instructions on borrowing by municipalities should cross-reference the appropriate RBI document for clarity’s sake. 36 CHAPTER 4 • POLICY RECOMMENDATIONS ON LOCAL BORROWING REGULATORY FRAMEWORK Recommendation 2: RBI should consider adding direct lending to ULBs for capital investments in obligatory functions to the list of lending defined as “infrastructure project lending�. It could happen even though the borrower is not an SPV, but rather a ULB. This no cost reclassification will reduce confusion as to the status of direct lending to municipalities. It will reinforce, directly, the link between legitimate purposes of borrowing and the asset classification system in use by RBI and banks that report to RBI. 4.1.2 Tax Treatment of Municipal Bonds The tax benefit for financial institutions accrues only to the difference between the cost of funds and a coupon rate capped at 8%. In the current interest rate environment, after-tax yields on taxable instruments are higher than the tax-free margin over the cost of funds allowed under the tax-free bond schemes. At higher cost of funds, there may be no margin at all under the coupon capping regime. This limitation on coupon rates applies only to municipal bond instruments (maximum coupon on tax free municipal bonds) – tax benefits for other “public purpose� bonds (REC, NHB etc.) are handled on a case-by-case basis, without interest rate limits, but rather within overall volume limits on the face value of these other types of tax free bonds to be issued. Furthermore, the tax approval process is time-consuming and is only given for the financial year of application, so there is substantial expiry risk. ULBs will be at a disadvantage until their tax-free qualification standards are similar to that of other public purpose entities competing for capital. In this context, it should be noted that capital market access through municipal bonds should go hand-in-hand with accounting reforms at ULBs to strengthen reporting and disclosure, as has already begun under various urban reform initiatives. Recommendation 3: Eliminate the interest cap and treat tax-free municipal bonds in the same manner as other tax-free instruments. A Union-wide annual volume limit should be used. In addition, approval should be valid for 12 months, not just within the fiscal year in which it was granted. 4.1.3 SEBI Bond Listing Requirements Bond listing and disclosure requirements are not appropriate for ULBs. Debt to equity ratios or descriptions of market share are likely only to become a significant issue should the municipal bond market become more liquid. The tax-free guidelines issued by GoI contain information requirements that are entirely appropriate in listing and/or initial issue disclosure documents25. Recommendation 4: SEBI should publish disclosure guidelines for public issues of municipal bonds. They should be in line with the GoI MOF tax-free standards and incorporate current international practice. This step would anticipate future public issues of bonds, and would be in congruence with requirements that provident funds and other institutional investors invest only in publically offered and traded instruments. International practices offer two options for India. It is appropriate to follow EMMA in the United States, where the SEC has no specific disclosure requirements and relies instead on the voluntary industry standard, though after suitably adapting it to the Indian context. For instance one thing to note in the Indian context is the protracted 37 DEVELOPING A REGULATORY FRAMEWORK FOR MUNICIPAL BORROWINGS IN INDIA litigation process. Conversely, some transition economies have developed specific municipal disclosure standards, as a part of their securities regulation. In addition, the investor base for municipal bonds should start including retail investors as well going forward. 4.1.4 Insurance and Provident Fund Regulations IRDA guidelines classify all municipal bonds as “non-governmental securities� that need to be rated at least A+ to be eligible for inclusion in an insurance company’s investment portfolio. PFRDA designates municipal bonds as Class C instruments26 while Class G is limited only to government securities, from which municipal bonds are excluded by definition. Municipal bonds thus have to compete with other Class C instruments, which may have higher yields than municipal bonds. These instruments make municipal bonds unattractive within this asset class. Municipal bonds are treated as being non-governmental, hence forcing them to compete with commercial issuers while excluding them from the class where states and GoI may issue debt. Recommendation 5: Add a new asset class called “rated municipal securities� to both IRDA and PFRDA’s guidelines. This class should have the same allowed proportion as corporate bonds, i.e. asset Class C, and should be substitutable for investments in Class C. This means essentially that municipal bonds are not government bonds, but may be substituted one for one against corporate bonds. Therefore an investor may replace riskier but higher yielding corporate paper with rated municipal securities. Thus, depending on an investor’s risk and maturity preferences, this recommendation would allow them to diversify into a recognized asset class without any disadvantage in terms of permitted allocations among classes. Other studies have recommended reducing the rating requirement for approved municipal securities from A+ to A – this too can be considered as the regulatory environment improves. 4.1.5 Central Vigilance Commission CVC guidelines and state procurement laws – where they apply to banking services – exclude public sector banking service providers. In other words, ULBs seeking finance may avoid a formal tendering process if the potential service provider, i.e. financial institution, is considered to be “non-private�. In this sense, entities in public ownership enjoy a procurement advantage. There is an apparent view among ULBs that public lending entities bear less risk, as CVC guidelines, and hence potential CVC inquiries do not apply. Recommendation 6: The state procurement laws should be extended to cover services offered by both private and publically owned financial firms. Applicable and exempted financial services should also be specifically defined. Bids for banking and financial services should be required using simple standard formats, with full costs easily defined on a comparable basis. 4.2 Recommendations on State Regulation of Borrowers: Ex ante Rules The recommendations below (i.e. relating both to the ex ante and the ex post rules and procedures) 38 CHAPTER 4 • POLICY RECOMMENDATIONS ON LOCAL BORROWING REGULATORY FRAMEWORK are based on a detailed analysis of regulations and legislation in four states (Madhya Pradesh, Maharashtra, Gujarat, Tamil Nadu). They outline a set of fairly concrete guidelines intended to improve the regulatory environment in furtherance of the broad policy goals outlined earlier, but would need to be adjusted somewhat to local conditions if adopted. 4.2.1 Definition of Debt All state laws define debt as borrowing in the form of term loans or the issuance of debentures (bonds). Other types of financing, such as supplier loans, financial leasing, PPP contracts and deferred payments can by-pass the state borrowing approval process since by definition they are not considered “borrowing�. The definition of debt determines the scope of state oversight on borrowing in all of its forms. If state review and approvals are considered to be a legitimate instrument for maintaining prudence, this very limited definition of debt is obviously a major weakness. Furthermore, from the standpoint of data collection on budget users, alternative forms of debt may be disguised as operational expenditure. So they might create a distorted picture of future obligations and creditworthiness in the medium term as debt limits are approached or even exceeded. Recommendation 7: State regulations should define “debt� in terms broad enough to include those multi-year financial obligations whose repayment depends on the continuing existence of operational surpluses. All future monetary obligations27, such as guarantees issued by ULBs, financial leases, promissory notes and such should be subject to the State’s review and approval procedures by being defined as “debt�. 4.2.2 Purpose of Borrowing All ULBs can borrow for “works, land acquisition, repayment of earlier loans, slum rehabilitation and executing legal mandates.� The last purpose may be interpreted broadly to cover operational activities, i.e. not necessarily capital investment in vital infrastructure. However, the restriction on borrowing to fund recurrent deficits that prevails in the examined States is in line with international best practices. Unfortunately, the distinction between long and short term borrowing is not made and this creates a potential soft budget constraint with long-term borrowing perhaps being used to fund ongoing recurrent deficits. Recommendation 8: States should introduce a distinction between long-term and short- term debt. “Long term� debt is defined as (i) serving capital investment, (ii) repayable in over one year, and (iii) subject to state authorization. Short term debt must be repaid within the current fiscal year and not rolled over. A 5% limit on short term borrowing for liquidity purposes is in line with international practices where such borrowing is allowed at all. Recommendation 9: Long-term borrowing shall be limited to long-term capital investment. Long-term loans have over one year tenure and assets are created in water, sanitation, etc. services, as defined by the GoI Guidelines for Tax Free Municipal Bonds. Allow refinancing of earlier long- term capital investment loans only if the new terms are significantly more advantageous to the borrower, if all costs such as prepayment penalties are considered. 39 DEVELOPING A REGULATORY FRAMEWORK FOR MUNICIPAL BORROWINGS IN INDIA 4.2.3 Borrowing Approval In all cases, ULBs require state approval to borrow, as described in the analytic chapter before. There are no clear criteria, processes or mandatory timing, creating uncertainty and transaction risk for both borrowers and lenders. Certain procedures and standard forms, as well as published approval standards, are also missing. The following recommendations pertain to higher (state) level authorization, after the local level decisions/approvals for borrowing have already been taken, including from the elected representatives of the city government as appropriate. Recommendation 10: Retain state powers to grant or deny approvals, but shift to a rules-based system. Approvals should be directed at providing certification that the applicant is in compliance with stipulated regulations on debt service. This would rule out arbitrary judgments about the other merits of the project. The stipulated regulations should provide clear guidelines regarding the quantum and character of municipal borrowing. They should include clearly defined allowable purposes, stock and flow limits as appropriate, allowable security, etc. Project-related approval can be part of a separate process of administrative sanction, if this is regarded as necessary. Recommendation 11: The authorizing role should be played by a standing Inter-Departmental Committee. Its members should be recruited from the urban and finance sectors. This Inter- Departmental Committee should have a clear mission and simple, time-bound operating procedures (including “deemed approval�) and standard approval request forms. Recommendation 12: Regulations should clarify that approvals do not constitute state underwriting of municipal loans. Bailouts and unauthorized debt do not constitute a legitimate claim so the courts will not accept a request of this type that has not been duly authorized under the appropriate state laws. 4.2.4 Limits on Long Term Borrowing Current limits have little relationship to “creditworthiness� or to actual borrowing capacity and so do not act as an effective means of managing fiscal risk and/or are inappropriately confining, as reported in the analytic section. The following recommendations pertain to one specific formulation for such debt limits linked to operational revenue surplus. In addition, there could be other formulations as well, linked for instance to debt service as a proportion of revenues, stock of debt as a proportion of revenues, etc. Recommendation 13: States should introduce an annual debt service limit of 50% of the average net operational surplus of the ULB over the past three completed fiscal years. This limit is often used internationally. Financial institutions certainly project these surpluses into the future and such calculations could be a part of the state approval process. The average operational surplus includes all revenue funding, but excludes one-off capital receipts such as those stemming from sales of assets. Debt service includes current payments of principal and interest on outstanding debt (including financial leasing), as well as anticipated additional payments of principal and 40 CHAPTER 4 • POLICY RECOMMENDATIONS ON LOCAL BORROWING REGULATORY FRAMEWORK interest regarding the new borrowing. The implicit annuity of given guarantees at 1/tenor need be added on a current basis. It creates a cushion for the contingency of paying out a guarantee. If the new borrowing has a grace period, then anticipated new debt service must include not only interest, but also a level annuity of principal that is to be set aside in a sinking fund. Recommendation 14: The debt service test is applied only whenever a new debt is to be taken. It may be used by stakeholders to monitor the financial condition and statutory compliance of a given ULB borrower over time. 4.2.5 Security and Collateral Legal analysis indicates that all States allow pledging and escrow of local revenues. In addition, some states allow the mortgaging of immovable property (e.g. Madhya Pradesh, Nagpur in Maharashtra); some do not (e.g. Tamil Nadu). Successful and creditworthy projects exist in all locales regardless of these differences. As a universal critique of state regulatory systems, nowhere can municipal property effectively be attached, a tendency that questions the utility of public property as a form of security. Sinking funds, however, are allowed, where cash flows may be deposited in accordance with the financing agreement, and closely monitored by all parties. This also does not provide ultimate security, but instead offers a fair warning of impending insolvency, and allows other state intervention mechanisms to be initiated. Recommendation 15: Municipal assets should be subject to mortgages, provided that the assets are non-essential for the provision of mandated services. State laws governing municipalities should provide for the distinction between essential municipal assets and commercial/non-essential municipal assets. Rules should be made to provide for the tests to determine which particular category a municipal asset falls under. State laws should then provide clearly that commercial/non-essential municipal assets will be subject to general debt enforcement laws and can be attached by lenders in whose favor security was validly created for loans taken in accordance with the relevant municipal law. 4.2.6 Guarantees State governments are asked to provide guarantees towards third parties for borrowing by ULBs. Both the borrowings as well as the given state guarantee require state approval. Approvals of state guarantees do not seem to have clear published criteria, processes or timing. Market stakeholders seemingly do not consider state guarantees to be of much value when it comes to enforcement. Other fiscal pressures on state finances may have a higher priority. In principle, both state and municipal guarantees create moral hazard and they also appear to diverge from the current broader policy orientation of GoI. Recommendation 16: State guarantees of municipal debts should be discouraged. Special situations, such as borrowing from international organizations or foreign lenders may be exceptions. In line with international practices, the borrower might pay a guarantee fee commensurate with the amount of public (state) funds at risk. Were state guarantees of municipal 41 DEVELOPING A REGULATORY FRAMEWORK FOR MUNICIPAL BORROWINGS IN INDIA borrowing to continue, clear criteria should be established, along with published approval procedures and forms. As stated above, municipal-SPV guarantees shall be accounted for and treated as debt until the guaranteed debt is repaid. 4.2.7 Loan Tenor Limits If tenor limits are stipulated at all, they are all very long (up to 60 years) and have little effect on transactions. Recommendation 17: Maximum loan tenor limits should be stricken from state legislation. They have little effect on creditworthiness and are not to be used in approval decisions. Tenor should instead be matched with the useful economic or accounting life of the asset being created; but this is not a formal recommendation, rather a basic principle that all stakeholders should follow. 4.2.8 Monitoring, Reporting and Database There is no known India-wide or state-wide accessible database of municipal borrowing on a historical or current basis. Data on municipal balance sheets or budget execution (debt service, capital revenues and expenditures, current borrowing, etc.) are not maintained on a comparable, available and user-friendly basis within and among state jurisdictions. RBI does not list municipalities as a separate borrower (asset class) and instead merges ULB data with other public borrowers, making even India-wide datasets unfriendly towards conducting analysis of ULB finances. Recommendation 18: Approved and realized borrowings and their terms should be in a public database maintained by each state. It will reduce risk and uncertainty, and offer a level playing field to all market participants. Data reported to MOUD should be published annually on an aggregate level and on ULB basis for transparency purposes. 4.3 Recommendations on State Regulation of Borrowers: Ex post Procedures None of the debt and security enforcement mechanisms used in the private sector apply to ULBs. Both the Debt Recovery Tribunal (DRT Act, 1993) and SARFAESI apply to corporate bodies in their dealings with the financial sector and with each other. Since ULBs are both statutory and constitutional entities, they cannot be wound up or liquidated in the usual manner, nor may their mandatory statutory functions be hindered by a private party or by court action. For these reasons, enforcement mechanisms and bankruptcy law as they currently exist can neither be used to intervene in case of municipal default, nor to head off protracted insolvency. The states, as the analytic section has pointed out, enjoy wide powers to intervene in the affairs of errant ULBs, including dismissal of elected officials and the ordering of new budgets and local fiscal policies. While the usual debt recovery and security enforcement mechanisms are not effective against ULBs, the state governments are vested with overwhelming but ill-specified powers to intervene in ULB operations, with no clear trigger mechanisms or intervention procedures. There is also no 42 CHAPTER 4 • POLICY RECOMMENDATIONS ON LOCAL BORROWING REGULATORY FRAMEWORK mechanism for identifying and prioritizing the claim of various lenders, who have then to petition the state government for some form of relief. 4.3.1 Short to Medium Term Recommendations Recommendation 19: Legislation at both Union and state levels should ease the enforcement of both security and collateral against ULBs. In the short and medium term, the role of DRT in enabling enforcement of debt against municipal authorities should be extended by suitable amendment to the DRT Act and formulation of corresponding rules in light of the existing framework, so as to provide clarity on the rights of lenders against municipal assets. The rules should classify municipal assets as “essential� and “commercial� and allow for enforcement through DRT against “commercial/non essential� municipal assets. This step would have to be undertaken by the GoI and Union Parliament in consultation with state governments. Recommendation 20: In the short to medium term, municipal insolvency situations will need to be dealt within the current state administrative structure. This requires the creation of a formal, institutionalized Administrator at the state level responsible for municipal insolvency resolution within existing laws. The Administrator should have clear functions and powers relating to intervention in a municipality and the assumption of its financial affairs in instances of insolvency, and should operate within a set of consistent and transparent rules regarding events that trigger intervention and procedures during the intervention period (including negotiations between the municipality, its creditors and other stakeholders and the development and implementation of a financial recovery plan). The overall objective of any such intervention should be to balance competing imperatives, specifically addressing the legitimate claims of lenders and creditors while maintaining critical public services, without encouraging moral hazard. 4.3.2 Long Term Recommendation Recommendation 21: In the long term (by which is meant anything after about five years from the implementation of recommendations 19 and 20, depending on how conditions develop), and assuming that municipal borrowing activity has expanded to the point that such measures are justified, it would be advisable to create an entity within each state that focuses specifically on issues related to municipal default and bankruptcy. In essence, this would involve shifting some of the functions which, in the short-medium term are proposed to be undertaken by an expanded DRT and Administrator in Recommendations 19 and 20 above, to a properly and formally constituted Municipal Debt Tribunal (MDT). This would work with, but with substantial autonomy from, the state government in order to enforce contracts and security, and oversee municipal insolvency interventions in a way that is more systematized and institutionally objective than is possible with interventions run entirely under the state government. In this model, the MDT would take over the municipal-related functions of the DRT with expanded powers, not only to enforce security and collateral, but also to “force� debt adjustment and to impose settlements if the voluntary negotiations do not succeed. At the same time, the state government would retain certain key obligations regarding the institution of an Administrator in cases of municipal 43 DEVELOPING A REGULATORY FRAMEWORK FOR MUNICIPAL BORROWINGS IN INDIA insolvency, but with certain oversight and triggering roles to be played by a semi autonomous MDT, to minimize the politicization of any intervention and to ensure balance in the way that competing imperatives and stakeholders are dealt with. In simple terms, this structure and a possible distribution of roles is outlined in the following diagrams: Chart 6: Trigger Events, Tasks, Recovery Plan and Debt Adjustment State Government (Urban Department) Administrator Triggers Municipal Debt Tribunal (and staff) (quasi-judicial body) Assumes Oversees Enforces Supervises Key powers Financial Investor security Debt adjustment and functions Recovery Plan and collateral and settlement Chart 7: Structure and Roles of State and Tribunal Intervention Enforcement Direct State Municipal Financial Debt Adjustment of Security Administration of Recovery Plan (FRP) Municipality Trigger Event Leader files report to MDT Arrears over 90 days Simultaneous to Failure of any voluntary to call security based on reported by creditors State Administration debt adjustment stipulations of contractual to MDT negotiations and/or arrangement failure of State to deliver MDT orders State approved FRP intervention and suspends creditor claim for 6 months Powers Security claim checked State appoints Financial Recovery Plan Supervises and enforces and Role Administrator and key developed, approved by debt restructuring and Asset and/or fund is staff to assume running state and implemented by settlement on basis of attached to lender of the ULB focussed on Administrator stipulated order of claims obligatory functions Administrator conducts Orders liquidation of non- voluntary debt adjustment essential ULB assets and negotiations payment of State fiscal transfers as basis of debt settlement Exit-Result Lender cured Solvency restored FRP implemented Debt settlement 44 CHAPTER 4 • POLICY RECOMMENDATIONS ON LOCAL BORROWING REGULATORY FRAMEWORK 4.4 Phasing In Reforms Extensive consultations and careful phasing in, based on local conditions, must precede full implementation of all recommendations. In particular, substantial dialogue between the national government and state governments is required on the overall recommendations developed above if they are to gain traction at the state level, which is where much of the actual regulatory adjustment will need to take place. Within this overall context, and subsequent to such an engagement: • The suggestions for Union-level changes are fairly minor and technical. As noted, most of these recommendations have been made earlier. MoUD should convene a dialogue involving the key stakeholders to explore the possibility of implementing at least some of these under current conditions; • So far as the state regulation of borrowers – which really constitutes the heart of this report – is concerned, extensive state-level dialogue and some additional technical work would be needed to develop the short to medium term suggestions for ex post state- level procedures, particularly those relating to the Administrator. As this progresses, state administrations could fairly rapidly formulate the actual detailed ex ante regulations, which are more straightforward; • This work would need to take place state-by-state on the basis of the proposals made in this report. Given the variation among the four states examined in this study, the ex ante rules for each state would need to be developed on a customized basis, as long as they offer comparable and clear thresholds, deadlines and procedures; • The technical definition of debt service, methods of its limitation, record-keeping standards, etc. could be suggested in a Union-wide template, then adapted to local conditions where appropriate; • Once the ex ante rules have been clarified, published and explained, the second tier of state-level reforms may take place, i.e. equipping the Urban department to handle emergency intervention, negotiations and financial workout plans; • Finally, on a parallel track, creation of the quasi-judicial Municipal Debt Tribunal with the powers to reduce assets and debt, and to impose settlements, needs careful deliberation. But this should start at a later point once the ex ante and the short- medium term ex poste proposals have progressed in terms of implementation. 4.5 Impact on Municipal Lending The regulatory gaps and problems identified earlier can be corrected through the above recommendations to reduce uncertainty on the part of both the investor and the borrower. In the medium to long term, India’s massive urban infrastructure gap addressed primarily by mechanisms such as JNNURM will be complemented with borrowing not only for cost sharing purposes, but also to cover the rest of the investment gap. The direct demand for credit will be determined by the 45 DEVELOPING A REGULATORY FRAMEWORK FOR MUNICIPAL BORROWINGS IN INDIA interplay of operational and capital surpluses and deficits that are beyond the scope of this study. Correcting uncertainties regarding permitting, default, asset classification, procurement, tax- free status and other aspects of borrowing, reduces perceptions of risk for both the lender and the borrower. If full competition is allowed or mandated, if municipal assets may be invested in and traded on markets, if municipal bonds may compete directly and fairly with similar public investment vehicles, and if lenders have a secure mechanism to enforce their rights, given sufficient demand, borrowing will increase in India. Without these regulatory changes, the order of magnitude increases in borrowing for cost sharing and gap filling will not take place with sufficient speed. These uncertainties steer investors away from ULBs and discourage borrowing even where it would be appropriate. With the proposed regulatory changes, policymakers may focus on aspects that make ULBs more creditworthy or are able to generate surpluses; and larger capital investment programs will succeed since cost sharing can be filled with outside capital rather than just with accumulated surpluses. The policy changes suggested attempt to address those aspects of the regulatory regime that seem to be biased against ULBs, or make lending to ULBs less attractive and certainly seemingly more risky. Both borrowers and lenders run substantial transaction risk unless the approval standards and processes are made clear and fair. Other regulatory hurdles such as misinterpretation of RBI circulars, or a tax-free bond regime that does not provide treatment equivalent to other tax-free instruments, artificially withhold flows to the municipal sector without any visible benefit to the overall public sector. The ad hoc and largely hidden “interventions� of state governments in cases of default, or potential default, which inevitably involve some sort of municipal bailout, tend to undermine hard budget constraints and entrench the lack of predictable accountability or the moral hazard of random bailouts. This is addressed by operationalizing existing intervention legislation, adding ULBs to the powers of DRT, and suggesting a Municipal Debt Tribunal in the long run. The regulatory changes alone are not sufficient to cause a jump in borrowing, but in their current state, regulatory problems discourage even the current low levels of credit demand and supply. As these changes are at a low or no overall cost to the treasury, they should be considered as a comprehensive package for moving towards a transparent, rules-based ULB borrowing framework where all types of lenders face a level playing field and transparent conditions. End Notes 25 In the United States, Rule 15c2-12 under the Securities Exchange Act of 1934 (“Exchange Act�) relates to municipal securities disclosure, but only regulates broker-dealers and other financial institutions. Thus there are no SEC (www.sec.gov) requirements for the content of municipal disclosure statements. The Municipal Securities Rulemaking Board is a self-regulatory organization responsible for regulating broker-dealers and banks in the municipal securities market. EMMA (Electronic Municipal Market Access) is an information source for municipal offering statements and provides training for both issuers and broker-dealers. (See http://emma.msrb.org). 26 They are called “credit risk bearing fixed income instruments� 27 PPP obligations are particularly complex as their definition as “debt� versus as long term purchased services has varying solutions internationally and no global standard exists yet; therefore this recommendation does not extend to PPP payments in general. This does not rule out states including the impact of PPP payments upon the ability to repay debt in their approval process. 46 CHAPTER 4 • POLICY RECOMMENDATIONS ON LOCAL BORROWING REGULATORY FRAMEWORK 47 48 (Change background colour as needed) THE WORLD BANK 70 Lodi Estate New Delhi-110003