WO R LD BANK WO R KING PAPER NO. 2 Private Participation in Infrastructure in China Issues and Recommendations for the Road, Water, and Power Sectors Michel Bellier 26057 Yue Maggie Zhou March 2003 - V~~Y THE WORLDBANK .. - ~_F THE WORD AN W OR L D B AN K W OR K ING PAPER NO. 2 Private Participation in Infrastructure in China Issues and Recommendations for the Road, Water, and Power Sectors Michel Bellier Yue Maggie Zhou THE WORLD BANK INTERNATIONAL FINANCE CORPORATION Washington, D.C. A Member of the World Bank Group Copyright 2003 The International Bank for Reconstruction and Devclopmcnt / 'T"he Worid Bank 1818 H Street, N.W. Washington, D.C. 20433, U.S.A. All rights reserved Manufactured in the United States of America First printing: December 2002 1234050403 World Bank Worlkng Papers are published to communicatc the results of tm: Bank's work to the development community with the least possible delay. The typescript of th,% papcr thcreforc has not been prepared in accordancc with the proccdurcs appropriate to journz'I Drintcd texts, and the World Bank accepts no responsibility for crrors. Somc sources cited in this papcr may be informal documents that are not rcadily available. The findings, interpretations, and conclusions exprcssed in this paper arc cntircly thosc of the author(s) and do not necessarily rcflcct the *iews of the Board of Exccutive Directors of the World Bank or the governments thcy represen.. The World Bank cannot guarantec the accuracy of the data included in this work. The boundaries, colors, denominations, and oulhcr information shown on any map in this work do not imply on the part of the vWorld Bank any judgment of the lcgal status of any territory or the endorsement or acccptancc of such boundaries. The material in this publication is copyrightcd. 'T'he World Bank encourages dissemination of its work and normally will grant permussion for use. Permission to photocopy items for internal or personal use, for the intcrnal or personal use of specific clients, or for educational classroom usc, is grantcd by the VWorld Bank, provided that the appropriate fee is paid. Plcase contact the Copynght Clearancc Center before photocopying itcms. Copyright Clearance Center, Inc. 222 Rosewood Drive Danvers, MA 01923, U.S.A. Tel: 978-750-8400 o Fax: 978-750-4470. For permission to reprint individual articles or chzpters, pleasc fax your :-quest with complete information to the Republication Dcpartmcnt, Copyright Clearance Ccntcr, fax 978-750-447C. All other queries on rights and licenses should be addressed to thc World Bank at the addrcess above, or faxed to 202-522-2422. ISBN: 0-8213-5439-6 eISBN: 0-8213-5440-X ISSN: 1726-5878 Michel Bellier is Lead Transport Specialist in the Transport Sector Unit for t^he East Asia and Pacific Region at the World Bank. Yue Maggic Zhou is Investmcnt Officer in the Private Sector Advisory and Transaction Unit at the International Financial C,orporation. Cover photos: Foshan-Kaipmg Exprcssway (Guangdong Provincial C'ommumication Department), Harbin water treatment plant (SAUR International), Laibin B power plant (EDF). Library of (Congress Cataloging-L-P-iubElesion Private participation in infrastructure in China : issucs and recommendations for the road, water, and power sectors / World Bank, Intcrnational Financc Corporation. p. cm. - (World Bank worling papcr; no. 2) Includes bibliographical references. ISBN 0-8213-5439-6 1. Infrastructure (Economics)- China-Financc. 2. Investments, Forcin-China. 3. China-Economic policy-2000- I. World Bank. II. International Fkinancc Corporation. IIL. Series. HC430.C3P73 2003 338.951'05-dc2l 2003041142 CONTENTS Foreword ............................................................... vii Abstract ......................... . ix Acknowledgm ents .. .... ... ... .........................................xii A bbreviations .... .............. .... ....................................xiii Executive Summary . ......................... xv 1. Introduction ............................................... ........... 1 Concepts and Models for Private Participation in Infrastructure .........1 Private Participation in Infrastructure in Chma ........ 6 2. Laws and Regulations for Private Participation in Infrastructure in China ........ 9 General Legislation for Private Participation in Infrastructure ....................... 10 The Approval Process ..................................................... 15 Granting Authorities .......................... ........................... 21 The Regulatory Framework ........................... .... ... 22 3. Financing Private Infrastructure Projects in China ............. 27 Risk Allocation ........ . ................................................ 28 Project Financing Regulation .................... 31 The Secunty Law .................... 33 Domestic Bank Lending ................ ...................... 38 Bond Financing ................ ............................ 39 Equity Financing ........ . ... ........................................... 41 4. Road Issues and Recommendations ................................ 45 The Situation .......................... ............................... 45 Recommendations .51 5. Water Issues and Recommendations .......... ............................ 59 The Situation .......................................................... 59 Recommendations .... ...... ................ .......................... 66 6. Power Generation Issues and Recommendations ........ ..................... 73 The Situation .................. ........... .... ...................... 73 Recommendations ............... .................................... 79 Annexes Annex 1 List of PPI Projects in the Road Sector in China . . 89 Annex 2 List of PPI Projects in Water Treatment and Waste Water in China . . 99 Annex 3 List of PPI Projects in the Power Sector in China . .......... ...... 103 Annex 4 Summary of Benchmarking Study ... 113 Bibliography ........................................................... 139 iii IV CONTENTS PDaURES Figure 1-1 Extcrnal Financing to Emerging Markets, 1996-2002. 4 Figure 1-2 Investment in Infrastructure Projects with Private Participationi in Emerging Markets, by Sector, 1996-2000. 4 Figure 1-3 Investment in Infrastructure Projects wNith Private Participation in Emerging Markets, by Region and Type, 1996-2000. 5 Figure 1-4 Investment in Infrastructure Projects with Private Participation in Emerging Markets, by Region, 199 1-2000 . 6 Figure 1-5 Foreign Direct Investment and Pnvatc I'ebt Flows to Selected Developing Countries, 1991-2000. 7 Figure 2-1 Basic Project Approval Process rfor Piiot BOT Projects .17 Figure 2-2 Approval Process for Projcct Proposal and Feasibility Study Rcport for a County-level Toll Road Project wvith 50 Million in Investment .17 Figure 2-3 Approval Process for Establislhing a Joint Vcnture or Wholly Foreign-Owncd Enterprise. 1 8 Figure 2-4 Operational Approval for Plant Location and Construction .19 Figure 3-1 Project Finance Loans by Country, 1997 .32 Figure 3-2 New York Water Revenuc Bonds ................................. 40 Figure 3-3 Global Private Equity Investncnt, 1995-2000 . . . . 42 Figure 3-4 Private Equity Investment in Chnma, by Source, as of 1999 . 43 Figure 3-5 Private Equity Investment in China, by Industry, as of 1999 .............. ... 43 Figure 4-1 High-Capacity Road Construction in China, 1982-2000 . . 46 Figure 4-2 Institutional Setup for Highway Development .49 Figure 4-3 Organizational Structure of a Provincial Toll Road Authority .56 Figure 4-4 Financial Structurc of a Provincial Toll Road Authority .57 Figure 5-1 Investments in Water Supply and Samtation in China, 1992-98 ..... 61 Figure 5-2 Water Tariffs in Selected Asialn Cities, 1998 . . 62 Figure 5-3 Water Tanffs in Selectcd Cities Worldwidc, 1998 . . 63 Figure 5-4 Water Tariffs in Different Cities in Zhejiang Province, 1998 .64 Figure 5-5 Investments in Water and Sanitation Infi-astructure by Source, 1992-98 .64 Figure 6-1 Projected Power Generation Capacity under Two Scenarios for Demand Growth, 2000-10 . 75 Figure 6-2 Foreign Direct Investment in Generation, 1990-99 .75 Figure 6-3 Private Power Projects bv Model Used .77 Figure 6-4 Training Models for a Pool Markct .81 Figure 6-5 A Vesting Contract for Differenccs with Single Buyers .84 Figure 6-6 Partitioning a Power Purchase Agreement from the Competitive Power Market . ......... 86 Figure 6-7 Tariff Structures in New South WVales (Australia), Argcntina, and Zhejiang Province .87 CONTENTS V TABLES Table 1-1 Models for Private Participation in Infrastructure. 2 Tablc 1-2 Real Investment in Infrastructure Projects wvith Private Participation, in Emerging Markets by Contract Types, 1996-2000. 5 Table 2-1 Inconsistencies between Key Regulations on Private Private Toll Roads .12 Table 2-2 Models of One-stop Shop Investor Services .21 Table 2-3 Tariff Approval Regimes for Road Projects in Selected Countries .26 Table 3-1 Risk Allocaton in Concessional Projects .29 Table 3-2 Issues Involved with Mortgages of Various Asset Types .36 Table 3-3 Local Share of Debt Financing for Power Projects with Private Participation, Various Asian Cotutries, 1997 .38 Tablc 3-4 Domestic Infrastructure Bonds Rated by China Chengxin International Credit Rating Company, by Sector, 1996-99 .41 Table 3-5 Market Capitalization of Infrastructure Assets on the Australia Stock Exchange, by Sector, 2001 .42 Table 4-1 Public and Private Highway Investments, 1981-2000 .47 Table 4-2 Highway Investments by Type, 1990-2005 .47 Table 4-3 Private Participation Development Models and Financing Financing Modes for Highways .55 Table 5-1 Selected Indicators on Urban- Growth and Water and Sewer Sewer Systems in China, 1991 and 1999 .60 Table 5-2 Average Water Tariff in China, 1986-98 .63 Table 5-3 Private Water Projects in China Closed by 2000 .65 Table 5-4 The Profit Sharing Scheme in a Negotiated Cooperative Joint Venture BOT Scheme in Harbin .67 Table 6-1 Selected Power Projects Using Different Models for Private Parucipation . 78 Table 6-2 Shares of Long-term Contracts and Spot Markets in Various Markets .83 Table A 4-1 Comparison of Philippine anld Hunganan Legislation (Summary) .114 Table A 4-2 Tariff Approval Regimes in Road Projects in Selected Countries .123 Tablc A 4-3 Sources of Private Equity for Infrastructure .127 Table A 4-4 Local Financing Porton .128 Table A 4-5 Companson of Four Asian Domestic Bond Markets .129 BOXES Box 2-1 International Experiences with the Legal Framework for Private Participation m Infrastructure .12 Box 2-2 International Experiences with Streamlined Approval Processes .20 Box 2-3 International Experiences with a Central Coordinating Office .22 Box 2-4 International Expenences with- the Regulatory Framework .25 Box 2-5 International Experiences with Tariff Regulation .25 VI CONTENTS Box 3-1 International Experiences with Security Issues .............................. 34 Box 3-2 International Experiences with Structuring Utility Bonds ..................... 40 Box 3-3 International Experiences with Domestic Bond Financing ..................... 41 Box 4-1 China's Experience with International Debt Financing ....................... 50 Box 4-2 China's Experience with the Zhejiang Expressway ...........................5 S1 Box 4-3 International Experience with Private Investment in Roads .................... 53 Box 4-4 International Experience with Provincial Toll Road Authority Partnerships with the Private Sector ...................................... 56 Box 4-5 China's Experience with the Shanghai Midway ............................. 58 Box 5-1 China's Experience with Cooperative Joint Venture BOT Schemes .............. 66 Box 5-2 China's Experience with a Negotiated Wholly Pnvately Owned BOT Scheme ................................................. 67 Box 5-3 China's Experience with a Tender BOT Scheme ............................ 68 Box 5-4 China's Experience with a Concession of a Water Distribution Network .......... 69 Box 5-5 International Experiences with Concessions of Water and Sanitation Systems .............................................. .... 71 Box 6-1 International Experiences with Competitive Wholesale Power Markets ........... 79 Box 6-2 International Experiences with Partitioning Existing Power Purchase Agreements ........................................... 85 FOREWORD since the late 1970s China's economic reforms have generated stellar growth, averaging 8 percent a year. As a result poverty has fallen considerably, with more than 200 mnllion people lifted out of extreme poverty. The economy is fueled by enormous flows of foreign direct investment-the highest in the developing world China has become a powerful and important player in the global economy, as well as an economic powerhouse in East Asia. Infrastructure has played a major role mn China's rapid development. Over the past decade the road network expanded by more thani 40 percent, water production grew by more than 50 percent, and power generation exceeded 300 gigawatts-making China the world's second largest energy producer. But while this performance is unprecedented, there is room for improvement. In absolute terms China has had considerable success in attracting foreign direct investment in infra- structure, including road, water and sanitation, and power projects. Still, the amounts account for a small share of foreign direct investment flows and for only about 10 percent of total investment in infrastructure. The government's plans to further liberalize infrastructure and to develop western China should ensure greater achievements-and challenges-over the next few decades. Inadequate infra- structure services arc a major obstacle to improving poor people's lives and to increasing economic activity in less developed regions In addition, to reap the expected benefits of World Trade Orga- nization accession, exporters and consumers will require cheaper, more reliable, and more efficient infrastructure services. Mecting this demand will require more than $75 billion a year in infrastruc- ture investment over the next decade. Thus the Chinese government must raise the resources to finance the additional investment while improving service quality and efficiency. Increasing the participation of the private sector-domestic and foreign-is an obvious policy option. Building on the successes and lessons of recent efforts, including pilot BOT (build, operate, transfer) projects, the governmient is further developing the framework for private participation in infrastructure. In doing so, it is drawing on best practices and useful experiences from other devel- oping as well as industrial countries. Above all, these other countries show that successful private participation in infrastructure requires appropriate umbrella policies and regulations that provide adequate comfort to investors and financiers while protecting the public interest. Approaches to private participation will likely vary by sector, depending on the level of reform and the strength of relevant institutions. But wvell-designed public-private partnerships can support government efforts to alleviate poverty and ease regional disparities. Most important, private participation in infra- structure can reduce the fiscal burden on pubhc agencies and Improve the targeting of subsidies to poor people, students, the elderly, and other disadvantaged groups. The report that follows was preparcd by the World Bank Group to help improve China's approach to privatc participation in infrastructure-focusing on roads, water and sanitation, and power generation-by expanding foreign direct investment and domestic financing in such projects. The initiative also compared China's experiences with those of other countries. Based on these analyses, strategic suggestions werc presented to the Chunese government. This report is the restult of numerous policy discussions and technical studies that would not have been possible without the support and cooperation of China's State Development Plannmng Commission and many othcr government agencies It is hoped that this wvork will help the Chinese government advance its infrastructure development plan. Jemal-ud-din Kassum Vice President, East Asia and Pacific Region The World Bank vii ABSTRACT S ince 1978, but especially over the past decade, China has undertaken extensive policy reforms Sand investments in infrastructure and has experimented with maniy models for private participa- tion in infrastructure However, none has been fully developed and implemented, and more than 90 percent of investments have come from government. A great challenge lies ahead with the huge infrastructure investments still needed Weaknesses in the legal and regulatory framework are obstacles to broader private participation in infrastructure, and a series of issues deserve separate attention in the transport, water and power generation sectors. This report reviews China's current framework for private participation in infrastructure, high- lights achievements from and issues with private investments in roads, water, and power generation, and offers recommendations for overcoming hurdles to wider private participation from academic research, cross-country experiences and domestic and international practices. ix ACKNOWLEDGMENTS The Private Participation in Infrastructure Framework Initiative was funded by the Public-Pnvate TInfrastructure Advisory Facility,' AusAID (the Australian government's foreign aid program), the World Bank, and the International Finance Corporation. The initiative encompassed three niam activities: [ Reviewing China's framework for private participation in infrastructure and identifying achievements in roads, water, and power generation-including assessments of a sample of completed transactions. LZ Comparing China with a group of developing and industrial countries and identifying uscful lessons. El Developing recommendations for improving pnvate participation in infrastructure in China. The initiative was managed by Michel Bellier, from the World Bank's Transport Sector Unit for thie East Asia and Pacific Region, and Yue Maggie Zhou, from the International Finance Corpora- tion's Private Sector Advisory Policy and Transaction Unit. This report was undertaken by the World Bank and International Finance Corporation at the request of China's State Development Planming Commission, in March 2000. Written by Yue Maggic Zhou and Michel Bellier, it is the result of collaborative research by many consultants and cxperts, including Australia New Zealand Bank, Bridge of Trust, Chwoon Ang Lim, Clifford Chance, Credit Agricole Indosuez, Economic Consulting Associates, Economue et Humanisme, Freshficlds Bruckhaus Deringer, Lmiklaters, Mitchell Stanfield & Associates, MottMacDonald, PhaceLift, SOGREAH, and White & Case. Colleagues within the World Bank Group advised the authors throughout the course of the initiative. They include: I] From the World Bank-Mats Andersson, Jitendra Bajpai, Noureddine Berrah, Elisabetta Capannelli, Phil Gray, Rainer Hartel, Raja Iyer, Ranjit Lamech, Edouard Motte, Alfred Nickesen, Neil Roger, Richard Scurfield, Keshav Varma, and Ming Zhang. El From the International Finance Corporation-Sujoy Bose, William Bulmer, Denis Clark, Jerry Esmay, Karn Finkelston, Denise Leonard, David Mackenzie, Philippe Marin, Jennifcr Wishart, and Wenqin Zhu. Special thanks go to Denise Leonard, manager of the Private Sector Advisory Policy and Transaction Ullt, for her generous support to the project and insightful comments on the report. This work would not have been possible without the full cooperaton of officials from the State Devclopment Planning Commission, particularly the Foreign Investment Division, who provided valuable guidance and comments An international Conference on Private Participation in Infrastructure was held in Beijing, Cluna, in 2001 to dcscuss the many issues raised in an earlier draft of this report. The conference was attended by internationally renowned experts and policymakers, private sector representatives- includinig legal advisers, technical consultants, strategic operators, and financial investors-and many state and local officials from the Chmnese government. Their views were taken mto account by the authors when prcparing this final report. The authors are also grateful to the many people who assisted the project team by providing source documents, participating in conceptual discussions, and performing peer reviews of drafts. I PPIAF is a multidonor technical assistance facility aimed at helping developing countries improve the quality of their infrastructure through pnvate sector involvement For more information on the facility, see the Wcbsite at htrp //www ppiaf org xi ABBREVIATIONS $A Austrahan Dollar ADB Asian Development Bank B Thai Baht BOO Build, Own and Operate BOT Build, Operate, and Transfer BROT Build, Rehabilitate, Operate, and Transfer BTO Budld, Transfer, and Operate CAI Cr6cht Agricole Indosuez CfD Contract for difference CFS Co-Financing Scheme CJV Cooperative Joint Ventures CPI Consumer Price Index CSOP Capital Structure Optimization Program DBFO Design, Build, Financc, and Operate ECA Export Credit Agency EIB European Investment Bank EJV Equity Joint Venture EPF Employee Provident Fund FDI Foreign Direct Investment FIE Foreign Invested Entity FSR Feasibility Study Report FX Foreign Exchange GDP Gross Domestic Product HK Hong Kong HK$ Hong Kong Dollar IAIDB Inter-American Development Bank IBRD International Bank for Reconstruction and Development IFC International Finance Corporation IPP Independent Power Producer IMF Intcrnational Monetary Fund IPO Initial Public Offering IPP Indcpendent Power Producer lV Joint Ventures KLSE Kuala Lumpur Stock Exchange LPVR Least Present Value of Revenue LRT Light Rail Transit MGS Malaysia Government Secunty MLA Multilatcral Agency NPL Non-Performing Loan NRW Non-Revenue Water NTHS National Trunk Highway System O&M Operations and Maintenancz PHP Philippmes Peso PPA Power Purchase Agreement PPI Private Participation in Infrastructure PPP Public Pnvate Partnership PRC People's Republic of China RAM Rating Agency of Malaysia RM Malaysian Ringgit xiii xiv ABBREVIATIONS RULIB Chinese Currency (Yuan) aRQT Rehabilitate, Operate, and Transfer SAFE State Administration of Foreign Exchange $DFC State Development Planning Commission 83C Securities and Exchange Commission $ESHX Stock Exchange of Hong KCong GET, State Economic and Trade Commission $MC Social Overhead Capital $DE, State-Owned Entcrprise IS? State Power $SP]? Small Power Producer T(ODP Tariff One Part TOT Transfer Operate and Transfer UK United Kingdom U$$ United States Dollar MFOE Wholly Foreign Owned Enterprise Vl:(O World Trade Organization EXECUTIVE SUMMARY S ince 1978, but especially over the past decade, China has made extensive policy reforms and infrastructure investments. China's highways now carry more than 15 billion tons of freight and 15 billion people a year. Since 1993 growth in traffic has averaged 8 percent a year, and since 1998 annual investment in roads has been more than 200 billion renminbi ($24 billion), or 2.5 percent of GDP. Between 1990 and 1998 the capacity of water production plants and the length of the water network increased nearly 50 percent-the result of 86 billion renminbi ($10 billion) in investment over the same period. China's power generating capacity exceeds 300 gigawatts, making it the second largest in the world. Though there has been great success over the past decade, a greater challenge lies ahead with the huge investment needed for the next decade. In his report on the new Five-Year Plan at the fourth session of the Ninth National People's Congress, Premier Zhu Rongji emphasized the importance of intensifying construction of water, transportation, energy, and other infrastructure facilities and of recognizing the enormous importance of strategic issues related to China's resources. To develop infrastructure and clear bottlenecks, some $750 billion will need to be invested over the next decade-with annual investments close to the peak level reached in the past. The task is even more challenging given the government's urgent plans to develop western China. Despite the enormous financing involved, over the past decade more than 90 percent of infrastructure investments came from the public sector. Private capital can be instrumental in building infrastructure capacity, allowmg government funds to be used for pressing social, educa- tuon, and health needs. Yet private capital accounted for just 10 percent of infrastructure invest- ments over the past decade. China has experimented with many models for private participation in infrastructure, but none has been fully developed and implemented. Most infrastructure projects involving the private sector have been negotiated and carried out through joint ventures between private operators and government or quasi-government entities. Although such arrangements provide political comfort, xv XVI EXECUTIVE SUMMARY they raise concerns about conflicts of interest between government's role as regulator and as the (indirect) owner of project companies. Xln accordance with the 1995 13CT C-rculaz, t pl pioz 307 (b4 c;:ate, transfer) projects have been procured thzough .. 2 -.i- b,! g-W'zi: mixed results. The projects are Chengdu Water, Laibin B Power, and. Changsha Power. Most project companies are wholly foreign-owned entities, which helps reduce conflicts of interest. T'hese projects have attracted a lot of attention from investors because they have strong political support from the cen- tral government as well as more stringent procuremenet procedures and tighter contract structures. However it is not clear how the BOT model can be uIscd more widely. ioreover, the pilot BOT projects have not had fully satisfactory outcomes, and BOST is not necessarily the most appropriate model for private participation in infrastructure. Thus China needs to implement other models on a pilot basis. Chinese laws and regulations are a majoir obstacle to boender privaze parEcipation in Llfrz- structure. Among the weaknesses: A comprehensive legalfiraimerv0-ok 2is ;zeede to co;zsolidate th-e va-,-iGFs pieces of legislartioni. In recent years China has made tremendous progress in improving the investment environ- ment, issuing a series of crucial laws-including the Bidding Law, unified Contract Law, Security Law, and Project Finance Measures. Still, a comprehensive, consistent body of legislation is needed for private participation in infrastructure. In particular, the BOT Circular has limitations and should be replaced bv a comprehensive framework law for private infra- structure projects. A framework law could establish a basic legal structure applicable to a wider range of private investment models (leases, concessions, and so on) and funding sources (foreign and domestic) than are covered by the BOT Circular. A framework law could also establish basic contract rights for private infrastructure projects, mcluding the right to collect tariffs, the right to receive termination payments, and the right for lenders to "step in" to a project in cases of default. 7The legal systern leaves many importmant, i'outi2e edecisioes to aw zinistwative authorities, througb an approval process that shonld be sweamlined. To start a project, an investor (through the local government) has to obtain approvals from many government ministries, bureaus, commissions, and agencies. Statc Development Planning Commission approvals are required for most project proposals and feasibility studies for projccts involving foreign capital and more than $30 million in investment. State Administration of Foreign Exchange approvals are required for foreign financing, security packages, and the like. Approvals are also required to establish and register tlhc project company, followed by a third layer of approvals for starting operations. The cumbersome approval process is sometimes redun- dant and overlapping, with unclear legal effcects, and has often created delays and uncer- tainty in project initiation and disputes in implementation. Many efforts are needed to streamline the approval proccss. Legislation on approvals should be tightened The number of post-award approvals should be minimized. Approval authorities should be given more specific mandates. Central and local government responsi- bilities and authority should be clarified. Approval requirements should be made more transparent. Some approval procedures should be standardized. Communication betwecn central and local approval authoritics should be improved. The $30 million threshold should be raised where appropriate, and critcria for excmption from central approval should be clarified. The practices of one-stop approval shops should be reviewed. A standard approval process should be introduced for small projects. Long-term commitments should be given to approvals of tanff adjustments. [r Increasing private participation in infrrastractzire has pnt pressure oz- local gover,nmeats to steengtben their capacity asg g,ating avsator7ities. Building capacity wvill require local EXECUTIVE SUMMARY XVII governments not only to share knowledge, skills, and best practices but also to secure strong commitment and technical assistance from the central government. The central government should prepare and disseninate standard materials for private infrastructure projects, promote and publicize the framework for private participation in infrastructuLre, develop pilot projects based on best practices, provide technical support, establish a net- work of expertise among granting authorities, and benclmark utilities. To perform these tasks, the central government could first estabhlsh a central coordinating office. El The rules, institutions, and instruments of independent regulation bave not been devel- oped. Independent regulation is especially important for infrastructure because of infrastruc- ture's direct influence on tlhe public and its monopolistc nature. Independent regulation should be guided by at least twvo basic principles: the institutional structure should clearly separate policy and regulatory functions from ownership and management responsibilities, and the regulator's decision making process should be independent and transparent. Most of the private investment in China's infrastructure has been foreign and has been pri- marily equity, with little debt financing. To mobdize domestic financing and to ensure sustain- able foreign capital inflows-which have slowed in the wake of the East Asian financial crisis-several issues should be carefully addressed: Ta The need for sustainable financing for infrastructure projects requires that risks be bal- anced between public and private parties. Some Chinese officials have expressed concern that certain contracts for private infrastructure projects provide investors with excessive returns anud require government to assume too many risks, especially commercial risks. Such contracts, to some degree, reflect China's tiadequate laws and regulations investors demand favorable commercial terms to compensate for legal and regulatory risks, which should instead be borne by government Only after these nsks have been mitigated will for- eign investors lower their requirements for commercial terms. This principle also applies to domestic private capital, as domestic investors will seek sirmilar protection against the risks that have hindered foreign investments. Although risks vary by project, international best practice is that risks should be carefully assessed and allocated prior to contract signing or bidding. Furtlhermore, some risks can be mitigated more effectively by introducing performance-based contracts El Project Finance Measures need to be better defined. Project financing has not been a major form of funding for private infrastructure projects in China. The main concerns involve the credit risk of the counterparts and the constraints and uncertainties arising from the 1997 Project Finance Measures and 1995 Security Law. To get a foot in China's market, foreign investors in the three BOT pilot projects relied on their corporate balance sheets to secure debt financing from banks. But this approach is not replicable or sustain- able for investors who wvant to pursue a lot of investment projects in China. Investors have come to realize that the interpretation and application of the Pro)ect Finance Mea- sures is not simply a matter of predictable lawv and regulation. Rather, the granting of project finance approval is an investment policy tool as well as a means of regulating for- eign loans. While governments have valid concerns about the key issues involved in a typ- ical infrastructure project-the scale of capital investments, foreign exchange obligations, potential effect on domestic inflation, possible risk exposure of state entities through foreign debt and guarantees, and so on-more clarity could be provided to investors to broaden the scope of pro)ect financing, better define the applicability of Project Finance Measures, and reduce dependence on the complex approval process. 0 Although the Security Law has played a crucial role in China's efforts to provide an enabling investment environment, its implementation should be strengthened. Investors, especially lenders, have encountered many problems with the creation and enforcement of xviii EXECUTIVE SUMMARY securities, whether guarantccs, mortgages, or plcdges. This is partly duc to underdcvclopcd property rights and courts, and partly to wealnesses m rcgulation and administrative systems. - The registration system needs to be strengthened by providing better training to rcgistra- tion staff, incorporating securities that existed beforc the system was establishcd, unify- ing registrations of security on land and buildings, and making r-egistraLion records available to the public (or at least potential lenders). - The prohibition on gva;-a;ztees in a number of laws should bc ciarified anld narrowved, with more flexibility for large infrastructurc projects. For examplc, guarantees on tcrmi- nation paymcnts should not be considcred repaymcnt guarantecs, wrhich are prohibitcd. - Security on contract rights, which is crucial for infrastructurc projects to obtain dcbt financing, should be providcd by granting a lcgal basis for the full assignment of contract rights for infrastructure projects, granting full security on such contract rights in thc Security Law, and establishing a registration system for such security. - The provision and enforcement of mortga,ges can bc improved by strengthening the rcg- istration system on grant fees for land and clarifying legal provisions for third-party rights, including easement rights; unifying land and building registrics; allowing mort- gages on future or after-acquired property, such as allowing a land mortgage to cover buildings to be constr-ucted on the land, or a mortgage against machinery and equip- ment to include after-acquwred equipment; permitting lendcrs to unilaterally carry out registration based on a prior mortgagc; providing a legal basis for floating chargc; and providing a legal basis for maximum amotunt security (security over accounts receivable, inventories, fluctuating credit balances, and the like). - Restrictions on domtestic e;zities providing secisrity to foreign entities should be relaxed, especially when foreign currency loans have already been approvcd and rcgistered. - Claims on secured interests needs to bc cffcctively enforced by disassociating asset appraisal firms from local governmcnts, giving Icnders opportunities to present their rcc- ommendations in front of a judgc in state entcrprise bankruptcy cascs, and granting lenders clear legal right of appeal to a hlgher court. OI Domestic banks play a limsited role in fin-o3ncing infirastvnctnisre p;'oqjets because of interest rate controls imposed by the People's Bank of China, project finance rcstrictions imposed by the State Administration of Foreign Exchange, the Icnding and risk assess- ment capacity of domestic banks, and other constraints. Given the natural hedge that domestic currency financing can offer, bank reform (though beyond the scope of this report) should be expedited. [Ii Bond,s also provide lisited fr;3diqg for i;?f rasitrctn 7'e projects. Bonds can be instrumcntal in facilitating private participation in infrastructure. Relative to banks and equity investors, bondholders take less risk and so require highcr financial disclosurc and a tighter security package. Development of the bond market is beyond the scope of this report, but it requires introducing mandatory credit ratings, relaxing pricing restrictions imposed by the People's Bank of China, and removing regulatory restrictions on primary market participa- tion by insurance companies and pension funds. In addition to the issues affcting aL. seczomrs, tHise c 2': s - g roadEs, mez, and power gen=- ation deserve separate attention, given sec=or-speci;c needs for .; - . s . q or and di- ferent stages of reforn. Private participation models for roads have bccn limited to joint venturcs and securitization of assets, and even these have experienced difficulties. To broaden private participation, including by domestic investors, government policy toward the public and privatc sectors must be clarified. As a first step toward increased financing for roads, espccially for projccts not attractive enough on their own (such as those in western China), provincial toll road auth1orities could 'be created to consoli- date projects into a viable pool for financing. EXECUTIVE SUMMARY XIX For water the key issue is the financial viability of water and sanitation utilites. Sector reforms, especiafly tariff reform and further corporatizatson of water and samtation utilities, is of particular importance These utilities shoLlld be operated on a commercial basis, with better investment plan- ning to improve service, independent management and professionally audited financial statements, accountability for performance and easy monitoring of performance data by the public, and more transparent subsidies (if needed) targeted at the poorest groups. Furthermore, most private invest- ment in the sector has involved BOT schemes for water treatment plants financed by foreign capi- tal. Despite potential savings, direct private partictpation in wvater networks has not been aHlowed. However, it was reported in recently that the government plans to end the state monopoly on the vast urban water supply system by allowing private investors to build water distribution systems- welcome news for investors This initiative could be carried out through pilot concessions for large systems and by awarding leases, management contracts, or small-scale concessions to smaller sys- tems to domestic private investors. Finally, although water and sanitation services are mumcipal responsibilities, strong technical assistance from the central,government is essential for sector reforms and private investments to proceed smoothly. For power, an ongoing policy dialogue between the government and the World Bank is pro- moting liberalization of the sector through market reform. Thus this report only examines weak- nesses in the current transitional market structure that are not conducive to pnvate investment, such as weaknesses in the legal and regulatory framework that prevent expansion of pilot pool mar- kets, the lack of a mature structure for market governance, limited access to market settlement information that could facilitate long-tern planning by investors, and so on. Just as important is identifying ways to adapt existing lon,g-term power purchase agreements to fit into the pool market structure, and to ensure investors' interest and confidence in the sector. Investors' interest in the power purchase agreements can be protected in the context of market reform by using an interme- diary (like the market trader in Australia pool market) that isolates the agreements from the obliga- tions imposed by the market rules, or by adopting a vesting contract for differences that provides for the transferability of the rights and obligations under such contracts (including the assignment or transfer of such rights as security for financial obligations). Furthermore, investors will not buy into the pool market concept unless cross-ownership of assets by the government is resolved. Thus new generation capacity needs to be procured through various private infrastructure models-but not with the traditional joint venture approach, which creates conflicts of interest between govern- ment's role as both owner and regulator. Divestiture ofgeneration assets by the State Power Com- pany also has to be accelerated This report studies the above issues in detail and offers recommendations for dealing with them. The recommendations are distilled from academic research, cross-country experiences, and domestic and international best practices. If the recommendations are consistently and systematically implemented, they should aid considerably in facilitating private infrastructure investment in China XX EXECUTIVE SUMMARY SUELIJA],Y OXF JECOMLFNJDATHTnNS Recommendations for the LeaS RC r = Recometndwtions Details Enact a framework law for private participatton in nifrastructure The law should: _ Establish a framework of laws for private inf-astructure investments. [ Apply to all private infrastructure projects. ' Cover both foreign and domestic financing. - Establish basic contract rights for private infrastructure projects, including the right to collect tariffs, the right to receive termination payments, and the right for lenders to "step m" in cases of default. LJ Provide flexibility. I Refer to model contracts to facilitate implementation, but not oblige parties to use standard contract terms. _ Grandfather all existing projects. '7 Be consistent with sector laws. Streamline the approval process The central government should: iE Tighten legislation on approvals. U Minirmze the number of post-award approvals. i Assign more specific mandates to approval authorities. U Clarify the division of responsibilities and authority between the central and local governments. EC Make approval requirements more transparent. El Stanldardize some approval procedures. C1 Improve communications between the central and local approval authorities. LU Increase the $30 million threshold where appropriate and clarify criteria for exemption from central approval. C Review the practices of one-stop approval shops. IImplement a default approval process for small projects. H Provide long-term commitment to approvals of tariff adjustments. Strengthen the capacity of,granting authorities The central government can help granting authorities by: Preparing and disseminating standard materials on private infra- structure investment. El Promoting and publicizing the framework for private participation in infrastructure. ' I Developing pilot projects based on best practices. Providing technical support. L, Establishing a network of expertise among granting authorities. E Benchmarling utilities. To perform these tasks, the central government should consider establishing a central coordinating office. EXECUTIVE SUMMARY XXI Establish an independent regulatory framework Transparent framework The framework should establish: El Regulatory rules. a Legal instruments containing the rules. El The institutional structure of the regulatory agency. Independence of Independent regulation requwres establishing at least two basic principles: regulator al Regulators should be established in a way that clearly separates policy and regulatory functions from ownership and management responsibilities. O Regulators' decisionmaking should be transparent and independent. Institutional structure Careful consideration should be given to whether: of regulator n Regulatory agencies cover an industry, a sector, or several sectors. El Regulatory agencies are established at the national or local level for each sector. Regulatory methodology Several types of economic and tariff regulation should be considered: El Rate of return. El Cost plus pricing. E Pnce-cap regulation. In many cases a combination of methodologies can be used. Recommendations for the Financing Framework Recommendations Details Establish an efficient mechanism for allocating risks Risk mitigation The party best placed to control a certain kind of risk should assume and manage that risk. E The government should assume legal, approval, and regulatory risks E The private sector should assume commercial risks. El Some risks should be shared between the public and private sectors. Some risks can be mitigated more effectively by introducing performance-based contracts. Negotiation of Risks should be carefully assessed and allocated prior to contract risk allocation signing or bidding. Modtfy the Project Finance Measures El The scope for using project financing should be broadened. El The applicability of the Project Finance Measure should be better defined. El The approval process for project financing should be streamlined. Strengthen enforcement of the Security Law General Property nghts should be reformed and the court system should be deepened. XXII EXECUTIVE SUMMARY Registration of security The registration system should be strengthencd: Training shouid be provided to registration staff. Sccurities that existcd bcforc the rcgistration system was establishcd should be incorporatcd into the system. Registrations of sccurity on land and buildings should be unifled. ' Registration records should be made available to the public. Guarantees The prohibition on guarantees in a number of' laws slhould be clari'fied and narrowed for largc infrastructure projects. ror examplc, guaran- tees on termination payments should not bc considered repayment guarantees, which arc prohibited. Secunty on General legislation should provide a legal basis for the full assignment contract rights of contract rights for infrastructure projects. The Security Law should providc for full security on such contract rights. A registration system should be established for such security. Mortgages Land-strengthen thc registration system for grant fees and clarify legal provisions on third-party rights, including easement rights. ' Buildings-unify land and building registries. Future or after-acquired property-allow land mortgages to cover buildings to be constructcd on the land, and mortgages against machinery and cquipment to include aftcr-acquircd equipment; and permit lenders to unilaterally carry out subsequent registration based on prior mortgages. Floating charge-provide a legal basis for floating charge. Mortgage of a maximum amount-provide a legal basis for security over accounts rcceivablc, inventories, and fluctuating credit balances. Security granted Once foreign exchange loans have been approved and registered, to foreign entities restrictions should be relaxed on domestic entities providing security for these loans. Enforcement of Separate asset appraisal firms from local governments. secured claims Provide lenders with opportunities to present their recommenda- tions in front of a judge in state enterprise bankruptcy cases. Provide lenders with- a clcar legal right of appeal to a higher court. Insurance [I Allow privatc infrastructurc projects to place part of the msurancc with international insurance companies rather than just domestic ones. Consider allowing lenders to take sccunty on reinsurance proceeds. Promote domestic bank lending Loosen People's Bank of China interest ratc controls. i Loosen State Administration of Foreign Exchange restrictions on project finance. Strength en domestic banks' capacity in risk assessment and structurcd financing. Promote participation of foreign banks in renminbi Icnding Promote bond financing General Encouragc high financial disclosure and tight financing and secunty structures. EXECUTIVE SUMMARY XXIII Domestic bond market O Require credit ratings. O Loosen People's Bank of China pncing restrictions. El Remove regulatory restrictions on primary market participation by insurance companies and pension funds. Recommendations for Roads Recommendations Details Policy Establish clear policy on government's role in the sector, encouraging wider private participation. Institutional structure Corporafize toll road entities into provincial toll road authorities Domestic participation Foster the development of domestic private sponsors Recommendations for Water Supply and Sanitation Recommendations Details Sector reform Pursue water tariff reform and further corporatization of water and sanitation utilities, which should be operated on a commercial basis with: E Better investment planning to generate specific improvements in service. El Independent management and professionally audited financial statements. a Accountability for performance and easy monitoring of perfor- mance data by the public El More transparent subsidies (if needed) targeted at the poorest groups. Private investment E Develop pilot projects for concessions of complete water and in water networks sanitation systems. E Promote domestic private investment in the water sector. Central government The central government should help municipalites carry out such commutment reform, including by helping to establish systems for benchmarking and sharing informationl among water and sanutation utilities, and by implementing pilot projects. Recommendations for Power Generation Recommendations Details Market structure The structure of the power market should be strengthened to accom- modate the forthcommng liberalization of the wholesale market. El Establish a consistent legal and regulatory framework, including by amending the Electricity Law and developing market rules that cov- ers, at minimum, the responsibilities of the market operator and effective regulation of the wholesale power market. E Establish a market governance structure, including an indepen- dent regulator, to ensure independent and economic merit order XXIV EXECUTIVE SUMMARY dispatch, open acccss by gcnerators to trarnsmissior assets, and a transparent and non-discriminatory approach to transmission pricing. Enhance information access by rcleasing information to the public on market rules and othcr legal instrumcnts, short- and long-term power plans of provincial powver companics, anld results of pilot market operations, including (at a minimum) the market price and actual systcm demand. Protect existing powvcr purchase agreemcnts (see below). Develop the financial securities market for hedging instruments such as contracts for differenccs. Improve the creditworthinecss of market participants by requinng provincial power companies to increasc disclosure of their financial in formation Dercgulatc ftiel markets. Existing poNver Existing powver purchasc agreements should be protected. This can be purchase agreements achieved by introducing a market trader as an intermediary to isolate the agreemcnts from the obligations imposed by market rules or by adopt- ing vesting contracts for differences that provide for the transferability of the rights and obligations under such agreements (uncludmng the assign- ment or transfer of such nghts as secunty for fmiancial obligations). Generation assets Continue to procure new gencration plants through private infrastructure modcls, but avoid the joint venture approach. Expcdite the divcsture of generation assets from the State Powver Company I INTRODUCTION P rivate participation in infrastructure provides significant benefits to governments, consumers, and economies. It: * Expedites investments in infrastructure and frees governments from heavy administrative and fiscal burdens. * Lowers the cost of public scrvices by increasinig efficiency and improves performance through output-based contracts and incentives. * Transfers capital, managenal expertise, and tcchnological innovations to a country or region, stimulatmig the growth of domestic private infrastructure industries and capital markets. * Shifts risks to private investors and operators, who are often better at handling them. * Supports social and economic policy objectives, including poverty alleviation These benefits explain the rapid worldwide growth ui infrastructure projccts involving private partici- pation since the 1980s, especially in such sectors as power, water, transportation, telecommunica- tions, natural gas, and waste management Chile, for cxample, has privatized its power, gas, and telecommunications industries. In addition, the pnvate sector provides most water supply, sanitation, port, and highway services. Airports and railroads have mixed ownership, with both public and private participation. The private sector also participates in sectors mostly owned by the state, such as irrigation and urban roads (Jadresic 2001). In Australia large-scale privatizations have attracted $A60 billion in private investmcnt-$A16 bllion of it in 1999 and 2000-in infrastructure assets including power facilities, tolI roads, water treatment, gas transmissionl and distribution, railroads, airports, pnsons, hospitals, telccommunucations, and ports (Porter 2001). Concepts and Models for Private Participation in Infrastructure In most models for private participation in infrastructure, a government or quasi-government entity responsible for delivering a public service contracts a project company to take on all or part 2 WORLD BANK WORKING PAPER of the responsibilities for operating or financing the service (or both). The public entity is the granting authority, and the project company is set up by or wNith private m cstofs. TIable 1-1 shows the main models along a continuum of private scctor involvement, from the lcast (servicc and man- agement contracts) to the most (divestiture). The model a government clhooses mainly decpens on its objectives for private participation-social and economic considerations, investment needs, ser- vice quality, technology transfer, and so on. Service and Management Contracts and Leases With service and management contracts, the private operator (project company) does not have a-, ownershlp or investment stake and is responsible only for managing a facilitv or utility. At the next level are leases (also known as affermage), wvhere the privatc contractor pays a Ieasing fee and is responsible for providing the service at its own risk. WVith leasing the project company operates and maintains the infrastructure but usually does not invest in necv facilities or mn&-Iastructure. Service and management contracts and leasing arrangcments are often used when: . Governments want to obtain the benefits o private management or teclnology but the pii- vate sector considers the political, financial, tech-nical, or economnc risks too high to oxvwn or invest in the system. i Governments are reluctant to relinquish ownership due to their desire to maintain control or unwillingness to implement needed reforms. Continued government ownership provides access to low-cost funds (such as tax-exempt or donor funds) or grants that otherwise would not be available to a private utility. A state enterprise is in bad shape, and a service or management contract is being used in an attempt to turn it around before privatization. Many water projects in Africa involving private participation use management contracts because technical and collection risks have been perceived as being too high and utilnies are not consider2d creditworthy. Modified leasing arrangements were recently used in the Philinpines to introduec private management of water services in small towvns. Concessions and BOT and BOO Schemes Higher private commitment can be achieved through a concession or a B' OT (build, operate, transfer) or BOO (build, own, operate) scheme. Under a concession a pnvate entity mnanages a public ser- vice for a given period during which it assumes significant investment risks and some commercial risks. The concessionaire is responsible for delivering services to users in accordaance with terms and conditions specified in the concession contract but is usually allowed to choose the means for T i- 0llF .- 11. 1 4J I ! 1g1hI .-.il 1II)1 02 -v<< '-vCsC t=Za 0wnership qf assets Public Public Public Public/Private Private CzRW iinvestments Public Shared Private Private Private perational ef:e,r Limited Yes Yes Yes Yes M (3)1 - /'/4 County granting authority |Project Proposol Approa | Feasibility Study Report Approval Source. Asian Development Bank 18 WORLD BANK WORKING PAPER Project Company Approval Numerous approvals are also needed to establish and register the project company in accordance with company laws, including the Company Law of 1994, and the Joint Venture Law or Wholly Foreign-Owned Enterprise Law. The Foreign Investment Advisory Service (FIAS), a joint agency of the Internal Finance Corporation and the World Bank, advises governments on how to improve the business environment for foreign investment. According to a recent study it performed for a provincial government in western China, up to 18 approvals are needed to establish a company in that province (Figure 2-3), some of which come from the same approval authorities but must be obtained sequentially on different visits. Operational Approvals Once the project company has been established, more than 10 other approvals are needed to acquire land and start construction (Figure 2-4). Many of these approvals could be eliminated or FIGURE 2-3: APPROVAL PROCESS FOR ESTABLISHING A JOINT VENTURE OR WHOLLY FOREIGN-OWNED ENTERPRISE State State Ministry of Industrial Technology Development Economnc Foreign Trade and and Supervision Planning and Trade Economic Commercial Bureau Commission Commission Cooperation or its Registry local branch '\) (3) I) 7a C) (65 (2)yZ (8)/ (5) - ~~~~~~~~~12) Atot /(8) /17) \85 Social Employment 11Expatnate Foeg 'AExhag Secunty Service Center Work Institution | Bureau Registry Permit lAddress l _ _ I ~~~~~~~~Registry l Personnel and Public Labor Dept Security lou: F.Bureau Source: FIAS. PRIVATE PARTICIPATION IN INFRASTRUCTURE IN CHINA 19 0**. . ~~~ ..*, 0 *S0 Investor Project Comipany Approvals Cultural and Antiquity Protection (Figure 2-3, (I) - (8)) \] H ~~~Agriculture Bureau Urban Planning r e SBure auS. expeitedthrogh adetaledzonig pln, pblised bth governet specifyings acptbeus Construction Finlly tCommission e p t ars C ans arealo t by~~~~~~ ~ sit inpcn frmnmroslclgoenetagniste poic,environme nta bueu wom ' aState Land Bureau on-tt cpanynora:bsnes tralinn h Aprv l Processmuntls China' cmEnvbronment o i underelod leal s h Protecsm on Bureaunt a atht T t Source. FIAS. expedited through a detailed zonung plan, published by the governmcnt, specifing acceptable uses of land This approach is common in many industrial countries. Finairy, the approval process is not over when the project starts. Comnpanles are also bothered by sdte inspectaons from numerous local government agencies-the police, environmental bureau, women's association, and so on-tcat creatp interference in the company's norlal bfasiness. Streamlnigng the Approval Process China's cumborsome approval process is a dircct outcome of its underdeveloped legal system, which leaves many important a-id routinc decisions to administrative authorities. Thus the first step in improving the approval process should be to tighten the legislatnon. Many laws are underdevel- oped due to uncertainty in econom--ic policies (see below). It remams extremely challenging for Chinese policymakers to develop well-defined, well-articulated economic pohlcies. Still, better approval procedures are possible even within the current legal framework. Post- bidding approvals increase uiicertainty and so reduce the benefits that could be gained from competition. When a projecr is procured under compentitve bidding, approvals should be advanced as much as possible to the pre-bidding stage, and the number of post-award approvals should be minimized Past projects show that a pre-bidding approval and consultation process could shorten project closing by 2-3 months with no loss of public oversight. In addition, responsibilties should be more clearly delegated between central and local approval authontles, and rules and regulations for the approval process must be made more transparent. The 20 WORLD BANK WORKING PAPER government could also consider standardizing some approval procedures-for example, usmg stan- dardized inquiry and submission formns. Investors are most concerned about uncertainty and delays with approvals at the central levcl. Such concerns are increased by local govcrnments' own confusion about the approval process and by their efforts to circumvent central rcquirements, only to learn later that a project is not permit- ted by the central government. Thus, better communication is needed between the central and local governments. In addition, the $30 million threshold for central approvals needs to be increased because all infrastructure projects cost more than that. This unrealistic threshold simply encourages local governments to Lmplement projects piece by piece, eliminating any benefits fromn economies of scale. Moreover, a clearer definition of "comprehensive balancing" is needed so that investors and local governments will know whether a project that falls undcr the "encouraged" category but involves more than $30 million in investment can be approved locally rather than by the central State Development Planning Commission. One-stop shops have been created in many parts in China, as in many other developing countries where the government is unxvilling to let foreign investments proceed without screening (Box 2-2). Since the late 1980s Shanghai has had an Investment Commission made up of staff from four agencies that previously shared power for approving foreign investments (municipal branches of the State Development Planning Commission, State Economic and Trade Commis- sion, Ministry of Foreign Trade and Economic Cooperation, and Ministry of Communication). The commission is the municipality's sole foreign investment approval authority, responsible for approving all stages of foreign investments. Moreover, its approval guarantees all subsequent approvals and permits required by other agencics (such as those in charge of land, environment, electricity, water, and telecommunications). There are many types of one-stop shops (Table 2-2). The setup of a shop is less important for its success than the government's political will to truly facilitate the screening process. One-stop shops are usually successful if there is strong support at the highest levels of government. But withl- &0)9 1@ n, -1l!. , --Ar,t - J.. 1 .- ,-21, i: A !Ii! 1 , F I q. , J The Philippines established a One Stop Action Center to help investors understand the process of obtaining consents and permits. The center houses representatives of different government agencies responsible for accepting and processing investment applications. The representatives have the authority to act on all invest- ment matters under their jurisdiction, which helps ensure thc center's effectiveness. Through Invest Australia, Australia's federal government runs a Major Projects Facilitation service that provides investors with advice and information on approval processes and facilitates major investment projects. A prol- ect is eligible for assistance if it involves capital spending of more than A$50 million, requires federal govern- ment approval to progress, and the developer can demonstrate the project's commercial viability and the developer's readiness to proceed through the approval process. If a project is granted Major Project status, Invest Australia identifies relevant federal approvals and prepares timelines for all approvals at all government levels. It also coordinates federal and state or territorial requirements so that they progi-ess in line with com- mercial requirements. It also identifies government policies and programs that may benefit a project. It has been suggested that the federal government make the Major Projects Facilitation service available to all transjurisdic- tional projects. Providing all relevant approvals under a singles statute is one way to simplify the approval process. This approach is used in some Australian states and was used in some early transport projects in the United KCingdom, with toe government passing legislation that enabled the private sector to avoid making separate applications to relevant agencies and local authorities. Source: Clifford Chance and Credit Agricole Indosuez. (Annex 4) PRIVATE PARTICIPATION IN INFRASTRUCTURE IN CHINA 21 PA M i' iO . i0 .1 True one-stop shops Investor servicing centers are granted, by law, the power to approve invest- ments, grant licenses, and make decisions related to investments. Few coun- tries have tried to establish entities with comprehensive legal power on investment issues, and most attempts to do so have failed for political and regulatory reasons. Delegated powers model In this model (used in Indonesia) government authorities delegate their approval powers to investor service agencies within certain limits. In most cases only limited powers, covering relatively minor approvals and licenses, are delegated. Hence this approach has marginal effects. Single-window model Empowered officials of relevant authorities are required to work together in a single office to provide investors with needed clearances, licenses, and approvals. This model is expensive and hard to maintain in countries with few investors and varying service requirements. In such cases the model becomes a clearance point for various agencies and is ineffective at delivering quality investor services. Account executive model This model can be effective when applied by full-service investment promo- tion agencies. Each account executive is assigned a portfolio of investors for whom the executive is required to provide a full range of services, including facilitation of approvals and licenses. The executive serves as a point of con- tact and information and uses a network of government contacts to aid the investor. This approach has been applied successfully in Ireland and Singapore. Source: FIAS. out adequate support a shop can become, in the worst case, a "one-more-stop" shop, creating an additional hurdle for investors (FIAS 1993). Some cities have adopted a default approval process for small projects, where a project is con- sidered approved if the relevant approval authority does not respond within a certain period. But the legal implications of this process are unclear. Because most applications for approvals are sub- mitted by local government officials on behalf of the project companies, the private participants are always conccrned about what was submitted for approval and whether an existing approval was granted properly. Finally, uncertainty about the long-term enforceability and adjustability of tariffs is a major deterrent to infrastructure investment. National and local price bureaus approve prices only on a short-term basis and are reluctant to formally commit to future pricing. With limited certainty on prices and revenues, investors either negotiate hard for a guaranteed rate of return or invest only in small, equwty-financed projects, where local political support is easier to obtain for future tariff adjustments. The BOT Circular addresses these issues in its model concession agreement-a step in the right direction. Granting Authorities Although the clear regulatory and contractual arrangements of the pilot BOT projects have gen- erated a lot of investor interest, local governments have not been as enthusiastic. The main reason is that the public tender approach requires local governments to prepare bidding documents- a time-consuming, challenging task. With the joint venture approach, a local government usually works with the foreign partner from the beginning to review and design the project, usually at no additional cost to the government. Thus the tender approach may entail higher transaction costs and greater risks that the project is not properly designed. 22 WORLD BANK WORKING PAPER Although the concerns of local governments are valid, public tendering is the only effective way to introduce fair competition and to curb corruption. Higher short-term transaction costs, before local governments build their capacity in handling infrastructure projects, are a price worth paying. Moreover, international competitive bidding will also motivate local governments to strengthen their capacity as granting authorities. To help strengthen the capacity of local granting authorities, the central government can: * Prepare and disseminate standard materials on private participation in infrastructure, including contract templates, technical indicators, and tariff adjustment mechanisms. In doing so, the central government must make it clear that these standard materials are only for reference, and that specific contract terms should be left for negotiation between local granting authorities and potential investors, in order not to pose unnecessary rigidity on local governments. * Promote, monitor, and publicize implementation of the framework for private participa- tion in infrastructure, and develop pilot programs based on domestic and international best practices. * Provide technical support to local governments, such as help with the feasibility study and advice or resources throughout procurement and project implementation. * Establish a network of expertise among granting authorities to facilitate the sharing of knowledge and best practices at the local and national levels. * Expand and improve the benchmarking of utilities, to provide incentives for privatization. This assistance could be provided by a newly established Coordinating Office, similar to the Trea- sury Taskforce in the United Kingdom (Box 2-3), staffed by individuals with strong private sector experience. The office could also be the focal point for general investor queries on private partici- pation in infrastructure. The Regulatory Framework Regulation is another important aspect of government involvement in infrastructure projects. Reg- ulation is especially important for infrastructure because of the sector's direct effect on the public and its monopolistic nature. iFor example, the regulator has to ensure that low-income users have * *_W .._* The U.K. Treasury Taskforce helped central government departments develop a better understanding of the private sector's requirements and expectations. It succeeded in doing so because it was staffed by individuals recruited from the private sector, though the taskforce operated on the side of the granting authorities. This approach contrasts with the traditional pool of civil service recruits in the United Kingdom and most other countries-well-educated individuals with no commercial training or experience working in the private sector. The Treasury Taskforce was so successful in promoting private participation in infrastructure that, at the end of its two-year life span, two bodies-the Office for Government Commerce and Partnerships U.K.-were created to take over its role. The Office for Government Commerce is intended to provide government departments with a central resource of procurement skills, promoting best practices in the public sector. One of the office's first tasks was to establish a common framework for all government procurement. Partnerships U.K. is a company that will eventually sell some of its shares to the private sector. It will work in partnership with the public sector but will charge for its services, and will help the public sector obtain the best possible deals in private investment programs by providing project advisory and implementation skills. Source: Clifford Chance and Credit Agricole lndosuez. (Annex 4) PRIVATE PARTICIPATION IN INFRASTRUCTURE IN CHINA 23 affordable access to privately operated public services. Thus an effective regulatory framework is crucial. It consists of thrcc elements O Rutles-governing tariffs and tariff adjustments, service standards, penalties for noncompli- ance, and so on. l Instruments-the laws, decrees, contracts, and licenses that contain the rules. O Instituttons-the agencies that enforce the ruiles and update them as needed. The instruments cstablish the prmciples of independent regulation, which clearly separates policy and regulatory functions from ownership and management responsibilities, and which ensures that the regulator's decisionmaking process is transparent and independent. This section focuses on the institutions and the rules, particularly for tariff ad)ustments. Institutional Structure Infrastructure regulators monitor and enforce regulatory and contractual rules on pnces and service quality. Among a regulator's most important tasks are awarding licenses and concessions, admin- istering rules in licenses and concessions (such as on prices), settling disputes between parties (governments, operators, consumers), monitoring operators' compliance with regulatory norms, and prosecuting operators for noncompliance, including imposing penalties. A regulatory agency's design is important in ensuring effective regulation. Important design issues include: E Independence or autonomy. E Coverage-does the agency regulate an industry (such as power or telecommunications), a sector (such as energy or communications), or several sectors0 El Location-is the agency located at the national, provincial, or municipal level in government? Independence. Agency independence is one of the important design issues for effective regulation. Independence is especially important in infrastructure because of the large-scale, long-term nature of the capital investments Investors require credible commitments from government that their funds can generate a reasonable return. Only an independent regulator can assure investors that their rights-based on fair rules and contract provisions-will be honored and impartially enforced through the term of the project. Kerf and others (1998) identify the crucial elements in designing the institutional structure of an independent regulator To ensure the agency's independence: El The regulator's mandate must be clearly defined by law, and not be subject to the discre- tion of political authorities. E The executive branch's discretion in making appointments should be constrained by legisla- tive provisions specifying certain qualifications E The tenure of individuals on a regulatory board should be staggered to reduce the influence of any one administration. E Arbitrary removal should be prevented to provide security of tenure. E Regulators can be given access to independent sources of funds, such as user fees. The authors also noted that the option of a completely independent regulator could meet sub- stantial political resistance. If that happens, alternative models or intermediary steps can be used to compensate for the lack of independence For example, mdependent agencies with an advisory role could be established. To increase transparency, their recommendations and the government's final decision (together with a justification for any deviation from the recommendations) should be published together. 24 WORLD BANK WORKING PAPER Coverage. There arc many diffecrnt approachcs to sector covcragc. For ex. nplc, rnany countrics have separatc power regulators-though because of the convergence of cncrgv forms, all energy is increasingly rcgulated by a single agcncv. Thel Unitcd Kingdom initiated this approach, creating separate regulators for powcr and gas, but thcse have since bccn merged in.o a regulator that covers the entire energy sector. Nearly cvery country has a telccommunications rcgulator, but in somc countries thesc are part of agencies covering all communications, including broadcast mcdia. In Australia, Brazil, Canada, and the Unitcd States, states and provinccs have cstablishcd cross-sectoral agencies that regulate all key infrastructure sectors at the statc levcl. In Ncw ,ecaland the Commerce C'ommission ovcrsees all infrastructure scctors. In developing countries Nsith limitcd capacity, a multiscctor regulator can achieve cconomies of scale and rcduce interfercnce from linc ministries. Location. Should the regulator be locatcd at the national, regional, provincial, or local level? The answer largely depends on a country's pohtical and legal structurc, and is worth further investigation in Chma. A centralized regulatory systcm may oficr consistcncy, but it could sacrifice flexibility and knowledge of local concditons and nceds. But w1hile a deccntralizcd sys,cm -would placc the regulator close to the service provider and the community, the regulator nmw not benefit from the larger pool of expertise availablc at the national level, and ma) suffcr from higher corruption. Design issucs are often related. For example, in fcdcrations single-sector or single-industry rcg- ulators are common at the national lcvel and cross-scctoral rcgulators are morc common at the state or provmcial level. U S. states havc estabLished cross-scctoral regulators, while singlc-scctor agencies exist at the federal level (such as the Fcdcral 'omrmunications Commission and Federal Energy Regulatory Commission). The responsibilities arc divided so that the federal agencics ovcr- see interstate issues and the state agencies tacldc local issucs such as setting local rates and super- vising utilities that fall within thicr purvivcw. Takng the watcr sector as an example, in England and Walcs the economic regulator works at the natonal level. In Canada and France watcr utilitics are regulated at the municipal level. There are also project-specific regulators, as in Manila (thc PIhulippincs), where the regulator reports to a Board of Trustees representing central and local government organizations (Box 2-4). Tariff Regulation There are many possiblc ways of regulating tariffs to achieve certain objectives, but most tariff adjustment procedures are based on the formula defied in the contract (Boxc 2-5). Disputes should be resolvcd by an independent panel of expcrts. The contract should also spccify alternative financial compensation when a request to adjust tariffs is refused becausc of poliiical or social con- cerns. Common alternatives include cxtending the cofntract, providing direct or indircct public financial support, and relaxing certain service requircmcnts. 'T'hc t-wo main modcls f,or regulating tariffs on infrastructure projects arc: Rate of return regulation or cost plIns pricing. With this method thc rec ulator cvaluatcs the capital and operatmg costs of providirg the senrice anel scts a rate of rcturn on capital or assets. This approach tries to hmit the profits of the opcrating compa.L and, in some cases, provide the investor wvith a minimum rcturn. Rate of return regulation is uscd in the United States, where public companics supply water to about 85 nercent oft' 'c population. Onc problem with this method is that it does not provide incentives for cflcicncy or technologi- cal innovation. Price cap regulation. With this method the price cap is linked to the %-tail price index increase minus a factor corresponding to efliciency or productivity gains expected or required by the regulator. Price cap regulation encourages the operator to manage its resources efficiently and obliges it to share the productivity gains wlith consumers through smaller tariff increases. This approach is used in Chile, England, and WVales (AD)B 2000). PRIVATE PARTICIPATION IN INFRASTRUCTURE IN CHINA 25 0 *_*01 After prvatizing its water companies and policing organizations, in 1991 the United Kingdom established the Offfice for Water Services to balance the demands of legislation, service levels, company profits, and customer protection. The office is an independent organization headed by a regulator, appointed by the secretary of state for a 1 0-year period, who is assisted by about 200 staff members Among the office's responsibilities are: D Periodically assessing the global tariff formula for each company using price cap regulation. This mechanism is supposed to share with consumers part of the productivity gains made by each operator. Q Every five years, setting the investment levels required of companies (taking into account their assessments) and deciding what share of these costs can be recovered through the water tariff. o Assessing each company's technical performance and gathering information on customer satisfaction. Com- panies face penalties if they do not achieve the performances levels defined by the regulator. 8 Publishing an annual report on the state of the water industry and water management. This report is discussed by all stakeholders, such as consumer associations and environmental groups. A cornerstone of the U.K. approach is the use of yardstick competition, which compares the technical, economic. and financial performance of the water companies. Having comprehensive data on the performance of each operator helps the regulator justify its economic and technical requirements. In the Philippines, after the successful concession of Manila's water and wastewater services, the city government developed regulation based on recommendations from the International Finance Corporation and principles defined in the concession contracts: 8 A ten-person Board of Trustees presided over by the Ministry of Public Works established the Regulatory Office. 8 A three-person Appeals Panel was created to solve conflicts between the regulators, the granting authorities, and the concessionaires. 8 The Regulatory Office has 5 members and a staff of 60. Members are appointed for five-year terms, but the term of two of the initial members (other than the director) will be three years. No member of the Regulatory Office may be removed unless agreed by a majority of the members of the Appeals Panel. The Regulatory Office mainly covers financial and technical contractual issues (tariffs, consumer complaints, achievement of technical performance, and so on) 8 The Board of Trustees must work with the concessionaire to implement tariff adjustments as instructed by the Regulatory Office or, as appropriate, by the Appeals Panel Source- International Finance Corporation; Economie et Humanisme; Mott MacDonald (2001). Among the countries studied there were three main approaches to regulating tariffs: tariffs can be set by gov- ernment departments, agencies, or regulators, by market forces; or by a contract. Whatever method is used, it must be robust against legal challenges If it is not, prospective investors will be wary. In the United Kingdom the water regulator (which is independent from government) increases tariffs based on a formula related to inflation. The regulator uses its power to fix prices as a means of creating "comparative com- petition" among the 10 regional water companies, which are effectively monopolies in their areas. The compa- nies are required to provide information on all aspects of their business, from water quality to capital spending. The companies are compared, and pressure is put on those that do not compare well by restricting price increases until better performance is achieved. Source Clifford Chance and Credit Agricole Indosuez (Annex 4) 26 WORLD BANK WORKING PAPER ~jAt4r L- J1! .,1 --I. * I' : 1 C -. I .I4j 1;' $.1' , iX CC- I~ -~( 2 ]ndanes!a Initial tariffs require presidential approval. There is no guarantee that tariff adjust- Concession company proposes tariff ments every 2-3 years will be approved. adjustments every 2-3 years based on a Uncertainty over toll approval and adjust- formula linked to inflation. ment procedures is unattractive to Tolls are collected by the operator based investors. on use. P3hflippinas Tariff set in concession contract. Revised Revised tolls must be published before each year based on a formula that takes becoming effective, and users can chal- into account inflation, interest rates, lenge the reasonableness of the revised exchange rates, and the cost of construc- tolls. tion materials. If toll road investors happen to receive a Tolls are collected by the operator based windfall, they keep all the upside. on use. LDDDed KinSdom Under the latest projects, operators About three-quarters of the tariff is based receive shadow tolls based on availability on availability, limiting the demand risk and use of the road. Payments are made (and making financing easier). by the U.K. Highways Agency to the Tolls are not collected directly from road operator. users, though contracts contain provisions for this to occur at a later date at the option of the U.K. Highways Authority (with mechanisms that ensure the opera tor's overall income is noz preludiced).' Source: Clifford Chance and Credit Agricole Indosuez. (Annex 4) There are many variations of these two models, and many private infrastructure projects around the world have used a combination of methods (Box 2-5). For example, an economically sustainable price cap can be achieved through negotiations with various investors or through com- petitive bidding. Pnce caps should apply throughout the term of the contract, but to ensure that the operator is not making excessive profit or that the project is solvent for debt service purposes, the regulator could review and rebase the price cap every 5-10 years. Another approach is to benchmark the vanous cost components against a group of comparable utility companies to deter- mine the revenue requirements of the utility involved. Finally, exogenous cost (such as the cost of raw water to a water distnbution company) should be passed on to consumers. Examples of tariff regulation regimes for roads are shown in Table 2-3. 3 FINANCING PRIVATE INFRASTRUCTURE PROJECTS IN CHINA I nfrastructure projects have high initial investment costs and returns that hinge on profitabil- ity over the long run. Accordingly, most private infrastructure projects are financed by equity and a significant share of debt. Debt accounts for 60-70 percent of financing for an average infrastructure project. Private infrastructure projects in China, however, have been financed mainly by equity. This sub-optimal funding structure results in a higher weighted average cost of capital that will ultimately lead to higher tariffs. Moreover, it is unsustainable. Limited debt financing prevents developers and sponsors from leveraging their capital and hinders their ability to invest in more projects. Because the revenues from road, water, and power projects are in local currency, assumption of local currency debt avoids currency mismatches and reduces foreign exchange risks-especially in countnes where foreign exchange is tightly controlled. Yet, domestic financtng for private tnfra- structure projects has been negligtble in Chlna, creating another concern for investors and an addi- tional risk premium in their cost of capital How a project is structured and how risks associated with the project are allocated among the participants ultimately determine the cost and availability of each type of funding for the project. One reason international lenders are skeptical about China's market is that they per- ceive excessive credit rtsk amon,g their Chinese counterparts. Many of these counterparts are quasi- government entities or recent corporate spin-offs of government entities, and there is almost no information on their financial viability. As a substitute, support letters have been sought from governments to ensure that the Chinese counterparts will meet their contract obligations These support letters are not legally enforceable and have become of little value with the fallout from the International Tmist and Investment Corporations (ITICs), which the government refused to bail out upon request from international lenders. With the lack of creditworthy counterparts, for- eign lenders will place more stringent requests for securtty on the project, including securities on project assets and contract rights. As discussed below, creating and enforcing securities is not easy for lenders. 27 28 WORLD BANK WORKING PAPER Investors are also frustrated by approvalv isk. For example, several toll road companies that have issued bonds overseas-Zhuhai Expressway, Greater Beijing ExpressNi av, Cathay Interna- tional, Traffic Stream-experienced continuing diFliculties in obtaining approvals from the State Administration of Foreign Exchange to repatriate earnings in foreign currerncy. These cases tar- nished the reputation of China's private infrastructure projects in international bond markets, making it harder to secure bond issues On the domestic side, Chiniese banks do iot yt play a signzzficant role in financing private infra- stnicture projects. Banik regulations, suited to standard corporate lending to state enterprises, have not been adapted for more complex project fnance The four state-owned commercial banks, which have provided significant funding to govermnent-led projects, have narticipatcd in only a few private projects. As for domestic bond issues for infrastructure projects, they have been limited to government projects such as the Three Gorges D)ams and some rail transport projects. Issues with risk allocation and securities, just like issues with laws and regulations, are not lim- ited to foreign investors. They are as important, if not more important, to the domestic financial sector. As the government cleans up non-performing loans and strengthens domestic banks, a sUs- tainable flow of well-structured private infrastructure projects wvill provide the banlks with solid and attractive lending opportunities, helping to lay a solid basis for bank reform. Infrastructure projects involve a complex range of risks. Private participation offers a major benefit by shifting some commercial responsibilities and risks to the private sector, mitigating risks for the public sector and maling resource allocation more efficient. In addition to commercial nsks, infra- structure projects face political and macroeconomic risks. All these risks should be carefully assessed and allocated among the parties prio;i to contract signing, to provide clarity to the potential investors and so enhance project procurement and implementation. A flow of X ell-structured proj- ects is crucial for achieving sustainable pnvate financing. In the past, risk allocation mechamsms for projects m China have often been inconsistent, even betxveen different projects in the same sector, and have been unsatisfactory to granting authorities, the central government, or investors While many issues related to project structuring are subject to case-by-case negotiation, there are standard mternational best practices. A balmced allocation of nsks is crucial to achieve the objec- tives of the various participants (the public sector, investors, and consumers) and wxill offer them the best value from private infrastructure projects. The guding principle in structuring and negotiating the allocation of nsks is that the party best placed to contr-ol a certain risk sboz!!r assumie and mantfage that risk. The public sector usually assumes political and environment risks, including risks involved in procuring approvals, risks of discriminatory changes in laws, land acquisitioin nsks, and so on (Table 3-1). Requiring the private sector to bear risks that could be better managed by the public sector would raise financing costs to the detriment of consumers, and reduce sources of funds.' In China investors are most concerned about legal risks, approval risks (in cluding tariff approval risks), anzd reguilatomy risks, all of which should be borne by the government. When a sound legal and regulatory framework mitigates these risks, investors are wvilling to assume more commercial risks, such as project completion risk, operating risk, and demand risk. When such a framework is lacking and investors are forced to assume legal, approval, or regulatory risks, they will either not participate in the project or be unwilling to assume as much commercial risk. Thlis point is demonstrated by the fact that investors have asked for a guaranteed l ate of return in many private infrastructure projects. 1 When neither the public nor the private sector has the capacity to bear some of thc nsk, the government could consider using the risk-mitigating instruments provided by multinational institutions In several countnes products offered by the World Bank, International Finanec Corporation, and Asian Dc%e,lopment Bank-such as partial nsk and partial credit guarantees provided by the World Bank-have secured private investments in projects where commercial banks were not prepared to take political or long-term commercial risks PRIVATE PARTICIPATION IN INFRASTRUCTURE IN CHINA 29 What is the risk? How does it arise? How should it be allocated? Design and development risk Design defect Design error in tender specifications Public sector should bear risk. Contractor design error Liquidated damages-a sum of damages to be paid by the contractor, as agreed upon in the contract-should be paid by the contractor. Once liquidated damages are exhausted, prolect company's returns should be tapped. Construction risk Cost overrun Within construction consortium's Contractor should bear risk through control: inefficient construction prac- fixed-price construction contract plus tices, waste, and so on liquidated damages. Once liquidated damages are exhausted, prolect com- pany's returns should be tapped. Outside construction consortium's Risks to be shared between the public control: changes in the overall legal sector and insurance companies if insur- framework (changes in laws, taxes, ance is available. and so on) or government actions that directly affect the prolect (delays in obtaining approvals or permits, and so on) Delay in completion Within construction consortium's Liquidated damages should be paid by control: lack of coordination among the constructor. Once liquidated dam- subcontractors, and so on ages are exhausted, prolect company's returns should be tapped Outside construction consortium's Insurer risk, if risk was insured Once control force majeure and so on insurance proceeds are exhausted, prol- ect company's returns should be tapped. Failure of prolect to Quality shortfalls, construction Liquidated damages should be paid by meet performance defects, and so on the constructor. Once liquidated dam- criteria at completion ages are exhausted, prolect company's returns should be tapped. Operating cost nsk Operating cost overruns Change in operator practices at Prolect company should bear risk. prolect company's request Operator failure Liquidated damages should be paid to the prolect company by the operator. Once liquidated damages are exhausted, prolect company's returns should be tapped Failure or delay in Public sector discretion Public authorities should bear risk. obtaining permissions, consents, or approvals Changes in prices of Higher prices Allocation of risk to the party best able supplies to control, manage, or bear it (supplier, prolect company, or users). Supplies not delivered Public sector failure Public authorities should bear risk. by public authorities (continued) 30 WORLD BANK WORKING PAPER What is the risk? How does it arise? How should it be allocated? Revenue risk Changes in tariffs In accordance with the terms of the Project company should bear risk. contract (for example, indexation of tariffs leads to lower demand) Government breach of contract terms Public sector should bear risk. Changes in demand Drop in demand Public and private sector should share risk according to the terms of the contract. Shortfall in quantity or Operator failure Liquidated damages should be paid by quality leading to the operator. Once liquidated damages reduced demand are exhausted, project company's returns should be tapped. Project company failure Liquidated damages should be paid by the project company to public authority. Finanaal risk Exchange rates; Devaluation of local currency; Project company should bear risk to the interest rates fluctuations extent that changes in exchange rates and interest rates exceed what is provided for in the tariff (and borne by users). Foreign exchange Nonconvertibility or nontransferability Public sector should bear risk. In cases of contract termination, compensation should be paid by government. Force majeure nsk Acts of God Floods, earthquakes, riots, strikes, Insurer risk, if risk was insured. Otherwise and so on risk should be borne by the project company. Changes in law Changes in general legal framework Normally, project company should bear (taxes, environmental standards, risk with proper compensation provided and so on) the government. Changes in legal or contractual Public sector should bear risk. framework that directly and specifi- cally affect the project company Performance risk Political force majeure Breach or cancellation of contract, Insurer risk, if risk was insured. Otherwise outright or creeping expropriation, risk should be borne by the public sector. failure to obtain or renew approvals In cases of contract termination, compen- sation should be paid by government. Environmental nsk Environmental accidents Operator failure Liquidated damages should be paid by the operator. Once liquidated damages are exhausted, project company's returns should be tapped. Preexisting environmental liability Public sector should bear risk. Source: Kerf and others (I1998). PRIVATE PARTICIPATION IN INFRASTRUCTURE IN CHINA 31 Another example comes from power sector reform. Part of the reason that power sector liber- alization in countries like Australia, Chile, and the United Kingdom has attracted private invest- ment is the strong regulatory framework established in these markets. As a result investors are willing to take greater commercial risk and participate in the competitive wholesale market. In China, without a credible regulatory framework, investors have been hesitant to enter the power sector, even without liberalization of the wholesale market. Besides allocatng risks, there is the issue of how to mitigate them morc efficiently. Performance- based contracts can help partics do so. Such contracts define the performance targets that the private operator is required to achieve within a given period and the consequences of compliance (or non- compliance), but the contracts do not stipulate how the operator should achievc the targets. The Chinese private infrastructure projects we reviewed often over specified technical issues, an approach that can inhibit the private operator from using its full expertise and most appropriate technology and from demonstraung innovation and creativity in solving problems. As a result service improve- ments were limited and tariffs were not minimized. In other countries pcrformance-based contracts have generated consistent savings in civil construction. In the United States such contracts have become the preferred approach for water investments in less than 10 years. Savings of up to 40 percent have been realized in civil projects with contracts based on performance output rather than materials input. Project Financing Regulation A common approach used by investors in infrastructure projects worldwide is to structure the proj- ect on a limited recourse project financing basis. This arrangement aliows investors to recoup their investments (through debt and dividend payments) only from the cash flows of the project, except in cases such as force majeure events or when the contract is terminatcd due to material breach by either party 2 The project financing structure allbws sponsors to invest in projects with limited risk exposure for their corporate balance sheets. For the largest international infrastructure companies, the project financing approach is the only way they can invest in a large number of projects all over the world (Figure 3-1). For various reasons, in the case of the Chengdu Water project, Vivendi had to offer a guaran- tee for lenders based on its own corporate balance sheet. Although Vivendi was willing to do so given the project's strategic value as the first BOT scheme in China, this structure is not sustainable as these sponsors look into other projects in China, and there are only a few international water companies in the world. To standardize the approach to international project finance, the State Development Planning Commission and State Administration of Foreign Exchange jointly issued the Project Finance Measures in 1997.3 But the much-needed explanations and implementing rules that usually follow regulation do not appear to be near completion for these measures. The biggest problems investors have with the Project Finance Measures are their definition and applicability Investors have to apply for an approval that their projcct will be granted project finance status. But they havc come to realize that the interpretation and application of the Project Finance Measures is not simply a matter of predictable law and regulation. Rather, the granting of project finance approval is an investment policy tool as well as a means of regulating foreign loans. This is due to government concerns about the important issues typically involved in a large-scale 2 In these cases the lendcrs will request the dcbt to be paid regardless of the source of the payment, even if it has to comc from cashflows outsidc the project itself 3 Including "Provisional Measures on the Administration of International Project Fmance," issued by the State Devclopment Planning Commission and State Admimustration of Foreign Exchange in 1997, and "Mea- surcs for the Administranoni of Borrowinig of International Commercial Loans by Domestic Organizanons," which includes four articles on project financing, effective as of January 1998, promulgated by the People's Bank of China and State Adminustration of Foreign Exchange 32 WORLD BANK WORKING PAPER (millions of U.S. dollars) o United Kingdom 9,930E Australia /o] Russia 22,664/ C D United States 7,728 * Indonesia o France M Italy 5,535 oBrazil Y*Saudi Arabia 2,072 * 1 / E Oman 2,200 3 - 1 8 o Others 2,500 2,900 4,277 I. Source: Project Finance International; Macquarie Bank; Porter (2001). infrastructure project, including large capital investmcnts, foreign exchangc obligations, potential impact on domestuc inflation, and possible exposure of state entities to risks through foreign debt, guarantees, and the like. In fact, some investors suspect that the Projcct Finance Measures were issued to stop Chinese companies from registering overseas and coming back to Chinia as foreign investors, then carrying out murky "project finance" transactions wvith domestic governments that were asked to offer excessive guarantees. One way of stopping such practices was to require additional review and approval by the State Development Planning Commission and State Administration of Forcign Exchange. But this requirement also delayed and discouraged qualified foreign investors from seek- ing true project financing. Another way to discourage the abuse of the project finance structure would be to procure projects through public tenders and set strict ownership, technical, and finan- cial criteria that could be met only by reputable strategic ilvestors. For example, the Project Finance Mcasures provide for foreign exchange convertibility, but to enjoy this privilege investors' projects must be granted project finance status by the State Develop- ment Planning Commission. The application for project finance status has to be part of the financing plan included in the Feasibilty Study Report (see Chapter 2). When the State Development Plan- ning Commission approves a Feasibility Study Report, including a plan for project finanicing, the approval document contains a statement such as: "The required investment, apart from the registered capital, will be obtatned throuigh foreign and local loans raised by limited recourse proiect financin,." However, such approval is very hard to get. In theory, the State Development Planning Commission evaluates every large-scale project to determinc whether it dovetails with state policy and whether the risks of large-scale foreign financing for a renminbi-earning project may be acceptable. When policy priorities cannot be sufficiently articulated or agreed, it becomes almost impossible to further define the rules-as appears to be the case with the Projcct Finance Mea- sures. Some investors have been told that thcir Feasibility Study Reports could bc approved faster if the "investors are responsible for raising the loans." In that case the State Development Plan- ning Commission's approval document contains a statement such as: "Investment beyond the PRIVATE PARTICIPATION IN INFRASTRUCTURE IN CHINA 33 registered capital wpill be obtained by the joint venture company through foreign and domestic loans, wvith secirity provided by the investing parties in proportion to the ratio of thetr respective contributions to re,gstered capital." The commission's approval is strictly followed by other relevant government agencies, most notably the IVhnistiy of Foreign Tradc and Economic Cooperation in its revicw of the joint venture contract and the State Administration of Foreign Exchangc, which registers foreign debt anld approves all securitics involving foreign parties. TIhese processcs result in delays and unnecessary rigidity in financing. In addition, the State Dcvelopment Planning Commission approves the Feasi- bility Study Report (including the final financin-g plan and concessioni agreemcnt) only after bid- ding (see Figure 2-1 in Chapter 2), which crcates a lot of uncertainty for the bidders because they do not know for sure at the time of the bidding what financing structure they can use and, accord- ingly, wvhether they iptll bc able to get foregn or domestic loans. Even so, the excerpts from the two approvals shown above point to two of the more fortunate types of projects. Although the approval proccss is not tralnsparent, once approval is obtamned from the State Development Planning Commission the investors in these projects know whcre they stand. In many cases the wording of the Feasibility Study Report approval is not so clcar, and investors must try to gauge for thcmsclves which financing structurc is acceptable. They must also try to ascertain the risks of carrying out a financing plan that has not been explicitly approved by government authorities-a risk that lenders nlow know too well because of the problems with International Trust and Investment Corporations The Security Law To control their nsk exposure, one of the first things that lcnders look at in assessing potcntial financing opportunities is the kind of security schcmc avaulable. A sccurnty schemc can provide vari- ous kinds of protection In the evcnt of default, including recourse to the project assets, the ability to take over control of a project by stepping into the operation or transferring it to a third party, and priority ranking in claims. The China Security Law has playcd a critical role In China's efforts to promote a favorable investment environment sincc its promulgation in 1995. Nevertheless, investors, especially lenders, have had many problcms with the creation and enforcement of the Security Law International comparisons showv that China's securit) regimc is morc restrictive and less attractvc to private investors (Box 3-1). For example, floating charge (tlhat is, effective security ovcr a company's mov- able assets such as its stock in trade) is recognized under common lawv systems and was recently adopted by Hungary, a civil law countiy. But it is not recognized in China. This section discusses some of the problems with creating and enforcing securities in China, and possible solutions. But beforc discussing the dctails of the legislation, it is wvorth noting two complemcntary aspects of the legal system without which a security lawv would not be effective One is the framcwork for defining an-d reguLlating property ri,ghts The difficulty of enforcing secu- rity in China partly lies in the fact that many borrowers do not truly own their assets Unlike other civil la-w jurisdictions, China does not have a civil code that establishes the basic principles of prop- erty law. China's first unificd Contract Law Just wvent into effect in 2000, and the draft property law is still under discussion. Because somc of the problcms discusscd below might be addressed by the draft property lav, it is important to ensure consistency between the two laws and subsequent implcmenting regulations The other aspect required for an cffective security law is the court system, which lenders must be able to rely on to enforce security rights against competing claims on the same property. This area also needs to be strengthened in China, to ensure timely enforcemcnt of security rights by local courts. The Security Law provides for three main types of secunties related to infrastructure projects- guarantees, pledges, and mortgagcs. The registration, creation, and enforcement of thcse security itcms are discussed below. 34 WORLD BANK WORKING PAPER The security regime in England and Wp_`zo (Scotland has a slightly different regime) is perccivcd to be lender friendly. Registration and enforcement are straightforward, and there is a well-establish_d regime incorporating a full range of security instruments, including floating charge. In addition, lenders are ablc to acquire step-in rights. In aungery it is possible to take security over most assets, and recent reforms have improved the processes for taking and enforcing security. For example, floating charges and mortgages over movable property have been possible since 1997. In addition, in 2000 a new registration system was established for these mortgages. Lenders are also able to take security assignments over project contracts. Most analysts believe that step-in rights are possible under Hungarian law, and they are gencrally provided to lenders to infrastructure projects involving private participation. The effectiveness of such rights has not yet been tested before H-,ungarian courts, however. Source: Clifford Chance and Credit Agricole Indosuez. (Annex 4) Registration of Securities The registration system for securities needs to be st7en,gthened. Varying sophistication and training among the staff rcsponsible for registering titles and secunties havc caused considerable frustration among lenders. Examples include demanding a two-page "standard form contract" even if the par- ties have negotiated a far more sophisticatcd agrecment, requinng a diefinite "expiration datc" even when the parties have agreed that the security will remain in effect until thc dcbt is fully paid off, and refusing to register a security if the value of the loan differs from the valuc of the asset Finally, registration offices around the counitry havc not dcvcloped a standard approch to indicating the priority of the mortgage. The public title and security registration arc not casily accessible by the public, which is incon- sistent with the Sccunty Law. It is often not casy for lenders to get sufficicnt information on bor- rowers' tides to assets and on securitics that other lcnders might already havc on the assets. Securities that existed before the registration rcquiremcnts also need to be clarified and brought into the current registration system Guarantees Chinese laws, includmg the Security Law, prohibit government agencies from providing repayment guarantees except for loans from foreign governmcnts. As a result, for the pilot BOT projects a great deal of effort was expended trying to insulate what was effcctively a gu rantec by the granting authority of the offtaker's payment and termination obligations from bcing clharacterized as a guar- antee. The prohibition on guarantees goes against international standard practice and wvill make future infrastructure projects impossible for private investors. We understand that the Project Finance Mcasurcs only prohibit guaran,ecs for the repayment of debt considered contrary to the "nature of project financing." Performance guarantees for ancillary project agreements (such as an offlaker agrccment) seem to be permitted as long as they are not from financial institutions, are properly approved, and do not "altc,- thc naturc of projcct financing " Yet, the prohibition needs to be clarzfied a;id narrowed downi fo; h e;ge-scale inii-a- struicture projects. For example, to the extent that tcrmination payments rcflect outstanding loan amounts and are payable directly to the lender, they should be allowcd as pcrformancc but not repayment guarantees. Pledges The biggest problem with pledges involves the creation of security over cont;'fct Z iglts. Chapter 2 discussed the need for the provision of some fundamental contract rights (such as "step-in" rights) in a general statute, such as the proposcd framework law for private participation in infi-astructure. PRIVATE PARTICIPATION IN INFRASTRUCTURE IN CHINA 35 Here we discuss hov to create security over such contract rights-a subject that is, at best, not at all clear in China's regulations. Due to a lack of specific provislons for assigning contract nghts for infrastruLcture projects, lenders traditionally havc relied on provisions of the Civil Law, the Contract Law, and foreign laws. These alternative Chinese laws do not clearly indicate what an assignment actually accomplishes- whether it constitutes a full assignment of all contract rights or just some. The Security Law is silent on this issue Besides movable property, it lists only four types of rights that may be plcdged. E Bills of exchange, checks, promissory notes, bonds, certificates of deposit, bills of lading, and warehouse receipts. a Shares and share ccrtificatcs that are transferable by law. C! Economic rights in patents, trademarks, and copyrights. a Odler rights that may be pledged by law The Supreme People's Court Interpretation4 added only one category of rights for inclusion under the fourth group. earnings from immovable property, including road bridges, tunnels, and ferry crossings. The list is far from complete for a comprehensive security over core contract rights The lack of clarity is not helped by the approval process for securities. The Mimstry of Foreign Trade and Economic Cooperation approves assignments of contract rights if they involve only owncrship transfer. If security is to be created on such assignment rights, a separate approval will be required at the tLme of the enforcement of the security (that is, when the title changes), which means the security is still uncertain-and this is certainly not acceptable to lenders. Thus it is urgent that the government develop a clear legal basisfor assigning security on con- tract rights, and regulatory procedures for establishing a system of registration for the same. The legal rights associated witl a pledge of contract rights need to be fully identified, including step-in rights and the point at which the rights become operative. Mortgages The most problematic area with mortgages involves the mortgages of land, although the problems have more to do with creatng and enforcing land mortgages than with the legislation. In China land belongs to the state, and land use rights used to be allocated to state enuties for free In 1986 a system was introduced granting long-term land use rights to private entities, although the fees for many such rights have not been properly paid Both allocated and granted land may be mortgaged, but the mortgagee cannot enforce the mortgage if the grant fee has not bcen paid in full at the time of the enforcement, and the grant fee often exceeds the value that can be realized from the land usc rights. To reduce front-end costs, in a number of BOT projects the project company has been encouraged or forced to use allocated rather than granted land, which will create problems later if lenders need to enforce the mortgage against the land use right In addition, allocated land does not afford the user security of tenure or a right to compensation for expropriation. The government should strengthen the system ofgrantfees, and the granting authorities should be prepared to be asked by investors to put aside suffictent resources in a secured account for paying thegrantfee if the mortgage is called. A separate issue on land mortgages of particular importance to water projects-given the lack of a property law-is the sketchy legal framework for casement rights owned by project companies. As Chlila introduces private investment in more infrastructure projects involving pipelines, a legal definition of third-party rtghts, tncludtng easement rights, needs to be clartfied. Mortgage issues also affect other types of assets; these are summarized in Table 3-2. 4 Thc Interprctationi of the Suprcme People's Court on Certain Question relating to the Apphcation of the Secunty La%N', issucd and cffective in December 2000 36 WORLD BANK WORKING PAPER Dteni Lssues Lieflcrnnoendxss [Bui0ding Chinese law generally does not allow land Unify the land and building registry systems. and the buildings on it to be separated in a transfer of ownership. Land use rights are granted by the Ministry of Land and Natural Resources, while the Ministry of Construc- tion is responsible for the registration of ownership and other interests in buildings, under a separate system with different rules. Although land and building registries in major centers have begun to share facili- ties, they still retain separate systems and report to separate bureaucracies. Ft.unre or The Security Law does not allow mortgages Allow a land mortgage to be drafted to zter-acquired for future or after-acquired property. A cover buildings that are to be constructed property lender who has obtained a mortgage on on the land, and a mortgage on machinery vacant land would need continuing cooper- and equipment that includes after-acquired ation from the mortgagor to register a equipment. mortgage on the new buildings on the land Permit lenders to unilaterally carry out sub- once construction reaches certain stage. sequent registration based on the prior The same problem applies to mortgages of mortgage so that the mortgagor cannot machinery and equipment. In a complex block such registration by not assisting. We infrastructure project where equipment is understand that the draft property law delivered over months or years, multiple takes such mechanisms into consideration. contract signings and registrations are required, as well as the continuing coopera- tion of mortgagors. Floating charge Chinese law does not allow floating charge, We understand that the draft property law which could create problems for power would expand the scope of mortgage rights projects that need to take security over to cover floating charge. fuel reserves, or for water projects that need to secure supplies of chemicals. Similar forms of asset security established by means of a general security agreement are available in numerous jurisdictions, including Hong Kong (China), Hungary, and Singapore. MHortgage of While the Security Law provides for the Provide a legal basis for taking security over a maximum giving of specific collateral to secure a accounts receivable, inventory, and fluctu- amount revolving debt, there is no provision for ating credit balances. the opposite-namely, providing changing collateral to secure a specific debt. Lenders who are persuaded to take such security in China must rely on foreign law, and incur the risk that a People's Court will not recognize the arrangement upon enforcement. Source: Freshfields. PRIVATE PARTICIPATION IN INFRASTRUCTURE IN CHINA 37 Granting Security to Foreign Entities Under China's exchange control system a separate layer of registrations and approvals is required when the secured party is foreign. Unless the approvals are obtained, the loan is invalid and the State Administration of Foreign Exchange will not authonze the conversion and repatriation of debt scrvice payments, termination payments, proceeds of any secunty, and so on. Under this system the State Administration of Foreign Exchange must first approve all foreign loans and the establishment of offshore escrow accounts. In addition, any security arrangement provided by a domestic third party for any such loan has to be approved and registered with the adminustration. For example, the Project Finance Measures requirc that for foreign loans taken on a project financing basis, the financing documents (including all security documentation) have to be submitted to the State Administration of Foreign Exchange to detcrmine whether the financial terms comply with the State Development Plannng Commission's approval of the Feasibility Study Report and are sufficiently "competitive." This process could be substantially simplified: once foreign exchange loans have been approved and registered, restrictions on domesttc entities provtd- ing securtty for those loans should be relaxed. Enforcement of Secured Claims The Security Law makes local courts primarily responsible for enforcing security claims if the debtors and creditors cannot reach an agreement, but the court system needs to be strengthened to limit interference from local governments, and legal training should be provided to judges. In enforcing claims, the first step is to confirm the borrowers' ownership of the assets used as secLtrity. With the restructuring and corporatization of state enterprises still under way, it remains challenguig to identify state assets held by state enterpnses, joint ventures in which state cnterpnses have a majority intercst, domestic collective enterprises, and limited liability companies. Once assets are identified as state assets, claims on them cannot be enforced without an authorized appraisal, which is often inflated (not least because the appraisal fees are based on a percentage of appraised value). After the appraisal the assets will first be auctioned, but for a mini- mum price based on the appraisal value. If the auction does not succeed, the court can order the property to be transferred to the creditor at a value based on the appraisal. Though, there are many ways for a debtor, particularly a state enterprise or otherwise well-connected entity, to delay or block these court actions for a long time, if not indefinitely The Ministry of Land and Naturai Resources has issued a circular requiring that all land appraisalfirms be disassociatedfromn local governments, but the circular has yet to be effectively implemented. The general rule for bankruptctes in China is that secured property is outside the bankruptcy regime, except for the special state enterprise bankruptcy procedure applicable in the 111 cities under the Capital Structure Optimization Program, where employee settlement payments come before secured creditors. This practice reflects China's current approach to state enterprise reform. One kcy issue, however, is that bankruptcy laws and regulations give local judges the power to decide nearly aHl questions that arise in a bankruptcy proceeding by "order"-a summary decision that generally does not impart a right of appeal. To increase creditors' confidence in the bankruptcy procedure (and more broadly, in the state enterprise reform program), issues relating to the substantive nghts of third parties who are not before the banikruptcy court-such as secured creditors-should not be decided by summary order. Wherc the local court decisions on such third-party rights contradict the law, the involved parties should have a clcar right of appeal to a higher court, as they would in any other court action. While it is alvays possible in China to petition a higher court to exercise its supervisory jurisdiction over a lower court, this is not by right-and unless special influence is brought to bear, there will gener- ally be no response. Finally, an issue not so much within the scope of security regulation but relevant to securities in general is insurance. First, it is required by law that direct msurance be placed with Chinese insurancc compames, which essentially forces lenders to rely on the creditworthiness of Chinese 38 WORLD BANK WORKING PAPER insurance companies. This is an excessive risk for lenders because in large-scale infrastructure proj- ects, insurance claims can run into billions of U.S. dollars. This could also impose lace unnecessary risk on the government. For example, in some cases the narrow range of products provided by domestic insurance companies compelled the government to assume force majeure risk to the extent that losses are not sufficiently covered by insurance proceeds. Second, lenders are not allowed to take security over reinsurance proceeds, except in the case of Laibin B. Neither of these restrictions is consistent wvith standard practice in many major markets around the world, and requires attention as China opens its insurance industry as part of its World Trade Orgamzation accession obligations. Domestic Bank Lending As a hedge against foreign exclhange risk, domestic lending is used extensively in the financing of infrastructure projects in a range of countries (Table 3-3). The domestic loans are even more valuable in China. This is not only because the revenue from infrastructure projects is primarily in local currency, but also because the government requires that a certain proportion of equipment be procured domestically. But while China's financial mar- kets have developed impressively over the past decade, domestic financing and particularly domes- tic lendmng have not played a significant role in financing infrastructure projects. One reason is the various rules on domestic financing and loan syndication. Traditionally, ventures with foreign equity were allowed to borrow renminbi only in proportion to the domestic equity in their shareholding structure. Although renminbi loans for foreign entities have been allowed since 1999, the old policies and practices of domestic banks still prevail. Indced, weak institutional capacity is another reason domestic banks provide only limited financing for infra- structure, as indicated by: * Limited availability of long-term local currency loans-the availability of domestic loans is limited, especially for large amounts, which may be further affected by government policy. The tenor of term debt is generally limited to eight years, and a 1997 notice from the People's Bank of China limits working capital loans to three years. * Restrictions on standard projectfinance practices-the State Administration of Foreign Exchange is reluctant to endorse certain standard project finance practices. For example, the administration and the People's Bank of China need to work out regulations that allow domestic banks to issue standby letters of credit for credit enhancement and foreign debt service, and that address intercreditor issues when lending alongside international banks (including service charge for offshore and onshore accounts). * Interest rate restrictions-China has started deregulating foreign currency lending rates and plans to liberalize renminbi interest rates in the near future. Renminbi interest rates are set by the People's Bank of China and prone to fluctuation. Accordingly, there is no such thing TABLE 3-3: LOCAL SHARE OF DEBT FINANCING FOR POWER PROJECTS WITH PRIVATE PARTICIPATION, VARIOUS ASIAN COUNTRIES, 1997 (PERCENT) Country Share Australia 100 Malaysia 94 Thailand 72 China 34 Philippines I Source: Clifford Chance and Credit Agricole Indosuez. (Annex 4) PRIVATE PARTICIPATION IN INFRASTRUCTURE IN CHINA 39 as a "fixed rate" loan, and a "floating rate" loan has to be within the boundaries set by the bank at any given point of time. Thus interest rates could be subject to significant influence from government policy and much less reflective of market condiuons and project risks. a Undeveloped lending capacity-Chinese draft loan documentation is often brief and does not contain the detailed, sophisticated provisions typical of infrastructure project loans. Domestic bank lending on a project finance basis (as for the Jingyuan Power and Shang- dong Zhonghua Power projects) was met with enthusiasm by private investors. In these cases Chinese banks share secunty with international lenders under common terms. Work- ing alongside international banks helps domestic banks develop expertise in credit evalua- tion, financial structuring, and risk management. And as in other countries where the commercial banking sector has not yet fully developed and government-sponsored policy banks provide long-term loans to infrastructure projects, the China Development Bank could be instrumental in nurturing longer-term lending markets. El Ltmited parttcipation offoreign banks in renminbt lending-foreign bank branches in China are not allowed to take renminbi deposits and so have limited capacity in extending renminbi funding to infrastructure projects. Allowing foreign bank branches to finance infrastructure witl renminbi would stimulate domestic banks and accelerate the building of capabilities to assess, underwrite, and manage the risks of private participation in infrastructure. Bond Financing Bond can be an important source of financing for infrastructure projects. In the United States, for example, local governments are able to mobilize significant private finlancing through the bond market even though 43 percent of the water sector (mostly systems in small commuwities, serving 86 percent of the population) is publicly owned. This is party due to the fact that the U.S. tax code allows political subdivisions to issue bonds that are exempt from federal, sate, and local taxes Interest rates on nontaxable bonds can be as much as 2 percentage points lower than the rates on conventional debt. Another important factor is the disciplined public financing market m the United States. As a separate investor group, bondholders take less risk than banks and equity investors, and accord- ingly require significantly higher financtal disclosure and tighter financing and securty structures for projects. U.S. government agencies have the sophistication and ability to structure the security so that utility bonds are isolated from the fluctuations in a mumcipality's general finances, allowing U S. cities to exploit the municipal bond route Denise Leonard (2000) provides a detailed expla- nation of the structure of the New York water revenue bonds (Box 3-2 and Figure 3-2). From this structure it is clear that to leverage the international bond market on a meaningful scale, Chinese bond issuers have to significantly improve their creditworthiness through much tighter regulations and contract arrangements. Credit can also be partly enhanced through security mechanisms offered by multilateral agencies, domestic banks (as with telecom bond issues in Malaysia), or insurance companies (as in the United Kingdom). Just as important is leveraging the domestic bond market China's bond market, with issues of about $40 billion a year and $419 billion outstanding, remains underdeveloped. Government secu- rities dominate tssues in the bond market. Corporate bonds account for just 4 percent of outstand- ing issues, while treasury bonds account for 62 percent and financial bonds for 34 percent. A signmficant amount of bond issuance is for infrastructure projects. Table 3-4 provides a breakdown of the mfrastructure bonds issued between 1996 and 1999 that were rated by China Chengxin International Credit Rating Company, which in 1999 rated 98.5 percent of the bonds issued. The State Development Planning Commission, in approving corporate bond issues, tends to prefer infrastructure projects sponsored by large state-owned enterpnses. But these bonds, like most of the other nongovernment bonds are tssued on a company basis, not a project basis. In addition to the small size of corporate bond issuance and limited types of debt securities, China's bond market suffers from incomplete laws and regulations on corporate governance, risk 40 WORLD BANK WORKING PAPER _0 '0~ * S.I The New York City Municipal Water Authority (the "Authority") was created in 1984 as a public benefit corporation for the purpose of issuing debt to finance capital improvements of the water and sewer system. A second public benefit corporation, the New York City Water Board (the "Board") was also formed to lease the water system from New York City and to maintain the system for a term of 40 years or for the life of the outstanding revenue debt The board was also granted the authority to set rates and charges as necessary to meet its operating, maintenance as well as debt service expenses. The arrangements were developed with a view to isolating the credit of the water and sewer system from that of the city (which was at the time rated Baa) by providing bankruptcy protection and automatic rate setting. To protect it from bankruptcy (i) the board and the authority were constituted as bankruptcy remote entities, neither had the ability to file for bankruptcy, (ii) the lease agreement between the board and the city estab- lishes the boards' ownership of the revenues, (ill) the revenues are pledged pursuant to the financing agree- ment to the bondholders, (iv) these agreements also provide for an operating reserve fund and a debt service fund, (v) the city's annual lease payment is subordinated and (vi) finally, legal opinions have been provided that the system's revenues many not be combined with the city's should the latter file for bankruptcy protection. With regard to tariff setting, the legal structure also provides for independent third parties, the rate consultant and the consulting engineer, to ensure that the rates are not subject to political manipulation. Additional protection includes: (i) if the board does not set adequate rates, the authority can petition for the appointment of a receiver to administer on behalf of the board and ensure that adequate rates are imposed; (ii) cash flow requirements for debt service are monitored on a monthly basis and rate adjustments can be made as necessary over the course of the year, (iii) the rate setting process is formula driven, rates and changes are set to equal 1I.15 times projected debt service of senior debt payable in the current year as well as 100 percent of operating expenses and required deposits, which include subordinated debt; and (iv) in order to take on additional debt (a) revenues for the last two years and (b) for the next five years, as projected by the rate consultant, have to pass the previous test. j t, , I ], * a * 11 , . ' S l | ~~~SeRat Cust*mers To a lock box' Accunt at Chas il at .. . . Water .B.o. Wate Boani * DtSv TntstAccount C am,ial Invrsumnm (3i NY Wate (Bond Issuer) nFmancig &nvcsttn ' .W | IP 8 | y ~~~~~..... Oprtmg NY City- Le sBondholders Source: Leonard 2000. PRIVATE PARTICIPATION IN INFRASTRUCTURE IN CHINA 41 TABLE 3-4: DOMESTIC INFRASTRUCTURE BONDS RATED BY CHINA CHENGXIN INTERNATIONAL CREDIT RATING COMPANY, BY SECTOR, 1 96-99 (millions of renminbi) Year Power Water Rail Road Telecom Others Total 1996 250 200 3,600 3,130 0 11,040 18,220 1997 1,960 15 1,690 120 1,000 5,032 9,817 1998 6,501 0 2,220 1,010 50 6,825 16,606 1999 3,187 35 4,110 700 0 4,177 12,209 Total 11,898 250 11,620 4,960 1,050 27,074 56,852 Source China Chengxin International Credit Rating Co., Ltd. management, bankruptcy, and the like, and from a lack of institutional tinvestors-the primary mar- ket is open only to banks, and closed to foreign investors and nonbank financial institutions such as domestic insurance companies and pensioni funds. In addition, credit rating is recommended but not manidatory. As with the lending rates of domestic banks, domestic bond pricing in China is based on deposit rates detcrmincd by the People's Bank of China-rates that do n(ot necessarily rcflect mar- kct conditions In industrial countries the spread between high-yield bonds (used to finance devel- opment projects) and govermenlt securitics has averaged 5 percentage points Ui recent years, while investment-grade corporate bonds often carry spreads below 1 pcrcentage point. Having such price flexibtilty in China would broaden the scope of bond Issues and cnablc maximum liquidity in all market areas. To make bond financing more attractive for infrastructure projects, regulatory reforms should aim at increasing transparency standards for bond issues, improving credit rating, dcepening the long-tcrm bond market, and devcloping the corporate bond market. Reform should also try to broaden the demandfor lnfrastructure bonds, includinig by allowiig nonbank financial institutionis to uivest in them. For instance, given their long-term liability structure, pension finds and insurance companies are well suited for funding private investments in infrastructure (Box 3-3) The develop- ment of the contractual savings sector entails reforms that are beyond the scope of this report, how- ever, it is important that this sector not be restncted to only a small number of eligible investments. Equity Financing There are thiree types of equity financing for infrastructure projects: direct investment by strategic investors, private equity financinig from institutional investors, and "public" equity Issued on stock markets. During the privatization of Chile's water companies, for example, $1.8 billion of direct investment went to purchase cxisting and new equity shares and $144 million of shares were listed on the stock exchangc (including $99 million sold to workers). Together these two types of equity financing account for two-thirds of the market value of the privatized water compames (Jadresic 20011) On the Australia Stock Exchange infrastructure assets account for about one- third of market capitalization (Tabic 3-5) (Porter 2001). Malaysia, with its Employee Provident Fund, and Australia, with a developed pension and mutual fund system, are able to source large amounts of local bonds for private participation in infrastructure. In Australia most companies involved in infrastructure (such as privatized airports or power generation units) raised their initial financing from banks, then refinanced in the bond markets after a couple years, having established investor comfort based on their performance. Thailand is moving in the same direction. Source Clifford Chance and Credit Agricole Indosuez. (Annex 4) 42 WORLD BANK WORKING PAPER -~~~~~w - 4 . Nt -- O - -- 2V~e [7-: 'l aOf' l 7elecom 60 3 Power generazon 35 S Power drsution 30 10 Water 20 4 Gas 5 2 Road transport 5 Rail transport 10 2 Airports 5 2 ~otal D( 32 Source: Porter 2001; Macquarie Bank Group. China has been successful in attracting foreign direct investment anrd listing infrastructure companmes on its stock exchanges, but it only recently began using pnvatc cquitv funds. In 2000 at least $177 billion, or 0.6 percent of global GDP, was invested worldwide-but only about S3 bil- lion, or 0.2 percent of China's GDP, was invested in China (Figure 3-3). Infrastructure has been the main sector attracting private equity investment in China, but tra- ditional sources of private equity financing-such as insurance companies and pension funds-have not yet been extensively mobilized (Figures 3-4 and 3-5). Strategic direct investment and private equity investments share some common concerns- including project economics (for example, tariff, currency, and security risks), the regulatory (billions of U.S. dollars) 200 180 - - - - - - - - - - - - - - - - - - _ - - 160 - - - - - - - - - - - - - - - - - - - 140 -- - - - - - - - - - - - - - -- 120- - - - - - - -- - - - - - - - 100 - _ 80 -- - - - - -_ _ - _ _ 60- _ ----- - - - _ 40 I2L~-- 20 - X I - -- - 0 1 ____ 11 __ 1995 1996 1997 1998 1999 2000 n Global n USA n Asia ri China/Hong Kong Source: Collated by AIF Funds Management LTD from "Asian Venture Capital journal 2001 Guide to Venture Capital in Asia," and "3i/PriceWaterhouseCoopers Global Private Equity 2001." PRIVATE PARTICIPATION IN INFRASTRUCTURE IN CHINA 43 (millions of U.S. dollars) 259 1 " Corporations c: Insurance Companies 444 /\ Banks o Govemment Agencies :-- 1,554 * Pension funds o Others 666 666 Source: Collated by AIF Funds Management LTD from "Asian Venture Capital Journal 2001 Guide to Venture Capital in Asia." (millions of U.S. dollars) o Infrastructure \3 Consumer products 1,190 3 Transport/distnbution o Computer related _ HeaW manufactunng 204\ _=--_ o Electronics -- - / \ t 13~~~~c3Others 238 406 272 Source: Collated by AIF Funds Management LTD from "Asian Venture Capital Journal 2001 Guide to Venture Capital in Asia " environment, foreign exchangc control, and the availability of debt financing. One critical con- cern for private equity funds is means of exit, which ranges from listing on local or offshore capi- tal markets to trade sales, mergers and acquisitions, and refinancing in industrial markets. But these options are extremely limited in China. China's financial sector reform needs to broaden the range of issuable sccurities such as preferred stocks, convertibles, and options to facilitate the exit by private equity funds. 4 ROAD ISSUES AND RECOMMENDATIONS O ver the past 10 years a national trunk highway system and a network of urban and provin- cial expressways have emerged in China. Still, an estimated $75-100 billion needs to be invested in highway construction over the next 15 years. Thus the challenge for the road sector is to raise large-scale financing, because government loans and tax revenues will not be suffi- cient to cover future capital requirements. Significant pnvate funds have been invested in roads, but the system developed to attract private participation depends on projects that are generally led by the public sector and does not ensure access to the widest possible range of private capital. This section reviews the situation in the road sector and recommends measures to increase private investors' interest in projects crucial for government road policy. The Situation Highway investments significantly increased in the late 1990s and reached 2.5 percent of GDP (around $27 billion). New construction has becn historically depcndent upon mainly public spending and the private sector has contributed less than 10 percent if total commitment to new construction, although total private funds acquired make China a leader among developing countries for highways. Few models have been used for private participation in roads and have experienced drawbacks. Capacity Planning and Development Rapid growth in passenger and freight traffic underscores the importance of highway development in China. According to tle State Development Planning Commnission, each year China's highways carry more than 15 billion tons of freight (equivalent to 704 billion ton-kilometers) and 15 billion people (636 billion person-kilometers), representing 14 percent of freight traffic carried by all means and 48 percent of passenger traffic. Since 1993 overall traffic on highways has been growing by 8 percent a year. 45 46 WORLD BANK WORKING PAPER Nearly all of the growth in roads has occurred since economic reform began in 1978. The road network has grown by more than 60 percent since 1980-and more than 40 percent since 1990-and covered 1.4 million kilometers in 2000, with expressways and high-capacity high- ways accounting for nearly 25,000 kilometers (Figure 4-1). During the 1980s annual investment in roads was less than 0.3 percent of GNP, but in 1998-2000 it reached $27 billion a year or 2.5 percent of GNP. China had only 940 kilometers of roads per 1 million inhabitants in 2000, far below the den- sity levels in industrial countries like the United States (25,326 kilometers) and Japan (9,096 kilo- meters) and even sparser than India (1,784 kilometers) and the former Soviet Union (3,335 kilometers). With increasing urbanization and economic development, the recent levels of invest- ment and expansion in China's road network are expected to be maintained in the near future. With considerably fewer technical and financial resources, and an economy and poverty level still within developing country standards, China is trying to duplicate the U.S. interstate high- way system (which took 40 years to develop) in no more than 25 years. The national trunk highway system is a priority, and 35,000 kilometers of the system are scheduled for completion by 2020 (33 percent by the end of 2000) at a cost of about $150 billion. In addition, a 20-year highway development program was announced in 2001 for the 12 central and western provinces, including an 18,000 klometer extension of the national trunk system with eight east-west and north-south corridors, 180,000 kilometers of national roads in western areas, and 150,000 kilometers of rural roads. The World Bank estimates that this program could cost about $200 billion over the 20-year period. Public and Private Financing New highway construction historically depended on public spending, through a combination of central grants, local budgets, and multilateral loans channeled through the public sector, and user charges collected by road authorities (Table 4-1). For example, since 1985 the World Bank has been a primary foreign lender to China's highway system. With 26 projects, the Bank had total exposure of $5.2 billion, contributing to 40 percent (3,500 kilometers) of the expressways built over the period. In addition, Bank funding helped upgrade, rehabilitate, and maintain 20,000 kilometers of secondary and rural roads. And until recently the Road Maintenance Fee and Vehicle Purchase Fee accounted for about 70 percent of highway funding. 30 .25 0 .! o15 00 O E 5 | S - W 1982 1985 1988 1991 1994 1997 2000 Source: Mitchell Stanfield & Associates. PRIVATE PARTICIPATION IN INFRASTRUCTURE IN CHINA 47 Revenue Source Period Billions of renminbi Share of total (percent) Public 1981-89 19 2 Public 1990-95 138 14 Public 1996-2000 805 84 Private 1990-2000 91 10 Total 1981-2000 963 100 Source Mitchell Stanfield & Associates. Today, however, the mix of hughway funding is shifting toward domestic loans and user charges These sources have increased the availability of public funding for expressways and high- grade (class 1 and 2) highways (Table 4-2). In the wake of the East Asian financial crisis, the Chinese government is encouraging public entities to take medium- and long-term infrastructure loans in renminbi. From 1998 to 2002, the government raised 660 billion renminbi ($79.5 billion) through public bonds, about 30 percent of which went to infrastructure development by local governments. In addition, some domestic com- mercial banks have provided provincial entities wvith medium-term (five to eight years) limited and nonrecourse project loans at annual interest rates just above 6 percent. Moreover, largely because of the ambitious growth targets for the national trunk highway system, great importance is being attached to road tariffs and user charges. The 1998 Highway Law permits tolls on almost all major highways, whether constructed by public entities or private concessionaires. The Highway Law also stipulates a fuel tax that could initiate an overhaul of the country's highway finance system, perhaps replacing the Road Maintenance Fee as the primary source of fundmng for operations and maintenance. But while the National People's Congress has approved implementation of the fuel tax, its exact size and structure have not been approved by the State Council Most national roads are expected to retaLn existng tolls, particularly those pledged to repay financing. Private financing has accounted for about 10 percent of highway investment during 1990- 2000 (see Table 4-1), making China a leader among developing countries in acquiring private 1990 2000 2001-2005 (est.) Investment Investment Investment (billions of Share of total (billions of Share of total (billions of Highway type renminbi) (percent) renminbi) (percent) renminbi) Expressways 0.01 0.1 56 28 280 Highways Class 1 0.02 0.3 30 15 150 Class 2 0.20 4.0 60 30 300 Class 3 0.84 16.5 12 6 60 Class 4 2.60 51.0 27 13.5 135 Unclassified Roads 1.43 28.1 15 7.5 75 Total 5.10 100 200 100 1,000 Source: Mitchell Stanfield & Associates. 48 WORLD BANK WORKING PAPER funds for highways. This capital is mainly invested in hiigh-grade highways 'because all are toll facilities and, along with rising traffic, are gencrating incrcasing rcvenues ior opcrations and mainteniance and dividcnds for investors. Even so, pnvatc funds have played a limitcd rolc in the funding of the lighway network and still lag in the race to mect China's ambitious construction requiremnents. In addition, over the past decade about 80 percent of the privatc invcstments in highways have been concentrated in coastal provinces Morc important, most private financing has been limited to govc -nment-controllcd toll road projects, particularly rchabilitat-ion or cxpansion of existing roads wvith proven traffic (for examplc, 3,500-5,000 vchicles a day for cxpressv,avs)-that is, broxvnficld orojccts. That means that in most cases all construction and traffic risks ace assumed by the public sector. Finally, there has been almost no long-tcrnm foreign dcbt raiscd on a proicct financc basis and very little foreign equity from sourccs other than FIong Kong (China). A fcv developers in Hong Kong have provided equLity and raised international debt on the strcngth olinhcir balancc sheets (instcad of the projccts). One of thc dcvclopers, New WVorld Infrastructurc, has also established a mainland subsidiary to complemcnt its cquity investments with borrowvings 'n rcnminbi for at lcast one toll road project. But recent changcs in central government policy on minimum rcturn guaran- tecs havc affected developers' capacity to securc government payments stipuiatCd in joint venture agreements. This has caused at lcast onc firm to announce in 2001 that it Is pulling back from sub- stantial investments in fuiture toll road projects. Policy and Institutional Structure The policy framework for private investmcnt in 'highways was picccd togctlhcr in the 1990s, complemented by cross-sectoral regulations .nd decrccs. Regulations anci decrccs spccific to roads and applicable to private, particularly forcigni, investment includc the "Provisions on the Establishment of the Forcign-Funded Construction Entcrprises" (1995), which permits Chinese-foreign joint ventures and cooperative cntcrp riscs but prohibits wholly foreign-owned enterprises from engaging in civil engineering projects. Moreovcr, the ne\ Highway Law has yet to be filly implcmented. The national highway program is admimstcrcd by a combinationi of central agcncics and mini- istries. It is often implemcntcd by provincial communications departmncnts with the involvement of many provincial and municipal cntities (Figure 4-2). For cxample, tarifFs must bc approved by provin- cial govcrnmcnts according to guidclines sct bv the State Development Planning Commission, and managed by provmcial pricc bureaus, oitcii Wvith the parLicipation of provincial committees. There is no arrangement at the cL ntral lexel For multivear Iunding of projects, nor is there a formula for allocating funds to provinces bascd on population density, lanc kcilomneters in opera- tion, number of vehicles, and other factors. Models for Private Participation The fcew models that have bcn uscd for private participation in roads havc c;ncncnced drawbacks in their application. Cooperative Joint Ventures. China has gencratcd significant highway rehabilitation work and some new expressway construction through joint ventures bctwccn developers and provincial or municipal agencies. During 1990-2000 therc haNc been more than 8() joint venture projects in 11 provinccs, along with the Shanghai and Tianjin municipalitics. 'I'hc total capital cost of thcss projects is cstimated at 112 billion renminbi, 61 billion of which is private fuiding (almost all from Hong Kong developers)-accounting for 54 percent of privatc investment in roads during this period. The coopcrative joint venture arrangemenet is in many ways a substitute for a broad legal and regulatory framework conducive to attracting foreign capital. In the absencc ot'-such a frame-work, foreign mvestors lack confidence that thcir Clhinese counterparts will compl) wkith contracts. The involvement of Hong KIong developers has been motivated by their capacity to ncgotiate arran-gc- PRIVATE PARTICIPATION IN INFRASTRUCTURE IN CHINA 49 l Central | |State Development Miityo |Plannmng Comnmission ||CommunicationsJ Govern ment -0 Commission Depomuiartions Expressway Company Source: Mitchell Stanfield & Associates. ments, including preferential treatmenit (guarantees, tax incentives, foreign exchange loss protection, and so on). As a result cooperative joint ventures on the mainlanid are closed to non-Hong Kong developers and havc been restricted to nonicompctitive transactions. The process could be made more efficient and perhaps cheaper for the project with competi- tive bidding for forcign partners, who could provide not just equity capital but also equipment and internauonial expcnencc in the commercialization of toll roads The current nutmber of players seems too limitcd. Toll equipment and operations and maintenancc services could be competitively procured on a contract basis A well-cstablished legal and regulatory framework would encourage more forcign investors to participate in competitive bidding and reduce public subsidies It would also allow existing roads with limited traffic risk to be partially refinanced with passive foreign cquity and rcgional infrastructure funds International Debt Financing in the 144A Market. China has had great difficulty developing workable regulations to permnt the financing of highway infrastructure with long-term debt secured in international markets. In the absence of a well-structurcd legal and regulatory framcvork, most mainland companies do not have access to large institutional investors (such as insurance companies) and are not able to secure significant rcnminbi debt from domcstic banks. Major Hong Kong developers with large project portfolios and relatively lower credit risk are able to issue corporate bonds in the US based 144A market. Other firms who pursue this approach with project bonds are penalized with late tariff adjustments and difficulties in gaining access to hard currency for repaymcnt of debt. The 1997 Project Finance Measures (see Chaptcr 3) effectively gives international creditors no access to any assets, revenues, or guarantees outside a projcct. In the lack of a clear regulatory sys- tem, such regulations have almost halted commercial project financing of infrastructure projects- even in coastal provinces with stronger creditworthiness. The limilted number of Hong Kong investors (New World, Road King, and Chcung Kong Infrastructure) who have securcd long-term 50 WORLD BANK WORKING PAPER debt on the strength of their balance sheets (instead of the projects) might see downturns in their financial results. The few non-Hong Kong foreign investors who have participated in China's transport sector also seem to be pulling back (see Box 4-1). Li ~ The entire system needs to be reorganized, with clear rules and lines of authority, and fairly administered approval procedures. Without these reforms, project financing in the 144A market wNill contnue to be difficult to sustain. Securitization of existing highway assets. Since 1996 eight expressway development companies- acting as financing vehicles for provincial governments-have been listed on the domestic stock exchanges, and five have been listed in Hong Kong, bringing in about $2 billion for the rehabilitation and expansion of existing transportation projects. Some provincial expressway companies have also explored overseas listings. Zhejiang Expressway Company, for example, has applied to the U.K. securities authority to have its H shares listed on the London Stock Exchange (Box 4-2). Mature expressways and bridges with well-defined traffic levels are injected into these share- holding companies to generate immediate cash flows and profits, to attract public investors on the exchanges. It is also common for these shareholding expressway companies to have preferential, if not exclusive, rights to develop expressway projects in the province. The companies have also been granted first right of refusal to accept any future road concessions offered by the provincial governments. The future level of asset securitization in the highway sector will fluctuate depending on mar- ket conditions and investor interest in and capacity for highway securities. In a more mature mar- ket, with limited share appreciation potential, the earnings and dividend performance of highway companies will also dictate how well new share issues are received in equity markets. Build-Operate-Transfer (BOT). Since the mid-1990s China has considered using BOT schemes for highway projects involving foreign investors, but no such schemes have been implemented. For example, the World Bank sponsored a BOT feasibility analysis of the Junshan Bridge over the Yangtze River in Hubei Province, to test the concept. In 1998 the State Council decided instead During 1996-98 four developers launched 144A financing for highway projects on the mainland, including two from Hong Kong (Cathay International and Traffic Stream Infrastructure) and two from the mainland (Zhuhai Highway Company and Greater Beijing First Expressways). In each case it appears that currency convertibility and repatriation of earnings to investors outside the mainland is a continuing problem. The central authorities have taken the view that the financing structures implemented by these developers were intended to circumvent regulations, especially approval from the State Administration of Foreign Exchange, and have penalized the companies at the expense of note holders. As a result three of the four developers (Traffic Stream, Zhuhai Highway, and Greater Beijing First Expressways) have defaulted on debt issues rated by Standard & Poor's. The remaining firm, Cathay International, has been downgraded to CC, with a negative outlook by Standard & Poor's. Other Hong Kong developers (Road King, Cheung Kong Infrastructure, New World Infrastructure) have had to use their balance sheets to secure international debt for their infrastructure joint ventures with mainland provincial entities. All these companies have reported higher earnings on a year-to-year basis, with timely road tariffs and conversion of renminbi revenues to hard currency for repayment of debt. But recent restrictions on the provision of minimum return guarantees to developers and a prohibition on loan guarantees by subsovereign government units could change the situation. The few non-Hong Kong foreign investors are also pulling back In May 2000 First Group, the large U.K. trans- port company, sold its 26 percent interest in New World Holdings to a New World affiliate. Source: Project team PRIVATE PARTICIPATION IN INFRASTRUCTURE IN CHINA 51 The Zhejiang Expressway Company develops, owns, and operates high-grade roadways in Zhejiang Province and engages in ancillary businesses such as billboard advertising and gasoline and automobile service stations Its primary asset is a 250-kilometer Class I highway connecting the cities of Shanghai, Hangzhou, and Ningbo. The company is a joint stock company held by agencies of the Zhejiang provincial government (65 percent) and H share investors through the Hong Kong and London exchanges (35 percent). The company raised about $425 million in its 1997 initial public offering (IPO) of H shares, making it the second largest joint stock issuer in the toll road sector. Since its incorporation the company has maintained a conservative financial structure, with a debt to equity ratio of generally no more than I to 9. The company's largest creditor is the World Bank, which extended $170 million in loan facilities for the initial phases of construction. The balance of the company's debt, approximately $25 million equivalent in renminbi loans, is provided by domestic banks. During 1999-2000 the company's shares traded at a discount following the East Asian financial crisis and in the wake of GITIC. But the share price recently bounced back for the following reasons: O The company's location in Zhejiang Province, which continues to exhibit strong economic performance and has emerged from the Asian crisis and GITIC period with a relatively clean market reputation. U The company's strong reputation with investors for disclosure and management practices. O The company's conservative financial structure. But the Zhejiang Expressway Project also raised questions for future prolects on the following issues: O The provincial government's ability to achieve competitive terms for follow-on projects where the original concessionaire holds a preemptive right. O Foreign currency funding for a business that derives all of its income in renminbi. O Conflicts of interest that may emerge in a transaction where government or quasi-government entities are playing the roles of regulator, manager, majority owner, and service contractor. O The considerable risk of patronage in toll road projects, even in the most economically robust regions. Source: ANZ Bank. to finance the project with public funding as part of the Third National Highway Project sponsored by the World Bank. But the policy issues raised by the Bank in its Junshan feasibility study are still applicable. Thesc include rcgulatory oversight, bidding process and evaluation, land acquisition and resettlement, permits and approvals, toll setting and adjustment, development rights, performance security, profit remittance, and dispute resolution. It is time to review the existing BOT framework for use in road projects This framework, dis- cussed in earlier chapters, shoild cover the allocation of nsk between public and private partners. A sound framework could open the way for new sources of foreign direct investment and project financing in the sector. Recommendations Premier Zhu, in his report on the outinie of the current Tenth Five-Year Plan, pointed out that "carrying out the strategy for western-region development to accelerate the development of the central and western regions is a major step taken to achieve the strategic goals of the third stage of the country's modernization drive. During the Tenth Five-Year Plan period, we need to place emphasis on key projects for a good beginning of the program Construction of infrastructure and protection of the ecological environment should take priority, and we should strive for major breakthroughs within five to ten years." China has massive highway investment needs over the next 20 years, indicating great potential for private investment. Most of the opportumties for government to attract investors are in devel- 52 WORLD BANK WORKING PAPER oped provinces and on corridors with significant traffic. Although road projects in less developed regions are of great social and economic developmcnt value, it is difficult to attract private financ- ing without some public financing, given the less profitable traffic and economic outlook. So, although developing a viable road transportation system in China's central and western regions is an urgent concern for the government, private financing of that system should be designed in a cautious, phased approach. Problems with existing road projects in the eastern region-including unclear government policy on public and private financing, lack of transparency in the approval process, weak contract enforceability, and defaults on loans and bonds-discourage mternational investment in the sector. Solving these problems would boost investors' confidence in China's road sector. In addition, broademng the scope and scale of private participation in infrastructure would enable the govern- ment to significantly reduce public funding in developed regions. To raise debt financing, govern- ment agencies in charge of toll road development need to be corporatized and legally separated. These reforms are needed if public funds and other resources are to be allocated more efficiently to the west. After a basic road network has been built with public funding and traffic has stabilized to a certain extent, the private sector can be brought in to the west to take over a "browvnfield" project, or a pool of projects-including greenfield ones, which may have fewer risks and lower returns. Government funding freed by these "takeovers" can be used for road systems in even less developed regions. Establish Clear Policy on Government's Role, Encouraging Wider Private Participation The dominance of public investment and ownership has dampened the interest of international investors in China's highways. Thus the government should establish a clear policy on the extent to which the pnvate sector should be involved in developing the highway network-especially in coastal regions, where properly structured projects can attract private investment. Broadening the private sector's role anid the scope of acceptable schemes for private participa- tion would also allow government to develop, on a case-by-case basis, a cooperative structure with private investors. The new structure would be intended to optimize the use of private participation in meeting project objectives, addressing financial characteristics and requirements, and defining an acceptable risk profile for the parties. International experience shows a variety of approaches (Box 4-3). For example, greenfield investments in the road sector are generally perceived as being too difficult to attract because of high construction and traffic risks. If the government can assume some risks that are under its direct control (planning risk, political risk, and so on) and participate in capital investment, the bar- ner to entry by the private sector will be lowered-especially in the first few years of the project, when traffic risk is highest. Greenfield projects could also be packaged with existing profitable proj- ects to improve profitability and lower the risks of startup roads. Cross-subsidies have been useful for network expansion in many countries, including Japan. Finally, even where private investors' interest is lnmited, models for private participation involving less pnvate capital investment but sig- nificant managerial and technical expertise-such as management or supply contracts, concessions, and TOT (transfer-operate-transfer) or ROT (rehabilitate-operate-transfer) schemes-would allow the public sector to allocate more commercial responsibilities to the private sector. Private responses to policy changes will depend on the method of implementation. Policy reforms should be widely publicized at the conceptual stage, and provincial governments should also be fully informed and asked for comment. Possible reforms could include private participation models that were once rejected, but that with refinement could be acceptable to the central gov- ernment. In addition, recommendations might be disseminated on the pros and cons of different models and practical solutions suggested to address key implementation concerns after testing them on specific projects. It is also important that policy issues be considered in parallel wvith the strengthening of the legal framework, to trigger renewed interest among investors after their main concerns on the structure for private financing of highways have been addressed. PRIVATE PARTICIPATION IN INFRASTRUCTURE IN CHINA 53 Argentina and Australia undertook programs to privatize their mature road networks, railways, and airport operations and have seen the development of some greenfield projects. Asian countries, by contrast, generally rely on prvate participation in infrastructure to build the new transport infrastructure required by their fast growth. In 1991 the Argentine government, lacking funds to properly maintain and rehabilitate its road network, simultaneously awarded twelve 12-year concessions of intercity highways through competitive bidding. Bidding documents required the consortiums to achieve prescribed service levels within a contracted period and to assume existing investment proposals. The decisive factor for selection was the ability to pay an up-front fee to the government on award of the concession. After 1992 the government awarded four 22-year conces- sions, including some greenfield prolects, for access roads to Buenos Aires. This time the sole criterion was the lowest toll offered under serviceability constraints. There are many privately run DBFO (design-build-finance-operate) roads in the United Kingdom, and the model for these concessions is being copied in several other companies, including Ireland and South Africa. Private operators generally acquire an existing road and are required to perform certain upgrading work or construct new sections of road and are subsequently permitted to operate the road under concessions, which usually run 30 years. U K. road operators are paid based on shadow tolls (determined by use and availability), and the government is comfortable that the revenue it must pay the private operators based on the shadow tolls will be less than it would have to pay were it to carry out the upgrading work and ongoing maintenance over the 30-year period. Chile has experienced rapid economic growth for more than a decade. The need to increase investment in infrastructure was recognized in the early 1 990s, when policymakers decided to introduce private capital in the transport infrastructure sector, covering roads and highways, bridges, tunnels, and airports. The chosen mechanism was a concession, whereby a private firm would finance and build a project and then operate the infrastructure for a set number of years. The concessionaire would recover its investment by collecting tolls from users. Chile's experience is interesting for several reasons. First, for its size and scope: more than 2,000 kilometers of roads have been concessioned. Second, unlike in Mexico, Chile's program has been mostly successful. Although it is still early in the program, the prolects that are in service have not confronted any major setbacks in terms of traffic levels, construction delays, cost overruns, or other problems. Third, the program has been the motiva- tion and testing ground for some interesting innovations in tendering mechanisms The main one is the least present value of revenue auction, which has been used in the tendering of two road concessions. In such auctions the bidding variable-rather than being toll levels or some other conventional variable- is the present value of revenue throughout the life of the concession that firms are willing to accept to under- take the project The firm that bids the lowest present value of revenue wins. The duration of the concession is then flexible and depends on the effective traffic levels encountered. If traffic levels are lower than expected, the duration of the concession is extended automatically, while if traffic is higher than expected the opposite occurs. Thus income uncertainty due to traffic variations is largely eliminated for the concessionaire. Other advantages to this form of concessioning include the reduction in traffic risk, and hence the reduced need for the state to extend income guarantees. Likewise, such concessions may reduce the likelihood of bidders low-balling their bids with the expectation of renegotiating the contract later. Nearly half of Chile's road infrastructure deficit has been covered by concessions (though not all through the least present value of revenue mechanism) The rest will be addressed through forthcoming projects funded by the private as well as the public sector. The program has generated, in present value terms, net revenues of $130-150 million (Gomez-Lobo and Hinojosa, 2000). Source Project team. 54 WORLD BANK WORKING PAPER However, the full impact of policy changes on private capital flows can b2 achieved only under the following conditions: The establishment of a consistent legal and regulatory framework as discussed above. Fair, open competitive bidding for private investors, whatever their role in project develop- ment and financing, including partners in joint ventures. A strong legal basis for each project structure, wvith particular attention to provisions for contract compliance by govermnent, regardless of whether the development method is con- cession or transfer of operating rights, BOT, lease, operations and maintenance contract, or another model for private participation in infrastructure. Corporatize Toll Road Entities into Provincial -roll Road Authorities Most tolled expressways constructed by provinces on a project-by-project basis arc operated by provincial communication departments or by shareholding companies controlled by the parent government. Consolidating thcse projects in provincial toll road authorities avith the capacity to manage large expressway networks would open opportunities to mobilize private participation in various roles, especially the funding of new investments. To the extent that provincial communication departmcnts are government dcpartmcnts instead of separate legal entities, it would be difficult in Ctina's system for these departments to access financial markets for debt capital to fund large individual projects. In addition, shareholding companies established to operate single corridor projects do not have strong enough balance sheets or financial performance to attract large-scale capital in financial markcts. Thus the high demand for capital in the road sector has created a need to consider new organizational methods for consolidatng and managing existing toll roads, thereby leveraging income generating assets to raise private financing for future expressway projects A survey of international experience showvs that publicly held toll road entities have been developed through a variety of structures and witl diversified financing modes (Table 4-3). All these schemes have succeeded in mobilizing private participation, though of a limited range relative to traditional private participation in infrastructure models, where private developers and lenders have wide latitude m construction and operations (Box 4-4). As outlined m Table 4-3, country experiences, when taken together, may provide several alter- natives for an institutional model suited to China's unique situation and constraints. The provincial toll road authority concept can be used by Chinese provinces to create a larger pool of assets and a broader base of project revenue flows to leverage new capital from the private sector. The model that ultimately evolves in China must identify the needs of key stakeholders at the policy, investment, management, and service delivery levels-namely, the government, users, fund providers, and operator. Hence the new structure should also have the capacity to: 7 Improve the cost efficiency of operations and maintenance of toll roads under management. . Raise new funds for construction by leveraging expressways in a network or system fashion. Devise new financial instruments for raising private capital in domestic and foreign markets without the need for sovereign guarantees. The move to provincial toll road authonties should evolve m two steps, continuing the move to an increased role for the market in the road sector as envisioned in the Tenth Five Year Plan. The first step would be to use the newly formcd provincial toll road authority as an exprcssway management company for publicly funded toll roads. In the second step stand-alone provincial toll road authori- ties would assume full ownership and operating responsibilities for expressways, consolidate their operations in a network, and access financial markets for capital to expand, maintain, and rehabilitate the network. In the first phase the provincial toll road authority provides utility-type scrvices for manage- ment and operation of provincial exprcssways through close interfaces wvith provincial communica- tion departments and other public agencies This model can mobilize private participation for PRIVATE PARTICIPATION IN INFRASTRUCTURE IN CHINA 55 TABLE 4-3: PRIVATE PARTICIPATION DEVELOPMENT MODELS AND FINANCING MODES FOR HIGHWAYS Development Structure Financing Mode Countries Comments Public entity Public toll road 1: Long-term revenue O Norway a Requires strong operations and authority bonds E United States maintenance capacity and solid Ol User fees development track record Expressway a3 Indirect user fees l Australia D Operations and maintenance management company. 3 Korea service contracts administered by 0 New Zealand public agencies but funded off budget Public highway D Sovereign-backed El Portugal Commercial financing supported by corporation debt or equity Q Slovenia user fees C User fees a External debt requires sovereign support Asset securitization a Equity offering D China n Market sensitive and costly to owner O Limited capacity to fund large-scale development Private participation models Joint venture a3 Public and private China Ol Satisfies need for limited shareholding company equity development capital a Bank loans D Asset base too small for debt 1 User fees financing Maintenance contracts El Commercial debt and O Argentina O Requires strong institutional or concessions World Bank loans El Brazil oversight O Government payments E Chile D Needs stable budget allocations Design-build-finance- n Projects implemented Finland Dl Limited commercial risk transferred operate contracts by private sector 0 Netherlands to concessionaire prolect debt/equity Portugal 13 Needs stable public funding base Q Shadow tolls paid by D United E Project credit related to strength of government Kingdom government payment mechanism Private highway Commercial debt Italy Best example is Italy's Autostrade, corporation 0 Private equity formerly owned by government, but sold in public offering in 1999 BOT concessions Ol Public and private El Chile D High level of public and private risk equity E3 Greece sharing needed to secure debt El Prolect debt Mexico financing El Government subsidy Dl Spain Source Mitchell Stanfield & Associates. management, operational skills, and innovative technology, but it is not intended to provide access to pnvate financing. In the second phase the provincial toll road authority becomes a stand-alonie concessionaire for the province, legally separated from provincial communication departments and with responsibility for operating a wide toll road network and developing new corridors (Figure 4-3). The rcvenue from tolls would finance maintenancc of expressways and new construction. Most important, a stable income flow would enable the provincial toll road authority to attract private 56 WORLD BANK WORKING PAPER Box 4-4: INTERNATIONAL EXPERIENCE WITH PROVINCIAL TOLL ROAD AUTHORITY PARTNERSHIPS WITH THE PRIVATE SECTOR The United Kingdom is probably Europe's most notable recent example of highway management by a strong, well-capitalized central agency, the U.K. Highways Agency, with a large contingent of mainly European contrac- tors, design firms, and toll operators who have formed private consortiums for a massive program of highway rehabilitation and expansion using the design-build-finance-operate technique. Rather than applying direct tolls to its most traveled roads, the Highways Agency compensates the consortiums for their maintenance activities through indirect user fees payable to concessionaires based on a set of road performance criteria. The design-build-finance-operate program is concentrated in several corridors of high-capacity highways. New developments in the program call for an independent managing agent to interface the Highways Agency and the consortium to monitor and assist in the enforcement of performance standards for road operation. In contrast to BOT-type roads, the design-build-finance-operate projects are more easily financed in capital markets because they rely on the strength of the Highways Agency balance sheet rather than direct tolls as the source of payment to the concessionaire. Source: Project team. capital through long-term debt in financial markets, well beyond the proceeds of occasional equity listings. The structure of the New York City Municipal Water Authority, is an example (see Chap- ter 3, Box 3-2 and Figure 3-1). In China the provincial toll road authority could borrow funds from commercial banks and issue bonds secured by future toll revenues as collateral for debt service (Figure 4-4), if various conditions are satisfied. Owner/Funding Province Banks Investors Owner/FundingII Sources Dividends Loans/Equity D S I ~~~Debt Service I. and Dividends Manager/Operator PTRA Funding Tolls Provincial Assets Expressway I Contract Service Contractors Provider PRIVATE PARTICIPATION IN INFRASTRUCTURE IN CHINA 57 Owner/Fundinig Sources Province/ PCD Capital Markets Project Sponsors Strategic Investors A Financial Investors Equity/Debt iiedDb . t Dl~~~~~~~evicendDb Provin ial TollConcession MVanager/Operator Road Authority Investments Revenues Investmnents |Tol Assets Toll Roa etwor The provincial toll road authority first needs to comply with disclosure requirements and achieve a sustainable level of creditwortlhliness, which vill be evaluated by undervriters and bond purchasers mainly on the basis of ratings issued by bond rating agencies, with a well-defined sct of criteria: 0 Scope of activity of the provincial toll road authority, as defined in its article of association and the provisions of the agreement or concession signed with the province and delegating operating rights on toll roads. S Current netwvork traffic volume, growth potential, and assumptions underlying both. Li! Historical cash flows [3 Aggregate toll revenues, tariff levels, and likelihood of future growth E3 Competing roads or bridges that may divert traffic. S Political wvillmgness to sustain needed capital and maintenance spending at an appropriate level to support network performance. Lenders and bondholders may also require that a certain level of the toll road authority's operating cash flow be held in a reserve fund. Such a trust account would help preserve the authorlty's finan- cial viability if, for example, contractually determined tariff increases are not approved by govern- ment and debt service ratios fall below a predetermined level. A creditworthy structure for the provincial toll road authority should significantly improve the breadth and cost of filancings that have been completed by provinces that have accessed financial markets. A prime example is the Zhejiang Expressway Company, where the provincial government collaborated vith local government entities to develop a provincial-level Class 1 expressway under a clearly defined concession arrangement. Zhejiang raised anchor financing from the World Bank and then tapped the H share market in Hong IKong for fiinds to complete the construction. But the company did not raise finds on the bond market. The situation is the same in Guangdong, where the Guangdong Provincial Expressway Development Company raised about $127 million from B and A equity share issUes in Shenzhen. 58 WORLD BANK WORKING PAPER Shanghai Midway Infrastructure (Holdings) Limited is one of the first private domestic companies principally engaged in investing in high-grade toll roads in China. The International Finance Corporation has invested in the company indirectly through Hansom Investment Limited (Hansom), which is registered in the British Virgin Islands. Midway owns substantial stakes in all its subsidiary companies, with ownership rights to six toll road concessions in the provinces of Zhejiang and,Jiangsu. The total length of roads included in Midway's concessions is 205 kilometers. The Narada Group, a private company in Zhejiang Province, directly and indirectly holds 70 percent of the equity in Midway. The International Finance Corporation and AIG Asian Infrastructure Fund 11 indirectly hold 15 percent. The International Finance Corporation was attracted to the investment for several reasons, including * While Midway has municipal or provincial partners for all its projects, for the most part it remains in a major- ity position, with responsibility for toll road administration, operations, and maintenance. The projects were expected to benefit from local governments' financial interest in the returns of the project companies and the efficiencies of private management. * Midway's strategic partnership with Narada Group, a strong local investor committed to the domestic toll road market, with international standards for corporate governance. * China's need for new infrastructure investment and the government's commitment to use the private sector to spur development of infrastructure projects. Although Midway's portfolio doubled in size between 1998 and 2000, the company has been able to limit the size and growth of management overhead. The company retains a small, highly skilled group of managers whose capacity is enhanced by the occasional use of experienced international consultants for financial and legal advi- sory services, traffic projections, and engineering and technical analysis. Subsequent to the International Finance Corporation's investment, the Chinese government made available sub- sidized sources of funding, generally through state-owned banks, which crowded out new private investment. In addition, local governments responsible for enforcing individual road concessions came under pressure from local citizens and higher-level government authorities to lower the costs (either directly to users or indirectly through tax policy) of the toll roads to society, at the expense of Midway's investment return. Source: International Finance Corporation. Sustain the Development of Domestic Private Sponsors A few large Hong Kong developers have captured a sizable share of the market in coastal provinces for expressway construction and highway rehabilitation and maintenance. But most of these pro- jects are for high-traffic roads linking major cities and activity centers. Larger developers generally have not accessed the market for Class 1 an-d 2 highways at the subprovincial level, leaving new opportunities for private domestic companies in local road networks. With a growing number of Chinese domestic sponsors experienced in developing, constructing, and operating roads, public agencies should consider transferring more of these responsibilities to the domestic private sector. Although the Shanghai Midway has had mixed results (Box 4-5), it may be suitable for private participation in infrastructure on a smaller scale. The model has the potential for indigenous appli- cation to China with public partners in a minority position, but who are willing to contribute right-of-way and necessary construction and environmental permits. Such smaller, focused compa- nies can be used to develop roads at the municipal and provincial levels with streamlined local approval procedures and minimum involvement by central agencies. 5 WATER ISSUES AND RECOMMENDATIONS T he spectacular population and economic growth in China over the past decade has signifi- cantdy increased the demand for water and samtation services. In response, enormous public fundcng has been channeled to the water and sanitation sector The government wants to make the sector more efficient through numerous reforms, including adapting the laws and regula- tions related to water supply, implementing tariff reforms for water and wastewater, restructuring and corporatizing water utilities, experimenting with various forms of private participation, and so on. This chapter reviews the current situation and recommends a multifaceted approach to advance sector reforms. The Situation China's water resources are unevenly dtstributed. According to a recent World Bank report (World Bank 2001a), northern China contains 37 percent of the country's population and 45 percent of its cultivated land-yet has only 12 percent of water resources. Water is particularly scarce in the so-called "3-H" catchment area of the Huai, Huang, and Hai rivers, where about 35 percent of national GDP is produced. Water availability in the Hai-Luan basin, for example, is as low as 355 cubic meters a person a year, far below the internationally accepted definition for water scarcity of 1,000 cubic meters per person per year. Over the past decade, population and economic growth, together with increased urbanization, have led to a sur,ge in the demand for water. The water supplied by municipal water companies to nonindustrial users has been increasing by 7.5 percent per year (Tablc 5-1). Per capita consump- tion in urban areas is about 200 liters per day, higher than the consumption level in industrial countnes. Leakages in the distribution networks of some old urban water systems are estimated to be as high as 30-40 percent. The growth in demand puts tremendous environmental pressures on water resources (water short- ages, pollution, falling groundwater table). The problem is rapidly approaching crisis proportions- particularly in the north, where downstream pollution of water resources is making further 59 60 WORLD BANK WORKING PAPER TABLE 5- 1: SELECTED INDICATORS ON URBAN GROWTH AND WATER AND SEWER SYSTEMS IN CHINA, 1991 AND I 999 Annual growth Indicator 1991 1999 (percent) Official urban population (millions) 305 389 3.1 City-based nonaggcultural 148 202 4.0 population (millions) Number of designated cities 479 667 4.2 City-based domestic water use 10,626 18,969 7.5 (millions of cubic meters), Domestic water use per city-based 197 258 3.4 nonagricultural resident (liters per capita per day)b Municipal wastewater treatment 1,703 5,779 16.5 capacity (millions of cubic meters) Treatment capacity / domestic II 21 8.4 water consumption (percent) City-based sewer network length 61,601 134,486 10.3 (kilometers) Notes: a. Includes all nonindustrial municipal water supply, including institutional demands, urban irrigation, and so on. b. Calculated as (city-based domestic water use/city-based nonagricultural population). The numbers seem to be twice as high as consumption levels in industrial countries. Part of the reason could be lack of data on "agricultural populations"-that is, rural immigrants dwelling in cities who have not properly registered the status of their residence with the municipalities. More accurate estimates for 1999 are believed to be close to 200 liters per capita per day. Source: World Bank 2001 a. investments in potable water supplies even more difficult and expensive, to the poinlt of affecting future economic development in cities with severe water shortages. Capacity Planning and Development In response to the growing demand, the government has declared the development of water and wastewater infrastructure a national priority. Over 1990-98 about $9.5 billion was invested in water supply infrastructure and $6 billion was invested in sanitation infrastructure (Figure 5-1). As a result the capacity of water production plants and distribution networks increased by almost 50 percent. Around the same period, municipal wastewater treatment capacity grew 17 percent a year in terms of volume and 10 percent a year in terms of sewer network length (Table 5-1). Still, the continuity of service and the quality of tap water remain major concerns for urban populations. In addition, more than 30 percent of water networks have been in use for more than 20 years. These aging networks appear to be operating at excess capacity and deteriorating. Official data for physical unaccounted-for water, though suspected to be understated, incrcased from 9 percent in 1990 to 13 percent in 1998. Although enormous effort has been made to develop urban sanitation infrastructure, the double- digit growth in capacity has not been enough to stop the growing contamination of water resources. The sewer network covers only half the population connected to the potable water network, which contributes to the contamination of underground water resources. With only 30 percent PRIVATE PARTICIPATION IN INFRASTRUCTURE IN CHINA 61 _WaterSupply -0 San.lmalor Investmenl| 1 8000 160000 1 4 000 12000 a 10000 so000 6000 4000 -9_ 2000 0 1992 1993 1994 1995 1996 1997 1998 Year Source SOGREAH data of wastewater flow being treated before discharged, pollution loads discharged from urban areas to receiving waters have actually increased. The government's investment targets for the urban water sector, as exprcsscd in the Tenth Five-Year-Plan, are unprecedented: ovcr the 2001-2005 period, more than $11 billion-cquiva- lent to the capital costs of 103 Chengdu BOT water supply projects-will need to be invested in watcr supply and $15 billion will be needed for sanitation infrastructure (including raising waste- water treatment coveragc to 40 perccnt of total sewer flow). Institutional and Tariff Framework Being such a huge country, China understandably has a complex institutional strutcture for the water sector, with strong interdependencies across departments and agencies. At the state level the main agencics are the Ministry of Watcr Resourccs, Mimistry of Construction, State Environmental Protection Admiiustration, and State Development Planning Commission. At lower government levels there are commissions and bLireaus responsible for planning, construction, finance, indus- try, and environmental protection, all of which could relate to water supply and sanitation. These commissions and bureaus report to both their local government bodies and to their state-level counterparts. Municipal governments are primarily responsible for providing water and wastewater treatment services, owing and managing morc than 60 percent of water capacity Unlike water service providers, which have been corporatizcd (albeit government-owned, Linder the supervision of muniicipal public utility bureaus), wastewatcr companies have only recently been establishcd in a milority of municipalities. Most wvasteivater systems conttnute to be managed separately from wvater systems by municipal management bureaus. Various studics have shown that wvater tartffs are very low in China, even lowcr than water pro- duction costs (not including the cost of distribution and commercial management) This reversed relationship bctween production costs and retail tariffs wvas evident in several water BOT projects, where the bulk tariff the project company charged the local water company was higher than the retail 62 WORLD BANK WORKING PAPER water tariff for end consumers. Tariffs are also low relative to those in both developing countries in the region (Figure 5-2) and in many countries in other parts of the world (Figure 5-3). More and more municipalities are starting to understand the importance of recovering costs and have steadily increased water tariffs. Between 1986 and 1998 the average tariff in China increased 18 percent a year (Table 5-2). Still, tariffs are not high enough to ensure the financial via- bility of water companies. Annual losses in the sector were estimated at $600 millions in 1998, requiring increased pubhc,subsidies for the sector. There are also significant discrepancies between water tariffs in different cities, even in the same provinces. For example, in 1998 water tariffs in Zhejiang Province varied from 6 U.S. cents per cubic meter in the city of Jinhua to 25 cents in the city of Zhoushan (Figure 5-4). The "Regulation on Management of Pricing of Urban Water Supply" was promulgated in 1998 (see below). It offered the first national guidelines on water tariffs and established the princi- ple of full cost recovery through tariffs. In 2000 the average residential tariff for Chinese cities was increased to 1.09 yuan (about $0.14) per cubic meter, 40 percent higher than the 1998 level. Because the wastewater tariff is even more problematic than the water tariff-few municipalities applied sufficient wastewater tariffs to cover even operational costs-in 1999 the "Circular on Strengthening the Collection of Wastewater Treatment Tariffs" came into effect. While these tariff regulations provided a legal basis for further tariff reform, they have not yet been effectively enforced nationwide. Legal and Regulatory Framework Several laws and regulations specifically govern water and sanitation services. The Water Law was promulgated in 1988, and various water quality and wastewater discharge standards have been issued at the national, provincial, and municipal levels-sometimes contradicting each other. The "Regulation on Management of Pricing of Urban Water Supply," the first national guide- lines on water tariffs, states that municipalities should be responsible for approving tariff changes and that water tariffs should cover all operations and maintenance, depreciation, and interest expenses, and allow for a 8-10 percent return on the net value of fixed assets. While the principle of cost recovery is stated correctly in the regulation, an 8-10 percent return on the net value of (U.S. cents per cubic meter) 160 . . . . ... .. ... . 1 40 - . . ............ 400 . . . . . .. . . . . . . . . ... 200 80 59tt gS <99 to att9 ttG GOO 0 00490 0 Source: World Bank and United Nations dau. Source: World Bank and United Nations data. PRIVATE PARTICIPATION IN INFRASTRUCTURE IN CHINA 63 (U.S. cents per cubic meter) 1 0 - - - - - - - -- - 120 - .-. . 100 - - - 80 - - -- 6 0 - - - - - - - - - - - -- - - - - - - - - - - 4 0 - - -- - - - - - - -- - - - - - 20 - - -- - - 0 1777 Source World Bank and United Nations data. fixed assets (which is roughly equivalent to return on equity) is too low for investments in China's water sector. The "Circular on Strengthening the Collection of Wastewater Treatment Tariffs" (1999) established similar economic principles for sanitation services. It cstablished a wastewater treatment fee that requires recovery of operations and maintenance costs and a moderate profit. This fee replaces myriad unregulated and ad hoc fees that were used to finance wastewater investments. Several decrecs and regulations on foreign direct investment also affect the water sector. For example, the "Interim Provisions on Guidmg Foreign Investment Direction" and the "Catalogue for the Guidance of Foreign Investment Industries" (1995) prohibitforeign tnvestment and management in urban networks of water supply, sewers, and water drainage. The government is considering allowing private investors to build tap water distribution systems, and this initiative has been well received by the pnvate sector. But until the private sector can operate and manage the network, including metering and billing consumption and collecting tariffs from end users, sigmficant improvements in service quality and cost savings are inlikely As with the road and power sectors, the effective enforcement of regulations has been con- strained by institutional complcxity and by the divcrsity of interpretations of regulations and direc- tives by various agencies and local governments. Some regulations and directives have been released only to local govcrnments and are not readily available to the private sector. Even veteran water devclopers may face prolonged project lead time, because different regions may have differ- ent approving authorities, procedures, and enforcement practices. This ad hoc approach increases development costs and financial risks for private investment in China's water sector. 1986 1996 1997 1998 Average tariff (industrial and residential, yuan per cubic meter) 0.14 0.63 0.78 0.92 Average annual increase (percent) 16 26 18 Highest annual increase among cities (percent) 136 130 Source. Asian Development Bank. 64 WORLD BANK WORKING PAPER (U.S. cents per cubic meter) 3 0 ff- - - - - - - - - . . .. 20 Note: "i" Indicates that there are World Bank-financed water supply projects in the city. Source: World Bank data. Private Sector Participation Private investment in water and sanitation has lagged behind that in other infrastructure sectors. During 1990-2000 private investment totaled about $700 million, represcnting just 4 percent of total investment in water and sanitation infrastructLre (Figure 5-5). Moreover, such investment has been concentrated in ivater supply because privatc investment in urban distribution networks was prohibited and investment in wastewater has been negligible. Even in water supply, private mivest- ment has fallen as a share of total (public and pnvate) funding-from 7.5 percent in 1997 to less than 3.0 percent in 2002. A detailed list of water projects involving foreign direct investment is provided in Annex 2. Only 25 projects had reached closure by 2000, and most were BOT schemes for potable water treatment (Table 5-3). Individual investment rarely exceeds the $30 million threshold at which State Develop- ment Planning Commission approval is required. This small number of projects contrasts with the huge need for capital investment in the sector. Major international water companies (mostly French and U.K. utilities) an-d Hong Kong developcrs have tried to gain footholds in China's water market. FIGURE 5-5: INVESTMENTS IN WATER AND SANITATION INFRASTRUCTURE BY SOURCE, 1992-98 Private Instiatlvos mostly in Water Supply Public 3 8 % Investment in Wast0 Water 38 0% Public 58 2 % Source: SOGREAH data. PRIVATE PARTICIPATION IN INFRASTRUCTURE IN CHINA 65 Number of Capacity (cubic Foreign Investment Foreign Developer Transactions kilometers a day) (millions of U.S. dollars) Sino-French (a) 9 3,260 III Cathay International 5 1,890 197 Vivendi Water 3 1,150 141 China Water 4 450 32 Thames Water International 1 400 73 Saur International 1 225 15 Cheung Kong Infrastructure 1 400 9 Giantmost 1 250 8 Total 25 8,025 586 Notes: a. A joint venture between Suez Lyonnaise des Eaux and the Hong Kong-based New World Infrastructure Source. SOGREAH data. But because of the legal, security, and contract issues noted above, these sponsors have manlaged to obtain bank lending on a limited recourse basis for only a few projects, and have to seek financing for most other projects through their corporate balance sheets-an unsustainable approach that has weakened their interest in financuig large-scale water projects in China (see Chapter 2) Models for Private Participation As noted, most private ventures in the water sector are BOT schemes for water supply sponsored by international developers. Wholly Chinese-owned ventures are the exception (Shenyang Water Company). But there have been some other experiments, notably witlh a management contract (combined with an initial public offering), a BOT scheme for wastewater treatment (Xuzhou), and a water concession (Tanzhou, in addition to the concession in Macau awvarded under Portuguese tenure) Mana,gement contract. The only known signed management contract is the Shenyang Water Supply Project. This was initially a 30-year BRtOT (buLild-rehabilitate-operate-tran-sfer) scheme starting in 1995 through a joint venture between Sino-French and the Shenyang General Water Company. In late 1999 the Shenyang Public Utility Holding Company was listed on the Hong Kong stock exchange. It purchased the shares in the water plant from Sino-French and signed a management contract witlh Sino-French. BOT contracts Most of the 25 private wcater projects listed in Annex 2 and Table 5-3 are BOT schemes for the construction of water supply facilities (treatimnent plants plus sometimes intakes and transmission lines) I Most fall under the $30 million threshold for State Development Planning Commission approval, with the exceptions of Shanghai Dachang ($73 million), Chengdu ($107 million), and Jinan ($90 million) These projects can be broadly divided into three categories First are negotiated cooperative joint venture BOTschemes, which do not involve competitive bidding. They are carried out by cooperative joint ventures controlled by provincial or muniicipal authorities, with limited support from the central government. Most water supply BOT schemes in China fall into this category I Several projects arc actually ROT(rehablhtation-opcration-transfer) schemes. In this variant of a BOT scheme, most investmcrits are in rehabilitating existing assets instead of buildng new ones. 66 WORLD BANK WORKING PAPER (Box 5-1). Most of these joint ventures provide the private partner with either a guaranlteed return (which was prohibited by the 1995 BOT Circular) or priority access to dividends (Table 5-4). Although this approach might have been justified for earlier projects due to the poor creditworthi- ness of off-takers and regulatory uncertainty, these projects have not been able to tap international financial markets and so cannot be promoted as a sustainable model. The second category is negotiated wholly privately owned BOTschemes. Here the local govern- ment acts as granting authority and regulator but does not take equity stake in the project com- pany, to avoid creating a conflict of interest by acting as both owner and regulator. There is only one such BOT scheme, Shanghai Dachang (Box 5-2), which was successful in accessing inter- national financial markets on a limited recourse basis. The third category is official or tender BOTschemes. They involved competitive bidding, substan- tial central government involvement (such as a state guarantee on currency convertibility and transfer), and off-taker credit support provided by provincial or municipal governments. Only two projects have been awarded: Chengdu Water Supply (Box 5-3), the pilot BOT project and Beijing Water #10. Whether this latter project fell under the scope of the BOT Circular is unclear. Together Shanghai Dachang and Chengdu Water secured almost 30 percent of the private investment in China's water sector to date. Concession contract. Concession contracts were permitted prior to the issue of the 1995 Foreign Investment Catalogue (see Chapter 2), which prohibited foreign companies from investing in or managing water distribution networks. Today there is only one concession contract in mainland China, as well as one in Macau granted during Portuguese tenure (Box 5-4). - Recommendations Water and sanitation services have some unique features that have traditionally called for higher government involvement than in other infrastructure sectors. First, the services have features of a natural monopoly, which limits the scope for competition, especially given the high cost of trans- portation and distribution (about half the cost of potable water). Second, water and sanitation ser- vices have considerable health and environmental externalities and are an essential public service that often raises serious social issues. Finally, water and sanitation are usually the responsibility of local governments, which tends to increase their political visibility and exposure to interference in service provision. With proper regulations and project structures, however, governments in most industrial and many developing countries have attracted private investment to this sector, thereby sharing with the private sector the responsibility for providing such "public services." * 0 The 30-year Harbin water supply BOT contract was negotiated and signed between SAUR International and the Harbin Municipal Water Company in 1992 without a competitive tender, under a joint venture arrange- ment. The total project investment was 165 million renminbi ($30 million), financed with an equity injection from the sponsor and an in-kind contribution (land use rights) from the local government Several lessons can be drawn from this project, one of China's first water BOT schemes. First, the profit-sharing mechanism was extremely unbalanced in favor of the private partner. Second, the project was financed entirely through an equity iniection by the sponsor and was not able to secure any debt financing on limited recourse basis. Third, the introduction of operational best practices cut operating costs while maintaining a high standard of treated water. These improvements permitted the plants to operate continuously, while city-run plants using the same resource suffered from intermittent shutdowns. Although the city of Harbin benefited from the project in term of access to private funding and foreign technology, the cooperative joint venture model adopted in this early deal appears unbalanced and did not allow full access to international private markets. Source: Project team. PRIVATE PARTICIPATION IN INFRASTRUCTURE IN CHINA 67 TABLE 5-4: THE PROFIT SHARING SCHEME IN A NEGOTIATED COOPERATIVE JOINT VENTURE BOT SCHEME IN HARBIN Period Profit Sharing Years 3-10 10 million renminbi a year for Saur International; the balance for Harbin Water Company Years 1 1-20 10 million renminbi a year for Harbin Water Company; the balance for Saur Years 21-30 Saur and Harbin Water Company split the profits evenly Source: Project team. The development of water and sanitation infrastructure has been declared a national priority by the Chinese government, and the enormous investment needs require broadening the sources of funding to include private foreign and domestic investors. To make the sector financially self- sustainable, the effcicency of existing infrastructure needs to be improved through modern man- agement practices and teclnology transfers. Thus a broader range of private participation models should be tested, including privatizmg small local water utilities to domestic firms, operations and maintenance contracts, BOT schemes modeled after the Chengdu project, and full concessions. Despite the enormous size of China's potential market, private participation in water and samtation faces four major challenges: Box 5-2: CHINA'S EXPERIENCE WITH A NEGOTIATED WHOLLY PRIVATELY OWNED BOT SCHEME The Shanghai Dachang Water Treatment Prolect is a 20-year BOT scheme that treats 400,000 cubic meters a day in Bao Shan District, Shanghai. It was negotiated and signed in 1996. Cosponsored by Bovis Limited and Thames Water Overseas, the prolect was the first wholly foreign-owned water project in China, avoiding the common joint venture model between sponsors and local governments. It was also the first water project in China to be financed on a limited recourse basis. Total investment was $73 million, with 70 percent in debt raised from international financial institutions without export credit agency support. Through a concession contract, the Shanghai municipality provided the investors with timely access to sufficient foreign exchange for debt service and repatriation of profits, authorized the use of concession rights as security to lenders, and gave an irrevocable off-taker payment guarantee backed by a wholly owned investment arm of the municipality. This project broke new grounds in a number of aspects: O It was the first water project in China to attract offshore limited recourse financing, and this was done without political risk coverage from multilateral or bilateral agencies. ci Despite the fact that the project was not approved at the central government level, and revenue was denominated entirely in local currency, lenders took comfort in the economic profile of Shanghai and the substantial support provided by the municipal government-underscoring the importance of the off-taker's creditworthiness for such projects. O The project showed that a well-structured, wholly privately owned prolect with a strong international sponsor could be an alternative to joint ventures for investment in the water sector. On the other hand, the prolect was structured with a guaranteed fixed return to the equity investment, limiting the potential downside to the sponsor and the incentive to achieve operating efficiencies. Given the project's size, the lack of competitive bidding probably resulted in higher than necessary costs (especially for the turnkey construction contract). The bulk tariff that the project company charged the local water company was higher than the retail water tariff for end consumers Source Project team. 68 WORLD BANK WORKING PAPER The Chengdu Water Supply Project is a greenfield 18-year BOT scheme for the construction and operation of a water treatment plant (400,000 cubic meters a day), two water intakes and a raw water transmission line, and a 27-kilometer transmission pipeline from the plant to the Chengdu distribution network. Cosponsored by Vivendi Utilities and the Japanese group Marubeni, it was the first water BOT project in China developed according to the guidelines of the BOT Circular. Total investment was $106.5 million, making it the biggest private investment ever in China's water sector, funded with 70 percent debt and 30 percent equity. The $74.5 million in debt was raised through limited recourse financing on international markets with the help of the Asian Development Bank (which provided a direct loan for $26.5 million). The project was structured around a concession contract that was signed in 1999. Construction is under way. Among the key aspects of the project: * It was the first water project in China awarded through international competitive bidding. Considerable competition took place among top-notch international investors during the tender. *. According to the BOT Circular, the central government will guarantee currency convertibility and transfer- ability. Nevertheless, since the city of Chengdu has no authority to commit the country's foreign exchange reserves and the BOT Circular was never formally ratified, the legal enforceability of this guarantee is unclear. * Project revenues were backed by a 400,000 cubic meter a day take-or-pay guarantee from the municipal water company. The bulk water tariff paid to the project by the Chengdu water company will be 0.98 yuan per cubic meter (subject to foreign exchange rate adjustments)-higher than the water tariff charged to residential users, which is 0.65 yuan per cubic meter. * The Chengdu municipal government provided backup support to the municipal water company's obligations under the offtake agreement, and assumed direct obligations under the concession agreement, including for termination payment. In addition, two specialized funds dedicated to the water sector were funded from consumer bills. Despite the uncertainty about the project's guarantee framework and the lack of information about the creditworthiness of the off-taker, the deal generated considerable interest during the tender process from both investors and lenders. The project's success showed that international capital markets can take a greater interest in China's water sector if some key international best practices are followed, especially conducting transparent international bidding and clearly separating operational and regulatory functions by avoiding the joint venture model, which created conflicts of interest for local governments. Source. Project team. * A tradition of low tariffs, below cost recovery levels, which affects the sector's financial sustainability and impede government goals of protecting the environment and conserving water resources. * A complex institutional framework, which prevents private developers, lenders, and users from assessing the performance and financial viability of water and sanitation systems. * A lack of proper regulation that clearly separates service provision and regulation. * A prohibition on private participation in the urban network, including full water and sanita- tion concessions, which limits the benefits from private participation for the government and the public. Since 1995, China has made considerable efforts to facilitate private participation in the water sec- tor, modernizing its complex legal framework and promoting reform of water tariffs. Given the extensive involvement of local governments and the complex social issues involved, promoting pri- vate participation in water is not easy, and it is understandable that progress has been gradual. PRIVATE PARTICIPATION IN INFRASTRUCTURE IN CHINA 69 The Tanzhou (Guangdong) concession was signed in 1992 through a joint venture with Sino-French. The prolect company is responsible for operating, maintaining, rehabilitating, and upgrading the distribution system, Including billing customers and constructing a new water treatment plant (60,000 cubic meters a day). The Macau concession was signed in 1985 under Portuguese tenure. The water system had previously been operated under private local management since 1936, without proper regulation by the government, and the service provided to the population was bad, with poor water quality and frequent disruptions The concession contract, signed with a consortium of Sino-French and the former private owner, was modeled on Western infrastructure concession contracts, with clear allocation of responsibilities between parties, strong performance obligations for the private operator, and a solid regulatory framework with the creation of a water regulator and the inclusion in the contract of key clauses for tariff adjustments, investment plan reviews, early termination, and dispute resolution. Public information suggests that the Macau concession achieved significant financial and operational results. The quality of drinking water was raised to EU standards and water soon became available on a 24-hour basis for all consumers. The sponsor was able to finance significant rehabilitation of the distribution network (70 percent of pipes have been replaced), and unaccounted-for water has fallen from 40 percent to 17 percent. More efficient operations helped reduce the need to invest in new production capacity despite the fact that demand has tripled since the contract was signed. The water company is now profitable, generating significant financial receipts to the Macau government in term of taxes and concession fees. Overall, the Macau experience demonstrates the benefits that can be gained by delegating to an expenenced inter- national operator, with a well-designed regulatory framework, the responsibility for financing and managing a complete water system Source. Project team. Although countries like Argentina, Chile, and the Philippines have had muich more success attract- ing private investors to the water sector, Chuna's situation is not uncommon. Several other big countrics (such as Brazil and Mexico) with similarly complex institutional frameworks and tradi- tions of low water tariffs have also seen limited private participation in the water sector. Our recommendations fall into two groups: El At the level of the municipal government: accelerate reforms, including implementing tariff adjustments, corporatizing water and sewerage services, and promoting private participation to separate regulation from service provision. O At the level of the centralgovernment: clarify the roles and functions of the Ministry of Construction and Ministry of Water Resources on water and sanitation activities; help local governments restructure water and sewerage services, including establishing systems for sharing information and bcnchmarking various water and sanitation companies; and imple- ment pilot concessions of complete urban water and sanitation systems Accelerate the Reform of Public Water Companies The central government has taken the right steps in introducing legislation for water tariffs. Despite the obvious need throughout China to raise water tariffs to full cost recovery levels, doing so is no easy task given the poor quality of water provided by most municipal water companies and the fact that tanffs are enforced by local governments International experience has shown that consumers are usually reluctant to pay higher tariffs when service is of poor quality and there is little trust in the management capacity of the public wvater utility Thus the question is not whether to increase tariffs, but how to do so in a way that receives sufficient support from the public. 70 WORLD BANK WORKING PAPER Thus the main recommendation for public water companies is to pvrsi!e hibge; water tariffs in combination wtth tr;ue corporatization of water co;;zpanies. This approachl would ensure that tariff increases are supported by a more business-like approach to the provision of water services, with: Better investment planning committed to specific improvements in service performance. Independent man-agement and professionally audited financial statements (including suffi- cient and accurate data on fixed assets and depreciation). i Accountable quality performance and easy monitoring of the periormancc data by the public. L More transparent and targeted subsidies, if needed, directed at the poorest part of the population. Through the corporatization process, small systems in different municipalities can be consolidated and less efficient players can be taken over by others. Small projects such as the Harbin water proj- ect can also be put into porifolios, with dcbt financing being introduced at che holding company level. On the other hand, large systems can be split for compctition-as in Shanghai, where the Shanghai Water Company was split into four companies. The fact that local govermments are responsible for providing water and sanitation services reduces the central government's ability to accelerate sector reform. This largcly explains why, although the state has initiated some major changes in legislation (as wvith wvatcr and samtation tar- iffs), local enforcement has sometimes not been as desired. Local governmcnts need clearer policy guwdance and technical support from the central government on what is and how to develop the "right" model for private participation in water. To strengthen local capacity, a technical support center at the statc level could help municipalities develop a database of best practices from water privatizations worldwide. The ccnter should also help municipalities develop a benchmarkuig system to report and compare utility costs and performance. These can then be published to allow consumers to compare services and tariffs. This approach, used successfully in the United Kingdom and Victoria Australia, and being tested in Manila (the Philippines), could make utllities (public or private) much more responsive to consumer needs. The government could also consider developing a utility rating system, as is bemg tested in Indonesia. Develop Pilot Concessions of Complete Water and Sanitation Systems There appear to be significant opportunities for optimizing the operations and maintenance of the sanitation system (sewer network and wastewater treatment plant), including improving compli- ance by wastewater plants with discharge standards, reducing leakages from deteriorated sewer net- works, and promoting pre-treatment of industrial wastcwater by small industrics. In the water sector the most significant savings usually result from the cfficicnt managemcnt and maintenance of the distribution network, not water production or treatrrcnt plants. This is because efficient management of the distribution nctwork can sharply reducc ihc levcl of nonrcv- enue water and so can delay the need to cxpand water production capacity, anrd bccause commer- cial management practices (billing and collection) usually increase revenues, gcnerating additional resources for infrastructure investment. In addition, proper management of the sewerage netvork by experienced pnvate operators can facilitate the operations of recciving wvastcwater trcatment plants, increasing compliance with environmental discharge standards. In China most private investment has been through BOT schemes for water treatment plants. Such schemes can solve production shortfalls but are ill suited to the poor creditworthiness and limited financial transparency of China's municipalitics. Moreover, BOT schcmes do not address fundamental operational deficiencies in a water utility-such as inadequate capacity of pipelines, high unaccounted-for water, and mstitutional or labor problems-and so are not capable of trans- forming financially wealk utilities into sound ones. Without careful planmng, a .3 OT contract may actually delay much-needed systemwide improvements. PRIVATE PARTICIPATION IN INFRASTRUCTURE IN CHINA 71 The announcement that the government is considering allowing private investors to build water dcstrnbution system (South China Morning Post, 23 August 2001) is welcome news. How- ever, until the private sector can operate and manage the network, including metering and billing consumption and collecting tariffs from users, significant benefits are unlikely. Conccssions of integrated water production and distribution systems are a successful model for private participation m infrastructure, one widely used in other countrics (Box 5-5). Given the enormous need for private investment in the sector, developing water and sanitation concessions for major cities should be a top priority. For example, the water and sanitation concession imple- mented in Buenos Aires, Argentina, allowed the country to tap more than $1 billion over the first five years of the contract and dramatically improve service quality-while passing over all financial, operational, and commercial risks to the private sector. Implementing similar arrangements for China's biggest cities would make a major contribution to the objectives of the Tenth Five-Year Plan (which foresees $26 billion in investment in water and sanitation infrastructure). Despite the potential benefits, concessioning the fuil operations of an urban water and sanitation system is a process frill of pitfalls. l The government's legitimate concern for continuous control over service provision, because of its social and environmental implications, requires a careful structuring of the concession's regulatory and contractual framework (with performance targets, tariff revi- sions, investment plan approval, and subsidy schemes). Box 5-5: INTERNATIONAL EXPERIENCES WITH CONCESSIONS OF WATER AND SANITATION SYSTEMS In many countries where concessions have been granted for complete water and sanitation systems (production and distribution networks, sewerage network, wastewater treatment), private participation has resulted in considerable benefits for the government and the public-especially when contracts were awarded competi- tively, with clear performance targets and enough flexibility to the private operator to raise financing and adopt innovative technical solutions. In Buenos Aires, Argentina, in the early 1 990s, water service was suffering from deteriorated infrastruc- ture and low operational efficiency. In 1993 a concession was awarded to a consortium of international water companies and local investors. The winning consortium, Aguas Argentina (headed by Lyonnaise), offered the largest tariff reduction-26.9 percent. The concession led to major gains for the government and the public, resulting in. O Unprecedented private funding, with more than $1 billion invested over the first five years. (The International Finance Corporation arranged limited recourse financing for $392 million.) C Significant new connections, with service coverage rising from 70 percent to 83 percent of the population. O Efficiency improvements, with unaccounted-for water dropping from 44 percent to 34 percent in the first five years. o Overall improvements in service quality and user satisfaction. In Manila, the Philippines, two concessions were awarded in 1997 (for the two halves of the city) to two consortiums of international water companies and local investors. The International Finance Corporation was the lead financial adviser to the government for this transaction, and proposed the division of the city into two concessions to address concerns about excessive monopoly power if only one consortium were to operate the entire system. Competitive bidding resulted in a 43 drop in the water tariff in the western zone and a 74 percent drop in the eastern zone During the first year of operations, unaccounted-for water was cut from 60 percent to 36 percent, while more than 200,000 additional connections were added to the network. Over the 25-year contract period $7 billion in private investment is expected. Source International Finance Corporation; MottMacDonald, Economie et Humanisme (2001). 72 WORLD BANK WORKING PAPER * The monopolistic nature of the service tends to limit competition, making competition for the market-that is, awarding the concession through competitive bidding-essential if acceptable financial conditions are to be obtained for the government and the public. * There are relatively few international private players in the water sector, so the transparency and credibility of the tender process are essential to ensure proper competition for the market. If well done, the development of a water anid sanitation concession for a major city could easily take 18-24 months until the contract is awarded. Thus it is recommended that, given the govern- ment's decision to explore private participation in the urban network, a ptlot water and sanitation concession be undertaken in a major Chinese city. This project would provide an; opportunity to ana- lyze all the issues of concern to the government and the public (tariff levels, regulation and control, social issues), allowing the design of innovative Chinese solutions adapted to government objec- tives, the socialist market economy, and social and institutional realities. The success of such a pilot project would have a strong demonstration effect for local governments throughout China, an-d its contractual and regulatory design could be used as a blueprint for other concessions. Promote a Broad Range of Private Participation Models in Water In addition to concession schemes, other models for private participation could be investigated in China to improve operational efficiency, including management contracts and leasing contracts. Although these models tend to rely heavily on public funding to finance infrastructure investments, they might be of interest for smaller systems with minimal capital investment needs, or in cases where political resistance to private participation prevents implementation of a concession scheme in the short term. Each of these models has benefits and drawbacks that must be kept in mind in the context of the water sector: * Management contract-because these contracts entail a limited transfer of risks to the pri- vate sector, they often bring limited benefits. It is difficult to structure a contract in which private investors have strong incentives to increase efficiency when they do not have signifi- cant money at risk. In addition, monitoring the performance of a private operator can be tricky when there is a lack of reliable historical data. * Leasing contracts-under this model the private sector is supposed to assume all operational and commercial risks, but assigning responsibility for service performance can be difficult in practice. For instance, it can be difficult to hold private operators liable for poor services if they have no control over mvestment decisions. Leasing contracts also tend to create incentives for private operators to favor rehabilitation of assets (to be financed by the public counterpart) instead of maintenance (for which the operator is liable), which can be economically inefficient. Another possible way to promote private participation in water would be to foster the development of private domestic operators. Recent international experience shows that it is possible to promote the creation of domestic water operators, avoiding the sometimes politically costly process of trans- ferring the operations of urban water services to foreign companies: O In Italy the public water company in Rome was corporatized and gradually floated on the stock market. It is now privately owned, with no foreign control, and the municipality maintains only a minority shareholding. The new private company, ACEA, is even pursuing international expansion. * In several Latin American countries (Argentina, Colombia) domestic water operators have been created to operate leasing and concession contracts in medium-size cities. 6 POWER GENERATION ISSUES AND RECOMMENDATIONS C hina is reforming its power sector with a view toward developing a competitive market. Reforms will occur in three phases. The first entails separating generation from transmission and distribution, and splitting generation assets into a number of market players (with lim- ited or no cross-ownership to minimize or eliminate market power). During this phase generators will bid for dispatch into a malidatory energy pool, from wvhich a single buyer-whichl is also the monopoly provider of transmission and distribution services in the market-purchases electricity. A common market-clearing price will be based on the bid from the last generator dispatched. In the second phase competition will be introduced in wholesale markets (by allowving direct contracts with distributors and large customers), and in the tlird phase in retail markets. Reforms are at an early stage, with a number of pilot schemes being tested The World Bank recently issued a report titled Fostering Competttton in China's Pow1er Markets and is actively advis- ing the government on the policy framework. Given these ongoing efforts, this report comments only on the power generation subsector. It does not cover the general framework for sector liberal- ization, including detailed designs for the market struLcture, pnvatization of the distribution net- work, and the like. The Situation China's power sector has achieved impressive size and growth, but it has problems similar to those in the road and water sectors-partcularly thc complex and inefficient institutional structure and lack of long-term commitment to tariff adjustnents. Capacity Planning and Development China's power generation capacity exceeds 300 gigawatts, making it the world's second largest system. Over the past two decades electricity demand grew by nearly 9 percent a year (except in 1997 and 1998, when annual growth slipped below 5 percent due to the East Asian financial crisis) The Tenth Five-Year Plan estimates that demand will grow by at least 5 percent a year 73 74 WORLD BANK WORKING PAPER through 2005. Assuming annual growth of 5-9 percent, China can expect to add more than 20 gigawatts of generation capacity each year through 2010 (Figure 6-1). At a capital cost of $800 a kilowatt (for a typical coal-fired power station), the projccted growth implies that investment in power generation will be at least $16 billion a ycar through 2010. The government hopes that foreign and domestic private investors will provide 25 -30 percent of this fumding, or about $50 billion. Institutional Structure China's power sector has been undergoing institutional and organizational reforms for several years. To separate policy and regulation from ownership and management, icn 1998 the Ministry of Electric Power was disbanded and its functions were distributed among the State Power Corpora- tion, State Development Planning Commission, anid State Economic and Trade Commission The State Development Plamnning Commission is primarily responsible for policy issues, investment planning, and pricing approvals. Both the State Development Planning Commission and State Economic and Trade Commission have subordinatc departments at the provincial level. Management of the sector was reorganized into national, regional, and provincial levels. The State Power Corporation owns the national transmission and distribution systems as well as nearlv half of generation capacity (which generates nearly 70 percent of the country's power). Seven regional power groups operate interlinked regional transmission systems and some generation, but do not operate any distribution networks. Finally, 32 provincial power companies manage provincial transmission and distribution systems on behalf of the State Power Corporation, as well as significant generation assets. At the county level, China used to have thousands of decentralized power companies with small coal or micro-hydropower generation plants.i Many of these decentrali/ed plants are con- nected to the centralized grid. To increase energy efficiency, develop new technologies and indige- nous manufacturing capacity, and protect the environment, the Tenth Five-Year Plan promotes the renovation and construction of large coal mines (300-600 megawatts) using clean coal technology and the development of hydroelectric power and some nuclear power, with an view to phasing out these old, small, inefficient power plants.2 The State Power Corporation wlfl ultimately divest its ownership in generation assets according to a plan recently announced by the government that encompass the creation of five competing power groups while the State Power Company would also surrender its grid assets in provinces in southwest China. Initially the gnd company will likely control transmission and distribution assets. Eventually, however, transmission and distribution assets will need to be separated, with the State Power Corporation and the other grid company retaining ownership of transmission. Pilot programs for separating generation from transmission are being carried out in Zhcjiang, Shanghai, Shandong, Heilongjiang, Liaoning, and Jilin. Zhejiang also leads tihe way in a pilot wholesale generation market program. The Private Sector's Role China permits private investment in power generation but not in transmission and distribution, though major investment is needed to expand the transmission and distribution networks. Pri- vate participation in generation rose in the 1990s until 1998, when it collapsed due to the drying up of financing and slowdown in demand growth in the wake of the East Asian financial crisis (Figure 6-2). 1. The county is the fourth administrativc levcl in China, below the state, region, and province. There are about 2,000 counties Each province has its own rural clcctricity bureau, which plans and manages rural electrification. At the county level there is usually a county electric powcr bureau, xx hich is a branch of the provincial body. 2. At the same time, other countries havc Iarncd that retrofitting old powcr stations Xwith new technology is an attractive way of improving thclr efficicncy and stimulating competition, becausc it directly countcrs the market power introduced by large power plants PRIVATE PARTICIPATION IN INFRASTRUCTURE IN CHINA 75 FIGURE 6- 1: PROJECTED POWER GENERATION CAPACITY UNDER Two SCENARIOS FOR DEMAND GROWTH, 2000-10 800 700 -GW o [5%pal 2~~~~~~E 9% pal1 600- 500- 400- 300-- 0L Source Economic Consulting Associates data. Until 2000, 95 generation projects had involved private participation an-d foreign direct invest- ment (see Annex 3). Forty-five of these projects involve plants generating more than 100 megawatts, with total capacity of about 36,000 megawatts-accounting for 12 percent of China's installed generation capacity. Although together the 95 projects provide capacity of more than 40,000 megavatts, most (wAith about 35,000 megawatts, or 68 percent of total capacity of the 95 projects) are concentrated in a small number of provinces, including Guangdong (19 percent), Shandong (13 percent), Zhejianlg (11 percent), Fu)ian (6 percent), Aihwu (4 percent), Jiangsu (4 per- cent), Henan (4 percent), Hubei (3 percent), Shannxi (3 percent), and Gansu (1 percent). (millions of U.S. dollars) 5000 2 4500- 4000- E3500- 0 3000- a 2500- 0 2000 - E 1500 1000 190 1990 1992 1993 1994 1995 1996 1997 1998 1999 Source: Economic Consulting Associates data. 76 WORLD BANK WORKING PAPER The decentralized power companies also have sigmficanlt private involvc,,-cnt. Up to 200 of these small plants may feature private investment, or about 3 pcrcect of total installed gcneration capacity.3 The decentralized plants cnablc the competitive rctailer to hcdgc against prices in the wholesale market and so play an important role ui market reform. Whcn the market is operating in all three phases of reform, the wholesale pricc will become the maun driver to stimulate reform of the decentralized power companies. Foreign direct investment in a plant wvith foreign investment typically averages 51 perccnt Although 39 foreign investors have been idcntificd in the 95 projects, nearly 40 pcrcent of foreign investment came from Hong Kong-based sponsors, followcd by 26 pcrcent i-om the United States, 19 percent from Europe, and 15 percent from other Asian sources. MSost dcbt financing came from commercial banks and export crcdit agencics. Models for Private Participation Most of China's power projects involving private part-icipation have bccn "nonconccssional" (see Chapter 1) and carried out through joint ventures bctwvcen foreign Investors and poxvcr companies owned by local governments. Reccntly, somc of the largest projccts have bcn implemented using the BOT model, with the project company being cither a joint venturc or w'nolly forcign-owncd enterprise. A few state-owncd power projects have also beeni partially privatiz.ed through initial public offerings (IPOs), divestitures, and TOTs [transfcr, operate, transfer] (Figurc 6-3). Joint ventires have accounted for about 61 pcrccnt of the amount of private investments in power and about 75 percent of the numbcr. Joint ventures allow Chinese sponsors to select foreign partners while enabling the governmcnt to rctain significant control over the projcct. Joint vcn- tures are also popular among foreign investors because of China's uncertain legal and regulatory framework. Investors expect their partncrs to help build conscnsus on projects among local and state organizations, reducmig legal and regulatory risks. The local partncr is often secn as a safec- guard against unfavorable price rcviews, dispatch risk, or both. For example, although most plants are contracted with a minimum dispatch, when the .otal system load is lower than cxpected (as has been occumng recently), local power companics tend to dispatch thcir plants first or at a highcr rate. Thus when an investor selects a local partner, significant considcration is given to whether the candidate is a power company or other local authority. BOTschemzes, including those implemented through joint vcntures or w1holly foreign enter- prises, have accounted for 24 percent of privatc powcr invcstments bv value but only 14 percent by number. There are several reasons for the limited number of projccts in this category: l The development of a comprehensive legal framcvork for BOT projects has been sub- stantially delayed, and different proccdures havc bcen uscd in the few R3OT projects that have been approved, resulting in mixed outcomcs. LI BOT regulations are pnmarily sccn as complcmcnting existing lawvs rather than superseding them. As a result the approval process for BOT projects is not nccessaily perceived as con- tributing to lower nsk relative to othcr forms of pnvate participation. The BOT tendering process has not resulted in cheaper projects because the cost of capital is higher wvithout government guarantccs and the shorter duration of contracts has rcsultcd in higher front-end costs. Becausc the BOT Circular precludes renminibi financing, foreign exchange risk has not been sufficiently mitigatcd. i BOT projects, with the exception of the successful Laibin B, are not cxpccted to rcceive sig- nificant attention and governmcnt support at all levels. 'Ncgotiatcd joint venture uidcpcn- dent power projects are still preferred for the rcasons notcd above. 3 Small projects are those Lnder 100 megawatts that do not require central government approval. They are mostly implemented by local organizations. Staustical data on pnvatc participation in these proiects are not as reliable as for larger projects. PRIVATE PARTICIPATION IN INFRASTRUCTURE IN CHINA 77 80% - -~- 70% - Number % 60% Volume % 40% 30% 20% JV BOT/ IPO TOT W PCE Source Economic Consulting Associates data. E4 BOT projects require sufficient pre-bid preparation and competitive tendering, which are not yet widely accepted by Chinese partics, who are used to preparing projects with assis- tance from their joint venture partners, then finalizinig contracts based on negotiations. Major aspects of three power projects witlh private participation are summarized in Table 6-1. The three projects involve issues common to power projects in China, including: Q Requirements for support letters from various levels of government due to the lack of a comprehensive legal framcwork for private participation in infrastructure. El Heavy reliance on international debt financing, which increases the projects' foreign exchange risks and financing costs. O State Admimstration of Foreign Exchange restrctions on the frequency of allocations to debt service accounts, which add to project nsks and the time required for project negotiations. O High levels of debt gearing (above 70 percent), particularly as China moves toward com- petitive dispatch. In addition, like most power projects in Chlina, as traditional joint ventures the Shandong Zhonghua and Hebei Hanfeng project companies faced potential conflicts of interests-an issue raised in many chapters of this report (see, for example, Chapter 1), This issue is a particular prob- lem in China becaLise contract enforcement remains weak and subject to political Interference. In addition, these projects suffer from weak lender security over reinsurajnce proceeds, government predeterminations on the sourcing of equipment, engineering, and construction (that is, the bal- ance between foreign and local suppliers); and the relatively poor financial disclosure of the elec- tricity purchaser On the other hand, the Laibin B project, China's first wholly foreign-owned and -financed major power plant, differs from other traditional power projects and has several positive features that could be used as tlhe framework is developed for private participation in infrastructure. First, it Involved a transparent, open tender. Second, its project documents and contract provisions can be used to develop national guidelines on key issues such as dispute resolution, termination, and force 78 WORLD BANK WORKING PAPER TABLE 6-1: SELECTED POWER PROJECTS USING DIFFERENT MODELS FOR PRIVATE PARTICIPATION Shangdong Zhonghua Hebei Hanfeng Laibin B Private Participation Joint venture Joint venture/"BOT" Foreign-owned Model enterprise/BOT Ownership Shandong Electric: 36.6% North China Power: 30% EdF International: EdF: 19.6% Hebei Construction and 60% China Light & Power Investment: 20% Alstom: 40% Hid. China Energy Hebei Elec. Power: 10% Investment Co.: 29.4% Siemens Huanenga: 40% Shandong ITIC: 14.4% Planned Capacity 5x600 megawatts for 2x660 megawatts 2x360 megawatts four plants Capital Cost $2.2 billion $1 billion $616 million Share of Debt Financing 70% 75% 84% Procurement Negotiation Negotiation Competitive tender Tariff Structure One-part cost-plus Cost-plus Bid on tariff (struc- (including basic tariffb, ture closer to a operation fee, fuel costs, two-part tariff) and taxes, funds, and adjustments). Note: a. Siemens Power Development Hanfeng GmbH, a subsidiary of Siemens. In 1998 it sold a 16 percent stake in the project to Hamburgische Electricitaets-Werke (HEW), leaving itself with a 24 percent stake. b. The basic tariff is essentially a capacity charge and compromises the most vital components to investors. The basic tariff is derived from depreciation costs, principal repayments, financing fees and interest, expected target profit, company overhead costs, and amortization of the company's development costs. Source: Project team. majeure. Third, the use of preissued support letters as part of the tender documents addressed certain concerns of lenders and sponsors (the letters could serve as a stop-gap measure until for- mal legislation provides more certainty on the impending industry restructuring). Fourth, the limited role of the provincial utility and government in the ownership and operation of the power plant (only through minority stakes in the operations and maintenance contractor) provides Laibin B with fewer conflicts of interest. This type of structure will support China's long-term goal of developing competitive power markets. Fifth, the contract discarded the cost-plus tariff structure in favor of a two-part tariff (with some distinctly Chinese characteristics). As will be dis- cussed below, a two-part tariff wiHl be useful as China moves its power markets toward merit-based dispatch. Sixth, the project showed that private investment and financing can be achieved (with some external support, in this case from the State Development Planning Commission) for projects in economically less developed regions. Finally, the project provided lender security over reinsur- ance proceeds. Securitizations through tnitial public offerin,gs (IPOs) have accounted for 12 percent of private power investments by value. Most of the proceeds of IPOs have been invested in existing rather than new plants, given investors' lower capacity for managing new project developments and lower appetite for construction risks. Since the first international IPO in 1994 of Huaneng on the New York Stock Exchange, three power companies-including the famous Beijing Datang- investing in and operating portfolios of generation plants have listed on both Chinese and foreign PRIVATE PARTICIPATION IN INFRASTRUCTURE IN CHINA 79 exchanges. Considerably more have listed solely on the Shanghai and Shenzen exchanges. Like IPOs in the road sector, opportunities for new IPOs in the power sector are influenced by the per- formance of companics already listcd in the sector and by general stock market conditions. Dramatic falls in share prices in 1998 and to a lesser extent in 1999 highlighted the risks of this approach. Divestztures of existing plants through trade sales and TOTs have accounted for 3 percent of private investments in power by value. Local power companies sell some or all of their holdings in existing companies, mainly to Hong Kong-registered subsidiaries of mainland companies, to finance their equity shares in new joint ventures Recommendations International expenence with competitive wholesale power markets has been mixed While some markets have seen supplies drop and prices skyrocket (as in the U S state of California), others have managcd to attract higher pnvate investment after the market reform, including Australia (Box 6-1). Throughout the reform process, it is crucial for the government to publish the overall policy on power market reform, and where possible the timetable and scope of each step in the process so that investors can assess risks before committing investment In addition, existing investors will seek reassurance from the government that their investments will be protected during the transi- tion period. A number of Chinese provinces are moving toward the first phase of power reform The strong development of generating capacity in the 1990s, coupled with the slower pace of economic growth since 1998, has providcd time for developments in these pilot reform markets to be observed .0 * ''0~~~kN'- ** _01940 S * SVAA In Australia private investors have actively responded to every government tender for the sale or lease of a government-owned power plant. Several features of power market reform explain this interest a Support for the main principles of the "gross" market design when they came under attack after the fallout from California's poorly enacted market reforms. Australia's market differs from California's in several ways. Off-takers have full rights to hedge any risks associated with their power purchases from the wholesale market. The wholesale market price cap ($5,000, planned to gradually increase to $20,000) provides strong signals to private investors in power plants. The single price, single settlement, single dispatch mechanism removes the opportunity for gaming by power plants. Full information disclosure is provided so that regulators have full, and the public some, access to market data on the day after the trading day. Regulators monitor the market weekly and make their reports public. Finally, Australia's environmental licensing requirements do not appear to be as stringent as those in California. o Active involvement in building and operating interconnector transmission lines that auction rights to inter- connector capacity on a regular basis These transmission lines have been built in parallel with regulated transmission lines and are successful where interconnector congestion occurs often. E In addition, the following areas of market design are receiving both short- and long-term attention: -Interconnector planning. -Streamlining multiple regulations. -Increasing civil penalties for generator abuse of market rebidding rules -Providing competitive markets for ancillary services. -Expanding the scope of existing laws on open access to infrastructure. -Developing the financial market to close the loop between it and the physical market. -Streamlining changes to the market code. -Developing firm access arrangements for transmission, -Reviewing the value of lost load to determine the effect of the price cap on investment in peaking plants. -Improving the relationship between the dispatch price and the settlement price. Source' PhaceLift (200 1) 80 WORLD BANK WORKING PAPER Economic disparitics among regions a-id the nccd to ensure adequatc trainingo f commercial, tech- nical, and regulatory staff suggest that thc reform should be adopted in phases and accommodate many different entry points. Morcovcr, because the reforms are likely to involvc some challcngcs and strain the appetite of private capital markets for Chinese elcctficity asscts, moving in phases may limit problems and enable lessons from early reformers to be applied to later rcformers. Thus it is possible that many of China's electricity markets will not cnter the first phase of power reform until adjustments are made to existing practices-such as improvements in market structure, adaptations of power purchasc agreements, rcstructuring of generation plants, changcs in financial management practices, and so on. As electriciry supply and dcmand converge across many of China's electricity markets, implementation of thesc adjustments durmg the transition period will take on some urgency. Strengthen the Market Structure There are two broad trading models for a compctitive power market. One is based on the "gross" exchange principle, where all encrgy is physically tradcd and settled through a common process (a pool market) and there is usually one common multilateral contract (thc market rules) for all generators. The other model is based on the "nct" cxchange principle, where physical settlemcnt can be established between a power plant and a consumcr as a private arrangement, with any of the plant's leftover (net) energy being settled on the spot market. In the net model, however, all energy is physically exchanged through a common process, as in the gross model (Figure 6-4). The pilot pool market introduced in Zhejiang is modcled on the gross market structure, which requires all existing and new power purchase agrccments to follow market rules (with terms and conditions substantially different from those of the power purchasc agreements) and all generat- ing units to be dispatched on price merit. Zhejiang's electricity market is in the first phase of reform and is moving toward the second. Although its trial operations only began in January 2000, a rccnt World Bank (2001b) review found that impressive achievements havc already becn madc. Sound trading mechanisms, operating rules, and risk management have been cstablished; information systems to operate a competitive energy pool have been well conceived, developed, and put in opcration, the rcform has been widely accepted and supported by govcrnment authorities; improvements have been made by gen- erators in cost control, internal performance managemcnt, and contract trading analysis and risk management; and, most important, averagc generation availability (92 percent m 2000) continued to improve as a result of competition. There are, however, problems that have prcvented the market from advancing further. Legal and regulatory changes are needed to intcgrate "off market" capacity (as high as 32 percent) with the pool market. Regulatory changes are also needed to integrate provincial markets with the regional market, so that generation plants in Jiangsu, for cxamplc, will be required to bid mto the pool market as well. In addition, information on the market settlement price, -which is critical for investors to make long-term capacity planning and investment dccisions, has been kept secret and is not available to investors. Long-term commitment has not been given to protect existing power purchase agreements, and vesting contracts for differences have not been long term or legally rcc- ognized (see below) 4 4 Contracts for differences commonly fcaturc a contractually agrecd strike price ,renminbi per megawatt-hour) and volume of dispatch (megawatts-hours per period of marlcct opcration) The strikc price represents the pnce at which the purchaser would likc to purchase and the sellcr would like to sell electricity over the term of the contract. If the market price is higher than the strike pnce, the gcncrator will pay an amount to the purchaser equal to the market pricc lcss the strike pricc, multiplied by the volume of electricity under the contract. Similarly, if the market pncc is lower than the stnkc price, the purchaser will pay to the generator an amount equal to the strike price Iess the market price, multiplied by the volumc of electricity under thc contract. PRIVATE PARTICIPATION IN INFRASTRUCTURE IN CHINA 81 Wholesale Retail Market % _ _ < Market (sellrD) ----------- Generator Cf o contract - ~ Distributor Consumer (seller) - .--(buyer) (off-taker) "Gross" Market Observation The "gross" market removes all use of the traditional PPA by replacing it with a multi-lateral contract and a CfD contract Wholesale Retail Market Market G + PPA Part MWh- X t ~~~M&SO ) () - (spot market) -- - - - - - -- - - - - -- - - - ------- -- - - --- - - - - W ~~Part Mwh jPAFatMh/ Distributor Consumer Generator (buyer) (seller) (off-taker) "Net" Market Observation: The "net" market makes use of. * the traditional PPA to interface with the buyer; and * a multi-lateral contract to sell surplus energy to the spot market Source: PhaceLift. 82 WORLD BANK WORKING PAPER A desired outcome of power scctor reform in Cliina-achicving a competitive markct while retaining investors' interest (both existing and future mvcstors)-can be achikved only if: K A consistent legal and regulatory fram o;eor.k is established based on the principles of effi- ciency through competition. T'his entails realigning the Electricity Layr: ead snipportinzg reqg- lations, and developing comprehensive za;*Iret rzles that provide for cnforccable multilatcral contracts, specifying the functions and responsibilities of the market operator and systcm operator. C A market governance structure is establishcd, includmng an i;bdepende;it egulato;; to ensure economic merit order dispatch, open access to powcr infrastructure by all parLies (govern- ment and private companies), and a transparent, nondiscriminatory approach to transmis- sion pricing. 7- Access to information is enhanccd. Private investment in a competitive market cannot be maintained without adequate transparency. Market rules and other legal instruments have to be publicly disclosed so that private invcstors can assess thcir role m each phase of the power reform process. Provincial powcr companies should regularly update and publish their short- and long-term powcr plans. Results of pilot market operations, including (at the least) the market price and system dcmand, should be publishcd daily. K Existing power purchase agreements are protected (scc below). K- The financial securities market is adapted to the competitive electricit) market so that hedging techniques (such as a secondary trading market for contracts for differences) can be used to manage the risk of movements in markct prices. K- The creditworthiness of market pa rticipants zs imizproved by first requiring provincial power com- panies (in their temporary role as single buyers in the first phasc) to incrcase disclosure of thcr financial information (preferably independently audited). Ultimatcly, appropriate requirements will need to be imposed on all m-arket participants to ensure acccptable crcdit quality. A divestiture program is implemented for genezrsaiou assets (see below). F-K Fuel markets are deregulated to cnsure that there is no dislocation between a deregulated power market and a regulated, inefficient fuel supply market.5 Since the transitional regulatory issues have been discussed on various occasions betveen the World Bank and the Chinese government, this papcr wvill not elaborate on morc detailed recom- mendations in this regard. Adapt Power Purchase Agreements for the Transition Period Given the long-term nature of capital investments in power generation, most of the world's competitive markets have a large portion of energy delivery hedged through long-term con- tracts-such as long-term power purchase agreemcnts, forward contracts, or contracts for differences (Table 6-2). Long-term contracts provide more certainty to powcr producers than do unhedged competi- tive multilateral contracts or spot market contracts. For China it is almost impossible, at least in the near term, that new financing for indepcndcnt power projects will be available solely on the basis of a competitive pool market without a significant part of power sales being hedged by long- term contracts. To the extent that the market for long-tcrm fornvard contracts and contracts 'or differences is nonexistent or illiquid, the government will need to provide a central default role- as a counterpart to long-term power purchase agreemcnts and vesting contracts for differences (see below)-until the power market and financial markct become more efficicnt. 5 For example, provincial power companies have often actcd as intcrmcdiaries bct-xccn fucl suppliers and independent power projects. If this structurc wcre to continuc in a deregulated powcr market, it would under- mine the benefits of such deregulation and impart undue markct power to the provincil! power companics. PRIVATE PARTICIPATION IN INFRASTRUCTURE IN CHINA 83 TABLE 6-2: SHARES OF LONG-TERM CONTRACTS AND SPOT MARKETS IN VARIOUS MARKETS (PERCENT) Share of Energy Hedged through Share of Market Long-term Contracts Unhedged Energy Market Model Argentina 80-85 15-20 Gross Australia 90a 10 Gross California (United States) 40-50 50-60 Net New England 80 20 Gross (United States) Pennsylvania 85-90 10-15 Gross (United States) United Kingdom 85-90 10-15 Net (was gross) (England and Wales) Note a. Estimate of negotiated long-term contracts for differences established between market participants. Source. World Bank data. In the net market a generator can establish power purchase agreements with end users (bilateral contracts) that have most of the terms and conditions of previous purchase agreements. But in the gross market it is more challenging for the govcrnment to protect investors' interests in existing power purchase agreements and at the same time integrate the power produced under those agree- ments into the pool market. Most independent generators have sought project financing based on such agreements signed wvith governments for a guaranteed level of dispatch and tanff. Stable cash flow is required for debt service; othervise the generator will have to refinancc its debt to reflect the different revenuc structure in a competitive market, probably at hugher rates. It is important that China maintain the sanctity of these power purchase agreements to avoid sending a negative message to capital markets and wcaken Investors' confidence in the govern- ment's willingness to abide by contract obligations. Invcstors should be offered three options: replacing power purchase agreements with spot contracts and proper compensation, replacing power purchase agreements wvith vesting contracts for differences, or partitioning power purchase agreements from the competitive powver market. Option 1: Replacing powver purchase agreements iv'th spot contracts. Undcr this option existing power purchase agreements would be renegotiated into spot contracts so that independent generators would participatc in the pool market. But under this opfion the government would have to spend a lot of time and cffort renegotiating every power purchase agreement (most likely with multiple parties, including banks) to reach an agreement on the level of compensation. Moreover, existing power purchase agrecments in China contain protcction for investors against unilateral alteration of the contract terms without agreement from the investor. In addition, the financial market, which is required to support hedging contracts for differences (explained below), is not strong enough to enablc private investors to manage their exposure to volatile pool prices. Option 2 Replacing poiver purchase agreements with vesting contracts for differences An alternative approach would be to replace existing power purchase agreements with vesting contracts for diffcrences in which the price and notional volume of the contract match the tariff and guaranteed dispatch under the purchase agreement. If accepted by the investor, the purchase agreement would be replaced by a registrationi withl the market operator (and in so doing be provided with a multi- lateral contract covering the opcration of the pool market) and a vesting contract for differences. 84 WORLD BANK WORKING PAPER For power producers in the Zhejiang market, market risks for 85 percent of sales are managed by vesting contracts for differences. There were 23 such contracts established at the start of the mar- ket, with each designed for a specific generation urnt belonging to the eight power plants registered as market participants. The single buyer was the counterparty to one side of each contract. Although the fixed price of each contract for differences was designcd to provide a financial return comparable to historical levels, the duration was set only for 2001-much shorter than a typical power purchase agreement, and so still creating considerable uncertainty for suppliers. Moreover, the contracts for differences in the Zhejiang market have not been fully recognized as legal contracts and are still sub- ject to monthly alterations for maintenance, transmission constraints, or other special events. While a properly enforced vesting contract for differences is an improvement over option 1, power plant owners still take full responsibility for trading in the market. The vesting contract would only restore owners' financial position to the position enjoyed under the power purchase agreement. Given that owners are likely to take more risk under contracts for differences than under power purchase agreements, they would ask for additional compensation (though substan- tially less than under option 1). If the market design provides new opportunities (and removes old grievances), power plant owners should find this option attractive. But it will be a considerable challenge for the government to develop a market reform policy that convinces power plant owners to willingly trade their power purchase agreements for contracts for differences. Replacing power purchase agreements with vesting contracts for differences could also involve multiple parties, including banks. In addition, it would require a party (as buyer) to be nominated to execute the contract with the indcpendent generator. If the payment stream is expected to be one-way over the life of the contract (for example, if the tariff set in the power purchase agreement was far above the expected pool market price), a mechanism would need to be established to com- pensate such a party. If not managed properly, contracts for differences and other such electricity derivatives can generate substantial financial risks. The advantage of a vesting contract for differences is that it can be designed and imposed by the government in parallel with the establishment of the finan- cial market. In most markets contracting has started with generators and large purchasers with considerable financial resources, then gradually extended to medium-size and small purchasers. In China contracts for differences could initially be signed by generators with the provincial power company (the single buyer; Figure 6-5), then with multiple buyers as reform progresses, until other intermediaries develop the skills and capacity needed to trade negotiated contracts for differences. h M&SO ( J Buyers -400nsumer contmcis \, Vesting CtD contrail- New contract established by the market rules G: Generators M&SO: Market and System Operator Source: PhaceLift. PRIVATE PARTICIPATION IN INFRASTRUCTURE IN CHINA 85 Option 3: Partitioning power purchase agreements from the competttive powver market. Another way to intcgrate existing and new power purchase agrecments with a mandatory pool market-in a way that will not seriously undermine the competitive market-is to move the power purchase agrecmcnts upstream of the mandatory pool market, through an intcrmediary. This option is based on experience from the liberalization of Australia's power sector (Box 6-2). Under this option powcr purchase agreements are executed between the generator and an intermediary, the market trader, and so are isolated from the market pool. The market trader provides an interface between the agreement and the competitive market. The market trader is a temporary organization owned by the government. It acts as a buyer to existing power purchase agreements but registcrs as a seller in the pool market and operates solely for the purpose of trading the agreements in the competitivc wholesale market. The private investor is required only to manage power plant assets in accordance witlh the power purchase agreement The government may allocate a vcsting contract for differences (without any compensation) to the market trader to help manage its cash flow. During the first phase of reform the smgle buyer would be the counterparty to the vesting contract. In subsequent phases the vesting contract would be distributed to multiple buyers in accordance with government policy. In addition, the market trader can decide whether to enter into negotiated contracts for differences if and when such other parties exist. Thus structure is shown in Figure 6-6 A review of sevcral powver purchase agreements in China reveals that no terms or conditions would prevent their reassignment to a government-owned market trader Attention should be paid to certain areas if the market trader is to enter into new purchase agreemcnts; these include the separate provision of ancillary services, the use of a two-part tariff structure, penalties for unavail- ability, unit dispatch instnictions, and communication of relevant performance and status informa- tion to and from the powcr plants. Establish a Multipart Tariff in Temporary Power Purchase Agreements While private projects in China have achieved lower capital costs and tariffs than those in other countrics in the region (Indonesia, the Philippines, Thailand), government authorities remain conccrncd that private investors are not bearing enough risks for the returns they are getting. Investors, meanwhile, are concerned about the uncertainty involved in initial and subsequent tariff approvals. The "nlCw plant new price" approach, issucd by the Chlinese government in 1985, was well received by private investors and led to a surge in investment in the 1990s. But the authorities feel Box 6-2: INTERNATIONAL EXPERIENCES WITH PARTITIONING EXISTING POWER PURCHASE AGREEMENTS In Australia all existing power purchase agreements have been accommodated through the market trader principle (option 3 in the main text). This approach has proven to be commercially and politically manageable for the government and the only option acceptable to the private investor. In one jurisdiction six power pur- chase agreements (representing about 2,400 megawatts, or 25 percent of the market) were managed this way. The six agreements included baseload coal-fired power plants, midrange coal-fired power plants, and gas turbine peaking plants. All the purchase agreements were bundled into one portfolio and managed by the market trader. The market trader was structured as a corporation under company law, including a Board of Management and executive managers. This greenfield corporation was able to introduce innovative risk management techniques that sub- stantially improved the cash position of the portfolio. In effect, the market trader was a way of partitioning the power purchase agreements from the market reform process. In this way the government could continue to offer protection for the private investor while benefiting from the early stages of the reform. Source. PhaceLift. 86 WORLD BANK WORKING PAPER FIGURE 6-6: PARTITIONING A POWER PURCHASE AGREEMENT FROM THE COMPETITIVE POWER MARKET (private investor) (Market Trader) (Retailers) PPA Seller PPA Buyer Market Buyer Market Seller (Wh (pool) Buyers Consumer PPA Multi-lateral(pool) // ~~~contatX_, Existing PPA is partitioned from market . _ CfD contract -- New contract established by the market rules G: Generators, M&SO: Market and System Operator, MT: Market Trader Source: PhaceLift. that this approach allows capital costs to be recovered too quickly (8-10 years in some cases), which results in front-loading of power prices. The cost plus approach also takes into account the project's actual financing cost, giving further assurances to investors but not providing enough incentives for efficiency. Figure 6-7 compares historical tariff structures in Zhejiang Province, New South Wales (Australia), and Argentina. The tariff for transmission and dispatch seems to be low in China, leaving the government with inadequate resources to expand transmission and distribution. But generation costs and taxes are higher in Zhejiang. The higher generation costs could be the result of higher fuel costs and higher capital costs. The government should determine the exact reason for these price discrepancies and remove potential impediments so that China can achieve world-class efficiency from its power sector. Many existing projects were given a one-part tariff and guaranteed minimum dispatch level. A one-part tariffprovides a purchase price only for units of energy produced or the guaranteed rmi- mum level, whichever is lower. The intention was to ensure that investors would havc a minimum revenue stream, mainly to facilitate project financing. But this structure puts pressure on off-takers, and with the recent slowdown in demand local off-takers have started to dispatch below the mini- mum level and to renegotiate the contracts. During the transition period all new power purchase agreements should provide for a two-part tariff, with a capacity component and an energy component. The capacity payment will usually be designed to recover fixed costs, including the cost of financing, and the energy payment will be designed to cover variable costs with a small safety margin. Both components would be determined through competitive bidding to form the basis of medirum- to long-term power purchase agreements. The capacity charge would be paid regardless of the actual dispatch for the billing period, while the energy payment would be subject to the level of production for the billing period. This structure will increase flexibility for dispatchers and provide economic certainty for investors. If a market trader was given responsibility for trading such power purchase agreements in the manda- tory pool market, the trader would price production (regardless of its purchase obligations under PRIVATE PARTICIPATION IN INFRASTRUCTURE IN CHINA 87 FIGURE 6-7: TARIFF STRUCTURES IN NEW SOUTH WALES (AUSTRALIA), ARGENTINA, AND ZHEJIANG PROVINCE 100% - 90% - 80% - 24% or 13 5 fen/kW h 66% 70% - 60% - 50% - 40% _ 1 i 40 3% - ~ . ( 70% or 40 fen/kW h 30% - 33% 20% - 10% - .. 10% NSW Argentina Zhejiang Tax Distribution Transmission Generation Source PhaceLift (2001) and Mercados Energeticos internal presentation to the World Bank. the agreements) to recovcr both the capacity charge anid the energy charge from the market's pool clearing price. Restructure Generation Assets Investors are less concerned about a transparent and competitive market than about the conflicts of interest created when governments and state power companues play multiple roles in owning gen- eration assets. The establishment of a competitive market is a step m the right direction to remove such conflicts of interest. But a lot more remains to be done. International experience shows that many countries engaged in power sector liberalization have actuaHy seen demand grow faster than envisaged whle supply lagged behind. During power reforms the government needs to continue to tender directly for power plant investment, to ensure that the supply and demand balanice is appropriately managed from year to year. The traditional project funding techniqucs of BOT and secuntization can continue to be used with the market trader as the government's vchicle. But there should be a reduction in traditional joint venture projects, because cross-ownership of generation assets, transmission assets, and the system operator role distorts the market. The government also needs to expedite the removal of its cross-ownership in existing power assets (particular gencration and network assets). Removing such cross-ownership will give private investors confidence that competitive neutrality exists in China's power sector. One way to remove cross-ownership is to divest government's interest in generation plants. These generation assets can be packaged to access a broader source of financing, in addition to traditional project-based financ- ing. For example, companies investing in a portfolio of power plants at various stages of implemen- tation (brownfield, greenfield, and so on) can seek financing from domestic and international security markets, in addition to strategic and institutional investors. Annex I LIST OF PPI PROJECTS IN THE ROAD SECTOR IN CHINA 89 Tariff and Cost Km Project JV Dev. Private FDI Equity/ Private Toll Rates Date Project Name (US$ m.) Lng. Type (Yrs.) Str. Own. Debt (US$ m.) Sponsor (per vehicle) Comments Anhui 1997 National Highway 206 51 90 O&M 25 CJ-V 60% 30.8 (E) RKI Rmb 5/ton 1997 Provincial Highway 307 42 21 O&M 32 CJ-V 60% 16.78 (E) RKI Bengbu Huaihe Bridge Highway 1997 Provincial Highway 307 120 59 O&M 32 CJ-V 60% 12.30 (E) RKI Bengbu Huaiyuan- Mengcheng Highway 1999 Bengbu Chaoyanglu 44 1.8 O&M 30 CJ-V 60% 13.3 (E) RKI Huaihe Highway Bridge 2000 Hefei-Yeji Highway 260 130 O&M 25 CJ-V 50% 66.39 (E) RKI Total 517 302 140 Guangdong 1990 Guangzhou City 173 22 O&M 33 CJ-V 60% 68.25 (E)/22 (D) NWI Rmb 10.21 (Avg.) Syndicated loan Northern Ring Road $97.5 mm 1991 Guangzhou-Shenzhen 1922. 123 O&M 30 BTO 48% 922 Hopewell Syndicated loan Superhighway - $800 mm w/ GITIC gtn. 1992 Beijing-Zhuhai Express- 472 62.2 O&M 37 C J-V 25% 17.46 (E)/ 100.54 (D) NWI Rmb 5.9 (Avg.) Guaranteed Repay- way (Guangzhou-Zhuhai ment Period section) 1992 Shenzhen-Huizhou 61 58 O&M 30 CJ-V 68.1% 2.41 (E)/21.31 (D) NWI Rmb 23.95 (Avg.) Exchange Loss Expressway (Huizhou Protection section) 1993 Foshan Guangzhou- 109 52 Greenfield 25 C J-V 35% 38 (E) RKI Rmb 0.25/km-Rmb Sanshui Expressway 2.375/km 1993 Roadway No. 324 27 24 O&M 22 C J-V 52% 9.8 (E/D) NWI Rmb 8.76 (Avg.) Guaranteed Repayment (Gaoyao section) Period 1993 Shantou Bay Bridge 86 11 Greenfield 35 C J-V 30% 26 (E) CKI 1993 Shen-Shan Highway 326 140 O&M 35 C J-V 33.5% 112.44 (E) CKI (Eastern section) 1993 Shunde Road System 415 102.4 O&M 30 BROT 50% 18 (E)/5 1.9 (D) Hopewell 1994 Boca Tigris Bridge 413 15.76 Greenfield 30 BOT 10% 3.2 (E)/28.3 (D) Hopewell Rmb 5-Rmb 180 1994 Jieyang Highway 124 110 7 O&M 30 C J-V 50% 62.1 (E) RKI Rmb 2-Rmb 30 Network 1994 Luoding-Chonghua 36 35 O&M 20 C J-V 61% 22 (E) RKI Highway 1994 Provincial Highway 268 14 26 5 O&M 20 C J-V 75% 13 (E) RKI Rmb 2-Rmb 23 1994 Qinglian Highway 360 215 Greenfield 33 BOT 51% 184 (E) Pinetree 15% year increase 1994 Roadway No. 1964 23 32 O&M 25 C J-V 74% 7.7 (E)/9.8 (D) NWI Rmb 709 (Avg.) Guaranteed Repay- (Zhaoliang section) ment Period 1994 Roadway No. 321 3 1 42 O&M 25 C J-V 60% 6.8 (E)/ 12 5 (D) NWI Rmb 25 (Avg.) Guaranteed Repay- (Fengkai section) ment Period 1994 Roadway No.321 59 79 O&M 25 CJ-V 55% 33 (E/D) NWI Rmb 32.56 (Avg.) Guaranteed Repay- (Deqing section) ment Period 1995 Roadway No. 1960 44 47 O&M 25 C J-V 50% 21 88 (E/D) NWI Rmb 19.06 (Avg.) Guaranteed Repay- (Sihui section) ment Period 1995 Roadway No. 1960 35 60 O&M 25 CJ-V 55% 19.5 (E/D) NWI Rmb 19.12 (Avg.) (Guangning section) 1996 Deqing Xijiang Bridge 16 1.4 Greenfield 25 C J-V 65% 4.41 (E)/6.19 (D) NWI Rmb 6.98 (Avg.) Guaranteed Repay- ment Period 1996 Gaoming Bridge 18 1 1 O&M 25 C J-V 51% 3.69 (E)/5 53 (D) NWI Rmb 5.39 (Avg.) Guaranteed Repay- ment Period 1996 Guangzhou Three 377 4.26 Greenfield 33 CJ-V 70% 181.88 (E)/26.35 (D) NWI Rmb 3.33- Exchange Loss Bridges 5.69 (Avg.) Protection 1996 Hui-Ao Roadway 97 86 07M 33 C J-V 50% 4.52 (E)/44 (D) NWI Rmb 5 94 (Avg.) (Hui-Dan and Hui-Ao Section) 1996 Jiangmen Road Network 99 43 O&M 30-32 C J-V 50% 49.52 (E) CKI 1996 Nanhai Road Network 330 140 O&M 24-28 CJ-V 49-64% 179 (E) CKI 1996 Roadway No. 1962 8 19.5 O&M 26 C J-V 55% 1.55 (E)/3.11 (D) NWI Rmb 4.20 (Avg.) Guaranteed Repay- (Guangning section) ment Period 1996 Roadway No. 1969 11 27 O&M 28 C J-V 58% 3.23 (E)/4.84 (D) NWI Rmb 5.73 (Avg.) Guaranteed Repay- (Gaoyao section) ment Period (continued) 7ariS and Cost M[m Pro.ect DV Dev. Pivate FM Equity/ pHvate T7ol Rates lDate Project Mame (US$ in.) Lng. Type (Yrs.) Str. Own. lDebt (US$ mn.) Sponsor (per vehicle) Comnments i 996 Shenzhen Airport-Heao 157 23.3 O&M 30 C J-V 45% 70.6 (E) RKI HK$0.6/km- Expressway, Eastern $3.6/km section 1997 Guangzhou Ring Road 542 39 Greenfield 35 BOT 75% 482 (E) CKI/ Hopewell 997 Laolong Bridge O&M 22 BTO Bridgecon Toll collection started 12/97 3 ? 7 National Roadway 30 33 O&M 25 CJ-V 51% 3.21 (E)/i 1.47 (D) NWI Rmb 19.21 (Avg) Guaranteed Return No.105 (7i anping Co. north section) : iv Roadway No. 1906 22 26.8 O&a 30 CJ-V 80% 17.6 (EID) NWI Rmb 7 (Avg.) Gu2ranteed Return/ (Qingcheng section) 2000 operational I Wi Roarway No. 1959 24 26.6 Greenfield 30 CJ-V 79% 19.5 (E)/4.9 (D) NWI Rmb 4.49 (Avg.) Guaranteed Repay- (Qingxin section) rment Period SQ7 Roadway No. 1967 12 25 O&M 25 C J-V 60% 7.2 (E/D) NWI Rmb 8.98 (Avg.) Guarznteed Repay- (Xinxing section) ment Period L * 99' Shenzhen-Huizhou oaM 26 C J-V 50% 7.23 (EID) NWI Rmb 7.4 (Avg.) Exchange Loss Roadway (Huizhou Protection section) :997 Shuangjian Roadway 34 26 61% NWI Rmb. 3.94 (Avg.) Guaranteed Repay- (Gaoyao section) ment Period i 79i Zengcheng l ixin Road 26 30 O&aI 23 C >-V 51% 13 (E) CKI : 9 Roadway No 32 23.8 25 6i% NWI Rmr. 9.86 (Avg.) Guaran=6zd Repay- (Gaoyao section) ment Period 999 Roadway No 1962 32.4 30 60% NW! Rmb. 6.68 (Avg.) Guaranteed Repay- (Gaoyao section) ment Period 599 Roadway No. 1958 30 25 65% NWI Rmb 8.06 (Avg) Exchange Loss (Deqing Section) Protection RC23 Panyu Beidou Bridge 22.5 3 O&M 25 CJ-V 40% 7.95 (') CKI Total 6522 1907 2996 Guangxi 1996 Guilin City-Liang Jiang 26 11.44 O&M 30 BTO 100% 26 (E) Guilin Eng. Toll collection Airport Toll Road started 11/96 1996 Yulin City Ring Roads 55 66 O&M 25 CJ-V 70% 38.3 (E) RKI Rmb l-Rmb 20 and Yulin-Gongguan Highway 1997 Beiliu City Roadways 29 40 O&M 25 C J-V 60% 18 (E/D) NWI Rmb 6.25 (Avg.) Guaranteed Return 1997 Cangwu County 20 10.12 O&M 25 CJ-V 70% 13 (E/D) NWI Rmb 7 18 (Avg.) Guaranteed Return Roadway 1997 Roadway No. 321 22 13 O&M 25 CJ-V 60% 13 (E/D) NWI Rmb 17.09 (Avg.) Guaranteed Repay- (Wuzhou section) ment Period 1997 Rongxian Roadways 25 27 O&M 25 C J-V 70% 17.4 (E/D) NWI Rmb 6.75 (Avg.) Guaranteed Return 1997 Yulin to Shinan 52 86.5 O&M 25 C J-V 60% 31 (E/D) NWI Rmb 8.10- Guaranteed Return Roadways II 26 (Avg.) 1997 Yulin Shinan-Dajiangkou 19 38.7 O&M 25 CJ-V 60% 4.61 (E)/6.96 (D) NWI Rmb 8.08 (Avg.) Roadway 2001 Yulin Shinan to Guigang 16 20 O&M 25 CJ-V 60% 9 (E/D) NWI Operational in i/0I Guaranteed Return Roadway Total 264 313 177 Hebei 1997 National Highway 307 40 40 O&M 20 C J-V 60% 24 (E) RKI Rmb I 0-Rmb 60 1997 National Highway 309 48 79 O&M 18 CJ-V 70% 33.8 (E) RKI RmblO-Rmb 60 1997 Tangshan-Tangle Road 24 100 O&M 25 C J-V 51% 12 (E) CKI Total 112 219 70 Henan 1997 National Highway 107 49 114 O&M 27 CJ-V 66% 39 (E) CKI (Zhumadian sections) 1997 National Highway 311 50 80 O&M 23 C J-V 50% 24.8 (E) RKI Rmb 5-Rmb 25 and Provincial Highway 01 Total 99 194 64 (contnued) I Co s~~~~~~~~~~ot mmr [FroDect DV [Dev. private F70b Esdaniy/ PriEvate 7cob Rates FSate Fr>tect Mame (US$ m.) Lng. 7ype (Prs.) Scr. Own. Debt (=$ Mt.) Sponsor (Fer veackCe) commnents Hube5 b993 Wuhan Airport 89 18 Greenfield 30 CJ-V 67% 3 (E)/60 (D) NWI Rmb 11.21 (Avg.) Guaranteed Return Expressway IM94 Wuhan Bridge 205 4 O&M 30 JSC 49% 101.2 (E/D) NWI Rmb 6.28 (Avg.) Tax Incentive Development Total 294 22 164.2 b 97 Changsha Wujialing 60 5 O&M 25 C J-V 44% 27 (E) CKI and Wuyilu Bridges '9N7 Changsha-Yiyang 168 75.6 O&M 27 Cl-V 43% 74.5 (E) RXI Expressway Total 228 81 10:.5 b 994 Suzhou-Shanghai 27 52.8 O&M 23 C J-V 50% 23.4 (E) RKI Rmb 3-Rmb 60 Airport Highway : S$t Provincial Highway 211 30 26.1 Greenfield 20 C J-V 60% 18 (E) RkI Rmb 2-Rmb 60 ' C9 Yangzhong-Changjiang 3 1.17 O&M 25 ROT 51% IJM Bridge Total 60 80 41 I 96 Shenyang Changqing 51 4 O&M 32 C J-V 54.2% 27.4 (E) CK; and Shenyang Gongnong Bridges 0 b P5 Shenyang Shensu 5I 12 O&M 32 C J-V 54.2% 28 (E) CKI Expressway M7 Shenyang Da Ba Road 195 23 O&M 30 C J-V 52% 100.64 (E) CKI Total 297 39 156 Shanghai Municipality 1996 Shanghai Inner Ring 1.7 b 56.11 Asset Sec. BOT 35% 600 (E) SIHL Co. receives 80% of Road/N-S Elevated revenues Expressway Total 1700 56 600 Shanxi 1996 National Highway 108 19 38 O&M 20 CJ-V 65% 19 (E) RKI 1997 National Highway 29 33.60 O&M 23 C J-V 65% 29.4 (E) RKI Rmb 5-Rmb 50 1 08-Taiyuan-Yuci Rmb 5-Rmb 20 Highway & Bypass 1997 Provincial Highway 104 27 41.50 O&M 23 C J-V 60% 16.4 (E) RKI 1998 Jincheng-Jiaozuo 174 32 Greenfield 30 C J-V 60% 104 4 (E/D) NWI Rmb 5-Rmb 50 Exchange Loss Protec- Expressway Rmb 1.45 tion/Op. 10/00 1999 Roadway No. 309 22.2 O&M 25 C J-V 60% NWI (Avg.) est. Exchange Loss Protec- (Changzhi section) Rmb 12 (Avg.) est. tion/Op. 12/99 1999 Taiyuan to Changzhi 18.3 O&M 25 C J-V 60% NWI Exchange Loss Protec- Roadway (Changzhi Rmb 12 (Avg.) est. tion/Op. 12/99 section) 1999 Xiaodian Fenhe Bridge 7.87 5.5 20 C J-V 25% 1 96(E) RKI 2000 Shanxi Taiyuan-Guliao 22 23.18 O&M 27 C J-V 60% 5.21 (E)/7.83 (D) NWI Roadway (Taiyuan Rmb 9.36 (Avg.) Section) Total 279 214 184 Sichuan 1994 Chengdu-Mianyang 170 90 Greenfield 30 BOT 60% 99.6 (E) New Exchange Loss Expressway China Protection Total 170 90 100 Tianjin Municipality 1997 Tangjin Expressway 398 60 O&M 30 C J-V 60% 239 (E/D) NWI Bank of China 8-year (Tianjin North section) Rmb 25.25 (Avg.) loan Rmb I bn. 1998 Tianjin Yonghe Bridge 0.5 25 90% NWI Tax Incentive Total 398 61 239 Rmb 5 85 (Avg.) (continued) bariqT and Cost Km probect IV Dev. Private F[ Enquity/ Private T70N Rates [Date Protect bmae (LS$ in.) Lng. lype (Vrs.) Str. Own. Debt (=$ in.) Sponsor- (per vehcle) Comments Zhetiang 995 Fuyang-Luzhu 47 O&M 30 BROT 55% 26 (E) Bridgecon Toll Road 099 Tonglu-Zhaiqi 19 18 O&M BROT 55% 10.6 (E) FACB Toll collection Toll Road started 3/97 Total 66 18 37 3 Asset Securitizations 3996 Anhui Expressway 113 164 Asset Sec. IPO 35% 113 (E) Anhui Company Limited HK$0.30-HK$ 1.50 996 Guangdong Provincial 223 130 Asset Sec. Share 20% 223 (E) GPED Expressway Develop- listing ment Co., Limited :95 dJliangsu Expressway 564 545 Asset Sec Share 25% 564 (E) Jiangsu NWI has 3.12% Company Limited listing equity share .9d7 Shenzfien -xpressway 150 60 Asset Sec. IPO 34% 150 (E) Shenzhen Company Limited I991D Sichuan Expressway 178 226 Asset Sec. [PO 10% 178 (E) Sichuan NWI has 10% equity j Company Limited share 2997 Zhejiang Expressway 425 400 Asset Sec. [PO 425 (E) Zhejiang Company Limited ' 99S Hainan Expressway 55 Asset Sec. IPO 55 (E) Company Limited 995 Hunan Expressway 101 Asset Sec. IPO 101 (E) Company Limited Total 1647 1140 1309 3g9 z GafLnay [Mte'mzt!Gnz]Dn $350 mm note financing Guangqing Highway 101.3 22.6 O&M 30 60% 56.8 (E)/4.0 (D) 16.5% mm. return (existing) Guangqing Expressway 98 22.3 O&M 30 BOT 60% 58.8 (E/D) 16.5% min. return/ (proposed) completed in 2000 Zhaoqing Highway 53.5 50.3 O&M 30 60% 32.1 (E/D) 16.5%mm return (existing) Zhaoqmng Expressway 78 46.8 Greenfield 30 BOT 60% 129.4 (E/D) 16 5% min. return (proposea) 1996 Guangzhou Bridge and 205 O&M 20 60% 123 (E/D) 17.5% min return Tunnel Project Total 536 142 404 3 Mainland Private Toll Road Company Shanghai-Midway Infrastructure, Ltd. National Highway 320, 85 80 BOT 32% 26.8 (E) Jiande section National Highway 320, Longyou-Hangbu section National Highway 330, 55 35.5 O&M BOT 16% 8.6 (E) Yongkang section/ext Zhenda Highway, Jiangsu 25 21 O&M BOT 92% 23 (E) Provincial Highway 61, 6 19.36 O&M 100% Jiangbei section National Highway 104- 25 144 O&M 50% Rui'an section Total 196 170 58 Grand Total (USD) 13,547 5,433 7,341 1,975 mm commercial notes and loans Grand Total (RMB) 112,440 60,930 16,393 mm A nnex 2 LIST OF PPI PROJECTS IN WATER TREATMENT AND WASTE WATER IN CHINA T r his list has been compiled from a number of sources, including the World Bank PPI data- base, and interviews with the various developers In compiling this list, two difficulties should be noted: O The value of the transactions is perhaps rather small compared to other sectors; therefore, there is not necessarily a register of such transactions at the central government level (i.e. with SDPC). To obtain information concerning these transactions requires investiga- tions at the Provincial/Municipal level or of the various developers. Due to the sensitive nature of much of this information, many parties have been rather loathe to provide detailed information. O Although the history of PPI in China is rather short in water sector (the first ventures were signed in 1992), changes have been occurring in these ventures involving in some cases the degree of private sector participation and in one known case the nature of the partici- pation That such latter changes should occur is not surpnsing given the experience of such ventures in developed countries. The rapidity by which this change has occurred is perhaps surprising. Supposing a liberalization of the market in the PRC in forthcoming years, one might expect such trends to continue. 99 0 0 Mzlnt lDueraVRon of IRaw Water- [3lk water EDomestUc Oate capacity Cost % contract Taruq taHff water tarif z (s6zinztUre) MYamae (Mn3/dZF) Ty[pe (Us$ rnii) pH1vate (Vezrs) RM[B/Mn3 MMG[B/M3 11RG-I[B/n3 EoevaNoper : 1992 Tanzhou 50000 Concession $13 58 35 na na 1.3 Sino-French 0 Water Supply z 1994 Harbin 225000 JV (BOT) $30 50 30 0.62 I SAUR Water Supply International > m 1995 Nanchang 50000 JV (ROT) $11 50 30 na 1.05 0.8 Sino-French X Water Plant 1995 Shenyang 1800000 Initially $32 0 30 na 1.09 1.3 Sino-French Water Supply BROT now O&M Contract 1996 Shanghai Da 400000 BOT $73 100 22.5 - - 0.7 Thames Chang Water Water/BOVIS Treatment 1996 Nanhai Water 250000 JV (BOT) $16 50 20 - - - Giantmost Ltd. Supply 1997 Tianjin Water 500000 JV (ROT) $30 65 30 1.2 - - Vivendi Supply 1997 Lianliang Water 100000 JV (BOT) $15 60 30 na 1.25 1.13 Sino-French Supply 1598 Zhongshan 500000 jV (ROT) 527 66 22 na 0.77 c. 1.2 Sino-Fr2nch Water Supply 1998 Zongshang 200000 JV (ROT) $30 66 22 na - - Sino-French Water Supply 1999 Chengdu Water 400000 BOT $106.5 100 20 - 0.98 0 65 Vivendi/Marubeni Supply Waterworks Company Ltd. Chang.u, Laoning 0 JV (BO-T) $13 70 30 - 1.1 - Sino-French 2000 Baoding, Hebei 260000 JV (BOT) $26 90 30 not 0.61 Sino-French applicable 2000 Zhengzhou, 300000 JV (ROT) $38 50 30 not 0 84 I Sino-French Henan applicable - Shenyang 100000 - - 50 - - - - China Water - Xiaoqing, 100000 - - 50 - - - - China Water Zhejiang 1997 Jinan Water 900000 JV $90* 80 25 Cathay (Dongiiao, International Nanliao, Xiliao plants) 1997 Jianan Water, 400000 JV (BOT) $30* 60 25 - - - Cathay Dayang Plant International Binzhou Dongiao 40000 JV (ROT) $9-9* 60 20 Cathay X International Binzhou Cathay 50000 JV (ROT) $6 6* 80 20 Cathay > Water Plant Ltd. International m Jiangmen Water 500000 JV (ROT) $60* 80 - - - - Cathay > International - Xuzhou Cathay - Wastewater International - 0 Treatment Plant z 1997 Xian Water 250000 JV (BOT) $30 50 Berliner Wasser z Supply z 1998 Yueyang Water 400000 JV (ROT) HK$140 49 18 Cheung King Supply Infrastructure 1998 Zhongcao WTP, 150000 JV (ROT) $24 50 China Water c Guiyang, Guizhou 1998 Beijiao WTP, 100000 JV (BOT) $24 50 China Water Guiyang, Guizhou Z 1995 Chongqing Water - ROT $25 Sino-French n Supply _ > (contnued) 0 0 z 0 z ID)nraon f Raw Water MuHt wazter Dcmest EDaeC2Ct:ost ~% contract I,rf (s2hsat Hmw (3-r3/ic]yD -:Y;p2 z 9 m-1) 77"vat 7 y 3I[nt3 [J/M KL3Ir3 mevc?ep 1996 Jiangsu Province 1000000 Partial $162 49 - - - CITIC Pacific divestiture 1997 Xiejiang Water 240000 JV (ROT) $25 49 18 - - - Cheung Kong Supply 1997 Fushun City 360000 JV (ROT) $24 69 18 - _ - Cheung Kong Water Supply 1997 Shenyang Shifos, 400000 JV (ROT) $24 50 20 - - - Cheung Kong Water Plant i 999 Wanzhou, IOCOC0 JV (BOT) $15 so 30 - I Sino-French Chongqing A nnex 3 LIST OF PPI PROJECTS IN THE POWER SECTOR IN CHINA 103 104 WORLD BANK WORKING PAPER Category Definition Comment BOT I Build-operate-transfer Strictly by Chinese pilot BOT regulations this is the Competitively tendered only type of BOT 100% or majority foreign owned Large coal (mainly), new plant with long term PPA BOT2 Build-operate-transfer BOT category is applied to projects which are 100% Negotiated (possibly includes private equity without a Government authority involved some risk sharing through JV (ie no state-owned power company or government co-operation agreement) authority is involved in the equity partnership), but not Other features as BOT I all BOTs to date have been competitively tendered. Also referred to as wholly foreign owned (WVFO) TOT Transfer-operate-transfer The (partial) sale of existing power plants, usually in order to raise equity for new plants. JVI Large coal projects, new plant Previously debt share needed to be similar to equity Majority foreign financing share, but more recent projects can have minority Minority foreign equity foreign equity JV2 Large coal (or oil-category Does share of foreign financing influence the approval JV2-0), new plant Minority time or perceived risk of project for financing? foreign financing JV3 Large coal prolects, new plant Maybe no need to distinguish between JV I and JV3; Majority foreign financing pre-'95 Central Government approval was not given Majority foreign equity for foreign share >50% JV4 Medium coal projects IPO I Listed on International and Portfolios of plant investments, mostly existing plants. Chinese exchanges Generally have a regional or provincial geographical focus IP02 Listed only on Chinese As IPO I exchanges SC I Small coal, mainly new plant Small defined to be IOOMW Majority foreign owned 02 Small oil plant Mostly carried out for Municipal or City power compa- Wholly or majority privately nies (where approval process is simpler), some for owned Provincial Various types, Diesel, ST, GT H Hydro, all types Politically sensitive due to social and environmental impacts Wholly foreign ownership not permitted >250MW N Dayawan is only Nuclear plant Special case (1982-86) E e 3c c 41 4.0~~~~~~~~~~~41( E B G C 70 ea 0 4 'S 4 In~~~ 0 % U 041 0 c U> Dayawan ~ ~~~~~~ Gungon NL80 NI 18L L K360 2 01 00 EA Macuhan 0ainan C 250 JV I 1988 Y H 0 9 Shaojiaog B Guangdong C 700 JVTI 1984 Y Hoeweold H K 373 10 8/10 373 Consoru_um_ 2 Infrastructure Shajiao C Guangdong C i1980 JV I 1992 Y HK i1300 40 = 552 1870= Zhangping Fujian C 200 JV I 1992 Y _ _____HK 14 25 8 56 Yangpu Power Hainan 0 320 01I 1992 Y Japan 240 100 282_ project _ _ Xinchang Zhejiang C 24 SCI 1992 Y Lllinova 6 54 11 Generation Songyu Fujian C 600 JV I 1993 Y HK 84 25 = = 421= Pengcheng Power Jiangsu C 600 JV2 1993 Y China HK 125 35 517 Station Resource Group Shenhao Liaoning C 400 JV3 1993 Y China HK 93 55 170 Resources Chenghai Shantou, 0 75 02 1993 Cheng Kong HK na 41 Guangdong C_600 ___ Infrastructure Chaoyang Shantou, 0 90 02 1993 _ Cheng Kong HK na 50_ ______________________ Guangdong Infrastructure Topou Shantou, 0 114 02 1993 Cheng Kong HK 63 G uangdong ___Infrastructure Jingyuan I FGansu C 600 JVI 1993 Y Comuiy US 15 30 70/30 73 360 Sponsor 20 ToP Generatio Alternative (contnued ) c ~~~~~~~~~~~~4;c E X S 3 u u | 6 & |~~~~~~~~~~~~~~~~~~~~~~~~~~~~ t E E a E * C C~~~~~~~~~~~~~~4 E C Power Development~ ~ ~ ~~~~~r Listing Co Ltd.x r . !t - Huaneng~~~~ Power Nainl20 P 99 NwYr 0 2 Internationaln Ic Ci Xiangci Hunan HC 2 H I 199 AE US 5n Ji (ayIL 2 AO PA..Co IGansu IL=&. =2 J AL Shunde De Sheng Guangdong O 283 66 0 1993 Sithe Enrg 255 Shandong Huaneng Shandong 1750 IPOI 1994 w(New York 330 333 Power Development Listing) Co. Ltd. Huaneng Power National 2900 IPOI 1994 (New York 600 625 International Inc. _______Listing) ____ Qinglan Electric Hainan 0 1SO BOT2 1994- Enron us 130 100 70/30 lOS ISO na 12 ToP Power Plant Xiangci Hunan H 26 Hi 1994 -HAES us 52 na jingyuan 2 Power Co. Gansu JVl 1994 Y _ _____148 ___337 Houjie Power Plant Guangdong 0 66 02 1994 | Sithe Energy 73 68 108 Zhenjiang Combined Jiangsu C 27 SCI 1994 -CPower Fin 12 85 14 Energy Companies Mengdian Thermal Inner Mongolia 425 SO 1994 - na Power Co. Ligang Jiangsu C 700 JV3 1995 Huayuan HK 206 56 75125 154 436 ECAs 20 ToP (related HK to CITIC?) Zhujiang II Guangdong C 600 1995- New World HK 276 35 72128 106 422 Consortium Infrastructure Xinxiang Henan C 400 TOT 1995 Y CITIC HK 124j 90 67/33 83 138 20j Yangchung Fuyang I & II Guangdong 0 21 02 1995 - AES us 60 -0 Wuhu Anhui C 250 JV2 1995 Y AES, China US/China 29 25 70/30 20 115 na ToP Power mnt Hldgs Suzhou Coastal Jiangsu 13 ? 1995 Coastal Power 29 Cogeneration Power Plant Nanhai Power Guangdong 400 ? 1995 228 Plant I Co Ltd. I_II I _I _ Yongqi Power Plant 1995 _ 27 Nanhai Jiangnan Guangdong SO 1995 _ 30 Power Co. Ltd. Wuxi-Carec Jiangsu 0 40 1995 Coastal Power 26 Rizhao Power Shandong C 700 JVI 1996 Siemens, UDI Germany, 387 25 75/25 347 640 ECAs (85%), 20 ToP Israel Commercial banks (15%) _ Zhuhai Guangdong C 1400 JVI 1996 n? CK/Hutchison HK 970 45 70/30 804 1229 ECAs (85%), 20 Top 0.424 Commercial _ __________ _ Banks (15%) _________________B ( Chengdu 11 Sichuan C 284 JV3 1996 HK 154 51 300 Houshi Fujian C 1200 BOT2 1996 Taisu America Taiwan 1278 100 75/25 959 1278 20 0.39 Corp. Zhabei Zhadian Shanghai 0 400 JV2-0 1996 Y GE Capital, US 146 30 75/25 127 250 Commercial I5 ToP Power Plant MEPC Banks _ Tangshan Union Energy Hebei C 100 JV3 1996 Sithe Energy US 174 100 174 Small Coal Fired Hefei Prosperity Anhui C 116 01 1996 Lake Power US 45 70 64 Lake Power I _ Chengdu Kaihu Sichuan G 42 G I 1996 AES US 6 35 18 Gas Turbine Jiaozuo Wan Fan Henan C 250 JV3 1996 Y AES US 74 70 106 Power Co. Ltd Wuxi-AES-Carec Jiangsu 0 63 02 1996 AES US 24 55 44 Nanjing City Power Jiangsu 0 72 02 1996 AES US 12 40 30 Fuling Aixi Power Sichuan C 50 SC I 1996 AES US 22 70 32 Co. Ltd. Wuhu Zhaoda Electric Anhui ? 1996 115 Power Development Company Henan Power Plant Henan C I I 996 -_-19 (continued) E VI C4J0 esbi B~ Poe Gunx 0 O 97 EF E, Fac/U 1 0 03 3 1 Cs(7) 1 o Hanfeng Power Plat ebe C 132 J2 197N Semn, Geman 53 0 7/ 39 6 A > Z c 0 u o~~~ ~~ ~~ ~~~~~~~~ L.0 0 la r 0 C r 0.36k 4.1 0 ,,.~~~~~~~~~~4 A f 4 =x VI0 & '~~~0~~~ "jo -~~~~~~ In c o 0 A. 40 _ ~ ~ ~ ~~~~~~ _ Powe bank (15%) o A. ~ ~ ~ ~ ~ Y) A. U O v~~~~~~~~Ck o.5 0 IL ,. Huaibei Ah C 60 55 418 C0. 19 Laibin B Power Guangxi C 700 BOTI 1997 EDF, GEC, France/UK 610 100 70/30 438 610 ECAs (67%) 15 Top Aisthom Commercial _____________________ ____ _______ _____________ ____________ ~~~Banks (33% ) Hanfeng Power Plant Hebei C 0320 JV2 1997 N Siemens, Germany 503 40 75/25 397 1065 ECAs (85%), 20 ToP Yuan 0.36 North China Commercial 0.36/kwh Power -banks (I5%) HuaibeiDatangh A uina C 350 JVI 1997 Dynamic HK 86 25 70/30 55 418 Commercial 19 Investment Banks ____________________ ~~~~~~~~~~~~Ltd. Weihe Shaanxi C 1200 TOT 1997 Y Chinese HK 330 51 68/32 224 646 Tourist Co. Hanshin Power Plant II Hubei C 600 JV2 1997 Zhonglian HK 329 25 75/25 314 334.6 2 1 0.397 Power Finan- ____________________ cial Company Beijing Datang North China 3150 PO 1997 HK, London 450 100 124 450 Power Plant region (listing) Zhejiang Southeast Zhejiang 1660 IPO 1997 - Shanghai, 208 24.2 966 Electric Power London _________ ________________ (listing)_ _ _ _ _ _ _ _ Hefei No. 2 Anhui C 700 JV2 1997 N Singapore Singapore 305 49 75/25 236 560 ECAs (58%) 20 ToP 0.42 Investment Commercial Corp (42%) ___ Yichang Yihua Hubei S0 1997 - National UK 70- _____ ____ Power Plant Power Yichang CMI Power Hubei C 250 JVI 1997 - Consolidated US 178 75 70/30 124 237 Development Minerals, Inc. Company Ltd. Yangcheng International Shanxi C 12100 JV2 1997 N AES US 897 25 398 1600 ECAs Power Company Nanyang General Light Henan C 250 JVI or 3 1997 AEP US 172 Electric Co Ltd. Luannan Coal-Fired Hebei C 100 LO 1997 Panda US 155 Power Plant Meizhou Wan Phase I Fulian C 700 BOTI? 1997 Intergen US/Indonesia 655 100 75/25 466 755 Commercial 23 ToP 0.57, (70%), Lippo Banks (42%), 0.464 (25%), ADB ADB (34%), (5%) ECAs (15%?) Tong Zhou Meiya Jiangsu C 30 JV 1997 Y Meiya Power US na Cogeneration Co Ltd Company Ltd. Zunhua Keppel Power Tianjin C 100 JV4 1997 Panda US 70 Plant (Zhenhua?) _ Jiaxing jin Jiang Zhejiang 59 SO 1997 na Power Plant jiaxing City Coal-Fired Zhejpang C 2400 SO 1997 1633 Power Plant Fuyang Power Plant Anhui SO 1997 - na Changzhou City Jiangsu SO 1997 20 Combined-Cycle Changhong Thermal Sichan SO 1997 20 Power Plant Sichuan Quinwei Dali Sichuan SO 1997 30 Power Plant SJZ Yong Tai Power SO 1997 11 plant _ __ __ Small Scale Western SO 1997 na Resources Power Plants Shandong China Shandong C 3220 JV2 1998 EDF/CLP France/HK 948 49 72/28 627 1935 ECAs, Com- 20 0.447 Power Company mercial Banks (continued) ~~ .2 ~ ~ ~ ~ ~ . I Huhehaote r I25 21 Power Plan~~~~~~~~~~~~t Exedn Ioldin vP2oCect _ C L 0 e N a. ~~~~~~~~~~~0~ 0 a 0. E W L A 4 Wenzhou IL_ C IL J 1998 0n1 1 20 Jiaozuo Danhe Henan C 200 JV3 1998 Xinyuan HK 27 60 67/33 18 44.31 0.24 Industry ______ ~~Co. Ltd._ _ _ _ _ Huhehaote Thermal Inner Mongolia C 400 JV2 1998? Chaobao HK 164 35 75/25 145 210.20 20 Power Plant Extending Holding Project Co. Ltd. Wenzhou Zhejiang C 600 JVI 1998 Telluride us 158 40 75/25 125 415 20 0.494 Yiyang Henan C 100 JLa3 1998 us 85 85 67/33 59 91.34 20 Shenhua-shenmu Shaanxi C 200 TOT 1998 Enserch us NA 49 75/25 Power Plant sI Puqi Hubei C 600 BOT 1998 Sithe China US/Japan 454 100 70/30 300 454 Commercial 20 ToP 0.419 (75%), Banks Marubeni _____ ~~(25%)_ _ _ Small China Light and H -1998 - -1323.8 18 Power Hydro Power Plants Pingdingshan Henan C 100 JV I 1999 - Fei Ba Lu Er Germany 47 60 67/33 31 7072 Fushi Meiya Guangxi H 54 H 1999 Meiya Power US 53 Co. Ltd. Zuojiang Meiya Guangxi H 72 JV 1999 Y Meiya Power US 62 ______ ~ - Co. Ltd I__ __ _ _I_ _ Enerch-Meishan Energy Shanghai C JV3 1999? Enerch US 55 60 65/35 36 92.08 20 Thermal Power Plant Weigang Meiya Shanghai G 50 JV 1999 Meiya Power US 50 ___________________ ____________ ~~ ~~Coi. Ltd_ _ _ _ _ _ Shandong International Shandong LO 1999 _ 260 Power Development Corp. _ _ _ _ _ _ Nantong Meiya Jiangsu C 60 SO 1999 Meiya Power US 45 Co Ltd. _ _ _ _ _ _ _ _ _ Zhuiiang Guangdong SO 1999 Meiya Power US 62 Co. Ltd ____ Henan Peak Pacific Henan SO 1999 28 Power Huali Power Guangdong __1999 - 116 ______ Lin Zhang Peak Pacific Lin Zhang ____1999 24 Changsha Hunan C BOTI n n National UK 100 I ~~~Power Heilongjiang Electric IPO Power ..nnex 4 SUMMARY OF BENCHMARKING STUDY Introduction A review of the powver, watcr and transport sectors along with certain financial, legal and regulatory ISSUCs rclainng to PPI was carried out in eight countries: United Kngdom, Australia, Hunlgary, Brazil, Phlilippinies, Argentina, Malaysia, and Thailand. Information compiled on the PPI sector in each of the countries, xvere reviewed across the grid of benchmarking criteria to identify a list of pertinent issues that is presented in this report. This Benchmarking Study is presented in two sections. The first section considers certain issues applicable across PPI in all sectors and the second section considcrs issucs spccific to cach of thc power, water and transport sectors Table of Contents 1 Cross-sector issues .............. 114 1.1 LEGAL AND REGULATORY ISSUES ................. .... . ..... . .............. ....... 114 1.1.1 Overall PPI Framework. ............................... 114 1.1.2 The Legal Capacity of Granting Authorities ................................ 117 1.1.3 Project Procurement.. ......11. ..............1............... .1....................... 117 1 1.4 Form s of PPI ................... ......... . ...................................4......... ..... 121 1 1.5 Financial Issues .................................. ....... ..1...... . .....2......... .... 122 1.1.6 Project Imiplemcntatioii The Period After the Preferred Bidder is Selected ................... ..................................... ... 126 1.1 7 Secunty Issues . .... ... ...... ........... ....... ...... ..... .............. 127 1.2 FINANCING ISSUES ........ ... . ............................................... .......... 127 1.2 1 Investor Base ....... 12. ....... . .7.............. ........ ............... ....... . ... 127 I 1 3 114 WORLD BANK WORKING PAPER 1.3 DOMESTIC LENDING ................................................ 129 1.4 BOND FINANCING ................................................ 130 1.5 TRANSPARENCY OF FINANCIAL INFORMATION .............................. ......... 131 2 Sector-specific Issues .................................................. 131 2.1 ROAD SECTOR .. ....................................... 131 2.1.1 Extent of Private Participation ........................................ 131 2.1.2 Greenfield Projects Versus Brownfield Projects ......................................... 132 2.1.3 Toll Adjustment ........................................ 132 2.1.4 Government Support ......................................... 133 2.2 WATERSECTOR .......................................... 133 2.2.1 Extent of Private Participation: From Supply of Bulk WVater to Privatized Distribution ......................................... 133 2.2.2 Wastewater/Sewerage Services ................................................. 135 2.2.3 Off-taker Creditworthiness ................................................ 135 2.2.4 Efficiency of Privatized Nenvorks (tariff, non-rcvenue wate- rate, ...) . ....... 135 2.2.5 Size of Projects ................................................ 136 2.3 POWER SECTOR ..................................... . 136 2.3.1 Offtaker Creditworthiness ..................................... 136 2.3.2 Power Purchase Agreement Tariff Formula ...................................... 136 2.3.3 Market Risks Allocation ...................................... 137 2.3.4 Extent of Private Participation ..................................... 138 31 C(ross-sec:toir issues L.E LEGAL ANID REGULATORY ESSUE S R.1.1 Overall PIEK Framework Private investors are attracted to countries wvhere clear f;iamteworks fo; privpte investment exist. This allows tnformed investment decisions to be made and ensures consistency and clairity in the development of PPIprojects. In Brazil and Malaysia, constitutional amptentdm;ents were rnade to help facilitate PPI. Some countries, such as the Philippines and Hu ngapy, have introduced PPI specific legislation wvhich has helped facilitate investment. However, there is no eqUivalent in China, whe;lc the legplframe- work and regulatory structures are comparatively weak. This contrasts with tlhe ;;ioe developed reguslatory frameworks in the UIC and Argentina. Of the countries which were reviewed, the United 1(ingdom and Australia have attracted the most private mvestment per capita. Both countries have stable legal systems, transparent and stable frameworks for investment and have established competitive bidding procedures. In addition, in the UK in particular, there is an established network of independent regulators in the energy, rail transport and water sectors. Argentina has also creatcd industry specific agencies for major utility sectors and established regulating commissions for each agency. PRIVATE PARTICIPATION IN INFRASTRUCTURE IN CHINA 115 The PPI sector in some countries has benefited from significant legislative measures introduced by host Governments. For cxampic, in Brazil the constitution was amended to opcn up certain sectors to private investment on a concession basis Likewise, in Malaysia, thc Fcderal Consutution and the Pensions Act were among the laws and regulations which were amended to allowv privatization to take placc A different approach was adopted in Hungary and the Phiilippines where specific PPI legislation has been introduced, intended to establish a legal basis for and provide a clear framework within which private investment can take place mi those countries Both the BOT Law in the Philippines and the Concession Law in Hungary contemplate pnvate sector investment in most sections of the economy. The following table sets out the key provisions of the Filipino and Hungarian legislation: HUNGARIAN CONCESSION FILIPINO ACT AUTHORIZING THE LAW 1991 FINANCING CONSTRUCTION, OPERATION AND MAINTENANCE OF INFRASTRUCTURE PROJECTS BY THE PRIVATE SECTOR ("BOT LAW") 1990 (as amended in 1994) Range of activities Wide ranging, including roads, passenger Non-exhaustive, including highways, water transport, telecommunications, public supply, transport systems, IT networks, edu- utilities and postal services. Does not cation and health facilities and power genera- include activities in the power sector. tion, transmission and distribution facilities. Maximum term 35 years (but can be renewed by up to 50 years. of concession 17 years) Structures Either the state/local government granting Several structures provided for and authority must maintain majority owner- described under the legislation including ship or voting rights in the relevant Build-Operate-Transfer, Build-and-Transfer, companies or the state/local government Build-Lease-Transfer, Contract-Add- entity can concede the right to pursue the and-Operate and Rehabilitate-Own-and- relevant activity under a Concession Operate. Agreement. Bidding regime - Public tendering unless national security Unsolicited proposals are permitted in interests would be prejudiced, some circumstances, e.g. when prolects - Local Chamber of Commerce to review include a new concept in technology, proposals before bidding; although such prolects cannot benefit from - Invitations to tender must be published Government support; nationally at least 30 days before bid Basic tendering procedure established for deadline and the period for submitting solicited proposals: tenders must be at least 60 days; - Invitations to tender must be published - Tender notice should include details of nationally for three consecutive weeks; the concession, including rights retained - Criteria set out for determining preferred by the granting authority, financial tenderer, e.g. under a B.O.T. scheme, provisions and conditions leading to the winning tenderer shall be the early termination, compliant tenderer who has submitted - Other information can be included in 'the lowest bid and most favourable the tender notice as required; terms"; TABLE A 4- 1: COMPARISON OF PHILIPPINE AND HUNGARIAN LEGISLATION (SUMMARY) (Continued.) - Granting authority must evaluate - Direct negotiation can take place when tenders within 90 days of the deadline there is only one compliant bidder; for submission; - Bidders found to be non-compliant have a - Winning tender is the one offering right of appeal (in respect of which the "most favorable conditions" (i.e. not relevant agency must respond within purely financial); 45 working days). [NB: In Hungary, European Union pro- - Detailed tendering procedures established in curement law may also be relevant in some the accompanying implementing rules e.g.: circumstances as a result of Hungary's - clearly defined minimum design and per- Association Agreement with the EU]. formance standards and specifications must be set out in the invitation to tender, along with the draft Contract; - pre-qualification requirements are standardized (e.g. public utility facilities must be operated by entities at least 60% Filipino owned, etc.); - financial considerations are established (e.g. eligible financing sources);- - bid security requirements are set out;- - process and criteria for selecting preferred tenderer is established (mechanics and timetable). Contractual provi- - A concession company must be - Tolls, fees, rentals and charges must esca- sions established by established within 90 days of signature late by reference to a formula fixed at the the legislation of the contract, or the contract will be outset and related to official Government terminated; price indices; - If a concession company does not obtain - Government will pay compensation on all required licenses within termination following no fault on the 6 months of the signature of the con- part of the Contractor. However, such tract, the contract will be terminated; amounts must be insured against with an - Parties are free to provide that disputes accredited insurer and the Contractor involving foreign counterparties may be must pay the premium. referred to international arbitration; - Concession company cannot assign its rights Regulations on rate Not provided for in the legislation. "Reasonable rate of return on investments of return and operating costs" permitted, to reflect the prevailing cost of capital in the domestic and international markets. Limited to 12% per annum in respect of negotiated contracts for public utility projects that are monopolies. Other notable Foreign companies expressly given the - Public utility facilities must be Filipino provisions same rights as domestic companies. owned or owned by a company which is at least 60% Filipino owed; - If Filipino labor skills are available, they must be used; - Projects over a certain financial threshold are entitled to fiscal incentives; - Provision for priority proposed projects to be publicized. PRIVATE PARTICIPATION IN INFRASTRUCTURE IN CHINA 117 Whilst the introduction of BOT legislation is generally beneficial and well received by private investors, it is important that such laws are well drafted. For example, the Hungarian Concession Law was not clear enough to prevent a successful challen-ge to the toll rate calculation method which was clearly fixed in the Concession Agreement for the Mi-Mi5 Motorway project. 1.1.2 The Legal Capacity of Granting Authorities In order to facilitate private sector investment, it is important that publtc sectorgranting authorities have the legal powver to enter tnto contracts with the private sector and perform their obligations under such contracts. The lack of clarity regarding capacity continues to stall investment in the Brazilian water sector and has previously slowed the development in PPI in some sectors in the UK. The lack of clarity regarding the capacity of municipalities to enter into contracts has resulted in vcry little pnvate sector investment in the Brazilian water sector (in contrast to other sectors). The Brazilian Constitution appears to establish the municipalities as the relevan-t authorities (although there is some ambiguity), but the Brazilian States argue that they are the relevant granting authorities because they have been investing in and maintaining most facilities in the water sector since the 1970s. Legislation is likely to be introduced to clarify the issue. Issues concerning legal capacity have also stalled the development of the PPI sector in the UK. There was doubt regarding the capacity of local authoriues and some health authority entities to enter into certain contracts with the private sector. As a result of this, the UK Government intro- duced specific legislation to clarify the issue and there was a significant increase in levels of invest- ment in PPI in the two years after the legislation was introduced, in comparison to the preceding two-year penod. Whilst the change of law was not the only reason for the increased investment, the increase does reflect the understandable importance which private investors attach to legal capacity issues, even when, as in the UK, the lack of capacity is far from certain. One way in which capacity can be established is through the overall framework legislation dis- cussed above. For example, in the Philippines, the implementing rules provide that contractual arrangemenits aulthorized under the BOT Law may be entered into by, amongst others, provincial, city and/or municipal Government entities which are authorized by law or under their charters to undertake infrastructure and/or development projects within their respective jurisdiction. 1.1.3 Project Procurement Most countries have introduced legtslation that establishes a competitive btdding process when assets are to be transferred, or concessions to begranted, to the private sector. However, such legislation does not extst tn China. Where frameworks for competitzve bidding do exist, they are not always ideal from the private investors' perspective, e.g. Australta and the UK. The former Treasury Taskforce in the UIC and the BOT Center in the Philipptnes both operated as central co-ordination bodies in the PPI process. Standard form documentation has been introduced in some countries, for example, the UK and Thailand, to help develop consistency. Framework for Competitive Bidding When operated effectively, a transparent and competitive bidding process can be beneficial in both attracting additional investment and ensuring better value for money for the public sector. The Brazilian Constitution requires a public bidding process in respect of any contract offered by the Government, individual states, muincipalities, state-owned companies, state agencies or any 118 WORLD BANK WORKING PAPER other public administration institution at any lcvcl. in addition, one of the positive fcatures of the Flipino BOT law and the Hungarian Concession Law are that thcy both attcmpt to establish a common approach to the bidding proccss in respect of projects to which they rclatc. However, both laws allow flexibility and so, in practice, the approach is not always consistcnt. The Hungarian and Filipino legisla ion also set out what bid documentation should include. In the Philippines, the granting authority must sct out minimum design standards. It is arguable that where there is scope for innovation, detailed design proposals should be lect for the private sector to develop because of the cost involved and also because of the perception that the private sector is more innovative than the public sector. This is ccrtamnly the vicew in thc UK where, in considering whcther a proposed projcct is suitable for PPI, the Governmcnt ta;kes into account the scope for mnovation in designing thc infTastructure and operating procedurcs. A common complaint amongst private invcstors in the UK is that the EU procurcment rcgime allows for a negotiated proccdure which can lead to public bodics rcqunng tw%o bidders to develop and negotiate full project and financing documcntation becore a prcfcrred bidder is selected. The cost involved is vicwed by the pnvate scctor as unsustainablc in the long-term, unlcss a system for reimbursing the bidding costs of the unsucccssful biddcr is introduccd. The situation in the UK contrasts with that provided for undcr the Filipino BRLT law which only allows direct ncgotiation when there is a singlc bidder rcmaining. In Australia where the Federal Government (i.c. national Governmcnt) has a less direct involvement in infrastructure developmiient, the Australian Procurement and Construction Council develops nationaLly consistent approaches to broadcr procurement policies, processcs anid practiccs, in particular in respect of assessing and mccting public sector infrastructure nceds, competitive tendenng and improving access to Government contracts for small to medium sized enterprises. However, more detailed procuremcnt regimes are designed at state level and rclcvant guidelines, policy statements and codes of practice vary from statc to state across Aiustralia. There are concerns amongst investors because of certain inconsistcncics and disparitiecs crcatcd by such cross-jurisdictional variability. This leads to strong calls from the industry for common approaches to project definition, presentanon of documentation, handling and processing of bids and criteria for assessment through cooperation between the Federal, statc and regional Governments. Management of the Bidding Process Notwithstanding the cxistence of a regime for competitive biddmg, it is important that that the criteria are transparent and that granting authorities arc able to evaluate and assess bids made by the private sector In order to do this, granting authoritics need to know what they are looking to achieve at the outset and have the resources and know-how to manage the bidding process. Even in circumstances wherc there is no competitive tender, the granting autlhority should fully define its objectives before procuring private scctor investmenz. China is not alone in having problcms in respect of thesc issues For cxamplc, some observers have suggested that one of the less commendablc featurcs of the Filipino BROTj law is the provision for the private sector to submit unsolicited bids. In these circumstances, the iiitiative for a project comes from the private sector and the public sector granting authorties arc often ill prepared to evaluate and assess the proposals of the privatc secto: biddcr. This can lead to a chaotic implemen- tation process. It is noteworthy that thcrc is ongoing discussion in the Philippines as to wvhether unsolicited bids should continue to be permittcd. Followving its expericncc with thc first round of toll road concessions, thc Argentinean Govern- ment designed simple, straightforwvard concession terms and bidding criteria for the Buenos Aires access roads in 1992. This contrastcd wNith the bid documcntation rclating to carlier concessions which required the provision of a substantial amount of data from prospcctix L conccssionaires which the granting authonty was ill-prcpared to evaluatc. Howevcr, bidders for the Buenos Aires concessions received a comprehensive concession contract dctalling the amotint and schedule of required investmcnt, the required servicc level and the risk-shanng arrangcments between the PRIVATE PARTICIPATION IN INFRASTRUCTURE IN CHINA 119 Government and the concessionaire. The only bidding criterion was the lowest toll offered. By using a single, unambiguous criterion, the awarding process became more transparent, and un- necessary complications resulting from tradeoffs between offers on multiple criteria by competing bids were also avoided. The problems of over burdenig central authornties have been seen elsewhere. In Thailand, investments amounting to more than US$25 million must be approved by central Government with invcstments of more than US$125 million being subject to Cabinet approval. The requirement that the Thai Government must receive a detailed report and analysis of proposed investments leads to a protracted procurement process which increases the costs for bidders and slows down the implementation of projects. An addinonal problcm has arisen in Thailand as a result of overlappmng responsibilities among the vanous agencies responsible for toll road planning, implementation and regulation. The Depart- mcnt of Highways, as well as each of the Rapid Transit Authority and the State Railway of Thailand, have established overlapping toll road development plans. Thius has led to the preparation of separate, somctimes competing, toll road projects. Clearly the procurement process would benefit from the estabhshment of an overriding Government agency or regulatory body, but it is difficult to persuade cxistng bodies to relinquish their powers. This is not a problem that is unique to Thailand. Central Coordinating Bodies In some countries bodies have been set up to help facilitate consistency and an improved procure- ment process. No perfect model can be identified and, of course, social and political factors must be taken into account when such bodies are established. Nevertheless, the existence of a centralized co-ordinating body can make the procurement process more efficient from the perspective of both the public and the private sector. In the UK a body call the Treasury Taskforce was established in 1997, principally to help the public sector become better able to procure, manage and implement transactions m partnership with the private sector. In the Philippines, the Coordinating Council for Private Sector Participa- tion (whuch embraces the Filipino BOT Center) is tasked with co-ordinating, monitoring, facilitat- ing and promotng infrastructure projects in the Philippines. Its form and function is not quite the same as that of the Treasury Taskforce in the UK, but is has nevertheless played an important role in the procurement process in several projects. The existence of the Treasury Taskforce as a co-ordinating body for the PPI sector within central Government acted as an effective stimulus for the growth of this sector in the UK. It is important to emphasize that, in the UK, local authorities and individual Government departments generally have sufficient resources to be able to evaluate bids and oversee a competitive bidding process. However, the Treasury Taskforce was established to support Government departments and other state entities involved in PPI, in particular by helping them test the viability of significant projects before the procurement process commenced and, thereafter, by providing advice and resources throughout the procurement process. The Treasury Taskforce helped central Government departments become more understanding of the requirements and expectations of the private sector. It succeeded in doing so because it is staffed by individuals who have been recruited from the private sector but operates at the heart of the public sector, on the side of the granting authorities. This contrasts with the traditional pool of Civil Service recruits in the UK (and indeed most other countries), i.e. well educated individuals who have no commercial training or experience of working in the private sector. Such was the success of the Treasury Taskforce in reinvigorating the PPI sector in the UK, that, following a review at the end of its two-year life span, it was decided to establish two bodies (Partnerships UK ("PUK") and the Office for Government Commerce ("OGC")), which have taken over the role of the Treasury Taskforce. The OGC is a distinct orgamzation within the UK Government that is intended to provide Government departments with a central resource of pro- curement skills. It is hoped that the OGC will provide a greater sense of direction in procurement 120 WORLD BANK WORKING PAPER and promote best practice within the public sector. One of the first tasks of OGC is to establish a common strategic framework within which all Government departments should conduct their procurement activities. PUK has been established as a separate company within which the UK Governmcnt will eventu- ally own only a 51% sharc. PUK will work in partnership with the public sector (but will charge fees for its services) and will help the public sector ensure it obtains the best possible deals in privately financed investment programs. There were criticisms of the Treasury Taskforce. For example, it was not sufficiently resourced to be involved in all significant projects in the UK and so its level of involvement and influence varied from transaction to transaction. The Filipino BOT Center acts as a central point for information dissemination and training. It forms part of the Coordinating Council for Private Sector Participation, which is tasked to coor- dinate, monitor, facilitate and promote infrastructure projects implemented under the Philippine Infrastructure Privatization Program or the Private Sector Participation Program. The Council is directly attached to the Office of the President and therefore operates at the heart of Government. The structure of the Council (and, within it, the BOT Center) has recently been reformed. Nevertheless, the functions of the BOT Center remain as follows. - Policy Advocacy (reviewing sectoral regulatory and policy guidelines and recommending reforms); - Project Development (assisting implementing agencies and local Government in formulating privatization programs); - Institution Building (providing education and training support to implementing agencics and local Government); - Marketing and Promotion (providing input into information and promotional materials relating to PPI); and - Monitoring (tracking projects and monitoring approaches to PPI in different sectors). The Center's staff includes many professionals, most-with graduate or post-graduate degrees in economics, accounting and business although, like the Treasury Taskforce (and PUK), they do not have the power to bind those departments and agencies which they advise. Nevertheless representa- tives from the BOT Center often attend meetings to evaluate prospective projects and, subsequently, may sit in on actual contract negotiations with private sector investors. This helps develop consis- tency in the procurement process, although the BOT Center has neither the resources nor the power to ensure that it is involved in all PPI projects. Nevertheless, it is estimated that the BOT Center has overseen the successful development of over US$20 billion in private sector investmenlt in the last decade. Standard Form Documentation The Treasury Taskforce has played an additional role in the UK PPI sector by assisting public bodies in the negotiation of documentation. It has attempted to develop a common approach to risk sharing across the PPI sector. Last year the Treasury Taskforce published detailed guidance oni contractual terms relating to privately financed projects. As well as providing specific recommended terms, the guidance contains detailed explanations of common provisions. A common theme is that risks should be taken by the party best able to manage them. The guidance is very ambitious because it is designed to be applied across different sectors. In Thailand, the Government used model power purchase agreements as part of an initiative to promote private investment in the power sector. Certain contractuial provisions became standard across the sector including in relation to the components of tanffs, force majeure and compensa- tion payments ansing out of early termination. The initiative was considered to be a success before it was stalled as a result of the East Asian financial crisis. However, existing private investors in the PRIVATE PARTICIPATION IN INFRASTRUCTURE IN CHINA 121 Thai power sector can take comfort that their contractual positions are generally no wvorse than those enjoyed by other investors in the sanme sector. It is important that investors are seen to be treated equally to encourage additional investment in the market. It is arguable that one way in which bid costs can be reduced and the procurement process shortened is for the granting authority to establish a standard form contract for prospective bidders to price. 1.1.4 Forrns of PPI There are three principal methods by wphtch PPI takes place in the couintries which wvere reviewed: joznt ventutres, privatization and BOT/concessions. In some circumstances (eg. powver projects), elements of all three methods can be seen. Joint Ventures The water sector has seen the most significant number of joint ventures, with there beng many more examples than in either the transport or power sectors (although there are examples of construction work being carried out by joint ventures in both these sectors). Joint venture partners may be either public sector bodies or state owned entities. Private sector entities may be less comfortable entering into contractual relations with the former because there is a perception that the political make up of the public sector joint venture partner may change every few years, reflecting changes in the political environment which may occur following elections. This has been an issue in Hungary where, since the Water Management Act (1995) sewage services can be performed by joint venture companies provided the local municipalities have a majority owner- ship. There are several examples of projects that have failed to be completed because the pnvate sector has been frustrated in its attempts to establish and agree strong contractual relationships with municipalities Nevertheless, in the Hungarian water sector there are several examples of private sector involvement in joint venture companies. For example an operating concession in Szeged. This structure is typical in the Hungarian water sector, withi the involvemiient of private investors being limited to operation and maintenianice contracts, rather than outright ownership. A further significant example of PPI through joint venture arrangements can be found in Manila (Philippines), where concessions have been granted to two joint venture private suppliers. This provides an example of PPI by way of a combination of a joint venture and a concession. Each concession is for 25 years and was awarded to separate foreign/Filipino consortia. Metropolitan Manila Waterworks and Sewage System ("MWSS") retains ownership of the fixed assets, but transfers operating and investment responsibility to the private investors. Standards of service and targets for investment are established in the concession agreements A regulatory office (which, oddly enough, is administratively unider MWSS) has been set up as an mdependent organ to regulate, monitor and enforce the concession agreements Privatization Since the 1995 Privatization Act, the Hungarian privatization program is now pnncipally under the auspices of the State Privatization and Holding Company (called "APV"), which is managed by a Board of Directors whose members are appointed by the Government. APV is authorized to exercise the state's ownership rights respect to assets in its portfolio, including the right to sell particular assets or companies. The Hungariani Government has substantially restructured and privatized the nation's electric- ity and gas industries. By virtue of a Government resolution issued at the end of 1994 which sets out certain privatization-related decisions, the legal ownerslhip of companies in the electricity sector 122 WORLD BANK WORKING PAPER (except the nuclear power plant at Paks) wvas removed from the state. Pursuant to this Government resolution, interests in all gas and electricity suppliers and generating companies have been offered to strategic investors. This approach contrasts with that in the UK where individuals could acquire shares that the Government offered for sale in gas and electricity suppliers and generating comparnes. In Thailand, the State Enterprises Corporation Act came into force in December 1999. This law facilitates the privatization of state-owned enterprises, by prescribing a mechanism for the conversion, either in whole or in part, of the state enterprises into limited companies with a share capital. The Prime Minister has overall charge and control of the execution of this Act. Under the Act, state enterpnses may be corporatized and established as one or more separate companies. Conversion can take place over time or can be done in a single proecss. The Government will guarantee the financial obligations of companlies that are established from state enterprises. Argentina started its privatization process from 1989 and has since privatized over 170 entities through sale or concession. The most notable example is the electricity sector where generation, transmission and distribution activities were separated vertically and sold off under very long con- cessions. Essentially, this method of privatization is a hybrid of traditional privatization where state assets are sold off irreversibly, and concessions which are discussed further below. The generation sector was broken up into twenty-five business units and sold separately to private owners. BOT/Concessions Concession structures are common in Hungary and the Philippines where the PPI specific legisla- tion discussed above has established a framework pursuant to which state assets are transferred to private sector operators on a temporary basis. In the UK, the Private Finance Initiative ("PFI") and Publhc Private Partnerships ("PPPs") programs establish schemes under which the private sector delivers services which would tradition- ally be delivered by the public sector in return for the payment of fees by the state to the private sec- tor operator. For example, a private sector operator will tale over the operation and maintenance of a road or the buwlding services requirements of a school or a hospital and will be paid a fee by the State for doing so. In some PFI/PPP structures in the UK, the private sector operator will pay the state a conces- sion fee for having the right to operate an asset that was formerly owned by the state/local authority entity. This is the structure of the Luton Airport concession under which the orivate sector operator has taken over the operation and management of London Luton Airport and wNas under an obligation to build a new passenger terminal. In Brazil, "privatization" tends to be carried out through concessions. Under Brazilian law, public services must always remain the subject of a concession in some form and for a determined period. Even if the state sells some or all of its interest in a public service company, the purchaser can only acquire the relevant company on a concession basis, so that ultimately the assets will return to state ownership. In Malaysia, BOT method is used as one form of privatization. It is often used for greenfield projects, in particular, road and highways projects. Under such structure, the private sector would construct the facility using its own funds, operate and maintain the facility fer a period and eventu- ally, transfer the facility to the Government at the end of the period. During the concession period, the private sector is allowed to collect tolls from the users of the facility. 1.1.5 Financial Issues Tartff Approval Regimes There is no single method for tariff approval and adjvstmnent that stands ont Ps being appropriate across all sectors PRIVATE PARTICIPATION IN INFRASTRUCTURE IN CHINA 123 Of the countries reviewed, three principal tariff approval regimes can be identified - Tariffs set by Government departments, agencies or regulators; - Tariffs set by market forces; - Tariffs established according to a contract. Characteristics and examples of each of the above methods are discussed further below. Tariffs set by Government departments, agencies or regulators. This process for determining tariffs can be seen in the UK water sector as well as the Brazilian energy sector. In the UK, the water regulator (which is independent from Government) fixes tariff increases based on a formula which relates to increases in inflation. Ten regional water compan-ies that are effectively monopolies xvithin their local areas dominate the water sector. The water regulator uscs his powers to fix prices as a means of creating a regime of "comparative competition." Under this regime, water companies are required to provide information on all aspects of their business from water quality to capital expenditure. Comparisons are made between the water companies and pressure is put on those companies who do not compare well by restricting price increases until improved targets are attained. In Brazil, the national energy authority ("ANEEL") establishes general criteria pursuant to which, in most circumstances, electricity tariffs must be deternined The national oil authority ("ANP") has a similar role to ANEEL within the oil sector Both agencies have the right to approve revisions and adjustments to tariffs and applications must be submitted to them before changes to tariff levels can take effect. In some circumstances the agencies will actually set tanff levels. Tariffs set by market forces. The most common example of tariffs being set by market forces are "pool" pricing methods established in the electricity sector In an emerging market it is difficult to allow tariff levels to be determined by market forces whilst, at the same time, giving pnvate investors confidence with regard to future returns. In Argentina, since 1992, there has been open access to the wholesale capacity and energy pool for generating facilities, and least-cost centralized dispatch. Transactions between sellers and large customers are normally based on long-term contracts, and bulk power transactions between generators are normally made at spot prices To fulfil their long-term contractual obligations, generators can use their own energy or purchase energy from other generators at freely negotiated prices, or from the pool at spot prices. Generators receive two types of payment from the pool: a payment for energy dispatched and a payment for capacity offered to the grid. They are paid only the market value of their output, i.e. full cost recovery is not guaranteed. Prices in the UK electricity sector are currently determined under a "pool" system. There have been difficulties with the UK's "pool" system and a new electricity tradmg regime is set to be introduced in Autumn 2000 which wlvl, in most circumstances, replace the "pool." In Hungary, a degree of liberalization of the energy market is being established, pending full liberalization pnor to actual EU membership. Prices establtshed according to a contract. It is common to establish initial tanffs and formulae pur- suant to which tanff levels are increased in concession agreements. This is particularly trLe in the transport sector and the table set out below provides some examples of how tariffs are established in road projects in particular countries. It should be noted that, whilst anniual tariff micreases may be provided for in concession agreements, it is often the case that a regulator or Government department must approval the initial tariff level and the formula pursuant to which such tariff is increased. 124 WORLD BANK WORKING PAPER _4 'S... * *031 * .0 p 0 Country Tariff Regime Comments Philippines Established in Concession Contract. - the revised tolls must be published before Revised annually based upon a formula they become effective and interested toll which takes into account inflation, road users can challenge the reasonableness interest rates, exchange rates and the of the revised tolls; cost of construction materials. Tolls are collected by the operator and - if toll road investors happen to receive a are based on usage. windfall, they keep all the upside. Indonesia Initial tariff levels require Presidential - there is no guarantee that the tariff approval. adjustments every 2/3 years will be Concession company proposes tariff approved; adjustments every 2/3 years based on a - uncertainty over toll approval and adjusted formula (which is linked to inflation). procedures is not attractive to private Tolls are collected by the operator investors. based on usage. United Kingdom Under the latest projects, operators - approximately 75% of the tariff is based on receive shadow tolls based on availabil- availability, therefore the demand risk is ity of the road and usage. Payments are limited (and financing is easier); made by the UK Highways Agency to - tolls are not collected directly from road the operator. users, although contracts contain provisions for this to occur at a later date at the option of the U.K. Highways Authority (with mechanics that ensure the operator's overall income is not prejudiced). In the electricity sector, powver purchase agreements usually establish a capacity fee (to cover a generator's fixed costs) and an energy fee which escalate according to a formula set out in the agreement. Invariably, regulators must approve the tariff levels. For example, in Hungary, energy payments are set on the basis of recommendations made by the Hungarian Energy Office which are implemented by Ministerial Resolutions, with the Ministry of Economic Affairs having overall control. Similarly, in the Philippines, whilst the formulae for both the capacity fee and the energy fee in power purchase agreements include mechanisms for periodic adjustments based on fluctuation based on certain indices relating to, inter alia, the peso/foreign currency exchange rate and the export price indices, the Energy Regulatory Board has overall control of pricing in the electricity sector. Whatever method is used to select and approve tariff levels, it is important that the method is robust against legal challenges. If this is not the case, prospective private investors will be wary. This situation arose in Hungary when the toll rates established in the MI/M15 motorway project were challenged in the courts for being too high. Under the Hungarian Civil Code, a judge can "adjust" a contract in "exceptional" circumstances in which the price is considered disproportionate to the service provided. Toll road projects developed by the private sector in Thailand have also been hindered by uncertainty regarding toll rate adjustments, the availability of toll-free alternative routes, and other contractual provisions. It seems apparent that such major provisions of concession agree- ments need to be fully disclosed and discussed from the outset-before the issue becomes a political crisis. PRIVATE PARTICIPATION IN INFRASTRUCTURE IN CHINA 125 Regulations on Rates of Return Cou7ntries have adopted a variety of approaches to regulate the rates of return available to prtvate investors. In order to help stimulate private investnent in Hungary, the Government promised intial private investors m the cnergy scctor a munmum aninual return of 8%, although subsequent investors were not given the same guarantees. Indeed, in 1996 a Hungarian Government resolution required half the amount of any profits of electncity supply companies that exceedcd 12% of shareholder funds to be returned to consumers. In the Philippines, the BOT law establishes a regime for regulating the rate of return permitted under contracts entered into pursuant to that legislation Investors are permitted a "reasonable rate of return" which, in respect of public utility projects which are operating on a monopoly basis, is restricted to an annual rate of return of 12%. In addition, it is not uncommon for the charters of some of the publicly owned corporations that are formed to develop and operate infrastructure faclities to contain ceilings on thcir rate of return. For example, the charter of MWSS restricts its rates of return to 12% of the aggregate amount of its assets in operation (revalued from time to time) and two months' operating capital. Accordingly, where Filipino PPI projects are structured as a concession by or a joint venture agreement with such type of publicly owned corporations wvhereby the sponsors are to receive payments directly from the users, the sponsors' return will be indirectly subject to the samc ceiling contained in their grantors' or partners' charters unless there are specific exceptions or top-up arrangements In some of the early transportation projects in the UK (e.g. the Second Severn Bridge Cross- ing), the rate of return of private investors was restricted by providing for a flexible concession period. The concession period would terminate once the private investor had repaid all its debts and had provided a pre-agreed return on equity to the project sponsors. This approach has only been used ui the few examples of transportation projects in the UI (all tunnels and bridges) where the operator is entitled to collect tolls directly from users. Such an approach is not used in current PPI projects in the UK, although in the road transport sector granting authorities are under a broad obligation to secure "value for money" for the public sector. The UK's Nanonal Audit Office ("NAO") has published criteria against wvhlch value for money can be assessed during the fcasibility stage of PFI projects, to help granting authorities assess whether a project will secure valuc for money if responsibility is passed to the private sector In Brazil, as the conccpt of public service guarantees citizens basic and affordable services in the community, granting authorities must assure citizens the principle of "value for money." In addition, the law provides for the maintenance of the initial economic anid finanicial equilibrium of the concession agreements, i.e., the economic and financial conditions of the parties upon signing must prevail through any ftiture changes, save wxhere changes rcsult as a consequence of the normal risk of the business assumed by the private sector. Foreign Exchange Issues Foretgn Exchange considerations are clearly more significant when the private investors are not resident zn the country where the investment is taking place. Investors witll be keen to ensure that they can receive dividends and make repayments under loan agreements tn a straightfbrvard manner. Some countries, such as Hungary, have introduced legislation to provide safeguardsfor private investors. Some countries have established favourable regimes wvith respcct to foreign exchange regulations. For example, in Hungary, m addition to a series of bi-latcral investment treaties, the Foreign 126 WORLD BANK WORKING PAPER Investment Act provides that any entitlement of a forcign investor to aftcr-tax profit, dividends, royalties, fees, proceeds of sale of sharcs or quotas, proceeds in the net assets of the company upon liquidation after discharging its liabilities, or considcration for the transfer of sharcs, may be freely transferred abroad by the foreign investor in the currency in which the original investment was made (provided sufficicnt cash is availablc to be repatriated). Indccd, Hungarian foreign exchange policy is generally considered to be compliant vith standards established by the International Monetary Fund and the Organization for Economic C'orporation and Developmncrt-there are no onerous foreign exchange requiremcnts and no known instanccs of delay irn repatriation. This creates an attractive regime for foreign invcstors. Other countries havc more stringent registration and approval rcquirements. For example, in the Philippines, forcign investments must be propcrly rcgistered with Bangko Sentral ng Pilipinas ("BSP") to ensure that the investor wil' bc entitled to full and immediate repairiation of capital and remuttance of earnings. Once registration with BSP has occurred, foreign currcncy loans can be serviced by foreign exchange funds available witlin the Filipino banking systcm. The Foreign Exchange regulations in Brazil are morc complex (and therefore less attractive to perspective pnvate invcstors). All forcign capital being introduced into Brazil must bc registered with the Banco Central do Brasil ("Baccn") (i.e. the initial procedure is not dissimilar to that in thc Philippines). Howevcr, in order to repatriate capital, approval must be obtaincd from Bacen wvithin 30 days of the proposed repatriation. T'he transaction can only proceed with a certificate of approval provided by Baccn. 1.;1.6 Project lnmplementation: Thc, eriod Afte th.e Preferred Bidder is Selected In the Phtlippines an oiganizAtziot has been establish;;zeut to help investors thr;oqh the itplementa- tion process. In some other counittries, suiclh as the UIK, other approaches have been adopted to make the implementation process ore efficient In Australha, Invest Australia provides essistanice to certain i qUalifyzizg projects and providesfor a co-ordinated i;nplewzentatioi process. The approval process in Thailand is particularly tircsome for private investors wvith a significant amount of approvals required from a variety of bodies who all have thcir own agendas to satisfy. Often private investors have found that Thai public agencies havc ovcrlapping rcsponsibilities. In Brazil, whilst fewer consents and authorizations tend to be required in comparison with Thailand, the legislation which needs to be complied with by successful bidders is diverse and pnvate investors must spend time ensuring that all rclevant statutes and regulations have been complied with. This is in part a consequencc of the structurc of Brazilian Government which provides significant legislative powers at both national and regional (state) level. In the Philippines, a body callcd the One Stop Action Ccntcr ("OSAC") has been established to help privatc investors understand the process of obtaining consents and permits. Under one physical location, the OSAC houses different Govcrnment agency representatives tasked to accept and process mvestment applications. The reprcsentativcs havc the authority to act on all investmcnt matters under their jurisdiction-this is significant m ensuring the effcctivcness of the OSAC. Therefore, the OSAC allows the investor to get a clcar view of the consent process so that this aspect of the implementation phase of a project can be managed morc efficicntly. In Australia, the Federal Governmcnt, through Invcst Australia, providcs a Major Projects Facilitation ("MPF") service to invcstors with advice and information on rcquirements for relcvant Federal approvals and facilitation of investment projects granted MPF status. A project is eliglble to be granted MPF status if it involves a total capital expenditure of over $A50 million, requires Federal Government approval to progress and if the prospective developer can demonstrate the project's commercial viability and readiness to proceed through the approval process. PRIVATE PARTICIPATION IN INFRASTRUCTURE IN CHINA 127 If MPF status is granted to a project, Invest Australia wvil identify relevant Federal approvals and prepare a system of comprehensive, cross-jurisdictional approval timelines for the completion of the approval processes at all levels of Government. It will also coordinate Federal and State/ Territory Government comphance requirements and processes so they progress simultaneously as far as possible to be consistent with commercial requirements. It will identify Government policies or programs which may benefit a project. Suggestions have been made to the Federal Government to make Invest Australia's MPF scrvices available to all trans-jurisdictional projects irrespective of whether or not it satisfies the "national" criteria. One way in which the consent and authorization process can be simplified is for all relevant consents to be provided under a single statute. This process is used in some Australhan states and was used in some of the early transport projects in the UK, with the Government passing legislation which avoided the private sector having to make separate applications to relevant agencies and local authorities (including in relation to the compulsory purchase of land). 1.1.7 Security Issues In some countrtes, such as the UIC, a full range of security rights can be granted to fund providers, including floating charges. In other countries, such as Thailand, taking security is a more complex process. Generally, the Benchmark Countries have security laws in place which are more wide reaching and effective than those in China The secunty regime in England and Wales (there is a slightly different regime in Scotland) is perceived to be lender friendly. Registration and enforcement are straightforward and there is a well-established regime incorporating a full range of security instruments, includmg the floating charge. In addition, lenders are able to acquire step-in rights. In Hungary, it is possible to take security over most assets and reforms have been introduced in recent years to improve the processes for taking and enforcing security. For example, since 1997, floatng charges and mortgages over movable property have been possible and, last year, a new registration system for these mortgages was established. Lenders are also able to take security assignments over project contracts. Most commentators believe that step-in rights are possible under Hungarian Law and they are generally provided to lenders to PPI projects. However, the effectiveness of such rights has not yet been tested before the Hungarian Courts. Taking security is a little more cumbersome and problematic in the Philippines and Thailand. For example, in the Philippines, the concept of a security assignment of contractual rights is not recognized under Filipino lawv. However, structures are possible which can provide some comfort to funders. For example, it is quite common to take an absolute assignment of the income stream to PPI projects (e.g. payments made to an IPP under ani energy conversion agreement) and a security assignment (which will have to be governed by the relevant foreign law) of the other contractual rights under the principal Philippine project documents. In Thailand, state-owned property cannot be secured and the concept of a floating charge is not recognized. In addition, it is extremely time consuming, cumbersome and expensive for a syndicate of banks (which include foreign banks) to take a mortgage in Thailand. Each bank will have to exhibit evidence of its own constitutional documents an-d its power to take the secunty. 1.2 FINANCING ISSUES 1.2.1 Investor Base Our research evidenced that the development of the domestic capital market and the enhancement of the local investor base for PPI are convergent. 128 WORLD BANK WORKING PAPER The following table summarizes the sources of private equity for infrastructure in some of the Benchmark Countries: Foreign (Volatile) Local (Stable) Developers (Stable) Philippines Malaysia Australia Pre-crisis Thailand Argentina Thailand Financial Investors (Volatile) Malaysia Australia In the Philippines, Argentina and Australia, foreign developers are predominant in PPI, although it is interesting to remark that in Australia, many domestic institutional investors, either banks or funds such as the Infrastructure Trust of Australia, have also been active. The Philippines paid the price of its reliance on foreign investment through generous foreign exchange guarantees. In Malaysia, on the contrary, investors are predominantly local and comprise conglomerates (such as United Engineers Malaysia/Renong or YTL), Government-related companies (Tenaga), financial institutions and the Employee Provident Fund (EPF). The EPF was created in 1951 in order to establish a social security and pension system for employed workers. The fund receives its contributions through a compulsory savings scheme, both from the insured person and the employer at defined rates (respectively 11% and 12% of payroll). Its total assets amounted to RM 46 billion in 1990, RM 97 billion in 1995 and RM 164 billion in 1999. When the Government had accumulated a financial surplus, it liberalized restrictions on EPF portfolio, whereby amount of funds to be invested in Malaysia Government Securities used to be 70%. The EPF has then invested approximately 14% of its assets in corporate securities. These invest- ments have included infrastructure projects, such as the Sikap Power project, the Kuala Lumpur International Airport, the Light Rail Transit, the YTL power project, the Lumut power project and the North-South Expressway. Another noticeable initiative in Malaysia was the introduction in 1996 of a special listing category on the KLSE dedicated to infrastructure. The Infrastructure Project Companies (IPC) meeting a profile including total costs, duration and acceptable rate of return could list their vehicles on the KLSE with exemption from the standard listing requirements on profit track records. In Thailand, since 1994, a US$ 150 million Infrastructure Fund has been operational. It is funded by the Thai Government; USAID and domestic and foreign financial institutions. In the early 1990s the Governlment also initiated a program of IPOs of its largest utilities. The move was considerable as it contributed to increasing the size of the stock market by 20%. Interestingly, local sponsors (such as Ch. Karnchang or Tanayong) progressively took the lead in infrastructure projects. The eagerness of the banking sector to lend and the booming capital market played a role of major catalysts in this evolution. This came to a halt when the Asian crisis occurred in 1997. Following this, foreign developers have naturally taken over from the financially distressed local conglomerates as the biggest investors. In Argentina, the US$ 500 million Capital Market Development Backstop Facility (World Bank Loan) aimed at encouraging longer term lending by commercial banks took off in 1995 and was cancelled in 1997 due to the: (i) risk inherent in rigid and narrowly defined products in a PRIVATE PARTICIPATION IN INFRASTRUCTURE IN CHINA 129 rapidly moving and volarile developing economy, (ii) risk of inadequate management of facilhty, (fi) risk of weak government ownership and (iv) risk of perverse signaling effects In Hungary, foreign developers are predominant in PPI Following a successful tender in 1995, strategic foreign investors acquired majority interests in all five of the regional gas distribution comparies and minority interests in the six regional electricity companies. Four of the seven electricity generation companies were also sold to foreign investors durnng 1995/1996 Investors included RWE Energie of Germany, EDF of France and IVO of Finland. 1.3 DOMESTIC LENDING Argenttna and the Philippines provide examples of countries wvhere the banking system has been deregulated without creating additionalfitndsfor infrastruictutre projects. Malaysia and Thailand are two examples of how a buoyant domestic loan market can be a healthy source for domestic infrastrutctutre projects Finally, at the extremity of the spectrum, the mature Australtanfinancial system, wvhere the role of the Central Bank is restricted to ensutring its stability only, has proved to be the strongest solurce of fundtng infrastructutre development requtirements. For the Benchmark Countries it is more usual to observe a direct link betveen the degree of development of the domestic capital market and the local financilg portion of a project. Country Local debt portion Australia 100% Malaysia 94% Thailand 72% China' 34% Philippines 1% 'Local portion of debt financing for power projects with private participation Today, in Argentina, the Philippines and Thailand, offshore funds continue to provide the core of PPI debt, in the form of MLA, ECA and foreign bank loans. This signals a lack of local capacity to ftund PPI. However, access to local debt in pre-crisis Thailand was casy, making it one of the most promi- nent Asian countries where financing was predominantly domestic (for example, the power sector was ftinded for 72% by domestic loans). This ended wvith the onset of the Asian crisis in 1997, where both the local capital market and the local mvestor base vanished At the other end of the spectrum, in Malaysia and Australia, debt for private projects, either developed by local or foreign sponsors, has traditionally come from the domestic market. In Australia, the highly deregulated bankinig system has satisfied the totality of demand for loans. As foreign banks have gained access to retail banking in as early as the 1980s, it is not surprising to observe that a significant amount of projects domestic debt is sourced from these foreign banks' Australian branches. Long-term financing is generally provided by syndicatcd loans from Australian and foreign banks Domestic lending is becoming increasing common in PPI projects in Hungary, although most of the banks involved in PPI have been privatized and some are foreign owned. 130 WORLD BANK WORKING PAPER 1.4 BOND FINANCING Countries without prominent institutional investors (such as the EPF in Malaysia) seeking long- term stable revenues and deep local bond markets cannot provide locally the suitable funding instruments to infrastructure projects. In the Philippines and Argentina, projects that were bond financed have resorted to international bond issues. In the Philippines, international bonds usually enjoy sovereign ratings due to financial guaran- -tees from the Government. For instance, in the case of the US$ 371.5 million Rule 144a high-yield offenng for the CE Casecnan greenfield hydroelectric power project, a Performance Undertaking supports the off-taker's obligations and project revenues are denominated in USD Quezon Power US$ 215 million senior notes enjoy the ratings assigned to the Government as privately held off- taker Manila Electric Company ("Meralco") financials are deemed solid. Newly privatized Argentine utilities have accessed the US bond market when conditions were favorable. New stakeholders have often refinanced their acquisitions through Rule 144a offering. In Malaysia, Thailand and Australia, the more advanced stage of development of the local bond markets or the existence of strong institutional investors has allowed an effective channeling of domestic savings to infrastructure projects. In Malaysia, the bond market has been developed through the establishment of Cagamas in 1986, the role of which is to repackage housing loans to improve banks' liquidity. On the investor side, the EPF (see Investor base above) has provided bond financing to a plethora of infrastructure projects. For instance, it subscribed to the 10-year RM 1,500 million bonds issued by the YTL power project (rated AA3 by RAM), constituting almost half of the debt raised by the project. Issues Outstanding as of 1999 Country Government Corporate Largest investors China 110 3.5 Banks and financial institutions Malaysia 50 8.2 Non-banking institutions (EPF, Social Security, etc.) Philippines 22 1.3 Government financial institutions (Social Security, etc.) Thailand 25 4.4 Commercial banks All amounts in billion USD In Thailand, financing of the Rayong project involved the local private placement of 12-year secured debenture for B 3.5 billion (US$ 140 million) while another US$ 140 million was raised through a US private placement of bonds with a 15-year door to door maturity. Bond financing of PPI projects in the UK is well established, in particular in respect of projects financed under the UK Government's Private Finance Initiative such as roads and hospitals. The first PFI project to be financed on the bond market was the Al (M) Motorway Project. In this transaction, the bonds were offered with the benefit of a guarantee provided by a AAA-rated monoline insurance company. In the UK, the bond market is increasingly being seen as providing a means for refinancing a project following the completion of the construction phase, once construction risk has been removed. PRIVATE PARTICIPATION IN INFRASTRUCTURE IN CHINA 131 1.5 TRANSPARENCY OF FINANCIAL INFORMATION Countries where capital markets have been developed up to international norms have enacted laws aimed at protecting individual investors notably by ensuring listed companies provide sufficient and timely (often quarterly or at least semi-annually) information, the quality of which is set by national accounting standards. It is clear that the lhsting on transparent stock markets of utilities such as the Manila Electric Power Co in the Philhppines or Tenaga Nasional Berhad in Malaysia provide more comfort to investors and lenders. The example given above of the Quezon power project in the Philippines demonstrates the extent of comfort a financially strong off-taker such as Meralco can bring to lenders by itself w ith- out having to rely on further Government guaranitees. 2 Sector-specific Issues 2.1 ROAD SECTOR 2.1.1 Extent of Private Participation While Argentina and Australia have largely przvatized existing roads and railways, Southeast Asian countrtes have,generally relied on PPI tofinance the development of their infrastructure. Argentina and Australia undertook programs of privatization of their mature road networks, rail- ways and aiLrport operations and have seen the development of some greenfield projects. Asian countries generally chose to rely on PPI to build the new transport infrastructure required by their faster growvth. In 1991, the Argentine Government, lacking funds to properly maintain and rehabilitate its road netwvork, simultanieously awarded twelve 12-year concessions of inter-city highways via com- petitive bidding. Bidding documents required the consortia to achieve prescribed service levels within a contracted timeframe and to assume existing investment proposals. The decisive factor for selection was an up-front fee payable to the Government on award of the concession. After 1992, the Government awarded four 22-year concessions, includtng some greenfield projects, for access roads to Buenos Aires. This time, the sole cnterion was the lowest toll offered under serviceability constraints. Argenuna's railway privatization is noteworthy to the extent it separated the freight services and commuter transportation into different concessions. VVhile the freight concessions were sup- posed to be profitable, the Buenos Aires suburban lines were subsidized The selection factor was the lowest amount of subsidy required In Austraha, privatization of Qantas (1995) and 17 airports (1997) raised US$ 3.3 billion. Sig- mficant financing concessions such as infrastructure bonds were key to private sector involvement. Road concessions also raised some US$ 2.4 billion. The pnncipal methods of transfer to the private sector were by way of public flotation and trade sales. In Southeast Asia, most toll road concessions are BOT. In Malaysia, a project can be identified by the Government in the Highway Network Development Plan or unitiated by the private sector. In either case, the concessionaire is usually selected following competitive bidding. There are many examples of privately run DBFO ("design build finance operate") roads in the UK and the model for these concessions is being copied in several other companies, such as South Africa and the Republic of Ireland. Private sector operators generally acquire an existing road and are required to carry out certain upgrading work or construct new sections of road and are subsequently permitted to operate the road under concessions which are usually 30 years long. In the UK, road operators are paid on the basis of shadow tolls (based on actual usage and availability) and the Government is comfortable that the revenue it must pay the pnvate sector operator based on the shadow tolls will amount to less than the Government would have to pay 132 WORLD BANK WORKING PAPER were it to carry out the necessary upgrading work and ongoing maintenance ovcr the 30 year period itself. 2.1.2 Greenfield Projects Versus lBrownfleld PzoJe