WORLD BANK DISCUSSION PAPER NO. 343

~~3Eq)~ ~ ~ ~

Work in progress 9,

for public discussion

Financing PrivateInfrastructure inDeveloping Countries

David FerreiraKanran Khatami

Pub

lic D

iscl

osur

e A

utho

rized

Pub

lic D

iscl

osur

e A

utho

rized

Pub

lic D

iscl

osur

e A

utho

rized

Pub

lic D

iscl

osur

e A

utho

rized

Recent World Bank Discussion Papers

No. 272 Adolescent Health: Reassessing the Passage to Adulthood. Judith Senderowitz

No. 273 Measurement of Welfare Changes Caused by Large Price Shifts: An Issue in the Power Sector. Robert Bacon

No. 274 Social Action Programs and Social Funds: A Review of Design and Implementation in Sub-Saharan Africa. AlexandreMarc, Carol Graham, Mark Schacter, and Mary Schmidt

No. 275 Investing in Young Children. Mary Eming Young

No. 276 Managing Primary Health Care: Implications of the Health Transition. Richard Heaver

No. 277 Energy Demand in Five Major Asian Developing Countries: Structure and Prospects. Masayasu Ishiguro and TakamasaAkiyama

No. 278 Preshipment Inspection Services. Patrick Low

No. 279 Restructuring Banks and Enterprises: Recent Lessonsfrom Transition Countries. Michael S. Borish,Millard F. Long, andMichel Noel

No. 280 Agriculture, Poverty, and Policy Reform in Sub-Saharan Africa. Kevin M. Cleaver and W. Graeme Donovan

No. 281 The Diffusion of Information Technology: Experience of Industrial Countries and Lessonsfor Developing Countries. NagyHanna, Ken Guy, and Erik Arnold

No. 282 Trade Laws and Institutions: Good Practices and the World Trade Organization. Bernard M. Hoekman

No. 283 Meeting the Challenge of Chinese Enterprise Reform. Harry G. Broadman

No. 284 Desert Locust Management: A Timefor Change. Steen R. Joffe

No. 285 Sharing the Wealth: Privatization through Broad-based Ownership Strategies. Stuart W. Bell

No. 286 Credit Policies and the Industrialization of Korea. Yoon Je Cho and Joon-Kyung Kim

No. 287 East Asia's Environment: Principles and Prioritiesfor Action. Jeffrey S. Hammer and Sudhir Shetty

No. 288 Africa's Experience with Structural Adjustment: Proceedings of the Harare Seminar, May 23-24, 1994. Edited by KapilKapoor

No. 289 Rethinking Research on Land Degradation in Developing Countries. Yvan Biot, Piers Macleod Blaikie, Cecile Jackson,and Richard Palmer-Jones

No. 290 Decentralizing Infrastructure: Advantages and Limitations. Edited by Antonio Estache

No. 291 Transforming Payment Systems: Meeting the Needs of Emerging Market Economies. Setsuya Sato and David BurrasHumphrey

No. 292 Regulated Deregulation of the Financial System in Korea. Ismail Dalla and Deena Khatkhate

No. 293 Design Issues in Rural Finance. Orlando J. Sacay and Bikki K. Randhawa

No. 294 Financing Health Services Through User Fees and Insurance: Case Studiesfrom Sub-Saharan Africa. R. Paul Shaw andMartha Ainsworth

No. 295 The Participation of Nongovernmental Organizations in Poverty Alleviation: The Case Study of the Honduras SocialInvestment Fund Project. Anna Kathryn Vandever Webb, Kye Woo Lee, and Anna Maria Sant'Anna

No. 296 Reforming the Energy Sector in Transition Economies: Selected Experience and Lessons. Dale Gray

No. 297 Assessing Sector Institutions: Lessons of Experiencefrom Zambia's Education Sector. Rogerio F. Pinto and Angelous J.Mrope

No. 298 Uganda's AIDS Crisis: Its Implicationsfor Development. Jill Arrnstrong

No. 299 Towards a Payments System Lawfor Developing and Transition Economies. Raj Bhala

No. 300 Africa Can Compete! Export Opportunities and Challengesfor Garments and Home Products in the European Market.Tyler Biggs, Margaret Miller, Caroline Otto, and Gerald Tyler

No. 301 Review and Outlookfor the World Oil Market. Shane S. Streifel

No. 302 The Broad Sector Approach to Investment Lending: Sector Investment Programs. Peter Harrold and Associates

No. 303 Institutional Adjustment and Adjusting to Institutions. Robert Klitgaard

No. 304 Putting Institutional Economics to Work: From Participation to Governance. Robert Picciotto

No. 305 Pakistan's Public Agricultural Enterprises: Inefficiencies, Market Distortions, and Proposalsfor Reform. Rashid Faruqee,Ridwan Ali, and Yusuf Choudhry

No. 306 Grameen Bank: Performance and Stability. Shahidur R. Khandker, Baqui Khalily, and Zahed Khan

No. 307 The Uruguay Round and the Developing Economies. Edited by Will Martin and L. Alan Winters

No. 308 Bank Governance Contracts: Establishing Goals and Accountability in Bank Restructuring. Richard P. Roulier

(Continued on the inside back cover)

WORLD BANK DISCUSSION PAPER NO. 343

Financing PrivateInfrastructure inDeveloping Countries

David FerreiraKamran Khatami

The World BankWashington, D.C.

Copyright X) 1996The International Bank for Reconstructionand Development/THE WORLD BANK1818 H Street, N.W.Washington, D.C. 20433, U.S.A.

All rights reservedManufactured in the United States of AmericaFirst printing December 1996

Discussion Papers present results of country analysis or research that are circulated to encouragediscussion and comment within the development community. To present these results with the leastpossible delay, the typescript of this paper has not been prepared in accordance with the proceduresappropriate to formal printed texts, and the World Bank accepts no responsibility for errors. Some sourcescited in this paper may be informal documents that are not readily available.

The findings, interpretations, and conclusions expressed in this paper are entirely those of the author(s)and should not be attributed in any manner to the World Bank, to its affiliated organizations, or tomembers of its Board of Executive Directors or the countries they represent. The World Bank does notguarantee the accuracy of the data included in this publication and accepts no responsibility whatsoeverfor any consequence of their use. The boundaries, colors, denominations, and other information shownon any map in this volume do not imply on the part of the World Bank Group any judgment on the legalstatus of any territory or the endorsement or acceptance of such boundaries.

The material in this publication is copyrighted. Requests for permission to reproduce portions of itshould be sent to the Office of the Publisher at the address shown in the copyright notice above. TheWorld Bank encourages dissemination of its work and will normally give permission promptly and, whenthe reproduction is for noncommercial purposes, without asking a fee. Permission to copy portions forclassroom use is granted through the Copyright Clearance Center, Inc., Suite 910, 222 Rosewood Drive,Danvers, Massachusetts 01923, U.S.A.

For a copy of Update describing new publications, contact the Distribution Unit, Office of the Publisher,The World Bank, 1818 H Street, N.W., Washington, D.C. 20433, U.S.A., or Publications, The WorldBank, 66, avenue d'Iena, 75116 Peris, France. A catalog and ordering information are also available on theInternet at http:/ /www.worldbar.k.org.

ISSN: 0259-210X

At the time this paper was written, David Ferreira was senior financial sector specialist n the WorldBank's Financial Sector Development Department. He is currently manager, Private Sector Investmentsand Cofinancing, Development Bank of Southern Africa, Midrand, South Africa. Kamran Khatami was aconsultant to the Financial Sector Development Department.

Library of Congress Cataloging-in-Publication Data

Ferreira, David, 1962-Financing private infrastructure in developing countries / David

Ferreira, Kamran Khatami.p. cm. - (World Bank discussion papers ; 343)

Includes bibliographical references.ISBN 0-8213-3743-21. Capital market-Developing countries. 2. Privatization-

Developing countries. 3. Investments, Foreign-Developingcountries. 4. Infrastructure (Economics)-Developing countries-Finance. 5. Bonds-Developing countries I. Khatami, Kamran,1964- . II. Title. III. Series.HG5993.F47 1996332'.0414-dc20 96-30149

CIP

Contents

Foreword v

Abstract vi

Chapter 1: Why Private Infrastructure? 1Economic importance of infrastructure IPublic financing of infrastructure 2Rationale for private provision of infrastructure 5

Public sector inefficiency 5Economic pricing and cost recovery 7Advances in technology 8Advances in regulation 9Need for private resources 10Potential infrastructure investment gap, 1996-2000 11

Chapter 2: Moving Toward Private Infrastructure 14Reforming public enterprises 14Privatizing public enterprises 15

Country experience 15Privatization and capital market development 17

Foreign investment in privatization 18

Chapter 3: Mobilizing International Financing for Private Infrastructure 20Need for efficient contact between government and investors 20International capital and project finance flows 22Raising long-term debt 23

Improving country creditworthiness 25Rule 144A: A window for long-term U.S. dollar debt 26Special investment and financing tools 26

Philippines' Private Sector Infrastructure Development Fund 27Colombia's Private Sector Infrastructure Financing Facility 29

Managing sovereign risk guarantees 31

iii

Chapter 4: Developing Local Financial Resources 36Implications of inadequate local financing 36

The inflation challenge 37Bond markets and infrastructure finance 38

Prospects in Latin America 39Emerging bond markets in East Asia 40

More-advanced markets 40Less-advanced markets 42

Role of credit rating agencies 43Developing pension funds for infrastructure finance 45Strengthening commercial banks to provide infrastructure debt 50

Chapter 5: Conclusion 54

Bibliography 57

iv

Foreword

In an effort to ease fiscal constraints and improve ture in Developing Countries David Ferreira andefficiency, the policy agendas of developing coun- Kamran Khatami look at the rationale for privatetries are increasingly concerned with promoting infrastructure and the challenges faced in financ-private development, financing, and ownership ing its development. In so doing, they drawof infrastructure. This move has created significant lessons from experience and suggest ways ofchallenges for governments who must change proceeding.policy, legal, and regulatory frameworks, design We hope that this paper will help policy-and implement privatization and concession makers in developing countries and others deal-programs, and negotiate a division of risk-taking ing with these challenges to understand thewith the private sector in infrastructure projects. issues they face and learn how some govern-

Given these complexities only a handful of ments have addressed them to date.countries, mainly in Asia and Latin America, have The authors would like to thank Robert Hillmade tangible progress in attracting private for his generous and invaluable input into sev-investment to infrastructure. Still, many lessons eral versions of this paper and Khalid Siraj forhave been learned. In Financing Private Infrastruc- his comments on an earlier draft.

Diana McNaughtonActing DirectorFinancial Sector Development DepartmentFinance and Private Sector DevelopmentThe World Bank

v

Abstract

Chapter 1 of this paper argues that public sector performance guarantees. Third, it will introduceresources in developing countries are insufficient commercial players into local debt and equityto finance the demand for investment in increas- markets, which will help develop the capacitiesingly integrated infrastructure services. More- that are most relevant to infrastructure financing.over, the public sector is less efficient than the In addition, allowing foreign investment and man-private sector in managing newly entrepreneur- agerial expertise to participate in the privatizationial infrastructure activities. With integrated glo- of public enterprises could fill resource gaps in thebal capital markets giving domestic and interna- rehabilitation and expansion of services and pro-tional investors a greater degree of choice about vide valuable experience from a deregulated orwhere they locate, the time has come for private differently regulated infrastructure industry.actors to provide what were once assumed to be Chapter 3 focuses on mobilization of interna-purely public services. This transition will be tional financing for new private investments.aided by new technologies and regulatory and Given the shortage of domestic savings in mostcontracting techniques that have created oppor- developing countries, international financingtunities for competitive provision in many areas will be important to the realization of plannedof infrastructure that previously appeared to be investment programs. Moreover, capital mar-natural monopolies. kets in most developing countries have insuffi-

Chapter 2 discusses the role of reforms in pub- cient depth to finance large private infrastructurelic enterprises providing infrastructure services. projects on a limited recourse basis. These coun-These reforms aim to enhance internal and exter- tries will, therefore, have to rely on foreign pri-nal incentives in public enterprises and, ulti- vate savings for some time to come. How readilymately, to privatize these enterprises. Where out- they can do so depends on the credit rating ofright privatization may be politically sensitive or both the host country and the project. Wheretime-consuming, corporatization could bear pro- sovereign credit ratings are weak, special cofi-ductive results in the short term and prepare the nancing arrangements can be set up betweeninstitutional and regulatory environment for a governments and multilateral institutions togreater private sector role in the financing and ensure that private investors have access to inter-provision of infrastructure. Restructuring and national capital. To attract foreign financing,privatization of public infrastructure services governments should address any negative riskwill help improve the investment climate in sev- perception that may affect foreign willingness toeral ways. First, it will expedite regulatory invest. Given the foreign exchange convertibil-reforms in the sector. Second, it will make the ity concerns of foreign investors and relativeenvironment facing private developers more unfamiliarity with local conditions, sovereignattractive and so reduce the need for government guarantees to mitigate perceived and actual

vi

nonmarket risks are especially important. (Cor- countries can create such a demand and channeltinued progress on macroeconomic and struc- large amounts of savings toward long-termtural reforms will gradually eliminate such con- investments in a relatively short period. Com-cerns, but appropriate guarantees may have to mercial banks are another important source ofbe provided in the interim. Since such guaran- Iinancing. These banks continue to hold the larg-tees can accumulate into significant contingent est share of financial assets among financial insti-liabilities, efficient financial accounting proce- tutions in developing countries and can contrib-dures need to be put in place in order to value ute to infrastructure finance in a variety of ways,and manage these liabilities. ranging from project appraisal, debt packaging,

Chapter 4 identifies the prerequisites for local and construction phase financing to longer termcapital market financing. Long-term bonds are lending and the underwriting and trading ofamong the most appropriate financing instru- securities used in structured financing.ments for infrastructure projects, which are gen- Finally, Chapter 5 outlines a set of approacheserally characterized by stable, long-term local and policy priorities that would underpin ancurrency revenue streams. Low inflation is a key efficient and expeditious transition from publicfactor in developing markets for these bonds. to private infrastructure. These conclusions areThe development of bond markets also requires derived from common threads that haveinstitutional investors who can take positions in emerged from success stories as well as from thelong-term financial assets that match the profile challenges faced by countries where the transi-of their liabilities. Pension reform in developing tion has already begun.

vii

1Why Private Infrastructure?

Over the past decade governments across the Economic importance of infrastructuredeveloping world have come to recognize theimportance of developing and maintaining ade- The economic literature has not reached a con-quate and efficient infrastructure services-as sensus on the link between infrastructure andwell as the severe implications of falling behind economic growth. To determine the direction ofin the provision of these services. Fast-growing causation, the link would need to be disaggre-developing countries are finding that their infra- gated for the various types of infrastructure,structure capacities are not keeping pace with since each infrastructure component has a differ-the growth in other sectors. Other developing ent impact on productivity. In any economy thecountries, further behind on the path toward productivity of labor and noninfrastructure capi-market reforms, face more severe constraints tal depends on the availability of adequate infra-because of past neglect in the creation and main- structure. 1 If a comprehensive package of infra-tenance of infrastructure facilities. These coun- structure services is viewed as essential fortries are now trying to develop policy and regu- underpinning the productivity of labor and capi-latory frameworks to ready their fledgling tal and facilitating growth, then the link andinfrastructure facilities to support economic direction of causation become even more diffi-reforms. cult to define. Moreover, the absence or inade-

As a result of this increased awareness, gov- quacy of one form of infrastructure may under-ernments are also looking increasingly to private mine the availability of another. This is becausesector resources and managerial capacity for the most infrastructure services are deliveredprovision of infrastructure services. Since 1984, through interlocking networks of investments86 industrial and developing countries have (roads and highways, gas and electricity distri-privatized 547 infrastructure companies worth bution networks, sewers and water pipes).$357 billion, and at least 574 private new invest- Unlike private investment in plant and equip-ment projects, worth $308 billion, are under way ment, the productivity of any part of the networkin 82 countries (So and Shin 1995). Roughly 50 depends on the size and configuration of thepercent of these privatizations and 70 percent of entire network.new investments are taking place in developing The dependence of productive, noninfrastruc-countries, mostly in Asia and Latin America. ture capital on infrastructure services suggestsWhat developments have led to greater private that in many cases investment in productivesector participation in the financing and provi- capital must follow the availability of infrastruc-sion of infrastructure services? ture. In other words capital investments in

I

manufacturing, agriculture, or services are Most infrastructure expenditures in develop-encouraged when their productivity-and thus ing countries have been funded directly from fis-return to capital-is greater. And greater produc- cal budgets. But several factors, such as macro-tivity depends on the adequacy and efficiency of economic instability and growing investmentinfrastructure services. Thus in the absence of requirements (particularly following the debt cri-effective financial and institutional structures sis of the 1980s), have shown that public financ-that ensure timely and adequate provision of ing is volatile and, in many countries, rarelyinfrastructure, a vicious cycle of capital decumu- meets crucial infrastructure expenditure require-lation can be conceived: low levels of investment ments in a timely and adequate manner.will jeopardize the financial viability and expan- The debt crisis and the adjustment efforts thatsion prospects of existing infrastructure stocks, it required generated a volatile and generallywhile the deterioration in infrastructure stocks declining investment profile in almost everyresults in lower investment, and so on. This phe- developing country it affected. The publicnomenon is particularly important in the evolv- finance problems of the countries in Table 1-1 areing regionalization and globalization of national representative of the trends across a large num-economies, since infrastructure services are an ber of developing countries during the 1980simportant element in retaining domestic and and, for some, into the early 1990s.attracting foreign capital. The volatility and decline in investment were

Expansion of infrastructure facilities may lead, less pronounced for Colombia, where prudentbe in step with, or lag behind demand. Where it economic and debt management helped avoid alags it is clearly inefficient; whether leading is debt restructuring program. The pace of domes-inefficient depends on the circumstances. A good tic investment in Chile also recovered after suc-mix of leading and synchronicity between devel- cessful debt restructuring in 1985. But betweenopment of infrastructure services and demand 1980 and 1985 the ratio of gross domestic invest-would be an efficient approach. Some lead time ment to gross domestic product (GDP) fell 7.7may be needed to develop certain infrastructure percentage points in Argentina, 4.1 percentagecapacities, since infrastructure development can points in Brazil, and 13.8 percentage points in thestimulate investment and increase demand for Philippines. By 1991 the ratio had fallen eventhe same service or other infrastructure services. further in Argentina and Brazil. For all Latin

American and Caribbean countries the netPublic financing of infrastructure investment coefficient dropped from nearly 23

percent in 1980 to about 17 percent in 1989.The World Bank's World Development Report 1994 Lower rates of investment were accompanied bydiscusses the role of infrastructure in develop- low or negative economic growth rates for Latinment and the capacities for and constraints on America as a whole. The Philippines had a simi-improving the quality and quantity of infra- lar downturn, as one of the exceptions to thestructure services in developing countries. Thereport identifies infrastructure subsectors asfollows:

Table 1-1: Gross domestic investment, selectedPublic utilities: Power, telecommunications, countries, 1980-91

piped water supply, sanita- (percentage of GDP)tion and sewerage, solidwaste collection and dis- Country 1980 1985 1991posal, and piped gas.

Argentina 25.3 17.6 14.7Public works: Roads and major dam and Brazil 23.3 19.2 18.9

canal works for irrigationand drainage. Chile 24.6 17.2 22.2

Colombia 19.1 19.0 16.8

Other transport: Urban and interurban rail- Philippines 29.1 15.3 20.4ways, urban transport, portsand waterways, and airports. Source: World Bank 1995j.

2

Table 1-2: Average annual GDP growth, 1980-89(percent)

Region/Country 1980 1981 1982 1983 1984 1985 1986 1987 1988 1989

Latin America andthe Caribbean 5.9 -0.2 -1.2 -2.4 3.9 2.9 4.9 3.3 0.7 1.0

Argentina 3.7 -5.9 -3.2 3.8 1.8 -6.6 7.3 2.5 -1.8 -6.3Brazil 9.1 -4.4 0.6 -3.4 5.4 7.9 8.4 3.3 -0.3 3.3Chile 8.2 4.8 -10.4 -3.7 8.0 7.0 5.6 6.5 7.3 9.9Colombia 4.1 2.1 1.0 1.6 3.6 3.3 6.1 5.4 4.1 3.5

East Asia and the Pacific 5.1 5.4 5.4 8.7 8.7 6.7 7.3 8.8 9.6 6.2Philippines 5.1 3.4 3.6 1.8 -7.3 -7.3 3.4 4.3 6.8 6.2

Source: World Bank 1995j.

strong economic performance in East Asia and and privatized during the 1980s, these trendsthe Pacific during the 1980s (Table 1-2). affected almost all infrastructure services, physi-

Since infrastructure accounts for 40-60 percent cal and social. In many countries noninfrastruc-of public investment in developing countries, the ture economic enterprises that were owned andtightening of resources in the 1980s took a heavy operated by the government were also affected.toll on the ability of public budgets to finance For some countries, like Argentina and Mex-much-needed infrastructure investments (Figures ico, public investment as a share of GDP contin-1-1 and 1-2). The weighted average of public ued to decline during the early 1990s, but theinvestment in Latin America and the Caribbean process of attracting private sector capitalfell from 7.6 percent of regional GDP in 1980 to through privatization of infrastructure facilities6.5 percent in 1989. Except for the electricity sec- had begun. As a result underinvestment andtor in Chile, which was successfully restructured neglect slowed in certain infrastructure sectors.

Figure 1-1: Public investment in selected Latin American countries, 1980-93(percentage of GDP) 14o

.-- Argentina12% -- -- Brazil

Chile

10% ........*.... Colombia

Mexico

8%

40/ _. .. .

2%

0 0/0 I I I I I I I I I I I

0' oCY , oa am oE \ ON O o a ' a\

Source: IFC 1995b.

3

Figure 1-2: Public investment in selected East Asian countries, 1980-93(percentage of GDP)

20%

18% _ - - - Indonesia

18% - ............ Korea16% ---- Malaysia

14%o >...... Philippines/ - Thailand

8%

6%

4% .........

2% -

0% I I I I I I I I I I I IC X. c 9 In IDoo o cH J c'ecc cc cc sf cc o c o c cc oi 0' 0 o a, a, a, a' a' a a, a, a'ON ' o '1 r4

414 r4 - 4 - r

4- f-

Source: IFC 1995b.

Evidence suggests that, possibly because of during the 1980s, vital current expenditure cate-deferrable impacts and lower immediate politi- gories such as repair and maintenance of infra-cal sensitivities, cuts in capital budgets consti- structure facilities were often sacrificed in thetuted a disproportionate share of the overall face of mounting interest payments on domesticspending cuts that were part of the economic and external liabilities. In Brazil the combinedausterity and adjustment programs of the 1980s. external liabilities of the central government, theAs the figures in Table 1-3 indicate, current central bank, and nonfinancial public enterprisesspending remained relatively stable-or even accounted for 60-90 percent of total long-termincreased-in several countries during the 1980s debt, with annual interest payments averaging(Chile is an exception). $7.6 billion between 1980 and 1989, equivalent to

Although current expenditures from general about 25 percent of average annual exports. Ingovernment budgets remained unchanged or Argentina average annual interest payments oneven increased in many developing countries consolidated public sector debt amounted to just

over 3 percent of GDP between 1980 and 1988(Easterly, Rodriguez, and Schmidt-Hebbel 1994).

Table 1-3: Cufrent government spending, The neglect of crucial maintenance andselected countries, 1980-91 upgrading of infrastructure facilities in Latin(percentage of GDP) America and the Caribbean during the 1980s left

much of the infrastructure in disrepair by thestart of the 1990s. A 1992 World Bank survey of

Count7ry 1980 1985 1991 road agencies suggests that only three of sixteen

countries with reliable expenditure data (Brazil,Brazil 9.2 9.9 14.4 Colombia, and Venezuela) had sufficient funds

Chile 12.5 13.4 9.7 to maintain their networks in good condition.Fiscal resources are an insufficient and unsta-

Colombia 10.1 10.7 10.3 ble source of financing for infrastructure projects,

Philippines 9.1 7.6 10.2 especially since they are part of the toolkit formacroeconomic management. If left unad-

Source: World Barnk 1995j. dressed, the volatility of public financing and the

4

frequent neglect of infrastructure expenditures Table 1-4: Potential savings to developingwould set off a vicious cycle of income and out- countries from increased efficiency in theput deterioration. Public and private investment provision of infrastructure servicesare strongly interdependent. A decrease in the (billions of U.S. dollars)

private capital stock, which may result frominadequate infrastructure or the crowding-out Annual

effect of excessive government borrowing, leads Sector savings Source of inefficienciesto a reduction in tax revenues and a matching cutin public investment. The subsequent decrease in Roads 15 Investment requirementsthe supply of social and physical infrastructure created by improperdepresses private capital accumulation still maintenance.more, resulting in a further loss of tax revenues, Power 30 Transmission, distribution, andand so on. Thus private and public capital decu- generation losses.mulation feed on each another in a destabilizingdownward spiral. Water 4 Leakage.

For countries that experience substantial debt Railways 6 Excess fuel use, overstaffing,overhang, the adjustment process is usually long and locomotive unavailability.and traumatic. Once growth decelerates as a Total 55result of higher fiscal outlays for debt service andlower investment in public infrastructure, fiscalproblems become systemic. In such situations Source:WorldBank 1994c.governments find it compelling, year after year,to make cuts in productive expenditures. Even in demand and supply, soft budget constraints, thecountries where the public sector has made suffi- absence of financial risk management, and thecient investments in infrastructure, the pace of entwining of financial management of publiceconomic growth and increasing importance of enterprises providing infrastructure servicesinfrastructure as a competitive means of attract- with macroeconomic management are only a fewing foreign capital have outpaced these coun- sources of the inefficiencies that have character-tries' public sector financing and managerial ized the provision of infrastructure services bycapacity and ability to absorb mounting financial public entities. World Development Report 1994risks. provides information on the annual costs to

developing countries of inefficiencies in the tra-Rationale for private provision of ditional structure of infrastructure servicesinfrastructure (Table 1-4) and the considerable gains that can be

achieved through appropriate reform (Table 1-5).The recent surge in private investment in infra-structure has many causes. Among the mostimportant are the inefficiencies of public provi-sion of services, the need for economic pricing Table 1-5: Fiscal burden to developingand cost recovery, technological advances countries of underpricedinfrastructureenabling greater private participation, advances (billions of u.s. dollars)in regulatory frameworks, the need for privateresources, and the potential investment gap Potential annual savingsdeveloping countries face. Sector from better pricing Source

Public sector inefficiency Power 90 Underpricing.

Water 13 Underpricing,The provision of infrastructure services through 5 illegal connections.traditional institutional arrangements-publicsector financing and operation-has been Railways 15 Underpriced passengerfraught with inefficiencies. Low productivity of service.labor and capital, weak incentive structures, Total 123neglect of timely maintenance, lack of sufficienteconomic and institutional links between Source: World Bank 1994c.

5

No matter what management structure a pub- road user charge receipts of $220 million (Worldlic enterprise adopts, authority often ends up in Bank 1992).the hands of politicians who are not appointed A combination of financial and administrativefor their entrepreneurial skills. Political consider- shortfalls has resulted in increasing inefficienciesations frequently favor allocating resources to in the provision of infrastructure services inconstruct additional infrastructure capacities at many developing countries (Box 1-1). Of somethe cost of maintaining existing facilities, since 420 million people in Latin America and the Car-maintenance is less attractive politically. Because ibbean in 1989, 18 percent had no access to publicthe liabilities of public enterprises are ultimately water supply and 42 percent did not have sanita-backed by the resources of the state, there are tion facilities. Some 30 percent of the region'sfew incentives for strict financial risk manage- population lacked electricity service. Thement. Moreover, the monopoly position occu- region's 2.2 million kilometers of roads, most ofpied by such enterprises diminishes the risks which were built during the boom years of theassociated with customer dissatisfaction. By the 1960s and early 1970s, were adequate to providesame token, since the state ultimately under- accessibility, but have badly deteriorated sincewrites the liabilities of public enterprises, the the 1990s began (World Bank 1992).possibility of undue judicial interference in favor There is also a lot of room for improving effi-of these enterprises is higher than where private ciency in Asia. For instance, the amount offirms are involved. In public enterprises the energy lost in power transmission and distribu-extent to which managers can reap the rewards tion-through illegal tapping into overhead linesof investment and operational efficiency are or poor maintenance-is at least four times theseverely restricted and in many cases are insig- average for efficient public utilities. Even amongnificant. The wage structure in public enterprises high-growth East Asian countries measuresis often linked to civil service wages, thus mak- designed to improve the incentives in publicing recruitment and retention of qualified techni- enterprises have proved that efficiency mustcal and managerial staff extremely difficult. increase.2 Also, countries in the region are falling

Even where public enterprises can provideinfrastructure services efficiently, privatizationfurther improves efficiency. For instance, Chile'spower sector was successfully restructured and Box 1-1: Some aggregate inefficienciesprivatized during the 1980s. As a result theprivate sector improved upon what was already * Some 40 percent of power generating capacityconsidered an efficient sector operation. Chilectra, in developing countries is unavailable forthe privatized distribution company that production.accounts for half of the country's electricity dis-tribution, was able to reduce its energy losses * Half the labor in Latin American railways isfrom about 13.5 percent of net generation in 1990to about 9.5 percent in 1994. The losses were * In Africa timely expenditure of $12 billioncaused by inefficient distribution networks and would have saved $45 billion in road recon-unauthorized connections. struction costs.

User charges for many publicly owned andoperated infrastructure services (especially in the * The average port facility in developing coun-transport sector) are often deposited in national tries moves cargo from ship to shore at 40 per-treasuries, which consider them a dependable cent the speed of the most efficient ports.source of revenue. Under such a system fundsfor crucial operations and maintenance costs are * The technical efficiency of power utilities insubject to a budgetary allocation process where fifty-one developing countries has declinedincreased operating efficiency results in lower over the past twenty years.allocations. Such distorted incentive structures * Water supply systems deliver an average of 70frequently govern the allocation process for new percent of their output to users, comparedinvestments as well. In Jamaica the government with design targets of 85 percent.spent about $40 million a year on roads during1985-89, of which about $12 million went to Source: Ingram 1995.

maintenance, compared with average annual

6

behind in meeting demand for infrastructure marginal costs of supplying electricity had beeninvestment. Public utilities in Thailand, for studied in most countries by the late 1980s, fewexample, are generally efficient, but telecommu- governments and power utilities adopted effi-nications services are scarce and unreliable. Tele- ciency pricing for electricity supply.phone penetration averaged 3 lines per 100 peo- In Asia, for example, the financial performanceple in 1992, with a waiting list exceeding 900,000 of state utilities deteriorated over the past two(World Bank 1995j). Only recently has private decades. Rates of return in state utilities fell frominvestment of about $6 billion in the sector about 9 percent in the mid-1970s to 5 percent inbegun to expand capacity and provide periph- 1991, and average power tariffs declined fromeral services, particularly wireless mobile and 5.2 to 3.8 cents per kilowatt hour. In the absencepaging services. of transparent cost recovery mechanisms, con-

sumers usually end up incurring a variety ofEconomic pricing and cost recovery other costs. Poor repair and maintenance of

roads, for example, increase vehicle operatingWhen the public sector provides infrastructure costs and require premature reconstruction.services, it rarely recovers from users the full cost These costs were estimated at $1.7 billion a yearof providing the services, whether for political or in Latin America and the Caribbean at the starteconomic development reasons. This inadequate of the 1990s (World Bank 1992). Subsidized pro-cost recovery is at the heart of financial con- vision often results in inefficient consumptionstraints for most infrastructure services, al- and overinvestment in new assets. And whenthough to varying degrees in different countries. these expenditure outlays come from the public

In the Republic of Korea, for example, the budget, they have inflationary effects.costs of various modes of transportation have Such costs are ultimately channeled to con-been kept artificially low as part of an anti- sumers through an unstable and inflationaryinflationary policy. Cross-subsidies are used to environment that erodes the real value of incomeoffset the resulting imbalances. As a result the and wealth. The lost efficiency and fiscal burdenreal prices of gasoline, tolls, and urban transit of mispricing are equivalent to about 3.5 percentmodes (other than subways) consistently of developing countries' 1995 GDP (using thedeclined between 1980 and 1990. Low prices are figures in Tables 1-4 and 1-5 and developingcausing financial shortfalls for transport agencies countries' GDP as reported in World Bank 19951).and carriers and are forcing private operators to The quasi-deficit nature of deteriorating infra-curtail services or disobey regulations in order to structure and insufficient expansion also contrib-survive (World Bank 1995h). utes to economic instability, lower growth, and

The World Bank has found that electricity tar- higher inflation.iffs in developing countries during the 1980s World Bank research on the losses fromremained relatively stable (in constant U.S. dol- increased vehicle operating costs and prematurelars) between 1979 and 1983 but deteriorated reconstruction of badly deteriorated roads showsmarkedly thereafter (World Bank 1990). Between that each dollar not spent on needed road main-1983 and 1987 tariffs fell by about 35 percent. tenance can increase vehicle operating costs byAverage tariffs during the late 1980s were only about $3 and require an additional $2-3 for pre-about half the average incremental economic mature reconstruction. Other losses are incurredcost of power system expansion during the through the indirect effects of low-quality ser-1990s. Moreover, the average tariff in developing vices, such as the health and medical costs ofcountries during the late 1980s was just over half inaccessibility to safe water or transportationthe average level in OECD countries. Thus elec- costs stemming from poor telecommunications.tricity consumers in developing countries paid Developing countries that are trying to pro-tariffs that were too low to encourage efficient mote private financing of infrastructure willuse of electricity, while power utilities were have to implement tariff reforms that allow eco-unable to raise sufficient funds from revenues to nomic cost recovery (Box 1-2). Prices may rise infinance expansion. The relative movement in tar- countries where heavy subsidies were providediffs, expressed in constant local prices, shows prior to reforms, but cost recovery does not nec-that electricity consumers in developing coun- essarily mean higher prices. When tariff reformstries faced greater fluctuations in tariffs than are undertaken in conjunction with other policytheir OECD counterparts. Thus even though the and regulatory reforms that aim to create an

7

Box 1-2: Benefits of sound economic pricing

Sound electricity and water pricing policy can gen- * Proper prices lead to better financial situationserate a number of benefits, among which the fol- for utilities, which lead to maintenance expen-lowing: ditures to improve operating conditions and

the quality of supply (that is, to reduced costs* Price is the best inducement for conservation. and improved availability and reliability of

Industries will invest in conservation to the point supply).where the cost of saving an additional unit ofenergy (or water) is the same as the purchase * Adequate prices are necessary to implement effi-price of that energy (or water). ciently and reliably a proper maintenance policy,

to lower forced outage rates, to increase effi-* The benefits of subsidizing energy (or water) ciency and availability, to reduce losses, and to

consumption are smaller than the benefits of improve reliability of supply.allocating the equivalent subsidy to infrastruc-ture, social, and human capital formation. * Sound pricing encourages efficiency in supply by

providing utilities with signals for an efficient* Sound pricing enables private sector participation. operation and investment program.

enabling environment for private participation, Advances in technologyconsumers can end up paying less for higher-quality services. Technological developments have allowed

There are many examples of lower prices unbundling, private entry, and competition inthrough higher productivity and competition in many infrastructure services once viewed as nat-the postreform provision of infrastructure ser- ural monopolies. In telecommunications, wire-vices. Regulatory reform of Argentine ports, less technologies such as satellite and microwavewhich included decentralization, privatization, systems are replacing long-distance wire-basedand deregulation, caused tariffs to fall and is esti- cable networks, and cellular systems offer anmated to have saved $156 million a year between alternative to local distribution networks. These1989 and 1994. Between 1989 and 1993 staffing in changes have gradually eroded the network-the sector (including dredging activity) dropped based monopoly in telecommunications andfrom more than 13,000 to around 5,500, while have made competition possible. Such techno-cargo movements increased from 40.0 million logical advances have influenced many restruc-tons to about 46.5 million tons. turing decisions, even among privately owned

In most cases allowing cost recovery in a more firms in industrial countries. American Tele-competitive environment results in lower prices phone and Telegraph's (AT&T) September 1995relative to inefficient public monopolies. The announcement that it would split into three com-experience in industrial countries in this area is panies (services, equipment, and computing)encouraging. According to the OECD, business was partly influenced by such evolving and tech-tariffs in countries where telephone service is nological regulatory factors. At the time of thestill provided by monopolies fell 3.1 percent announcement AT&T's chairman cited new leg-between 1990 and 1994-but they fell 8.5 percent islation that would allow local telephone compa-in countries that had introduced competition nies to compete directly with AT&T in long-(The Economist, 9 December 1995). And prices are distance and local telephony, with the local abil-generally lower in Britain, Scandinavia, and the ity arising from new cost-efficient technologicalUnited States, where industries are competitive. possibilities (The Washington Post, 26 SeptemberMoreover, the regulated prices of all of Britain's 1995).privatized utilities (except water, where higher New technology also has made competitionprices are allowed for new investments) have among electricity suppliers technically feasible.fallen in real terms since privatization, even The 1992 Energy Policy Act in the United Statesthough the firms have been making large profits. took advantage of these possibilities and allowed

8

dozens of independent power producers nation- Advances in regulationwide to sell bulk electricity to distribution com-panies over existing transmission lines. More- Regulation has always been closely tied to indus-over, rapid technological development in tries with natural monopoly characteristics.metering systems has made full retail competi- When such services are provided by public enti-tion possible. New regulatory regimes can com- ties, many regulatory issues become immaterial.plement technological advances in changing the Effective regulatory techniques and institutionscompetitive landscape of infrastructure services. are, however, central to the creation of a compet-

In the transport sector containerization has itive and incentive-based climate for the privatefacilitated competition in port services and provision of infrastructure services. Regulatoryremote electronic pricing systems have intro- improvements provide a better operating regimeduced cost-efficient and accurate determination for existing facilities or components and emergeand collection of user charges for major road net- in response to technological developments thatworks and urban transport systems. provide new options and structures for the pro-

New technologies also enable infrastructure vision of infrastructure services.firms to diversify from single sectors into previ- New regulatory and contract arrangementsously separate industries. Optical fiber networks, have alleviated certain concerns arising from thefor example, are being managed as part of gas monopoly nature of infrastructure services.and electricity transmission systems. Cable, Many infrastructure services traditionally pro-telecommunications, and electricity distribution vided through a public monopoly-such asservices to households are expected to adopt ports, power generation, and long-distance tele-common delivery, metering, and billing systems. communications-now have the potential to beIn Europe electric utilities have recently entered very competitive.into telecom markets, and in Chile electric I The regulatory techniques related to theutilities have become involved in the water unbundling of infrastructure services have insector. many cases lowered the intrinsic barrier to entry

Finally, innovations in financial technology stemming from the economies of scale that char-and the globalization of financial markets have acterize infrastructure investments. Some exam-introduced a larger pool of resources and a more ples of competitively provided infrastructuiediversified array of instruments that better services are shown in Table 1-6. The unbundlingmatch the financing needs of infrastructure of services in the electricity sector, which canprojects. These subjects are discussed in greater open competition in generation and even in dis-detail in Chapters 3 and 4. tribution, is one of the best-known examples of

Table 1-6: Infrastructure services provided in competitive markets

Sector Activity Examples

Electricity Supply to grid or large Argentina, Chile, Norway, United Kingdom,consumers and distributors United States

Gas Supply to large Argentina, Canada, United Kingdom, United Statesconsumers and distributors

Telecommunications Long-distance services Australia, Republic of Korea, New Zealand,United Kingdom, United States

Cellular telephony Australia, Hungary, New Zealand, Sri Lanka,United Kingdom, United States

Rail Passenger and freight services Sweden, United Kingdom

Water Bulk water supply Chile, United Kingdom, United States

Source: Smith and Klein 1994.

9

such regulatory reforms. In Britain any person and thus the welfare benefits of deregulation-who consumes more than 100 kilowatts per bill- for certain infrastructure subsectors in theing period may purchase power directly from a United States are shown in Table 1-7.generator or another distribution company act-ing as a "second tier supplier" (Stewart-Smith Need for private resources1995). The key to this system is metering toenable disaggregation of flows and settlement. ln In addition to its capacity for enhancing effi-the United States some state regulators have pro- ciency and viability, the increasing role of the pri-posed that residential customers be given a vate sector in providing infrastructure serviceschoice of power suppliers-just as they can arises from financial necessity. Public budgets inchoose long-distance telephone providers-and developing countries simply cannot be expectedindustrial customers are bargaining for lower to provide timely and adequate expansion of infra-rates from competing power suppliers. structure capacities or to save financially trou-

In sectors where duplication in the provision bled public entities providing infrastructure ser-of services is not feasible, such as roads and vices. Many developing countries that have emergedwater, various modes of bidding for concessional from debt crises have adopted austerity and macro-arrangements have introduced competition that economic adjustment programs that leave very littlepromotes efficiency and raises the quality of ser- room for raising public spending without jeopardiz-vice. Specifically, where natural monopolies are ing the basic goals of economic reform.unavoidable, awarding concessions through In stronger emerging market economies, espe-repeated franchise bidding subjects the monopo- cially in high-growth East Asian economies, thelies to competition. Auctioning monopoly fran- demand for infrastructure is outpacing thechises to the lowest-cost providers also results in financing capacity of public resources (Boxes 1-3lower prices for consumers. To ensure efficient and 1-4). East Asia has a healthy annual growthservice for users, however, several other factors forecast of more than 7 percent over the next tenmust also be taken into account when conces- years-and an infrastructure bill of $1.2-$1.5 tril-sions are awarded to private suppliers, including lion. In the power sector alone the combined newtechnical merit, bidder experience, operations supply requirements of China, India, Indonesia,expertise, and financing capacity. Malaysia, the Philippines, and Thailand are

Deregulation can also stimulate competition, 25,000-30,000 megawatts a year.providing substitutable services such as various The annual level of infrastructure investmentstransport modes and different energy carriers. required in developing countries in the 1990s isAs mentioned earlier, regulatory reform of projected to amount to more than $200 billionArgentina's ports stimulated considerable com- (1993 dollars), with $100 billion needed for power,petition and productivity. As a result the ports $60-70 billion for water and sewerage, $25-30 bil-were able to recapture cargo traffic that had been lion for telecommunications, and $15-20 billionlost to cheaper land transport services and morecompetitive port facilities in neighboring Chile.

Regulation of a natural monopoly is intended Table 1-7: Estimated annual gains fromto eliminate, curb, or allocate economic rents. deregulation of selected U.S. infrastructureInappropriate regulation carries considerable sectorscosts for the governments administering the reg- (billions of 1990 U.S. dollars)ulation, for the firms subject to it, and for theeconomy as a whole. For instance, rate-of-returnregulation in many infrastructure projects is Extent ofintended to cap the economic returns that a pri- deregulation Gainsvate firm can accrue from an infrastructure con-cession. One side effect of this form of regulation, Airlines Complete 13.7-19.7however, is that firms have an incentive to invest Trucking Substantial 10.6excessively, inflate costs, and engage in cross- Railroads Partial 10.-12.9subsidization. In cases where the purpose of reg- Telecommunications Substantial 0.7-1.6ulation is to allocate economic rents, all parties Total 35.4-44.8involved will spend resources to obtain the poten-tial rents. The costs of inappropriate regulation- Source: Winston 1993.

t0

Box 1-3: Thailand: Public sector efficiency Box 1-4: Growing demand for ports inmay not be enough China

The Thai state enterprises in charge of infra- Between 1980 and 1993 the cargo handled instructure services are generally profitable, pro- China's main seaports more than doubled, toviding a sizable net financial contribution to the 679 million metric tons. In 1994 containertreasury (about $781 million in 1993). Of eigh- throughput grew 28 percent. In terms of trans-teen public utility state enterprises, fifteen are port demand, ocean shipping is expected toconsidered major. These fifteen enterprises are maintain annual growth of more than 8 percentefficient and perform well financially except for through 2000, while annual domestic waterwaythree whose tariffs have been kept low for social traffic is expected to increase by 7 percent. Notreasons (the State Railway Authority, Bangkok surprisingly, export-oriented shipping is grow-Mass Transit Authority, and Expressway Trans- ing faster than domestic shipping.portation Organization). Thailand's industrial- Over the past decade China invested moreization has been rapid, however, and a major than $2.3 billion in the modernization andgovernment concern is ensuring financing for development of its ports. In 1994 alone Chinaextremely high increases in infrastructure built twentv deepwater berths in coastal portsinvestments. Capital expenditures by the fifteen with a combined capacity of 25.8 million metricmajor public utilities rose by an average of 5.4 tons, renovated 200 miles of navigation chan-percent a year during 1989-93, reaching about nels, and increased handling capacity at inland$10 billion in 1993. Investment requirements are river ports by 2.74 million metric tons.expected to grow by 8 percent a year until 2000, By 2000 China aims to have ten ports withwith average annual investment requirements capacities over 100 million metric tons a year,reaching about $15 billion. To ensure that suffi- including those at Dalian, Qinhuangdao,cient resources are attracted to infrastructure the Nanjing, and Guangzhou. Shanghai alone willgovernment has embarked on a privatization spend 10 billion renminbi ($1.2 billion) on portstrategy and is promoting increased private improvements during 1996-2000.participation in the provision of infrastructure This scale of investments and development isservices. expected to require considerable foreign capital

and managerial expertise. Foreign investors canSource: World Bank 1995i. own up to 49 percent of a port, which makes

=__ _ _ _ _ _ _ _ ports, along with power plants, one of the mostaccessible areas of infrastructure for foreigninvestors.

for highway rehabilitation (Santos 1994). Atthese levels infrastructure investments will Sourcc: Oxford Analytica Asia Pacific Daily Brief, Sep-account for between a quarter and a third of all tember 26, 1995.fixed investments in developing countries (esti-mated at $700-$750 billion a year). By compari-son, international aid (grants and official loans)for infrastructure was only about $15 billion in tion and capital mobility have increased the1993, as was aggregate annual private invest- growth potential of developing countries. And asment in infrastructure in developing countries. the number of developing countries adhering to

sound macroeconomic and structural policiesPotential infrastructure investment gap, 1996-2000 increases, higher growth rates will continue in

the next few years.As noted earlier, infrastructure investment in While higher growth clearly should be accom-developing countries needs to increase in every panied by a concomitant increase in infrastruc-region. The pent-up demand for infrastructure ture investments, the increase required acrossservices is substantial, whether because of low different regions and countries is difficult tolevels or inefficient past investment or because, measure. This exercise is not made any easier bydespite higher and more efficient levels of infra- the erratic and inconsistent relationship betweenstructure investment, supply has been outpaced infrastructure investment and economic growthby growth (as in East Asia). In addition, global in developing countries. Inconsistencies ariseeconomic opportunities brought about by freer from differences in how macroeconomic andtrade arrangements and deeper financial integra- structural policies are managed. Those policies

211

have a direct bearing on the efficiency of invest- various sources. To start with, Table 1-8 showsments and the impact on growth. the average level of infrastructure investment

It is, however, possible to compute aggregate across developing regions in 1993.figures for the incremental increase in expendi- Projected regional GDPs for 1995-2000 areture levels using the latest available data from shown in Table 1-9. Table 1-10 shows estimated

levels of investment in infrastructure for 1-3 per-

Table 1-8: Infrastructure investment in centage points of GDP higher overall infrastruc-ture investment, using the data in Table 1-8 as a

developingregions,1993* base. Except for East Asia and the Pacific the

changes in the last three columns of Table 1-10are basically chosen for arithmetic convenience;

Region Public Prioate' Total that is, they produce the desired average. 3 These

figures are simply used to assess the investmentSub-Saharan Africa 2.31 0.07 2.38 gap relative to current investment levels.South Asia 2.73 0.21 2.94 To compute the potential investment gap thatEast Asia and the Pacific 5.01 0.83 5.84 would exist if the overall average infrastructureLatin America and the level in developing countries were increased, the

Caribbean 1.98 0.62 2.60 figures in Table 1-10 are applied to Table 1-9; theMiddle East and North

Africa 4.14 0.07 4.21 results are shown in Table 1-11. According toEurope and Central Asia 3.68 0.19 3.87 these results, each one percentage point increase

in the level of infrastructure investment (as aAverage 3.64 percentage of GDP) would require about $300

billion of additional investment during 1996-a. 1984-94 average. 2000, or an average of about $60 billion a year.

Table 1-9: Projected gross domestic product, 1995-2000*(millions of U.S. dollars)

Region 1995 1996 1997 1998 1999 2000

Sub-Saharan Africa 313,346 322,747 332,429 342,402 352,674 363,254South Asia 372,242 387,132 402,617 418,722 435,470 452,889East Asia and the Pacific 1,548,424 1,703,267 1,873,594 2,060,953 2,267,048 2,493,753Latin America and the Caribbean 1,597,251 1,664,336 1,734,238 1,807,076 1,882,973 1,962,058Middle East and North Africa 673,020 703,305 734,954 768,027 802,588 838,705Europe and Central Asia 1,088,232 1,109,996 1,132,196 1,154,840 1,177,937 1,201,496

Table 1-10: Regional infrastructure investment rates at different average overall investment averages*(percentage of GDP)

Region 3.64a 5.00 6.00 7.00

Sub-Saharan Africa 2.38 3.50 4.50 6.50South Asia 2.94 4.50 5.50 7.00East Asia and the Pacific 5.84 6.50 6.80 7.50Latin America and the Caribbean 2.60 4.00 6.00 7.00Middle East and North Africa 4.21 5.78 6.50 7.00Europe and Central Asia 3.87 5.50 6.50 7.00

a. 1993 average.

' Sources for Tables 1-8,1-9 and 1-10: IFC 1995b, IMF 1995b, World Bank 1995k, So and Shin 1995, Kohli 1995.

12

Table 1-11: Additional investment required to finance higher average infrastructure investment indeveloping countries, 1996-2000(millions of U.S. dollars)

5 percent of GDP 6 percent of GDP 7 percent of GDP

Annual Annual AnnualRegion 1996-2000 average 1996-2000 average 1996-2000 average

Sub-Saharan Africa 19,234 3,847 36,369 7,274 70,639 14,128South Asia 32,756 6,551 53,724 10,745 85,177 17,035East Asia and the Pacific 69,008 13,802 100,204 20,041 172,994 34,599Latin America and the Caribbean 126,496 25,299 307,510 61,502 398,017 79,603Middle East and North Africa 60,592 12,118 88,205 17,641 107,443 21,489Europe and Central Asia 94,279 18,856 152,043 30,409 180,926 36,185

Total 402,365 80,473 738,054 147,611 1,015,194 203,039

Given the constraints on public budgets, to play a role in ensuring that incentive struc-increases in infrastructure investment will have tures remain attractive across infrastructure sub-to be achieved through private sector participa- sectors, so that infrastructure services willtion. According to the latest available data, the develop uniformly. Under these conditions infra-private sector accounts for less than 10 percent of structure services can effectively support eco-overall investment in infrastructure, and is nomic growth and then grow in step with it.mainly concentrated in selected countries inLatin America and Asia. Governments therefore Notesneed to carry out policy, institutional, and regu-latory reforms to ensure that private resources 1. Social aspects of infrastructure such as education and health are

are directed to infrastructure in sufficient and assumed to be embodied in the labor force.

sustainable amounts. 2. See the examples of Indonesia and the Republic of Korea in Chapter 2.

In addition to promoting private participation 3. The required infrastructure investment levels in East Asia are esti-in the development and provision of imfra- mated at 6.5 and 6.8 percent of GDP, respectively, for low case and

structure services, governments must continue baseline growth scenarios as reported in World Bank 1995g.

13

2 l 2Moving Toward Private Infrastructure

Recognizing that the inward and protectionist Kingdom have tried, with some success, toeconomic polices of the 1960s and 1970s led to reform their state-owned enterprises by impos-the severe economic distress of the 1980s, many ing on them the same framework of internal anddeveloping country governments have increas- external incentives that applies to successful pri-ingly changed economic course toward market vate corporations.mechanisms, private incentives, and export ori- Corporatizing public enterprises should leadentation. As a result public enterprises, indus- to the development of a financial and regulatorytrial promotion, and trade protection are out; environment that is more responsive to theprivatization, industrial deregulation, and free financing and regulatory concerns of privatetrade are in. infrastructure projects. As discussed in Chapter

In the infrastructure sector, where public sec- 3, debt issues by efficient public enterprises cantor overspending during the 1970s and under- make an important contribution to the develop-spending during the 1980s played an important ment of bond markets. This role is important inroie in economic instability and structural ineffi- setting benchmark rates in debt markets sinceciencies in many developing countries, greater government issues may gradually decline as aemphasis has been placed on privatization and result of fiscal discipline or macroeconomicprivate participation. The results of the deregula- improvements.tion policies of the 1970s in the United States and Some East Asian countries have implementedthe privatization experiences in Chile, New specific measures-including organizationalZealand, and the United Kingdom during the restructuring and performance agreements-to1980s encouraged this new thinking. improve the efficiency of public enterprises pro-

viding infrastructure services. When perfor-Refokiming public enterprises mance agreements were used in Korea the rate of

return on the assets of public enterprises (inDepending on country and enterprise condi- power and telecommunications) rose from lesstions, reforming the operation of the public than 3 percent before 1984 to more than 10 per-enterprises that provide infrastructure services is cent by the end of the decade. In 1983 Indonesiaan efficient strategy prior to privatization. The decentralized the management of ninety ports bymain objectives of such reform should be to max- creating four public port corporations. Effortsimize the internal incentives of managers and were also made to control overregulation andworkers and to create external incentives make managers more accountable for theirthroughi a competitive environment (Box 2-1). actions. As a result of these measures, as well asCountries as diverse as Chile, the Republic of an effective cost control program, costs were 5Korea, New Zealand, Sweden, and the United percent lower and revenues 20 percent higher by

14

Box 2-1: The public enterprise as a modern corporation

Internal incentives External incentives

Clarify the principal and agent incentives Encourage competition* Property laws should be defined, used, and Competition is perhaps the most important exter-

accounted for under the same rules that apply to nal factor an enterprise faces. With proper internalmodern private corporations. incentives, public enterprises will respond to pri-

* The corporation and its owners should have sep- vate competitors by increasing their efficiency inarate legal identities to insulate management order to survive commercially, particularly in thefrom political interference. absence of large state subsidies.

* Share transferability should be allowed in orderto diversify public enterprises ownership and Improve financial discipline; the role of debtenable residual risk bearers to effectively partici- Creditors can impose discipline akin to thatpate in management decisionmaking. imposed by shareholders and improve commercial

discipline in investment decisions. GovernmentsSeparate commercial from social objectives should, however, refrain from guaranteeing publicGovernments should set commercial objectives for enterprises' debt.public enterprises and give them incentives similarto those applied to private firms. Where social objec- Improve performance through equity marketstives are imposed, the cost should be identified and The market exerts an important discipline on man-the enterprise should be fully compensated. agement, demanding information flows and,

through pricing of equity, evaluating managementProvide incentives to corporate participants performance.Effective internal governance of corporatized publicenterprises can be achieved if the state provides Avoid complex monitoringadequate incentives-such as profit sharing and equi- Elaborate monitoring systems usually includety distribution-to boards, managers, and employees. management controls and complex formulas for

various markets. These can distort incentives andPut the private sector on boards are difficult to enforce. An efficient internal gover-A more effective way of ensuring that boards per- nance system coupled with the discipline of exter-form their strategic and monitoring role is to nal incentives works better than complex, central-include private sector representatives. ized monitoring schemes.

Avoid large holding company structuresHolding structures create additional layers ofbureaucracy, fail to shield the operating companiesin the group from political interference, allowcross-subsidization between companies, and dis-tort signals and incentives for management. Source: Muir and Saba 1995.

1988. Between 1987 and 1992 revenues grew projects that increase efficiency and expand ser-almost twice as fast as expenses. vices highlight the long-term economic advan-

tage of privatization even in countries where thisPrivatizing public enterprises objective played a minor role in launching the

program.The factors motivating privatization efforts differfor each country. In some countries the initial Country experiencepush for privatization comes from its potential togenerate government revenues while eliminat- In Latin America infrastructure services wereing a major source of public sector deficit. In privatized partly as a result of the emergingother countries privatization is seen as a neces- recognition of private infrastructure's advan-sary component of economic restructuring. tages (Chile) but mainly as a result of severeWhatever the motivation, private infrastructure macroeconomic distress (Argentina and Mexico).

15

Infrastructure privatization in the region has private financing and operation of new infra-often been used to generate revenues for govern- structure capacities. Most governments in thements and retire external debt while eliminating region did not face the dire macroeconomic con-a substantial source of fiscal drain. Severe ineffi- straints that were so pressing in Latin America.ciencies in state-operated infrastructure services Also, state utilities and agencies providing infra-also encouraged privatization. For instance, gov- structure services were more efficient and finan-ernment subsidies to the railway sector in ciallv viable than their counterparts in LatinArgentina averaged about $1.4 billion a year America. As noted earlier, the promotion of pri-during the 1980s. The railways, privatized in the vate infrastructure in Asia is primarily seen as aearly 1990s, now receive less than $100 million a way of complementing public sector efforts toyear from the government, and only for urban keep pace with economic growth.commuter railroads and subway concession- Malaysia began privatizing infrastructure asaires. As Table 2-1 shows, Latin America is far early as 1983 in the context of its New Economicahead of other regions in terms of infrastructure Policy. The objectives of the privatization pro-privatization. gram were similar to those in other countries: to

In Mexico the divestiture program began in relieve the government of the financial and ad-1983 as part of the macroeconomic stabilization ministrative burden of undertaking and main-program adopted with the support of the Inter- taining a vast and continually expanding net-national Monetary Fund. In Argentina electricity work of infrastructure; to promote competition,companies accounted for as much as half of the improve efficiency, and increase productivity;total deficit of state-owned enterprises by 1989, to stimulate private entrepreneurship and in-so privatization was critical. The government vestment to accelerate the rate of economicencouraged private participation in the sector to growth; and to reduce the public sector's sizeprovide needed capital, reform inefficient com- and presence.panies, and upgrade and maintain neglected Since 1989 the privatization effort has beenequipment. The government also used a debt- guided by a Privatization Master Plan that iden-equity swap mechanism in the privatization of tified a larger set of state-owned enterprises forENTel, the state telecommunications company, privatization. Roads, railways, the national air-bringing in cash proceeds of around $2.2 billion. line, telecommunications, and power generatorsThese funds were then used to reduce the gov- have already been privatized. Privatization isermnent's debts to foreign commercial banks. estimated to have saved about $2.5 billion equiv-

Among Asian countries (except Malaysia), alent in salaries and pension payments for 93,000privatization has not been emphasized as a first former state employees who are now in the pri-step toward promoting private provision of vate sector. In addition, the government wouldinfrastructure. Instead emphasis has been placed have had to provide some $16 billion equivalenton providing the necessary frameworks for for infrastructure projects had it not been for

privatization. Partly as a result of passing thehuge costs of infrastructure development to the

Table 2-1: Revenues from divestiture in private sector, the government has achieveddeveloping countries, 1988-93 budget surplus and a low level of external debt(billions of U.S. dollars) since 1993 (Finiancial Tinmes, 19 September 1995).

In Thailand the recently adopted strategy for

Infra- All privatizing infrastructure assets was partly aRegion structure sectors response to a rapid rise in investment-from

$2 billion in 1989 to $10 billion in 1993, or 8 per-

Africa 0.1 3.2 cent of GDP-by the fifteen major public utilitiesAsia 7.4 1917 (World Bank 1995j). To keep pace with the invest-Latin America 22.5 55.1 ment demand in infrastructure the governmentEastern Europe and Central Asia 2.0 17.9 is privatizing public utilities in energy, transport,Total 32.0 95.9 water, and telecommunications and creating newNumber of divestitures 267 2,279 regulatory frameworks for private investment inTotal divestitures, 1980-87 51 456 these sectors. In the power sector, for example,

the Electric Generating Company was set up as aSource: World Bank 1995a. subsidiary of the Electricity Generation Authority

16

with an initial generating capacity of 600 mega- and thereby diversify the currency mix of theirwatts. In 1994, 48 percent of the company's earnings profile. The privatized Chilean utilities,shares were privatized through a public offering, for example, have become regional players,and it now competes to build new generation acquiring assets in other countries. Chilean com-capacity. The national authority will in turn iden- panies have played a major part in privatizationtify and bid out generation projects to indepen- of infrastructure in Argentina, and Enersis, thedent power producers using build-operate-own Chilean power company, owns power generat-(BOO) or build-operate-own-transfer (BOOT) ing assets in Argentina and Peru.contracts. The Petroleum Authority also priva-tized a subsidiary, PTI' Exploration, in 1994. By Privatization and capital market development2000 the Electricity Generation Authority, theMetropolitan Electricity Authority, the Tele- In addition to relieving the public sector's mana-phone Organization, and the Petroleum Author- gerial constraints and alleviating fiscal burdens,ity will all be corporatized and up to 50 percent privatization of infrastructure projects helpsof their shares will be offered to domestic and develop capital markets and create a companyforeign investors. In 1993 Thailand's fifteen base that will then seek additional opportunitiesmajor public utilities held combined assets to provide infrastructure services.amounting to about $28 billion, compared with For example, Chile and Malaysia, whichtotal capitalization of the Stock Exchange of embarked on early and extensive infrastructureThailand of about $130 billion. Thus the privati- privatization programs, have been more success-zation of public utilities will give a strong boost ful than other countries in mobilizing domesticto development of Thailand's capital markets funds for new investnent in infrastructure.(World Bank 1995j). According to the International Finance Corpora-

Finally, investments in new infrastructure tion's Global Composite Index, infrastructureprojects often require foreign financing. Devel- stocks account for the largest share of stock mar-oping private domestic infrastructure companies ket capitalization in both countries (Table 2-2).will allow these companies to participate in In addition to development of equity markets,infrastructure projects and investments abroad private infrastructure companies or public enter-

Table 2-2: Industry concentration, end-1994(percent)

Sector with largest Sector's share ofRegion/coutntry share of capitalization market capitalization

Latin AmericaArgentina Mining 38.2Brazila Infrastructure 42.5Chile Infrastructure 50.0Colombia Manufacturing 57.7Mexico Manufacturing 27.5

AsiaIndia Manufacturing 81.7Indonesia Manufacturing 69.3Malaysia Infrastructure 25.6Philippines Manufacturing 35.1Thailand Manufacturing 50.9

IFC Global Composite Index Manufacturing 37.3

a. State-owned monopolies in telecom and electricity accounted for about 70 percent of market capitalization in 1992, with thegiant Telebras accounting for 46 percent of the Bovespa price index.Note: Infrastructure is defined as transportation, communications, and utilities.Sourcc: IFC 1995a.

17

prises with enhanced external incentives make a does not immediately create new capacities andconsiderable contribution to the development of services. Such sensitivities are even strongerdebt markets. These entities can rely on their sta- when infrastructure services are being priva-ble and long-term revenue profile in issuing debt tized, especially when privatization is accompa-securities, especially long-term instruments. nied by price adjustments and shedding ofSuch debt instruments help set important bench- redundant staff. In Argentina, for example,marks for the longer end of debt markets and Ferrocarriles Argentinos, the company responsi-provide attractive investment opportunities for ble for railways and subways, was unbundledcontractual savings institutions such as funded and privatized using twenty-year concessions. Inpension schemes. (The impact of debt issues by 1989, before it was privatized, the companyprivate and public infrastructure companies on employed 92,500 people. By 1994 the privatecapital markets is discussed in greater detail in concessionaires and the authorities that assumedChapter 4 in connection with development of responsibility for its services together employedlocal financial resources.) just over 19,500 people. Of these, only about

7,500 worked for the private concessionaires thatForeign investment in privatization had taken over most freight and urban passenger

services; the rest worked for the three commuterForeign investment in privatization transactions lines then awaiting transfer to a concession-takes place through either portfolio equity aire. Foreign companies from Belgium, Japan,investment or foreign direct investment. In port- Portugal, and the United States are shareholders infolio investment the foreign buyer engages in a most of these concessions along with Argentinepurely financial investment with individual partners.shares not exceeding 10 percent. For investment Foreign investment in privatization should notto qualify as foreign direct investment the inves- be viewed as threatening, however. Properly bal-tor has to acquire 10 percent or more and is usu- anced with mobilization of local funds, it hasally interested in having substantial influence several merits. Because infrastructure facilitiesover the operations of the company (Sader 1993). are usually very expensive and local capitalBox 2-2 describes two innovative portfolio equity markets may be too thin to mobilize sufficientinstruments that foreigners can use to invest in funds, foreign capital can help jump-start thedeveloping country companies. privatization process. In addition, foreign enti-

Foreign participation in a privatization pro- ties that are interested in purchasing substantialgram may encounter greater political obstacles equity in privatized infrastructure are usuallythan new investment undertakings that include corporate entities that are already involved inforeign capital. This is because foreign participa- managing infrastructure investments and facili-tion in privatization is sometimes perceived as a ties, normally in more advanced countries. Insell-off of domestic assets and a transaction that fact, privatization programs can require that key

Box 2-2: Foreign portfolio equity investment

Foreigners can make equity investments in devel- In 1991 Argentina became the first country to useoping countries either directly, by buying shares in GDRs when it privatized Telef6nica de Argentina,a company, or indirectly, by purchasing equity issuing securities with a nominal value of $364 mil-instruments traded in international securities mar- lion. In March 1992, $270.3 million in ADRs andkets. The most commonly used equity instruments GDRs were issued in the sale of 30 percent ofare American Depository Receipts (ADRs) and Telecom Argentina. The largest single issue ofGlobal Depository Receipts (GDRs). ADRs are ADRs was carried out by Mexico in May 1991,negotiable equity-backed instruments that are pub- when the remaining 15 percent of Telefonos delicly traded in the U.S. securities markets, attracting Mexico (Telmex) was privatized for $2.4 billion. AU.S. investors whose ability to directly invest in year later Telmex offered another $1.2 billion inforeign stock markets is limited by law. GDRs func- ADRs.tion like ADRs but can be traded simultaneously indifferent securities exchanges all over the world. Source: Sader 1993.

18

investors have to be just such "strategic" investors. services in a competitive and deregulated envi-Bolivia's recent experience is an example of this ronment was unheard of in most developingapproach (Box 2-3). countries less than a decade ago.

Regulatory changes in industrial countries are The participation of foreign corporate entitiesmaking it easier for their companies to partici- in privatized infrastructure companies in devel-pate in privatization programs in developing oping countries also enhances the possibilitiescountries. The 1992 Energy Policy Act, for exam- for raising the additional capital that is usuallyple, allowed certain types of U.S. utilities to needed to remedy accumulated neglect andinvest in foreign enterprises. As a result U.S. expand service capacities. In 1990 the govern-companies have become both partners and ment of Mexico began selling its majoritv stakeadvisers in privatization programs, especially in in Telmex, the state telephone monopoly, to athe oil and electricity sectors. Elsewhere, French consortium of investors that included Francewater companies are using their experience in Telecom, U.S.-based Southwest Bell, and a groupoperating private water concessions at home to of Mexican investors led by Grupo Carso. Asparticipate in newly private ventures abroad. part of the terms of the sale the new ownersInfrastructure companies from industrial coun- of Telmex committed themselves to investingtries can also help developing country regulators $12-14 billion and to making specific improve-and their private local partners make the transi- ments in phone density and service qualitv. Intion to a less regulated or differently regulated Argentina private companies in the natural gasenvironment. Such assistance may be crucial sector are committed to investing $611 milliongiven that private management of infrastructure during 1994-98.

Box 2-3: What is capitalization?

Capitalization is privatization in which sale pro- investor and the government hold equal stakes inceeds stay with the company to finance future the new company. Under the original plan the gov-investment. If, for example, the net fixed assets of ernment was going to immediately distribute itsan electricity company have a market value of $100 share in equal parts to all adult Bolivians-aboutmillion, a strategic investor would pay $100 mil- 3.2 million people. This component was later modi-lion into the company. The company would now fied in light of logistical difficulties. Instead thebe worth $200 million: $100 million in fixed assets government used the shares to endow pensionand $100 million in cash. This approach helps miti- accounts that were set up for each adult citizen.gate some of the popular doubts about foreign To date the Bolivian airline has been capitalizedinvolvement in domestic privatizations, and by by a Brazilian airline for $47.5 million. The electric-leaving the proceeds with the company helps ease ity company has been unbundled and its three gen-the shortage of cash for working capital and eration companies have been capitalized for $139investment that commonly afflicts popular partici- million (compared with an initial reference price ofpation schemes. $99 million). The telephone company was capital-

The Bolivian government decided to use this ized for $610 million (compared with a referenceapproach in 1993 when it announced that it was price of $135 million) by STET, an Italian telephoneprivatizing six large state enterprises in key sectors, company. Railways were divided into East andincluding infrastructure. Under the plan the enter- West branches and divested in November 1995.prises are divested by international tender. The Mining is set for divestiture by late 1996. Mean-successful bidder pays the agreed price not to the while, new laws and regulations in the power andgovernment but to the company, doubling its net telecom sectors have been passed by the Bolivianworth. The cash must be used for investment in the Parliament to address the postprivatization era.sector, stimulating expansion and efficiency improve-ments together with job creation. Initially, the strategic Source: Ewing and Goldmark 1994. j

19

3Mobilizing International Financing forPrivate Infrastructure

As discussed in Chapter 1, there is a large sector must be a priority. Government agenciesdemand for investment in infrastructure in every that traditionally have received direct budgetarydeveloping region over the next decade. At the allocations and contracted out project construc-same time many developing countries either do tion need to develop the skills required to oper-not have sufficient domestic savings or lack the ate in the context of private investments. Thesemacroeconomic record or institutional and legal agencies also need to acquaint themselves withinfrastructure that are needed to transform the myriad financing sources and sophisticateddomestic savings into appropriate financial financial and legal literature embodied in projectinstruments for financing private infrastructure. finance under build-operate-transfer or compa-As a result foreign financing will continue to play rable arrangements.an important role in new private investments in The longer negotiation time required toinfrastructure. In the long run, economic growth, develop infrastructure projects relative to moregreater volumes of savings, and well-developed traditional forms of direct investment is one offinancial intermediation mechanisms should gen- the factors limiting investment by many of theerate sufficient domestic resources for invest- infrastructure development funds that have beenments in infrastructure. set up to date. Differences in expectations about

Although foreign financing of private infra- actual and perceived risks, reasonable returns,structure projects poses project as well as macro- revenue sharing, risk sharing and management,economic risks,1 these should be weighed against future pricing policies, and so on have led tothe risks of neglecting investments in infrastruc- high transactions costs. Indeed, in Latin Americature. Poor infrastructure is no longer only a project developers face considerable difficulty inretarding factor for fixed investments in domestic negotiating concessions, often working as longresources, but also severely disadvantages devel- as five years without reaching closure. An exam-oping countries in attracting foreign capital. ple is the Carbon II independent power producer

This chapter focuses on how governments can in Mexico, where after years of discussionsfacilitate the mobilization of foreign resources. between the government's energy monopolyWays of managing the financial liabilities that and the project developer both parties walkedgovernments often have to bear in the process are away from the deal (Hamilton, Rabinovitz, andalso discussed, along with ways that such liabili- Alschuler Inc. 1995).ties can be mitigated. For industrial country firms investing in

developing countries, project size combined withNeed for efficient contact between government the complexity of dealing in foreign marketsand investors have made for long and expensive predevelop-

ment phases prior to contract awards. 2 As aImproving governments' ability to negotiate result development costs average about 5 per-infrastructure project financing with the private cent of total project costs. Moreover, failure rates

20

are high, with only 20 percent of private projects government support to private projects. Lessonsunder development reported as likely to come in this area can suggest policy actions andon line (compared with an average of 35 percent reforms that not only help mobilize greaterin the United States). Thus the mere launching of domestic resources, but also keep a close eye onproject development in a complicated adminis- the contingent liabilities of governments.trative and legal environment can be discourag- Finally, these institutions can coordinate theing, especially for small and financially weak development of a project pipeline and allocatedomestic or foreign investors. This trend favors funds for prefeasibility or feasibility studies ofhighly capitalized firms that have the technical priority projects.and financial resources to sustain longer, more Although single-window institutions as de-costly development cycles. scribed above are generally lacking, this concept

In India state governments have been negoti- has been partly executed in a number of devel-ating with private power developers for several oping countries. The Build-Operate-Transferyears, and eight power generation projects have (BOT) Center in the Philippines, which operatesbeen identified as candidates for private invest- under the provisions of the 1990 BOT Law, dealsment. The biggest of these is a 2,450 megawatt with private sector participation in all infrastruc-gas-fired power plant in Dabhol estimated to ture sectors. This law was recently revised tocost $2.5 billion. The project development was include more diverse forms of private sector par-steered through thirty government agencies, 170 ticipation (Table 3-1). The revision of the law toformal approvals, nine court cases, and thirty- make it more flexible in terms of private invest-nine months of grueling negotiations. Despite all ment structures partly reflects the BOT Center'sthis, project construction was halted by the accumulated experience in dealing with inves-newly elected Maharashtra state government in tors and domestic institutions and in identifyingAugust 1995 due to criticisms of the contract issues of mutual concern.negotiated by the previous government; only re- Given the emphasis that it has received fromcently was construction restarted. the highest authorities (including the office of the

In Latin America government agencies with President), the BOT Center should be able tomonopoly power in a particular sector have play an important role in streamlining the pro-resisted negotiations that would force them to cess for approval, financing, construction, andrelinquish their authority to operate and fund operation of private infrastructure projects. Thenew capital expenditures (Hamilton, Rabinovitz, success of such an institution hinges on the wayand Alschuler Inc. 1995). In some countries its relations and responsibilities are defined anddifferent line agencies become quite disorderly executed relative to the line agencies responsiblein trying to promote private sector participation for sectoral planning and regulation.in their respective sectors.3 Such behavior is con- Another example of a unified negotiatingfusing and frustrating to investors, and tends to agency comes from Pakistan, which recentlylead to duplication of similar resources and func- negotiated and finalized several private powertions across different public agencies. project deals. Until 1994 foreign investment in

To alleviate some of these concerns, develop- Pakistan's power sector was hobbled by the needing countries can set up single-window or uni- for protracted negotiations of power purchasefied institutions to negotiate project finance agreements and other project contracts. More-deals with private investors. Unified institu- over, pricing was not competitive and there wastions benefit from economies of scale in terms no central source of reliable information aboutof employing expensive technical resources. concessions and facilities. In response to theseThrough continuous interaction with investors, weaknesses a special commission was created inthese institutions also can obtain a better per- November 1993 to develop an energy policy thatspective on the approval process of domestic would attract foreign investment. The subse-institutions and propose streamlining measures quent policy reforms have drawn many foreignthat lead to more efficient and expeditious pro- developers and resulted in a number of financialcesses. They can arrange or directly provide closures. One element of Pakistan's policy re-training for relevant line agencies and help them form was the establishment of the Private Powernegotiate contracts. They also can act as central- Board within the Ministry of Energy. As the sin-ized information units where crucial informa- gle point of contact between developers and gov-tion is pooled and analyzed. This function is ermnent agencies, the board speeds the approvalparticularly important in terms of managing process and eliminates seemingly endless treks

21

Table 3-1: Schemes included in the Philippine Build-Operate-Transfer Law

Build-lease-transfer * Private sector finances and constructs facility* Government leases facility for fixed period* Government owns facility upon expiration of lease

Build-transfer-operate * Government finances project facility* Contractor builds facility* Contractor operates facility on behalf of government agency

Contract-add-operate * Private sector leases existing government facility* Private sector undertakes expansion or improvement of facility* Private sector operates the facility

Develop-operate-transfer * Private sector undertakes facility* Facility raises property values of adjoining property* Private sector obtains right to develop facility

Rehabilitate-operate-transfer * Private sector rehabilitates, operates, and maintains existing governmentfacility

* Government retains ownership upon expiration of contract

Rehabilitate-own-operate * Private sector rehabilitates existing government facility* Private sector operates facility for indefinite period on the condition that it

does not violate the terms of its franchise

Source: World Bank 1994b.

through layers of government bureaucracy that In 1995 project flows to finance infrastructurewere once typical in Pakistan (Infrastructure projects in developing countries accounted forFinance, February/March 1996). about 1 percent of international flows and about

10 percent of flows to developing countriesInternational capital and project finance flows (Table 3-3). Overall project finance flows to

developing countries were 21 percent higher inWith greater global integration, a significant 1995 than in 1994. East Asia and the Pacificamount of capital is now internationally mobile. attracts the largest share of these flows, withDespite short-term and incidental setbacks,international flows increasingly seek investmentopportunities in developing and emerging mar- Table 3-2: International capital market flows,kets due to lower yields in industrial countries 1993-95and the advantages that come with risk diversifi- (billions of U.S. dollars)cation. Long-term government bond yields in theG-3 markets (Germany, Japan, and the United Instrument 1993 1994 1995States) have fallen by half since 1990. As a resultfund managers are willing to accept greater risksin order to meet the income demands of increas- Bonds 481.0 426.9 460.6ingly sophisticated investors. But although inter- Equity 40.7 44.9 41.0national financial markets have the capacity and Other 160.2 278.9 388.3drive to meet the project financing needs ofdeveloping countries, economic and project level Total 818.6 953.4 1,258.4risks must be sufficiently manageable. Although Flows to developingthe global flow of capital is growing, the share of countries 75.3 89.6 113.3flows to developing countries has remained at Share of total (percent) 9.2 9.4 9.0around 9 percent of total capital market flows inrecent years (Table 3-2). Source: World Bank 1996.

22

Table 3-3: Project finance, 1995 1995 and about half those made for the entire(millions of U.S. dollars) year were for infrastructure projects, with a focus

on power, transportation, telecommunications,Destination 1995 1994Q4 water, and construction.

Region Raising long-term debtEast Asia and the Pacific 15,105 4,690Europe and Central Asia 4,781 2,745 The growing experience with internationalLatin America and the Caribbean 2,750 897 project finance flows suggest that equity is aMiddle East and North Africa 2,083 1,093 more readily available component in the financ-South Asia 1,616 161 ing of infrastructure projects. InfrastructureSub-Saharan Africa 1,068 54 projects usually have high leverage (debt toTotal 27,403 9,639 equity) ratios, and securing large amounts of

Sector long-term debt remains the main financing chal-Power 8,923 1,701 lenge. The number of international emergingTelecommunications 1,117 592 market equity funds in Asia grew from 50 inTransportation 2,189 804 1989 to 242 in 1994; in Latin America the numberOther infrastructure 1,251 1,083 of funds grew from 2 in 1989 to 108 in 1994Noninfrastructure 13,923 5,461 (Figure 3-1). These funds include specialized

infrastructure funds in both regions. Moreover,

Source: Euromoney Loanware and World Bank data. there are numerous project sponsors in mostinfrastructure industries that are willing and ableto take equity risks for equity returns. Match-

China and Indonesia accounting for more than ing funds to projects is not always an easy task,85 percent of the funds committed in the region. however.

Growing private sector participation in infra- For investors to achieve returns on equity it isstructure has been a stimulating factor in global often essential that a project be able to mobilizecapital flows. More than 40 percent of the financ- long-term debt for anywhere up to 80 percent ofing commitments made in the fourth quarter of its cost. Leverage ratios and capital recovery are

Figure 3-1: Net assets of international emerging market equity funds, 1989-94(millions of U. S. dollars)

35,000

30,000 -

25,000 - Asia

20,000 -

15,000 -

10,000 / Latin America ,10,000~~~~~~~~~~~~~~~~~.0

5,000

0- …------… I

1989 1990 1991 1992 1993 1994

Source: IFC 1996.

23

also subject to specific commercial characteristics phase financing usually comes from local andin various subsectors. In telecommunications, for international commercial banks. Except inexample, where shorter technological horizons Malaysia, the role of local commercial banks inand less regulated markets result in greater com- developing countries in financing greenfield pri-petition and price volatility, investors seek and vate infrastructure projects has been very limitedusually can achieve higher returns over shorter due to weaknesses in credit appraisal and financ-periods. Higher returns result in more equity in ing techniques. In addition to institutional weak-the financing composition and therefore a lower nesses, commercial banks in developing coun-leverage ratio. Debt-equity ratios of about 50:50 tries are usually unable to make long-term loansare common in telecommunications investments. because the profile of their liabilities is mostlyRoad and hydroelectric dam projects, on the short-term. This short-term profile of bank liabil-other hand, require large investments with slow ities (deposits) is largely the result of macroeco-capital recovery. The markets for these services nomic instability in many countries, especiallyallow a greater degree of predictability in reve- during the 1980s (see Chapter 4).nues, albeit with less flexibility in tariff adjust- Accordingly, international commercial bankment, thus lowering cash-flow risks. These char- lending through syndicated or individual loansacteristics result in higher leverage ratios, with is an important source of project financing, espe-debt tenors that can stretch as long as twenty- cially during the construction phase. Commer-five years or more to ensure expected returns on cial banks are able to appraise project risks andequity. can provide financial flexibility in response to

Private infrastructure projects usually access construction delays or cost overruns. Commercialdebt financing on a nonrecourse or limited bank lending is usually guaranteed by exportrecourse project finance basis. In nonrecourse credit agencies, which provide pre-completionproject finance creditors rely solely on the cover for varying degrees of commercial andincome and assets of the project itself for repay- political risk.5 Coverage of pre-completion risksment, rather than on the credit of the project helps secure financing for projects.sponsors. Under limited recourse finance the Still, lending secured by export credit agenciesproject sponsors have some limited liability with is not without risk. Most such agencies limit ten-respect to the project company's obligations, ors to ten years (twelve years for power plants),such as timely completion of the project, but do mandate semiannual repayment of principal,not provide a full guarantee of the project com- and allow only a six-month grace period follow-pany's debts, especially after commercial opera- ing project completion. These conditions forcetion begins. The inherent risks associated with the brunt of debt repayment during the earlyfinancing of infrastructure projects are com- phase of a project's life, when cash flows arepounded by the fact that most investments are most uncertain. As a result projects may not gen-very large and capital recovery is achieved over erate sufficient income to retire commercial banka long period. As a result most projects require debts if loan maturities are shorter than required.debt tenors of ten to twenty-five years. More- For instance, financing a thirty-year asset, suchover, a slow rate of capital recovery coupled with as a power plant, over ten years results in powera high expected rate of return on equity (20-30 rates that are not economic and limits creativitypercent) leads to high leverage ratios. In China, in structuring tariffs. In addition, the exportwhere economic performance is strong and the credit agency guarantees often required by com-country enjoys an international investment mercial banks must be backed by the host gov-grade rating, private investment in power has ernment's counter-guarantees. An importantbeen stalled by debate between the government drawback of such arrangements is that govern-and investors about appropriate rates of return. ment counter-guarantees end up covering bothThe government has insisted on capping after- commercial and political risks. To improvetax profits at 15 percent of equity-financed net projects' ability to attract long-term commercialfixed assets, while foreign investors are seeking bank project financing with fewer guarantees,at least an 18 percent return to compensate for policymakers must reduce the elements of riskthe risks involved in investing in the country.4 associated with the construction phase and

Debt financing for private infrastructure improve the regulatory and market environmentprojects is provided either before or after a in which projects will operate. Project contractorsproject's construction is complete. Construction are responsible for guaranteeing the timely

24

completion of projects, but governrnents have to Improving country creditworthinessensure that the operating environment allowsthem to assume such risks. Long-term capital for limited recourse infrastruc-

As mentioned above, some projects require ture projects is also abundant in the internationallonger-term financing than commercial banks bond markets. But the universe of accessibleusually provide to match their revenue profile capital is very narrow for projects that are unableand repayment ability. In such cases commercial to obtain international investment grade creditlenders usually try to ensure that projects are ratings. 6

capable of obtaining takeout financing once the The global market for investment grade debtconstruction or seasoning period ends. In Indo- was $14 trillion in 1994. By contrast, the world-nesia's Paiton private power project, for exam- wide private market for unrated or below invest-ple, commercial lenders extended $360 million ment grade project debt in emerging marketswith a guarantee that once the project was opera- was only about $12 billion (Hamilton, Rabino-tional the U.S. Export-Import Bank would take vitz, and Alschuler Inc. 1995). In the long run,over for the banks and make a direct loan to the then, the degree to which infrastructure invest-project. Multilateral development banks can also ments in developing countries achieve invest-design instruments to mitigate refinancing risks ment grade ratings will determine their ability to(Box 3-1). access international capital markets. Such ratings

To reduce the need for concessionary multilat- certify a project's ability to meet its repaymenteral or bilateral support, projects must be able to obligations under a specific debt issue. An in-demonstrate their ability to raise long-term vestment grade credit rating also lowers the costfunds in capital markets once commercial opera- of debt. Price differentials of 175-350 basis pointstions begin. Thus governments in developing occur in the international capital markets be-countries will not only have to ensure that local tween investment grade paper (BBB- and above)capital markets are strong enough to meet part and anything less. These differentials have criti-of the refinancing requirements, they also have cal implications for the tariff structure of infra-to maintain a macroeconomic framework that is structure projects.not perceived as jeopardizing the ability of Two conditions must be met for infrastructureprojects to float debt on international capital projects to achieve an investment grade rating inmarkets. international capital markets. First, projects

Box 3-1: The World Bank guarantee

A partial risk guarantee covers risks arising from governments take full responsibility for financingnonperformance of sovereign contractual obliga- and thus bear the entire risk in projects.tions to a private sector project or from force The World Bank charges two fees for the guar-majeure events affecting a project. By covering antee cover: a standby fee and a guarantee fee.risks that the market would not bear or would These fees are charged either to the borrower orprice very high, the guarantee lowers the cost of the lender. The standby fee, applied for the periodfinancing. A partial credit guarantee covers events when the guarantee is in force but not callable, isof nonpayment for a designated part of the financ- currently 25 basis points a year on the outstandinging. These guarantees extend maturities beyond amount guaranteed. The guarantee fee, appliedwhat private creditors would otherwise provide- for the period when the guarantee is callable, isfor example, by guaranteeing late payments or by set at 40-100 basis points on the amount coveredproviding incentives for lenders to roll over by the guarantee. This rate includes a fee of 25medium-term loans. basis points plus a premium varying from 15-75

Partial risk guarantees reduce a government's basis points, depending on the level of risk cover-contingent liability to the minimum required to age. The World Bank receives a counter-guaranteemake the project feasible, with the private sector from the relevant member country, and above 25taking on all or much of the commercial risk. This basis points the guarantee fee is passed on to thecontrasts with the traditional pattern in which government.

25

have to be investment grade in local currency. a 1990 amendment to the 1933 U.S. SecuritiesSecond, the credit rating of the country in which Act. The rule established a nonexclusive exemp-a particular project is being financed must be tion from the registration requirements of theinvestment grade. Securities Act so that investors can resell pri-

To illustrate the effects of an investment grade vately placed securities to eligible institutions,rating on private debt issues, Table 3-4 shows referred to as qualified institutional buyers,Standard and Poor's sovereign credit rating for a under certain limitations. By doing so, it enablednumber of Asian and Latin American countries, issuers to access the resources of institutionalas well as the average maturity and average buyers in the U.S. market-insurance compa-interest spreads of bond issues by private sector nies, investment companies, and pensionissuers in these countries. 7 The overall pattern is funds-without the attendant inconvenience ofclear-the higher the ratings, the longer the registering with the Securities and Exchangematurity and the lower the interest spread. 8 Commission or the limitations of the traditional

Thus achieving an investment grade rating is private placement market.essential to raising long-term debt on interna- In a Rule 144A placement an issuer normallytional markets in volumes that are sufficient to sells its securities to one or more investmentmeet investment demands. A government's sov- banking firms, which can then resell the securi-ereign credit rating is decisive to the outcome of ties to a large number of qualified buyers. Byprivate efforts in this regard. Policymakers must facilitating the resale of securities issued in pri-secure a favorable rating by sustaining policies vate placements, Rule 144A creates liquidity inthat ensure macroeconomic stability and efficient the secondary market, making the U. S. privateexternal asset and liability management. Policies placement market more attractive to foreignalso should promote stable and transparent reg- issuers, including those offering debt securitiesulatory frameworks and create market environ- to finance infrastructure projects. Prior to thements that allow economic cost recovery and adoption of Rule 144A, privately placed securi-provide reasonable parameters for determining ties were subject to significant restrictions onproject cash flows (in local currency) over the life resale, making them more illiquid.of projects. Securitizations of the operating revenues of

Mexican toll roads, beginning with the TolucaRule 144A: A window for long-term U.S. dollar debt toll road in 1993, were among the first develop-

ing country transactions in this market. The firstNearly all bond issues by private infrastructure greenfield project in a developing country to tapprojects in developing countries have taken this market for long-term debt was the Philip-place in the U.S. capital market under Rule 144A, pines' Subic Bay Power Plant, owned by the

Subic Power Corporation. This project raised$105 million in debt through a bond issue in 1994

Table 3-4: Interest rate spread and maturity for at a spread of 385 basis points over the U. S. Trea-unenhanced private sector bond issues, 1994 sury rate, with a maturity of fifteen years. Ihe

bond issue was closely timed with the start of theStandard Average Yield spread project operation and the proceeds went to repay

and Poor's maturity at launch a portion of the advances made by sponsors dur-Country rating (years) (basis points) ing the development and construction phases.

The Centragas natural gas pipeline in ColombiaMalaysia A+ 10.0 89 was the first greenfield project in Latin AmericaThailand A- 5.9 134 to issue bonds in this market. The $172 millionChile BBB+ 5.0 125 Centragas debt issue in December 1994 had aColombia BBB- 10.0 641 spread of 300 basis points over the U. S. TreasuryPhilippines BB 8.7 300 rate, with a final maturity of sixteen years. TheArgentina BB- 4.5 410 proceeds of this issue were used to financeBrazil B 2.9 387 project development and construction.

Note: Enhancement refers to such features as bond-equity Special investment and financing toolsconversion options, collateralization, guarantees, and putoptions. Given the issues and challenges involved in

Source: IMF 1995a. mobilizing international financing, developing

26

countries may need to design special financing lier stages of reforms with higher credit riskschannels to improve their access to capital for these vehicles act as intermediaries of fundsprivate infrastructure projects during the transi- from the international financial institutions and,tion to reduced sovereign risks and secure local in the process, help leverage additional flowsmarkets. This is where international financial from other sources (Box 3-2). Most of the fundsinstitutions, bilateral or multilateral, can step in. channeled through these mechanisms are raisedThese institutions help countries secure access to on concessionary terms. In countries that havelimited recourse project financing by providing made more progress with reforms (the Philip-either direct loans or guarantees of syndicated pines, Colombia; see below) support is either inbank loans or bond issues. As policy and institu- the form of guarantees or contingent lines oftional reforms take hold, the need for support credit, and normal commercial markets are pro-from these institutions diminishes (Figure 3-2). posed as the main sources for raising funds. TheLoans and guarantees may both be necessary financial products offered also become moreearly on, but the share of guarantees will gradu- complex since they are designed to mitigate risksally increase and loans can be provided through that may be slowing market flows.contingent commitments (for example, for take-out financing). THE PHILIPPINES' PRIVATE SECTOR INFRASTRUC-

To address the financing needs of limited TURE DEVELOPMENT FUND. Prudent economicrecourse projects in developing countries, espe- policies have fostered stability in the Philippines.cially those with non-investment grade sover- During 1990-95 the average annual inflation rateeign ratings, the World Bank has designed guar- was about 9 percent, allowing a gradual fall inantee instruments (see Box 3-1). World Bank real interest rates. The Philippine economy grewguarantees have been used for a number of pri- by an average of 5 percent a year during 1994-95vate infrastructure projects, where they have while posting strong export growth. Althoughmade financing more affordable and helped the country does not yet have an investmentextend the maturity of loans and bond issues. grade rating, the government is actively promot-

In several cases the World Bank has also ing private infrastructure and plans to privatizehelped member countries develop special infra- public enterprises under new laws that allowstructure financing vehicles. By providing finan- private investment (see Table 3-1).9cial support to develop private infrastructure The initial impetus for private participation induring the transition to strong domestic markets, infrastructure development came from a need tothe"e vehicles act as intermediaries for extending overcome crisis conditions. A severe electricitygovernment and international financial institu- shortage peaked in 1992-93, when brownoutstion support to infrastructure projects financed averaged seven hours a day in many regions andon a limited recourse basis. These vehicles also power shortages caused about $3 billion a yearhelp to attract sizable cofinancing from other in losses to the economy. A fast-track programsources. The structure of support provided by introduced in 1993 allowed independent powerthese vehicles varies. In countries that are at ear- producers to expand generation capacity. By the

Figure 3-2: The transition to an investment grade credit rating

Unrated/Non-investment grade Economic stability

j i . ~~~~Structural reforrns ~~~~~~.................................. ........ ................

PerceivedIFI support non-market

________ z ; ri ~~~~~~~~~~nks

. ~~~Access to limited!recourse financing Investment

grade

27

Box 3-2: Preliminary financing tools for private infrastructure in Pakistan, Sri Lanka, and Jamaica

Pakistan. Pakistan's Private Sector Energy Devel- opment Association (IDA) and $14 million fromopment Fund was created in 1988 to support insti- Germany's KfW. Operations began after the IDAtution building and to provide subordinated debt credit was approved in June 1996. A pipeline offinancing for limited recourse private power projects recommended for funding includes the 150projects. The fund was initially capitalized with a megawatt Kelanitissa power plant, a container ter-World Bank loan of $150 million cofinanced with minal, a wharf, and a 30-kilometer expressway.$150 million from the Export-Import Bank of Japan Loans made in U.S. dollars to private sponsors will(Jexim), as well as $99 million in loans provided by be at variable and fixed rates and will have maturi-France, Italy, and the United States for equipment ties of up to twenty-two years, including up toto be sourced from these countries. The fund, eight years' grace.administered by the National Development Financ- The company should give momentum to pri-ing Corporation, was replenished in January 1995 vate financing of infrastructure in Sri Lanka. Thewith a $250 million loan from the World Bank and a country's strong export potential in textiles is held$110 million Jexim loan (raised to $250 million in back by infrastructure bottlenecks, and 75 percentMay 1996). France also provided an additional $10 of industries and hotels produce their own power.million loan toward purchases of French equipment. Independent power producers could easily satisfy

The fund provides subordinated debt financing this demand at lower cost.(with up to twenty-three years' maturity with eightyears of grace) for up to 30 percent of the financing Jamaica. Jamaica's Private Sector Energy Fund wasof private energy projects. Project sponsors are also designed to promote limited recourse privateexpected to mobilize 20-30 percent of project funds investments in infrastructure. The World Bank andin equity and to raise the remaining funding as the Inter-American Development Bank each pro-senior debt. Although private power projects were vided $40.5 million loans to help set up the fund.the fund's initial focus, it is now financing other This money was used to provide a commitment toprivate infrastructure, and was recently designated refinance the commercial construction debt of theas Pakistan's Long-Term Infrastructure Credit Rockfort private power project. The project isFacility. In 1997 a newly created financial institu- slated to begin operations in August 1996.tion with a majority shareholding by the private In 1998 the project can call a takeout loan fromsector will be assigned to administer this facility. By the fund that will have a twelve-year maturity1998 the private sector will control about a third of with no grace period and a fixed rate equal to apower generation capacity and supply nearly half thirty-year U.S. Treasury bond plus 300 basisof Pakistan's power. points. This arrangement does not prevent the

project from testing the market for more favorableSri Lanka. The World Bank is helping the Sri Lankan takeout financing, however. The development ofgovemment promote limited recourse project the Rockfort power project and the accompanyingfinancing through a Private Sector Infrastructure changes in the policy and regulatory environmentDevelopment Company. This company, modeled for the private provision of power have had anafter Pakistan's Private Sector Energy Develop- important demonstration effect, thereby improv-ment Fund, has an all-debt capital structure that ing the prospects for future private power projectsincludes $70 million from the International Devel- in Jamaica.

end of 1995 thirty-one projects (worth $4.5 bil- Philippines in early 1995 as part of an ongoinglion) were providing 4,500 megawatts of new Private Sector Infrastructure Initiative that aimsgenerating capacity. Seven projects ($3.3 billion) to improve the policy and regulatory environ-currently under construction will add 2,750 ment for private infrastructure, improve the gov-megawatts of new generating capacity in 1996 ermnent's facilitation and promotion of privateand 1997. By 1998 nearly 80 percent of power projects, and enable equitable assignment ofgeneration will be in private hands. This is a project risk between the government and the pri-marked transformation from 100 percent public vate sector. The World Bank and the Philippinegeneration in 1991. government worked together throughout 1995

The idea of creating a privately managed developing the facility, called the Private Sectorinfrastructure financing facility emerged in the Infrastructure Development Fund (Figure 3-3).

28

Figure 3-3: Proposed structure for the Private Sector Infrastructure Development Fund in thePhilippines

Foreign DomesticInvestors InvestorsDoUlar Bonds Peso Bonds

World Bank/ tConvertibility Guarantee PSIDFOther Multilaterals

Long-term Debt

Greenfield transportation andwater projects, and.powiblepower project refinandngs

As originally designed, the fund was to be an COLOMBIA'S PRIVATE SECTOR INFRASTRUCTUREinvestment grade financing vehicle. It would FINANCING FACILITY. Weak planning and bud-have generated a portfolio of projects with an geting and inefficient public managementinvestment grade rating in local currency that a resulted in the deterioration of infrastructure ser-World Bank convertibility guarantee would have vices and undermined economic productivity inhelped achieve an investment grade rating on Colombia as the 1980s came to an end. This trendinternational capital markets. The fund's basic has been reversed to a considerable degree byobjectives were to mobilize affordable long-term regulatory reforms and economic plans that haveforeign debt for relatively small infrastructure divested public infrastructure assets and enabledprojects (costing $150 million or less),1 0 address private investment. Colombia is in a transitionthe financing needs of infrastructure projects in phase in terms of attracting long-term limitedsectors other than power where export finance recourse financing for infrastructure projects.credit may not be as readily available (such as Local banking and capital market capacities aretransport and water), and become an efficient gradually developing as a result of considerableintermediary for local funds, thus stimulating reforms in the 1990s, but these markets lack suffi-the development of long-term local debt markets. cient depth. Moreover, interest rates in local

The original concept of the fund is being reex- markets are high and the ability to manage risksamined in light of the fast-growing strength of in long-term financial assets is limited.normal funding channels in the market. Contin- Helped by Colombia's investment grade creditued progress in financial sector and capital mar- rating, private developers have accessed interna-ket reforms-including the granting of permis- tional banking and capital markets for severalsion to ten foreign banks to open branches in the projects, but the current and projected demandPhilippines, reforms in taxation of financial for new investments exceed current flows.1 1

instruments, and the issuing of government Since 1993 the volume of private flows has beensecurities with up to five years' maturity-have about one-fifth the levels targeted in the govern-improved the ability of local markets to meet ment's infrastructure development program. Aspart of the financing demand for infrastructure a result the Colombian government, in coopera-projects. Most of these reforms took place in tion with the World Bank, is developing a Private1995. The continued strong performance of the Sector Infrastructure Financing Facility to cata-Philippine economy will further ensure the lyze private flows into infrastructure. To meet itsavailability of foreign project financing while the objective, the facility is being developed to offertransition is made to fully developed local capital a range of products that address the concernsmarkets. Progress on these fronts has reduced and needs of private investors in Colombia'sthe fund's priority for the Philippine government. investment environment (Figure 3-4).

29

Figure 3-4: Private Sector Infrastructure Financing Facility in Colombia

Bilateral & MultilateralAgencies

World Bank PossibleCoffinancing

+ PrivateEL..L-_ _ l __Line of - Financial infrastructure

Credit jPSi7I'F Products Projects

t Repayment ManagementGuarantee

G6vernmontlo f Colomb:;i: Bacod:

The facility's capitalizationi and managenment. more efficient mechanism of extending and man-World Bank financing for the facility has been set aging sovereign risk guarantees to private projects.at $300 million. A contribution by the Colombian Products. As noted earlier, private investment

government, if made, could ease the facility's in infrastructure is taking place, but the pace iscurrency risk exposure to the extent that it pro- insufficient to achieve targeted investment lev-vides financing in local currency. A consortium els. Moreover, despite financial sector and pen-of Salomon Brothers and McKinsey & Co. is sion system reforms that hold out the promise ofworking alongside a government-World Bank increased local financing, foreign financing isteam to advise the government on the structur- still dominant. The facility's product mix will being and implementation of the facility. The gov- designed not only to address concerns relevanterminent has assigned management of the facility to foreign financiers but also to capitalize onto Bancoldex, which is a second-tier, largely local financial sector and capital market reformsgovernment-owned entity specializing in export and leverage existing capacities. It will do so byfinance. Among goverunent entities, Bancoldex providing products that are currently unavail-has the strongest financial position and the most able to interested foreign financiers and fledglingadaptable charter for accomplishing the objec- local long-term debt markets. The facility willtives of the facility. Its mandate was recently provide products that address political risksexpanded to promote infrastructure develop- (such as exchange risk) or systemic risks (such asment as part of efforts to improve export perfor- pervasive illiquidity in the debt market) but willmance. The project will include training to refrain from taking purely commercial project-strengthen Bancoldex's ability to administer the specific risks. Two specific products that havefacility, notably in the area of project finance, been identified are refinancing commitmentsGiven the progress to date, the facility could and foreign exchange liquidity support.begin operations in early 1997. The governlment As in many other developing countries, lend-is also working with the World Bank to develop a ers in Colombia are sometimes reluctant to

30

extend credit to infrastructure projects because In the event of a sudden and sharp deprecia-of uncertainties about refinancing possibilities. tion, projects may not be able to adjust tariffsThis uncertainty prevents credit markets from rapidly enough to keep pace with higherextending long-term debt. Most locally available exchange rate costs. Thus they may face a periodcredit must be repaid within three years, and of illiquidity. Foreign exchange shortages andloans from international banks are generally inconvertibility at the national level also maylimited to seven years. Capital recovery takes prevent projects from meeting their obligationslonger for infrastructure projects and, to ensure to foreign financiers. To address these concerns,competitive returns on equity, debt financing the facility will rely on foreign exchange lines ofusually constitutes up to 80 percent of the credit from the World Bank and other multilat-financing package. As a result repayment of eral or bilateral sources to provide commitmentsdebt, under feasible tariff levels, may take much of foreign exchange liquidity at times of severelonger than the maturities currentlv available in depreciation shocks. The lines of credit would bethe market. Once projects pass the construction drawn down as the commitments are called andphase and are operating successfully, their the proceeds made available to projects to ser-chances of accessing longer-term funds increase. vice financial obligations in foreign currency.The facility will mitigate the concerns of lenders P3rojects that call their commitments will con-to greenfield projects by providing refinancing tinue to be responsible for repaying their finan-commitments. Projects or their lenders can pur- cial obligations to the facility in foreign currency.chase these commitments and thus obtain assur-ance that refinancing will be available. Refinanc- Managing sovereign risk guaranteesing terms and conditions will be specified in thecommitments. While providing refinancing Private investors will continue to seek sovereignsecurity, the preagreed rates would be deter- guarantees for a variety of market (commercial)mined such that the market is the first resort for and nonmarket (noncommercial) risks that theyre-financing. Also, to be able to exercise their feel incapable of mitigating or managing. Therights under refinancing commitments, projects long-term nature of infrastructure investmentswill have to maintain appropriate operational increases the exposure of investors to such risks.and financial standards (for example, debt- Market risks are difficult to manage in under-equity and debt service coverage ratios). This developed markets, especially when they stemwill ensure that refinancing commitments from public enterprises that are unable to meetmitigate the risks associated with major sys- their supply and offtake commitments totemic disruptions in normal market funding projects. An uncreditworthy public enterprisechannels. that dominates demand for an output such as

The risk of disruptive shocks in foreign power cannot be regarded as a viable demandexchange markets can limit the willingness of market. Similarly, private investors often dealforeign investors to lend to domestic projects, exclusively with public entities on the supplyespecially when these projects only earn revenues side without much confidence in the uninter-in local currency. The December 1994 Mexican rupted performance and quality of service pro-peso crisis, for example, ended efforts to issue debt vided by these institutions. In both cases privatein a Rule 144A placement for the $752 million investors would likely seek a sovereign guaran-Termobarranquilla power generation project, and tee before committing funds. It may also bethe financing was secured elsewhere. Colombia's necessary for governments to provide sovereignforeign exchange reserves are adequate relative guarantees to stimulate investment in areasto its import levels. The depreciation rate of the where market demand is uncertain. [nvestors inColombian peso has been relatively stable. greenfield toll road projects, for example, may seekExcept for 1994, it ranged between 15 and 30 per- government guarantees on minimum traffic levels.cent every year between 1990 and 1995. Thus the In many countries the laws and regulationsrisk of abrupt and disruptive depreciation and that are being introduced to promote privateinconvertibility does not appear to be significant participation in infrastructure have no historicalin the short to medium term. Infrastructure precedent and the long-term political commit-investments span over an extended time hori- ment to maintaining relevant reforms remainszon, however, and foreign investors perceive this untested. As a result the role of governments andas a real risk over the longer term. regulatory bodies in influencing the timing of the

31

construction phase or maintaining a stable and local currency depreciation and convertibilityreasonable regulatory framework during the risks. It would be optimistic to assume that a riskoperation phase are usually perceived as non- such as sharp and sudden depreciation can sim-market risks for which investors will seek miti- ply be offset by compensating tariff adjustmentsgation through sovereign guarantees. without wreaking sociopolitical havoc and jeopar-

Reforms aimed at strengthening the market dizing the pattern of demand. Political circum-discipline for infrastructure services will gradu- stances that may result in civil unrest or national-ally improve the credit assessment of infrastruc- ization of projects also cannot be disregarded. Theture projects in local currency and obviate the various risks that foreign investors in infra-need for sovereign guarantees. These reforms structure projects are concerned about are listed ininclude transparency and efficiency in the laws Box 3-3. Except for currency risks, these categoriesand regulations governing the ownership and of risk also affect the incentives of local investors.operation of private infrastructure facilities. Tariff The risks other than commercial risks are fre-reforms intended to allow economic pricing are quently referred to as political risks, which isalso key to the commercial viability of projects. meant to capture the idea that they are controlledPublic enterprises need to play by market rules more by governments than by the private sector.and obtain impartial credit assessment. Govern- A positive investment climate is essential forments need to develop active plans for restructur- countries that hope to attract long-term foreigning and privatizing public enterprises in order to capital. Since perceptions of a country's invest-reduce the need to guarantee their commitments. ment climate are based on both current circum-

At the national level an improved sovereign stances and future prospects, it is essential notcredit rating can be achieved through a stable only that legal and policy conditions are condu-political and macroeconomic framework. Strong cive to the attraction and utilization of foreigneconomic fundamentals should be established capital, but also that foreign investors are confi-through macroeconomic policies that ensure mon- dent that there is little or no possibility that an un-etary stability, macroeconomic transparency, and favorable legal environment will develop. Inves-export-oriented growth. Efficient structural poli- tors gain such assurance when a favorable legalcies will benefit from macroeconomic stability and situation has existed for a sufficiently long time orsupport it through efficient allocation of resources. when a country's economic and political structure

Private investors that make long-term commit- is sufficiently stable that there is little probabilityments to infrastructure projects face considerable of radical change in the foreseeable future.

Box 3-3: Risks in infrastructure investments

Risk type Refers to

Commercial Cost overruns in construction and operation, uncertainties in demandfor services.

Noncommercial

CurrencyConversion The ability and willingness of the sovereign government to convert its

domestic currency into foreign exchange.

Exchange rate Changes in the value of one currency relative to the other over time.

Policy Adverse conditions that are imposed on a project due to arbitrary or ad hocchanges in the regulatory, legal, and economic policy framework.

Political force majeure Wars, political unrest and turmoil.

32

Many government efforts to influence the com- These guarantees must be issued in a way thatmercial viability of public enterprises will have does not hard-code institutional or sectoralidentical effects on private investment. Govern- arrangements that may later be subject to reform.ment performance in making and keeping impor- Other guarantees, such as convertibility guaran-tant promises influences private investors' per- tees, could be removed when countries achieveceptions of the environment in which they will be investment grade credit ratings.working. In many developing countries govern- In issuing guarantees, governments must con-ments have implemented programs designed to sider the expected value of their contingent liabili-remove nonmarket barriers affecting the opera- ties. Large liabilities raise perceptions of countrytional efficiency of public entities. But a World risk and imperil the sovereign credit rating. ThisBank study of governments' adherence to the outcome can be avoided if contingent liabilitiespromises that were made as part of these pro- are issued in a fiscally responsible framework, 12

grams suggests that in many cases governments with appropriate distribution of risks among par-fail to keep their word (Box 3-4). When govern- ties that are best positioned to mitigate them.ment prornises are not credible, it undermines the By valuing and budgeting for their contingentcreditworthiness of the public or private provid- liabilities, governments can maintain the initialers of infrastructure services. value of past and future guarantees and reduce

When providing sovereign guarantees, govern- effective demand for guarantees. Guarantees thatments can reach agreements under which they are not properly accounted for have several nega-effectively end their guarantee obligations by tive implications that are rarely considered. Thelinking them to reforms. In this way progress in main role of guarantees is to mitigate risks thatimplementing reforms could eliminate the under- would otherwise increase the cost and tenor oflying need for guarantees. Such agreements could capital beyond feasible levels.stipulate that guarantees will be lifted as public Investors that receive sovereign guaranteesagencies whose commercial undertakings affect place a value on them and expect this value to bethe financial viability of private projects are priva- preserved over time. Every subsequent guaranteetized or obtain a good independent credit rating. that a government issues will dilute the value of

Box 3-4: Government attempts to reform public entities

Entity Promises on which government reneged in whole or in part

Ghana Electricity Support collection of revenues and arrears; timely approval of tariffs.

Ghana Telecoms Timely approval of tariffs.

Ghana Water Prompt payment of bills by government and public enterprises.

India Electricity (promises Support collection of arrears; increase autonomy to invest; streamline red tape;by state governments) prompt approvals; increase tariffs.

India Oil Increase autonomy to invest; streamline red tape; prompt approvals.

Mexico Electricity Provision of support.

Senegal Electricity Prompt payment of bills by government and public enterprises; contribute toinvestment program; increase tariffs.

Senegal Telecoms Prompt payment of bills by government and public enterprises; increase tariffs;abolish subsidy to postal service by 1989; limit investments in rural areas.

Source: World Bank 1995a.

33

previously issued guarantees. This dilution effectnot only undermines the value of assets held Box 3-5: Sovereign guarantee reform inby earlier investors, 13 it also poses a bias against Colombiafuture investors who have access to less valuableguarantees As part of its efforts to develop a facility for

It is also important to note the impact of infrastructure financing, the Colombian govern-reforms on the value of contingent liabilities. ment is working with a World Bank team and, . ~~~~private consultants (Oliver Wyman & Co.) toWhen reforms are implemented in areas where pnat coslat (Oie.ya o)t

design an efficient system for issuing sovereignguarantees were issued to compensate for percep- guarantees and managing the accompanyingtion gaps and uncertainties, they effectively obligations. The objective of this exercise is toincrease the value of each guarantee. As reforms value the exposure to the government of finan-are implemented, the possibilities that previous cial guarantees, create a budgeting process thatguarantees will be invoked is reduced. Box 3-5 makes agencies offering the guarantees econom-describes an ongoing effort in Colombia that aims ically accountable, and ensure that adequateto develop an efficient mechanism for issuing and funds are available to pay the guarantees.managing sovereign guarantees. One product of these efforts will be a model

that enables the government to assign values toits guarantees of distinct risks, market and non-market, in each project. These values will beused to identify and prioritize reforms, theabsence of which requires government guaran-tees and accompanying budgetary allocations.Moreover, with better-defined liabilities the gov-ernment can work proactively to prevent eventsthat trigger the need for guarantees. This exer-cise will also introduce advanced risk manage-ment concepts into the government's budgetaryprocess.

34

Notes

1. The aggregate foreign exchange conversion demand of a large num- To further clarify ratings, Moody's assigns numbers from 1 (highest)ber of projects could become substantial. This demand should be bal- to 3 to differentiate borrowers within a given grade. Standard andanced through policies that promote export-oriented growth as dis- Poor's ratings of foreign currency debt are subject to a sovereigncussed in an earlier section. credit rating ceiling, which reflects the risk that, in a crisis, a country

may impose currency controls or impose restrictions on the ability of2. Development costs include travel, accommodation, communica- the project to pay foreign debt holders.tions, and international offire facilities as well as other materialexpenses that are incurred in the process of maintaining contact with 7. The interest spread is the difference between the interest rate on thelocal and international institutions and financiers, plus staff and con- bond at the time of issuance and the rate on U.S. Treasury securities ofsultancy costs incurred in management, legal, and financial work. comparable maturity

3. Line agencies are the government institutions, such as ministries, 8. Perceptions about greater risks in Latin America, the specific debtthat have traditionally been in charge of direct investment in infra- issues (which may have been concentrated in riskier undertakings),structure projects and delivery of services through parastatals. These and country conditions in 1994 partly explain why Colombia's overallagencies will most likely remain central as far as planning, prioritiza- standing was less favorable than that of the Philippines despite itstion, contracting, and sometimes regulation of private infrastructure higher credit rating.projects are concerned.

9. Moody's upgraded its rating from Ba3 to Ba2 in May 1994; Standard4. A new build-operate-transfer law expected to pass in 1996 will and Poor's upgraded its rating from BE- to BB in June 1995.likely help resolve the differences between the government and inves-tors. 10. Given the fixed costs associated with issuing debt on international

capital markets, the transactions costs for debt issues of less thin $1005. The U. S. Export-Import Bank is an exception and insists that com- million are usually prohibitive.mercial risks during construction are best handled by the private sec-tor-that is, by commercial banks and project sponsors. 11. Colombia was upgraded to Bal by Moody's in 1994 and to BBB- by

S&P in 1995.6. The credit ratings of the two best-known rating agencies, Moody'sand Standard and Poor's, are described below: 12. That is, the guarantees have to be valued and capital to back them

has to be allocated through the budgetary process.Moody's Standard and Poor's

13. Creditors, such as holders of fixed-income securities issued by theInvestment grade Aaa, Aa, A, Baa AAA, AA+, AA, project, price their funds according to their expectations of risk. Cove-

AA-, A+, A, A-, nants are never signed with governments restricting them to maxi-BBB+, BBB, BBB- mum exposure to liabilities resulting from future guarantees. When

governments overexpose themselves, the risks in assets held by credi-Noninvestment grade Ba, B BB+, BB, BB-, B+, B, B- tors increase. This will affect the value of these assets in seconldary

markets, whether bonds or loans.Default grade Caa, Ca, C, D CCC+, CCC, CCC-,

CC, C

35

4Developing Local Financial Resources

During 1966-94 funds from foreign sources their obligations to creditors. Any negativeaccounted for at least 70 percent of the financing movements in the exchange rate bear directly onfor greenfield infrastructure projects in develop- the financial viability of the project. If tariffs areing countries (IFC 1994). Because the revenues linked to the exchange rate, movements maygenerated by infrastructure projects are usually force tariff adjustments to generate adequate rev-in local currency, the foreign investors that enues for debt service. Sharp exchange ratefinance these projects often face sizable and vola- movements result from a variety of shocks,tile risks such as exchange rate and convertibility including political. These shocks can force tariffrisks. These risks affect a project's ability to meet adjustments that could otherwise be partlyits financial obligations. avoided if economic fundamentals were

In the long run local resources should provide attended to. In Mexico, for example, the tariffsmost or all of the financing needs of private for most toll roads constructed during 1988-94infrastructure projects. Developing countries were indexed to inflation but not to the exchangeshould recognize the risks associated with over- rate. After the peso crisis most projects werereliance on foreign financing and work to unable to service their obligations to foreigndevelop a long-term domestic debt market. The creditors due to the sharp devaluation of theemphasis on debt markets in this chapter does peso. Higher tariffs substantially reduced trafficnot in any way suggest that equity markets in all volume on many roads, and the governmentdeveloping countries are sufficiently developed, eventually stepped in to bail out nearly all thethough some countries are quite advanced in this consortiums involved in road concessions. 1

regard. However, mobilization of long-term local In some cases tariff levels are quoted and pay-currency debt remains a major challenge for able in foreign exchange by a purchasing statefinancing private infrastructure, even in coun- enterprise, such as in the power purchase agree-tries where local equity is more readily available. ment between the National Power Corporation

of the Philippines and the Subic Bay Power Cor-Implications of inadequate local financing poration. The Philippine government has guar-

anteed the payment of the capacity charge underThe use of foreign resources and the risks that it the agreement and therefore shoulders a sub-entails have implications at both the project and stantial exchange rate risk over the life of thenational levels. At the project level the main project.2issue is the potential mismatch between assets At the national level large-scale reliance onand liabilities. Even with a realizable stream of foreign savings could become a source of eco-local currency assets, foreign-financed projects nomic volatility. Increasingly integrated interna-face considerable exchange rate risks in meeting tional financial markets have given a substantial

36

degree of mobility to capital. This means that infrastructure investments have achieved ancapital flows can reverse in the event of a per- important degree of insulation from exchangeceived crisis, forcing countries into financial and rate risks and will not place a long-term convert-economic turmoil. This concern is particularly ibility demand on foreign exchange markets.relevant to foreign portfolio investments, where Not all developing countries seeking to pro-a drying up of foreign flows can jeopardize the mote private infrastructure can copy Malaysia'srealization of targeted investments. experience. But they should recognize the impor-

When foreign savings are mobilized to finance tance of promoting domestic financial marketsinfrastructure projects, the foreign exchange con- within a stable macroeconomic framework. Stepsversion required over the life of the projects may to develop and deepen local financial marketsbecome a source of rigidity in a country's exter- may not immediately increase savings rates innal accounts. Foreign funds that are used to countries where current savings rates do notfinance equity and debt components of large meet the demand for investment in infrastruc-infrastructure projects require that substantial ture. But these reforms will help transform exist-proportions of project revenues be converted ing savings into financial assets with longerand expatriated over the periods of debt matu- maturities. This transformation will promoterity and concession rights (some as long as thirty medium- to long-term investments that will inyears). The required outflow is effectively equal turn generate larger pools of savings throughto the foreign equity plus the net return on that higher and steadier income growth. To moveequity and the foreign exchange debt service along this path, developing countries must pro-payments over the life of the project.3 mote the role of commercial banks, develop

The foreign exchange demand that is gener- long-term bond markets, and reform the contrac-ated by foreign currency financing of infrastruc- tual savings sector. A stable, low-inflationture assets is not necessarily compensated by environment is essential to developing thesematching net inflows from exports. Many coun- capacities.tries that lack developed local capital markets(and thus depend on foreign financing of infra- The inflation challengestructure) are also implementing economicadjustment and reform programs. These pro- Economic instability is high on the list of factorsgrams are intended to improve transparency in that discourage the private sector from commit-resource allocation and emphasize economic ting its resources to commercial activities. Clear-comparative advantages, improving signals for ly, the longer is the duration of an investmentexport-oriented investment and growth. The undertaking, the greater is the need for economicresponse of the export sector may not, however, stability and reasonable predictability. Infra-be as swift as expected or required. structure investments require a much longer sta-

In fact, large inflows of foreign financing for bility horizon than other commercial activities,infrastructure privatizations or new investments such as agriculture, manufacturing, and nonin-may actually hurt the very export sector that frastructure services. To be financially viable atthese projects depend on for foreign exchange. affordable tariffs or user charges, most infra-Mexico, for example, had sizable inflows in the structure projects require financing of betweenearly 1990s, including funds directed to infra- ten and twenty-five years. The private sector'sstructure privatization and new infrastructure willingness to commit its resources and under-investment. But the peso appreciation caused by take long-term financial obligations dependsthese inflows resulted in imports growing faster directly on the stability of the policy and institu-than exports during the first half of the 1990s. tional framework of the country where theFor Latin America as a whole exports grew by 9 investment is to take place.percent in 1994 while imports increased by 12 High and volatile inflation rates, whichpercent (Naim 1995). shorten the investment horizon and severely dis-

By contrast, Malaysia, which has comple- rupt economic activity, are a major sign of insta-mented sound economic policies with timely bility. High inflation not only makes long-termfinancial sector reforms (including in the capital economic and financial planning less certain, itmarkets), has been able to mobilize substantial also creates major problems for long-terndomestic resources to finance large private infra- finance. Although long-term investment in astructure projects. As a result locally financed high-inflation environment can be funded by

37

long-term loans provided a high nominal inter- rate bonds are among the most suitable. Infra-est rate is charged, inflation-adjusted interest structure projects tend to have stable earningsrates are not a viable solution for long-term profiles over extended periods of time, provid-financing. Such rates do not provide investors ing a degree of predictability for future earnings.with a realistic repayment schedule-one that The earnings profile may not, however, be ablecan be supported by the cash flow from invest- to follow sharp swings, as may be required byments financed with the loan. Although high sudden shifts in interest rates, since in manynominal interest rates in high-inflation econo- cases tariff levels are regulated or tariff adjust-mies provide adequate real compensation to ments require official approval. These featureslenders and reasonable real total costs to borrow- are compatible with the financing profile ofers, they create problematic cash flows, espe- fixed-rate bonds.cially for infrastructure projects. Front compres- North American and European capital mar-sion of the loan repayment profile in a high- kets were formed when large public companies,inflation environment is generally incompatible most of which are now considered public utili-with the stable cash-flow profile of infrastructure ties, first issued securities. Developing countriesprojects. In fact, high rates of inflation transform have generally avoided this possibility by financ-all long-term nominal interest loans into short- ing these public works with government funds.term loans, making them unsuitable for financ- The recent shift toward privatization and mar-ing long-term investments (Box 4-1). A record of ket-based financing of such services in develop-low and stable inflation is required to extend a ing countries shows how the financing demandscountry's financial horizon. It is also a decisive of companies investing in or providing infra-factor in developing long-term fixed-income structure services can benefit from well-developedsecurities (see below). bond markets while also contributing to the

development of these markets by providingBond markets and infrastructure finance long-term investment outlets.

Fixed-rate bond prices are primarily deter-Of the various instruments that can be used to mined by interest rate movements in the cur-finance infrastructure projects, long-term fixed- rency of the bond's denomination. Thus a stable

Box 4-1: Long-term finance in an inflationary setting

The following formula gives the nominal interest nominal interest for the lender to achieve a 10 per-rate (i) in the presence of an inflation rate (f), and an cent real increase in the buying power of the loanedexpected real interest rate (Ri): amount. That is, $1 to retain the buying power of

the principal amount and $0.2 to provide the buy-(1+ i) = (1 +f(l+ Ri) ing power of the $0.1 interest at the end of the first

year.Thus an expected real interest rate of 10 percent

in the presence of no inflation requires a nominal Distribution of loan repayments under tworate of 10 percent, while an expected real interest inflation profilesrate of 10 percent in the presence of 100 percent (percentage of total)annual inflation rate requires a nominal rate of 120percent. 100 percent

A high nominal rate causes two loans of equal En100 intamounts and with similar maturity profiles ($5,000loans, repaid in five end-year $1,000 payments) tohave different repayment profiles, as shown in the 1 23.0 60.3table. 2 21.5 25.0

The front compression of repayment cash flow in 3 20.0 9.9the presence of 100 percent inflation is due to the 4 18.5 3.6fact that each $1 of loan would have to earn $1.2 of 5 17.0 1.2

Total 100.0 100.0Source: Mozes 1995.

38

macroeconomic environment in that currency's establishing and strengthening professional andhome country is important to the proper func- widely owned credit rating agencies, creating effi-tioning of a long-term bond market. Inflation is cient secondary market trading institutions to en-perhaps the biggest worry for investors who hance liquidity, and lowering transactions costs.purchase fixed-income securities such as bonds. Until 1988 Argentina's unstable macro-Not only does inflation erode the purchasing economic environment and hyperinflation pre-power of the principal portion of the investment, cluded the development of long-term bondit also affects interest rates, which can cause sub- markets. But recent improvements in economicstantial capital losses. As a result investments in management-through macroeconomic andlong-term fixed-rate bonds such as those used to structural reforms, as well as regulatory and taxfinance infrastructure projects are subject to both reforms specific to bond market development-the performance of the specific project and to a have created the potential for growth of thesehost of macroeconomic trends over an extended markets. Even before Mexico's peso crisis bondperiod. One way investors protect themselves maturities in Argentina were short of the rangesfrom excessive risk is by limiting the terms of the suitable for financing infrastructure projects;bonds they buy-for example, three to five years most of the 105 bond issues registered with therather than ten to fifteen years. Floating rate National Securities Commission by Septemberinstruments also protect investors against inter- 1993 had two- to four-year maturities. Moreover,est rate and inflationary movements. But these corporate debt issues in the bond markets wereinstruments need to be complemented with limited to $50 million, with maturities of noderivatives such as interest rate swaps in order longer than five years (World Bank 1994a). Theto offer a predictable cost profile for borrowers, recently reformed pension and social securityespecially for highly leveraged infrastructure system should, however, become a major sourceprojects. Long-term swap markets are, however, of demand for long-term debt instruments.naturally as rare as long-term bond markets in Between mid-1994 and early 1996 private pensiondeveloping countries. funds accumulated about $2.5 billion of assets,

mostly in government securities and bank timeProspects in Latin America deposits (Vittas 1995a).

Bond markets are also growing in Colombia,Given these criteria, it should come as no sur- although the total size of the corporate bondprise that domestic bond markets are either market is still far short of even the $2 billion inabsent or at an infant stage (Argentina, Chile,Colombia, Mexico) in Latin America. In 1992 thecorporate share of bond issues was 0.82 of GDPin Argentina and 0.11 percent of GDP in Brazil.Still, the Mexican peso crisis notwithstanding. Box 4-2: Better policy responses in Latineconomic management in Latin America has Americaundergone a considerable transformation. Fiscal When Mexico defaulted on its internationaland economic reforms such as budget cuts, debts in 1982, the flow of extemal funds to theprivatization, improved tax collection, and trade region came to an almost complete halt. Latinliberalization have improved the chances of sus- American governments responded by closingtained economic stability The policy response to their economies even more, tightening their gripexternal shocks also has undergone a marked on the private sector, imposing pervasive eco-shift (Box 4-2). As a result seventeen of twenty- nomic controls, and in some cases by nationaliz-two Latin American countries had single-digit ing banking systems. In 1995, however, the ini-inflation rates in 1995, and average inflation in tial reaction of the Argentine, Brazilian, andthe region (excluding Brazil, which began an Mexican governments to the backlash producedadjustment plan in 1994) dropped from 130 per- by Mexico's crash was to deepen marketcent in 1989 to 14 percent in 1994. Given the eco- reforms, accelerate privatization plans, boostcentmin 1989ili to at 14 s percen n 199Gieven futhre efiscal accounts, fine-tune foreign exchange re-nomic stability that has been achieved, future gimes, step up efforts to promote exports, andefforts to develop long-term bond markets strengthen private banks.should focus on the regulatory framework andinstitutional infrastructure, including eliminating Source: Naim 1995.tax disincentives for holding and trading bonds,

39

private annual investments targeted in the toward privatization of infrastructure services1994-98 infrastructure development program. and new investment by the private sector notThe interest rate on the latest issues of five-year only reduces the demand for budgetary outlaysgovernrnent bonds has fallen to 25 percent (from that might generate deficits, it also facilitates and27 percent in 1995), benefiting from declining accelerates the pace of corporate issues and theinflation. Still, this represents a real interest rate development of bond markets. Both Malaysiaof around 5 percent. But prudent fiscal policies, a and Thailand have used their bond markets tostable and declining inflation rate, growing raise substantial local financing for their privati-demand for long-term bonds by infrastructure zation programs as well as for new privateprojects, and recent reforms in the pension sys- investments in infrastructure.tem hold promise for the development of long-term bond markets in the next few years. MORE-ADVANCED MARKETS. In Malaysia the

In Mexico the prospects for developing long- issuance of debt securities increased by 160 per-term bond markets depend on the restoration of cent to $3.2 billion in 1994, mainly as a result ofmacroeconomic stability in the wake of the the increase in huge projects (such as power gen-December 1994 peso crisis. From 7 percent in 1994, erating facilities and roads) undertaken by theinflation jumped to around 35 percent in 1995. private sector. The $3.4 billion North-SouthMoreover, the exchange rate depreciated by about Expressway, which was financed entirely by90 percent in 1995. The resulting fluctuations in local funds, was able to raise the ringgit equiva-the treasury bill rate are shown in Figure 4-1. lent of $400 million by issuing convertible bonds.

This issue gave investors a low-risk utility-typeEmerging bond markets in East Asia asset with healthy capital gains prospects once

the concessionaire company became listed. TheIn emerging East Asian economies the prospects purchase by the Employees Provident Fund offor further developing and deepening bond mar- the ringgit equivalent of $550 million of fixed-kets benefit from much stronger economic fun- rate bonds as part of the financing for the $1.5damentals. In some East Asian countries histori- billion Lumut combined-cycle power project iscally low inflation and prudent fiscal policies another example of a bond market satisfying thehave already created a viable capacity in bond financing needs of large infrastructure projectsmarkets (Table 4-1). while providing an investment outlet for institu-

Government bonds still dominate most East tional investors looking for alternatives to now-Asian bond markets, but fiscal balance or sur- infrequent issues of government securities. Theplus in many countries has allowed rapid expan- rest of the financing for the Lumut projectsion of corporate bond volume. The move included the ringgit equivalent of $400 million in

Figure 4-1: Mexican treasury bill rates(percent)

70

60-

50-

40 -

30 -

20-

10

0

1994Q1 1994Q2 1994Q3 1994Q4 1995Q1 1995Q2 1995Q3

Source: IMF data.

40

Table 4-1: The emerging East Asian bond market, end-1994(billions of U.S. dollars)

ShareNational State State Central of GDP

Country government government enterprises bank Corporate Total (percent)

China 24.4 0.0 9.0 0.0 0.0 33.3 6.5Hong Kong 0.1 0.0 0.0 6.8 4.7 11.5 8.7Indonesia 0.0 0.0 1.5 6.8 0.7 9.1 5.8Korea, Rep. of 27.9 2.7 37.4 32.1 60.8 161.0 42.8Malaysia 29.0 0.0 3.7 2.0 4.9 39.5 56.0Philippines 24.8 0.0 0.1 0.2 0.0 25.1 39.3Singapore 42.3 0.0 0.0 0.0 2.5 44.9 72.4Thailand 2.6 0.0 7.6 0.0 3.5 13.7 9.8

Total 151.0 2.7 59.3 47.9 77.1 338.0

Source: World Bank 1995b.

equity from Malaysian sponsors and $550 million restrictions, as well as the decline in outstandingfrom a syndicate of Malaysian commercial banks. government issues as the government's fiscal

As recently as 1990 almost 91 percent of the position shifted to surplus in 1988, caused thebonds outstanding in Thailand were government Thai bond market to shrink in the late 1980s. Inissues; the rest were state enterprise issues. Cor- 1992 a new Public Companies Act and Securitiesporate bond issues were severely restricted by and Exchange Law allowed all public and pri-the Public Company Law, which allowed only vate companies to issue bonds. As a result thepublic enterprises listed on the Stock Exchange size and composition of the Thai bond marketof Thailand to issue debt instruments. Private changed, with the share of corporate bondscompanies and unlisted public enterprises were growing from 3 percent of the market in 1992 toprohibited from selling bonds to the public. Such 26 percent in 1994 (Table 4-2).

Table 4-2: Thailand's domestic bond market, 1988-94(billions of U.S. dollars)

Year Government State enterprises Corporate Total

1988 8.42 0.42 0.00 9.83(95) (5) (100)

1989 7.82 0.47 0.00 9.29(94) (6) (100)

1990 7.62 0.71 0.00 9.33(91) (9) (100)

1991 5.91 1.97 0.00 8.88(75) (25) (100)

1992 5.27 3.00 0.20 9.47(62) (35) (3) (100)

1993 3.98 5.33 1.04 11.35(38) (52) (10) (100)

1994 2.63 7.56 3.49 14.67(18) (56) (26) (100)

Note: Numbers in parentheses are percentages.Source: World Bank 1995f.

41

The rapid growth of the corporate segment of raised 1 billion baht ($25 million) using a couponthe Thai bond market has enabled greater local of 9.62 percent and was followed by the Metro-financing of infrastructure privatization and politan Waterworks, which raised 1.6 billion bahtinvestments. In 1994 the Thai government set up ($64 million) at 10 percent, and the Rapid Transitthe Electricity Generating Company with an ini- Authority, which raised 2 billion baht ($80 mil-tial generating capacity of 600 megawatts. That lion) at 9.99 percent. The maturity of bondssame year, half of the company's shares were issued by state enterprises ranges between threeprivatized through a public offering that raised and ten years. Some of these bond issues are inthe baht equivalent of $180 million.4 The com- the form of floating rate notes.pany is now bidding for new generation capac- Thailand demonstrates how positive shifts inity, with the Rayong Power Plant as the first of the government's fiscal position plus regulatorysuch ventures. To acquire the Rayong plant the reforms can improve the climate for long-termElectricity Generating Company borrowed about financial instruments and free up resources to$470 million (including bond issues and syndi- finance more productive sectors. The recent suc-cated loans) from local and international sources. cess of public enterprises and private investorsOf this, the baht equivalent of $240 million was in accessing a greater volume of resources inraised in local markets, half of it through a ten- Thai markets is the result of continued surplusesyear fixed-rate bond issue. in the government's fiscal position, which have

Public enterprises, especially in infrastructure fostered macroeconomic stability and low infla-and utilities, have been central to the development tion. Thailand's inflation rate has almost halvedof Asian bond markets. As noted earlier, except since the fiscal deficit turned into fiscal surplusin the Philippines, state enterprises in East Asia in 1988, with average inflation of 4.94 percentare more efficient than their counterparts in most during 1989-94. The Thai bond market is alsoother developing countries. They also have bene- preferred to commercial banks, whose fundsfited from a sound economic framework that has are more expensive. Whereas domestic bankskept inefficient pricing to a minimum. As a result charged an average interest rate of 13-14 percentthese entities can participate in capital markets a year in 1994 to prime corporate borrowers, thewithout requiring preferential treatment while annual cost to similar corporations of issuing aalso helping to set benchmarks for long-term five-year bond was about 12 percent. Bonds alsosecurities. offer higher returns to investors-an A-rated

The Korea Electric Power Corporation, for bond with a five-year maturity yields about 12example, periodically issues bonds to raise reve- percent a year, compared with a 6 percent returnnue to expand power generating facilities. These on savings deposits and 10 percent interest onbonds have three- to five-year maturities, and one-year deposits.their interest rates are similar to corporate bonds.The nominal value of Electric Power Corporation LESS-ADVANCED MARKETS. A lack of appropriatebonds issued between 1983 and 1994 was about benchmark rates is a major drawback of the$7.1 billion. Similarly, highway construction underdeveloped bond markets in China, Indone-bonds are issued with three-year maturities and sia, and the Philippines. In the absence of reliabletrade like corporate bonds. Other such bonds, benchmarks it is difficult to price long-term cor-referred to as special bonds, include Korea Tele- porate debt and impossible to price floating ratecommunications Corporation bonds, Korea Gas debt. The prevalence of U.S. Treasury rates andCorporation bonds, and Public Waterworks the London Inter-Bank Offered Rate (LIBOR) inbonds. pricing U.S. dollar debt issues in the U.S. and

Thai state enterprises are also major players in Eurodollar markets illustrates this. Market-basedthe bond market. Faced with massive capital benchmarks enable market participants to priceexpenditure bills and the government's strategy bonds in both primary and secondary markets.of financial self-sufficiency, these enterprises Such benchmarks may include securities issuedhave turned to local capital markets for funds. by a national mortgage corporation, state enter-Between 1990 and 1991 the value of state enter- prise bonds, or repackaging of existing debtprise bonds quintupled. The growing number of issues of major state-owned entities.state enterprise bonds being issued serve as In the Philippines private sponsors haveimportant benchmarks for the rest of the market. invested in many greenfield infrastructureIn April 1995 the Government Housing Bank projects, mostly in power generation. At least

42

Figure 4-2: Outstanding Philippine government securities by maturity, 1983-94(percentage of total)

100

80 - Short-term _ _ _ _ -_

* 60 - -

Long-term

2020

1983 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994

Year

80 percent of private infrastructure finance has resulting in a slightly lower interest rate (15 7/ -

come from foreign sources. The share of exter- 16 7/8 percent) than was projected. Rates movednally financed debt is even higher-more than 90 down again for the July issue (12 5/8 -13 percent).percent-because the local financial sector and Longer-term government issues have helpedcapital markets are unable to offer long-term extend the maturity of the benchmark yieldfunds at appropriate cost and maturities (Price curve in the Philippine debt market.Waterhouse 1995). Government short-term paper The tax regime has been a major impedimentdominates domestic securities issues, though to the development of the bond market in thesome longer issues have recently taken place Philippines. Specifically, a 0.5 percent documen-(Figure 4-2). In the past the high interest rates tary stamp tax made borrowing through bondspaid on risk-free assets, such as short-term gov- more expensive than loans from commercialernment treasury bills, prevented the develop- banks. Furthermore, the tax inhibited the devel-ment of a long-term debt market because private opment of a secondary market for debt becauseissuers could not compete. every sale of a debt instrument was subject to the

Prudent macroeconomic management has facil- tax. The government recently replaced the stampitated the gradual development of the Philippines' tax with the newly expanded value-added tax.bond market. Expenditure controls and in- Since borrowers can deduct the value-added taxcreased resource mobilization have significantly as an input tax credit, the distortions caused byreduced the public sector deficit, with the fiscal the documentary stamp tax regime have beenposition running a surplus since 1994. In addi- removed.tion, restructuring enabled the Central Bank to Nondiscriminatory treatment that levels theconduct more effective monetary management. playing field between bonds and other financialAs a result of such measures annual inflation fell investments (for example, in the taxation of divi-from 18.7 percent in 1991 to 8.5 percent during dends, interest, and capital gains and the levy of1993-95. withholding taxes) is essential to promoting the

Since 1993 the government has issued floating development of bond markets as an efficientrate notes with two- to three-year maturities. financing source for corporations. Some institu-These notes, priced at par with an interest cou- tional issues relating to bond market develop-pon based on ninety-one-day bill rates plus a ment are discussed in Box 4-3.fixed spread, have been very successful and themarket is now well established. In June 1995 Role of credit rating agenciesdebt markets took another important step for-ward with the first issue of five-year fixed-rate Infrastructure projects often issue bonds thattreasury notes by the government. The 3 billion stretch out to fifteen years or beyond. Purchas-peso ($117 million) issue was oversubscribed, ers of such long-term instruments must have

43

Box 4-3: Institutional aspects of bond Box 4-4: The first rating agencymarket development

The rating system was introduced in the U.S.Bond markets cannot grow without the institu- bond market in 1909, when John Moody pub-tional infrastructure needed to trade accurately, lished the first debt ratings as part of his Manualquickly, and securely. Institutional arrange- of Railroad Securities. The ratings were basedments should usually be provided by private on reports on the financial quality of some 250market participants but might, in the initial major U.S. railroad companies. To allow users ofstages, need some official encouragement and the manual to quickly compare the quality ofsupport. There are several key elements of such each of the several thousand bonds that theinfrastructure. First, payment, transfer, and cus- companies had outstanding, Moody assignedtody systems should be reliable and incur mini- each bond a rating. His A-through-C system hasmrum transaction costs for investors and issuers since become the world standard. The impor-of bonds. Second, market regulations for issuing tance of ratings increased greatly in U.S. capitaland trading bonds should require full disclosure markets after the Great Depression of the 1930s,of financial information by issuers. Pricing when the perception of risk of investing in fixed-information on bid and offer prices is also desir- income securities was underscored by high lev-able. Third, training and education are impor- els of defaults.tant. Public and private entities in developingmarkets will need to educate the public and Source: Moody's Investors Service 1992.train market professionals about such conceptsas yield to maturity, duration, and convexity.The task is more demanding if the intention is toestablish a market in derivatives or repurchases. For regulators, ratings can serve as an inde-

pendent yardstick for determining the eligibilitySource: World Bank 1995f. of some issuers to offer new debt securities.

Many regulators also use independent credit rat-ings to help them monitor the financial sound-

adequate assurance about liquidity risks. Second- ness of the organizations they are responsible for,ary market trading plays a crucial role in this ranging from banks to insurance companies andrespect. In addition to a sound policy frame- public utilities. Finally, rating agencies enhancework, a well-functioning secondary market the transparency of the market by providing aneeds an efficient supporting institutional infra- common language for credit risk evaluation and bystructure, including market-based benchmarks, contributing to the market's common body of ana-well-capitalized market-makers, efficient clear- lytical information on the quality of debt issuers.ing and settlement systems, credit rating agen- In many countries central banks or govern-cies, liquidity supports, and central information ment agencies check the credit quality of com-systems where all parties can access pertinent panies to determine their eligibility to enter themarket information. nation's fixed-income markets or to monitor the

Qualified independent credit rating agencies credit quality of market participants over time.are one of the most important elements of a Transferring these functions to an independentcomprehensive bond market (Box 4-4). Investors credit rating system helps lower regulatoryuse ratings to quickly determine which securities costs while removing an obstacle to marketmeet their risk guidelines. By using the ratings to development.assemble their portfolio, investors are exposed to The major rating agencies in eight Asian coun-less uncertainty about investment risks for a tries are listed in Table 4-3. The quality and per-much wider range of securities than the credit formance of these agencies depend on their own-staff of any single institution could follow. Debt ership structure and the general structure of theissuers use ratings to provide wider and more securities market, particularly the mode of gov-stable access to capital and to reduce their bor- ermnent regulation and intervention.rowing costs, since investors wi]l accept lower Rating Agency Malaysia is the most successfulyields for higher-grade ratings. Credit ratings local rating agency in the region. By March 31,also help merchant bankers and other financial 1995, it had issued ratings for about $10 billionintermediaries in the structuring, pricing, and equivalent of private debt securities. Insti-placement of securities on behalf of their clients. tutional investors in Malaysia, such as the

44

Table 4-3: Major local credit rating agencies in sentation from a more diverse set of institutionsAsia -especially leading financial intermediaries

who will distribute the debt issues in the pri-

Country Agency mary and secondary markets-would signifi-cantly improve the bureau's operations. The

China Che.xing Securite Rbureau also needs to strengthen its monitoring ofChinaaCheong StecuritieRal ting rated issues, since by mid-1995 only 5 percent of

Rating Company the initial ratings had been upgraded or down-Hong Kong Moody's (local office) graded. Until recently Korean rating agencies5

Standard and Poor's (local office) were constrained by the requirement that allIndia Credit Rating Information Services of bonds be guaranteed by financial institutions. 6

India Limited Such requirements create an investment cultureKorea, Rep. of Korean Investors Service that emphasizes a reliance on guarantees ratherMalaysia Rating Agency Malaysia Berhad than on measuring and managing credit risks,Philippines Credit Information Bureau Inc. and so marginalizes the rating agencies.Singapore Moody's (local office) Finally, government support is essential to the

Standard and Poor's (local office) establishment and development of rating agen-Thailand Thai Rating and Information Services cies. Rating agencies are service-oriented enter-

prises. Their commercial success depends on thedemand for their services and the price users are

Employees Provident Fund, recently have bene- willing to pay. In industrial countries demandfited from more flexible investment guidelines for rating agency services is firmly established,allowing them to invest in rated private securi- and almost all rating agencies have been profit-ties, including investments in the North-South able. In Asian emerging markets only two ratingHighway and the Lumut power project. For agencies-Credit Rating Information Services offuture development of the Malaysian bond mar- India Limited and Rating Agency Malaysia-ket, which has reached a relatively mature level were profitable by mid-1994, while the Philip-and size, it is important that more competition be pine rating agency posted its first profit in 1994introduced in the rating business. A competitive after being in existence for more than a decade.rating industry would help ensure continued Governments can help promote rating agenciesobjective ratings and prevent rent-seeking. using indirect means, such as by instituting regu-

Two Chinese rating agencies have received lations that use ratings as a measurement of risk.approval from the People's Bank of China to As noted earlier, government regulatory authori-operate at the national level. The bank also has ties will in turn benefit from the rating agencies'approved eighty-two local credit rating agencies; credit assessment and monitoring activities. Rat-some of the local agencies are essentially ing agencies are particularly useful in helpingaccounting or consulting firms. Rating agencies regulatory authorities to construct prudentialare used to help select the enterprises that are guidelines for investment by the contractual sav-granted permission to issue bonds. Since the ings sector, especially pension funds.parameters of bond markets in China are still setby the government through a credit plan, the rat- Developing pension funds for infrastructureing agencies are of limited value to the govern- financement and investors. Furthermore, investors donot place much weight on the rating agencies' Pension funds are among a class of relativelyassessments because of an excess supply of new nonbank financial intermediaries in mostinvestible funds and because state-owned enter- developing countries. 7 They sell employees andprises rarely fail. As long as socialized owner- self-employed people secondary securities in theship allows enterprises to operate under soft form of contractual agreements that provide forbudget constraints, the risk assessment role of benefit payments upon the participant's retire-rating agencies will remain marginal. ment. Because their benefits are to be paid in the

In the Philippines the Credit Information future, the secondary securities (that is, the liabil-Bureau has been in operation since 1982. The ities of the pension funds) are effectively longbureau is controlled by a group of companies term and the primary securities (their assets) canwho are in effect also the issuers. Allowing repre- and should also be long term. Pension funds are

45

part of the contractual savings sector. Contrac- growth of pension funds and other institutionaltual savings are any transaction in which eco- investors. Figures on the total pension assets ofnomic agents enter into a binding arrangement selected industrial countries are shown in Table 4-4.with institutions, public or private, to trade cur- The emerging market economies in Asia haverent consumption for future income. Contractual already begun to implement the policies thatsavings institutions (life insurance companies, made contractual savings institutions so promi-occupational pension schemes, national provi- nent in saving and investment in industrialdent funds, and funded social security systems) countries. In 1994 domestic institutional inves-have long-term and fairly predictable liabilities. tors in Asian countries, including mutual funds,

The institutions comprising the contractual held about $109 billion (World Bank 1995c).savings sector are often referred to as institu- Malaysia and Singapore accounted for 70 percenttional investors because of their role as investors of this amount, which suggests the enormousin capital markets. They are potentially good potential for saving and investment throughsources of finance for investment in corporate these institutions in other countries of the region.1 0

bonds and equities. Contractual savings institu- Trends in the asset accumulation of contractualtions acquire funds from the public at periodic savings institutions in selected developing andintervals on a contractual basis, transform these industrial countries are shown in Table 4-5.contributions or premiums into assets, and use Once legal and regulatory reforms have led tothe earnings to pay out the benefits and claims the establishment of public or private institu-on the policies. The savings mobilized, particu- tions such as funded pension schemes, it islarly in life insurance and funded pension funds important that these institutions be subject toschemes, are long-term, illiquid, 8 stable, regular, investment policies and regulations that allownoninflationary, 9 and to some extent accurately them to maximize their returns for the benefit ofdeterminable, depending on the precision of the their members. In many developing countries,actuarial base. As such, contractual savings insti- however, the preemptive use of these funds bytutions can play a key role in capital market governments (through requirements to invest indevelopment. government securities and low-return social sec-

In higher-income countries these institutions tors) has been a major impediment to the devel-are major investors in the securities market, espe- opment of contractual savings as a source ofcially in long-term debt instruments. They pro- long-term corporate finance. Government bor-vide savers with opportunities to diversify risk rowing from contractual savings institutionsand with the benefits of investing in a portfolio deprives markets of long-term funds, limitingselected by professional investment managers. A equity investment, stock market growth, and1993 report, commissioned by the U.S. Congress, credit to the private sector.on financing future infrastructure in the UnitedStates cited further mobilization of the resourcesof institutional investors as a priority (CPIAI Table 4-4: Pension assets in selected industrial1993). The report recognized institutional in- countries, 1989 and 1994vestors not only as potential sources of capital (billions of U.S. dollars)

but also as players in infrastructure finance thatcan bring the discipline of investment risk and 1989 1994return evaluations to infrastructure decision- Countrymaking. Also, when institutions such as pensionfunds grow in size and relative importance, new United States 2,426 3,760

instruments are developed to meet their needs Japan 513 1,118, . . . United Kingdon 453 775and to fill perceived gaps in the market. For ited Kingdom. . ~~~~~~~~~Netherlands 202 264instance, the development of securitization and C anda 101 238

financial derivatives in the United States has been Switzerland 133 191

attributed, at least in part, to the investment and Germany 84 124

risk management needs of pension funds and Australia 42 82other institutional investors. The emergence of Sweden 60 78block trading and the reform of stock exchanges Denmark 33 53around the world, including the abolition of fixedcommissions, can also be partly attributed to the Source: Intersec Research Corporation, Stanford, CT.

46

Table 4-5: Trends in asset accumulation incontractual savings, 1970-94 Box 4-5: Savings potential of funded(percentage of GNP) pension schemes

Country 1970 1990 1994 A social security system's potential for mobiliz-ing savings depends on the financing technique

b adopted, the maturity of the system, the demo-Chilea 1 26 43 graphic structure of the affiliated population,Malaysiaa 18 41 47 the short-term or long-term nature of the risksNetherlands 45 107 covered, and the impact of the economic envi-Singaporea 28c 76 ronment, such as the effects of inflation on ero-South Africa 20 84 sion of the real value of reserve funds.Switzerland 51 110 If labor income represents 50 percent ofUnited Kingdom 43 97 national income, 40 percent of the labor force isUnited States 43 75 covered by social security, and the accumulation

rate is 1 percent, pension schemes will accumu-

a. Figures are percentages of GDP. late funds equal to 2 percent of national incomeb. 1981. each year. If the nominal rate of return on fundc. 1976. assets is equal to the nominal growth rate ofSource: Vittas 1995b. GNP, accumulated resources will amount to 10

percent of GNP over five years and 20 percentover ten years. After the first ten years the accu-

Under prudential investrnent regulations pen- mulation rate will be affected by the growingsion funds have a great potential to collect and volume of benefits payments. But expandedaccumulate savings (Box 4-5). If the growth in coverage, an increase in the share of laborassets of these institutions in developing coun- income in total income, a higher contributiontries that have implemented economic and insti- rate, or a higher investment return will all tendtutional reforms is any indication, the volume of to increase the accumulation rate.accumulated resources can more than meet thesecountries' infrastructure investment gap within Source: Vittas 1995b.five to ten years. As an example, funding all thepensions in Latin America to the level necessaryto pay retirees (64 and older) in 2025 the 1996 per development of national capital markets andcapita income would require 9 percent of re- more generally to the financial and monetarygional GDP. This amount, nearly $110 billion a development of developing countries. When theyyear, is well in excess of projected infrastructure are free to choose among assets-within practicalfinancing needs (Chrisney 1995). and efficient prudential regulation-contractual

These benefits are, of course, realized within savings institutions try to maximize their risk-the confines of the prudential regulations that adjusted returns, for which long-term fixed-define the investment parameters of these insti- income instruments are the most logical choice.tutions. Prudential thresholds will have to be They also attempt to diversify their investmentsapplied to the level of investment in higher-risk in these instruments, including into infrastructure.securities relative to government securities or Countries that have already reformed theirbank time deposits. The status of capital markets pension institutions are reaping considerablein each country will help guide the development benefits in terms of mobilizing local savings toof regulations that allow diversification of finance infrastructure investments. For instance,investments across a broader array of assets. Chilean pension funds, which were privatizedLegislation on the investment portfolio of pen- and reformed in the early 1980s, had total assetssion funds should reflect efforts to maximize exceeding $22 billion (43 percent of GDP) byyield, or at least to invest in secure and profitable 1994. Holdings included 55 percent of govern-assets. Relaxation of investment rules involves ment bonds (including securities issued by thelowering the ceiling on government bonds and central bank), 62 percent of mortgage bonds, andgranting permission to invest in equities, and 59 percent of corporate bonds. The funds alsoultimately to invest in overseas assets. accounted for 11 percent of corporate equities

As mentioned earlier, contractual savings insti- and 9 percent of bank deposits. The pensiontutions could contribute substantially to the funds played an instrumental role in the

47

privatization of several utilities in the mid-1980s,and now hold 10-35 percent of the equity of Box 4-6: Investment rules for Chileanprivatized utilities. These equities, worth nearly pension funds$6 billion, account for 83 percent of the fund'sequity holdings. The funds were also an impor- Chilean pension funds' investments are subjecttant source of financing for the Santiago subway to limits by class of instruments and issuer,system and the national telephone company. By expressed as a percentage of total assets of the1995 pension funds and insurance companies fund or as a percentage of the liabilities (or

owe 45pretecompanis equity) of the issuer. These limits have beenowned 14.5 percent of the telephone company's revised frequently. Initially no investments inshares (out of total equity of $1.5 billion equiva- corporate equities were allowed, but later suchlent). Chilean pension funds recently received per- investments were permitted subject to increas-mission to invest abroad because they have more ing limits (currently 30 percent of the fund'sor less run out of domestic opportunities (Box 4-6). assets). Pension funds are not allowed to place

Malaysia also has done well in promoting the more than 5 percent of their assets in the equityrole of contractual savings institutions in mobi- of a single company or to acquire more thanlizing domestic savings and directing it toward 7 percent of the equity of a single companylong-term investments in infrastructure (Table 4-6). These limits are substantially lower for compa-The Employees Provident Fund dominates this nies with concentrated ownership and illiquidsector, holding more than 70 percent of total con- securities.seacto, holing mr than7rcn Investment in overseas assets was originally

prohibited but recently has been allowed up to aAs a major holder of Malaysian government rather low 10 percent limit. Actual holdings of

securities with maturities up to fifteen years and overseas securities amount to less than 1 percentbeyond, the Employees Provident Fund's invest- of total assets.ment profile is quite compatible with the financ-ing requirements of infrastructure projects.11 Source: Vittas 1995a.Accordingly, with the removal of certain regula-tory constraints on investment rules, the fundhas become an active player in financing infra-structure projects (Box 4-7). The contractual savings sector in the

The North-South Expressway and the Lumut Philippines comprises social security, occupa-Power Station are two prominent examples of the tional pension, and insurance institutions. At thefund's participation in the financing of private end of 1993 the sector's total assets wereinfrastructure projects. Along with a syndicate of equivalent to 17.6 percent of GDP, or more thanforty-six other financial institutions, the fund one-quarter of the market capitalization of thehelped finance the ringgit equivalent of $1.8 bil- issues listed on the country's two stocklion of debt for the North-South Expressway, its exchanges; current assets are estimated at 21first involvement in project financing. As part of percent of GDP. The assets of the Social Security$1.1 billion of debt financing for the Lumut System and the Government Social Insuranceproject, the fund purchased the ringgit equivalent System account for about 60 percent of theseof $550 million of fifteen-year fixed-rate bonds. assets (Table 4-7).

Table 4-6: Assets of Malaysia's provident, pension, and insurance funds, 1989-94(percentage of GDP)

Type of fund 1989 1990 1991 1992 1993 1994

Employees Provident Fund 40.00 40.34 41.00 42.09 44.54 45.68Other provident funds 4.27 4.31 4.08 4.19 6.07 6.65Life insurance funds 5.92 6.03 6.46 6.08 7.06 8.05General insurance funds 1.84 1.98 2.31 2.64 3.01 3.30Total 52.04 52.67 53.85 55.00 60.67 63.68

Total (billions of U.S. dollars) 21.44 24.44 28.0 32.56 39.56 47.12

Source: Bank Negara Malaysia and World Bank data.

48

Box 4-7: Malaysia's Employees Provident Fund: From noninflationary financing for thegovernment to financing the private sector

The Employees Provident Fund (EPF) is Malaysia's issues of government paper fell from $12 billion inlargest provident and pension fund and largest 1986-90 to about $4.5 billion in 1991-94. As a resultinstitutional investor. By the end of 1994 it repre- government issues can no longer fulfill the EPF'ssented the second largest group of financial institu- 70 percent statutory requirement. Recognizing thetions, lagging behind only the commercial banks. investment constraints faced by the fund, the scopeIn its early years the fund was required to invest its of the fund's permissible investments was widenedresources as deposits with the commercial banks. It in 1991 to cover new areas, including real property,was later permitted to invest in Malaysian govern- privatized projects, joint ventures, and bills ofment securities, deposits with banks, and loans to exchange. In addition, the new act requires theapproved companies in the form of private debt fund to invest only half of its annual investiblesecurities and trustee shares. At least 70 percent of funds in government securities, provided that theits annual investible and cumulative funds had to outstanding amount of government securities doesbe invested in government securities. not fall below 70 percent of its cumulative

The government's borrowing needs have fallen investible funds.in recent years, however, due to its improved finan-cial position and downsized operations. Gross Source: World Bank 1995d.

Although the role of the contractual savings * Exempting the social security funds fromsector in generating savings has improved, these withholding tax.institutions have considerable potential for gen- * Abolishing the 5 percent tax premium onerating additional savings. Some of the World the life insurance industry and allowing taxBank's recommendations for contractual savings deductions for contributions to private pen-sector reforms in the Philippines include: sions (subject to an overall ceiling).

* Increasing equity investment and reducing A recent report on capital markets in the Phil-below-market member loans, coupled with p p .divestiture of the Govermnent Social Insur- ippis esimated thatr8-0ubll psos ($360-

anc Sy ,e' lif inuac buies 600 millilon) from the contractual savings sectorcould be mobilized for investment in long-term

* Fully or partially privatizing asset manage- bonds (World Bank 1995e). At this stage new andment, with a provision for investing abroad more flexible asset allocation guidelines areif prudent. required to enable the contractual savings sector

Table 4-7: Contractual savings in the Philippines, 1980-93(billions of pesos)

Average annualType of fund 1980 1985 1990 1993 growth (percent)

Social security 18.4 43.1 98.6 154.1 17.8Social Security System 9.2 26.3 62.5 103.8 20.5Government Social Insurance System 9.2 16.8 36.1 50.3 14.8

Occupational pensions n.a n.a. 20.0 40.0 n.a.Insurance 11.1 18.1 45.5 62.0 14.1

Total 29.5 61.2 194.0 256.1 18.1

Share of GDP (percent) 12.1 10.7 15.3 17.6

Source: World Bank 1995n.

49

to become a more active player in capital mar- structure of their liabilities. Most of the liabilitieskets and project financing. in these institutions are composed of demand

In Thailand pension and provident funds play and short-term savings deposits. Making long-a limited role in capital markets and infrastruc- term loans would create a serious maturity mis-ture finance. The government pension fund is match between the assets and liabilities of thesenot funded, and pension payments are made institutions. This mismatch is even more danger-from the government's annual budget. The gov- ous in the absence of efficient and liquid moneyernment is planning a Central Provident Fund markets that would otherwise provide bankswith an initial government contribution of $100 with some tools to manage their liquidity andmillion equivalent. Total capitalization of exist- interest rate risks.ing provident funds is still small ($1.2 billion), In addition to requiring a low-inflation macro-since they cover only 1 percent of the nonagricul- economic environment-which would helptural workforce. banks mobilize longer-term liabilities such as

Argentina and Colombia both enacted pension extended time deposits or negotiable certificatesreform laws in 1993. Asset accumulation by new of deposit-most banks in developing countriespension systems is still small but is expected to need to strengthen their technical capacities ingrow quickly. In Argentina the new fully funded project finance. These institutions' limited expe-system became operational in July 1994 and in its rience with analyzing the various risks involvedfirst eleven months mobilized $1.88 billion in limited recourse financing and lack of know-equivalent in long-term financial resources. For a how on mitigation methods are the main reasonsone-year period this would correspond to domestic banks have played an insignificant roleslightly more than 2 billion pesos, or 0.7 percent in project financing activity.of GDP. This weak base of knowledge is the result of

Bolivia is advanced in its plans to enact a simi- collateral-based lending, which guides the exten-larly broad and fundamental pension reform, sion of credit in most developing countries. Inwhile Mexico introduced a compulsory comple- the United States the requirement that a bor-mentary retirement savings scheme in 1992 but rower post collateral or secure a guarantee fromhas yet to reform the social pension system. As a third party generally means that their credit-explained in Box 2-3, Bolivia's pension reform is worthiness is otherwise insufficient. Thus theclosely tied to infrastructure privatization since very need for this form of support implies a highthe new pension accounts will be endowed by level of risk. Where credit management is strong,the shares of several public infrastructure agen- however, collateral is not considered a substitutecies. The pension account shares were placed for creditworthiness-it merely provides anwith Citibank as trustee in July 1995 and will additional margin of protection for a loan that isremain there until July 1997, allowing time for already acceptable.the enactment of the pension reform law. The weaknesses of commercial banks in devel-

oping countries are rooted in interventionist andStrengthening commercial banks to provide repressive government policies that resulted ininfrastructure debt financial disintermediation. Such policies were

often pursued in inflationary environments,With the exceptions of Malaysia and recently which not only made financial assets unattrac-Thailand, commercial banks in developing coun- tive to private savers but also forced banks totries have played a very small role in project allocate a good part of their credit to priority sec-finance lending to infrastructure projects. In the tors-often at subsidized, low-interest rates.Philippines, for example, domestic banks pro- Many developing country governments havevided less than 5 percent of the financing for $1.4 required that 60-70 percent of bank credit bebillion of debt for a group of eight power projects allocated, compared with about 20 percent in(Price Waterhouse 1995). The situation is even industrial countries. India is a good example.worse in Latin America, where high inflation, Until financial reforms were initiated in 1991, 90exchange rate risks, and political uncertainty percent of the country's commercial banking sec-have made long-term financing extremely scarce. tor was owned by the state. Commercial banks

In many developing countries the inability of were required to invest 15 percent of their fundscommercial banks to provide medium- to long- in cash to fulfill the cash reserve requirement andterm finance stems directly from the maturity 38 percent in government and government-

50

approved securities. In addition, 40 percent of commercial banks use repurchase agreements tothe remaining funds (that is, nearly 19 percent of finance their securities inventories.total funds) were required to be provided as In Korea bonds held by banks grew fromloans to priority sectors at somewhat conces- about $3 billion in 1985 to $37 billion in 1993,sional rates (Vittas and Cho 1995). accounting for 43 percent of bond holdings by all

Despite severe setbacks during the 1980s, financial institutions in 1993.13 Corporate bondscommercial banks still played a prominent role accounted for most of this growth; their totalin the financial sector in developing countries in growth during this period was 1,938 percent,the early 1990s, holding 40-75 percent of depos- compared with 936 percent for governmentits and accounting for 25-60 percent of lending bonds (Table 4-8).(Germidis, Kessler, and Meghir 1991). Commer- The share of bonds held by banks in the totalcial banks hold 50-90 percent of the assets of all market capitalization of bonds in Korea has beenfinancial intermediaries in most developing declining after rising considerably between 1985countries and will continue to be at the heart of and 1990. This decline is mainly the result of thetheir financial markets for the foreseeable future. rapid growth of the Korean bond market in

Financial sector reforms that revive or estab- recent years. In addition to holding and tradinglish the role of commercial banks in long-term government and corporate bonds, Korean banksfinance are essential for increasing the share of issue debentures to raise long-term funds. Fordomestic resources in infrastructure finance. instance, the first private bank to issue bondsCommercial banks can play an important role in was the Korea Long-Term Credit Bank. Estab-screening and monitoring the behavior of lished under the Long-Term Credit Bank Law, itprojects. By developing and maintaining long- began issuing bonds in 1980.term relationships with their customers, com- In Malaysia commercial banks account for amercial banks can have superior information to sizable portion of the financial institutionsthat of outsiders, support expansion plans, and involved in bond markets. These operations arereduce the costs of temporary distress by provid- conducted under prudential regulations thating lines of credit to their customers. establish and monitor risk-weighted capital ade-

An efficient and deep commercial banking sec- quacy ratios and statutory reserve requirements.tor is important not only in terms of appraising Thus if a financial institution invests in a corpo-and financing infrastructure investments but is rate bond issue it must adjust its capital fundsalso a necessary prerequisite for the develop- and statutory reserves in order to maintain thement of the securities and eventually derivatives capital adequacy ratio and reserve requirements.markets that are essential to structured financial The central bank also limits the amount of loansmarkets. Bonds, for example, are not easily a commercial or merchant bank can lend to aabsorbed by individual investors. Thus most single customer. Finally, since the purchase ofbonds are absorbed by financial institutions such corporate bonds is considered to be an extensionas banks. Banks can also play a major role in of credit to the issuer, such operations are subjectenhancing the liquidity of bond markets by to the single customer limit, with the first ninety-executing repo transactions where regulatory day holding period exempted.frameworks permit the offering and trading of In Malaysia's corporate bond market commer-such instruments. The trading departments of cial banks mainly act as lead managers and

Table 4-8: Bond holding by banks in the Republic of Korea, 1985-93(billions of U.S. dollars)

1985 1990 1991 1992 1993

Total market capitalization of bonds 22.57 59.40 86.64 117.41 166.81Amounts held by banks, of which: 2.89 24.71 28.26 31.13 36.74

Government and public bonds 2.21 15.03 15.17 17.42 22.38Corporate bonds 0.68 9.68 13.09 13.71 14.36

Source: World Bank 1995c.

51

Table 4-9: Commercial banks' investment in marketable securities in Malaysia, 1991-94(billions of U.S. dollars)

Share Share Share ShareInstrument 1991 (percent) 1992 (percent) 1993 (percent) 1994 (percent)

Government securities 866.18 12.94 512.64 7.27 467.41 6.17 819.92 10.65Negotiable instruments

of deposit 3,256.99 48.67 3,438.70 48.74 3,008.15 39.69 2,438.28 31.66Other 2,568.38 38.38 3,103.83 43.99 4,102.96 54.14 4,443.36 57.69

Total 6,691.54 100.00 7,055.17 100.00 7,578.52 100.00 7,701.56 100.00

Source: World Bank 1995d and IMF data.

underwriters of new corporate bond issues, deeper and more complex. The main goal ofinvestors for their own accounts, and intermedi- financial regulation is to achieve stability with-aries in the market, matching the funding out undermining efficiency.requirements of issuers with the needs of inves- Deregulation should be implemented withtors. The composition of securities held by com- due consideration of a country's underlyingmercial banks in Malaysia is shown in Table 4-9. macroeconomic conditions as well as the techni-

The ultimate objectives of banking sector liber- cal capacities of its financial institutions (Box 4-8).alization are to stimulate competition and allow In Venezuela, for example, the banking systemgreater freedom for banks to respond to market was relatively easy to deregulate. Upgrading thesignals, choose their own customers, set interest regulatory framework was a much slowerrates, and determine the location of branches. process, however-and that lag proved fatal.Allowing foreign institutions to open branches, More than half the country's banks failedstart joint ventures with local institutions, or pro- because of massive corruption that went unde-vide specialized services from abroad can be tected by shoddy supervision and because ofanother source of competition, enabling access to economic instability resulting from misguideda larger pool of resources and introducing policies. The bank failures led to a bailout thatadvanced technologies in financial intermedia- was proportionally more than five times the sizetion. Deregulation, however, exposes financial of the U.S. savings and loan crisis (Naim 1995).institutions to greater risk exposure that requires Mistakes are made under any system ofcareful supervision. The need for prudential reg- finance. Market-based financial systems, likeulation increases as financial systems become public ones, are subject to fraud and instability.

Box 4-8: Necessary conditions for financial deregulation

Interest rate deregulation should proceed in stages, When these conditions are not satisfied realwith complete deregulation awaiting later stages of interest rates may rise to exorbitant heights, threat-reform. Experience in many countries suggests that ening the net worth of borrowers and ultimatelythe following criteria should be satisfied before the soundness of the financial system, as happenedcomplete deregulation: in Chile and Turkey in the 1980s. Malaysia and

Korea adhered most closely to these criteria, both* Macroeconomic conditions are reasonably stable. waiting until 1991 to achieve complete interest rate. The financial condition of banks and their bor- deregulation. The cost of observing these recom-

rowers is sound. mendations generally outweighs the considerable* A minimal base of financial skills has been risks associated with liberalized interest rates when

attained. necessary conditions are not in place.* Checks are in place to limit collusion among

banks in the determination of interest rates. Source: Caprio, Atiyas, and Hanson 1993.

52

The goal is not to create a perfect system, but one age firms. Such profits may not be conducive tothat mobilizes resources efficiently, minimizes efficiency and lower intermediation costs in theallocation mistakes, curbs fraud, and stops insta- long run.bility from turning into crisis. The resulting role Although East Asian commercial banks fre-for governments is to minimize their influence quently issue bonds in order to raise longer-termover pricing and allocation of credit and to focus funds, this instrument is rare among banks inon regulating and supervising financial institu- Latin America. Banks in Chile and Colombia aretions and markets in a stable macroeconomic raising funds through bond issues, although theenvironment. effort is very recent in Colombia.

Still, there is no single answer to how liberal- In Argentina banks raise funds in the marketized or tightly controlled a banking sector by issuing bonds with maturities of one to threeshould be. The results of government interven- years. Their ability to provide long-term loans is,tions in setting interest rates and directing credit however, hampered by uncertainty about thehave not been uniform. Some countries that have possibility of rolling over their maturing bonds.achieved economic growth and rapid develop- A recent World Bank loan for capital marketment, such as Korea and Malaysia, did not liber- development attempts to address this issue.alize their financial sectors until recently. Simi- Under the project a backstop fund will allowlarly, in postwar Japan a major objective of Argentine banks that meet certain operating cri-government interventions was to achieve term teria to purchase commitments for refinancingtransformation toward longer-term lending. their maturing bonds at a preagreed rate shouldSubsidized interest rates and directed credit pro- systemic disruptions prevent them from rollinggrams can be justified when the investment risk over their bonds in the market at maturity. Qual-of a particular activity is high because of such ifying disruptions are confined to events affect-features as large scale and long gestation period. ing all banks rather than those affecting a single

In Malaysia no banking license has been issuer. Thus the backstop facility supports rathergranted to a foreign bank since 1973, yet 30 per- than supplants evolving private markets. Oncecent of domestic lending is done by foreign banks purchase commitments, they can onlybanks. International banks are allowed to enter exercise their rights if they maintain a credit rat-joint ventures with domestic lenders, but they ing above a certain minimum standard, as deter-may own no more than one-fifth of the capital of mined by at least two approved local credit ratingsuch an institution. All banks must comply with agencies. After the exercise of any commitments,the rule that nonresident companies must source the backstop fund actively manages its inventoryat least 60 percent of their banking needs from of bank debentures. When the markets stabilize,local lenders. Restrictions on foreign entry have the fund sells its bond holdings in domestic orresulted in unusually large profit margins for international markets, creating additional liquid-domestic institutions such as banks and broker- ity in secondary markets (Habeck 1994).

53

Notes

1. Another cause of the problem was that the selection of the conces- 6. Including banks, the Korea Credit Guarantee Fund, merchantsionaires was based on the shortest concession period bid, with the banks, securities firms, and the Guarantee Insurance Corporation.maximum acceptable period having been 20 years. This resulted invery short concessions (under 10 years in some cases), which put 7. Including money market funds and mutual funds.upward pressure on tolls from the outset.

8. Savings are placed with these institutions for extended periods and2. In private power projects the capacity charge is designed to recover are not frequently withdrawn.the capital or fixed costs of the plant. The power purchaser is usuallyonly obliged to pay for capacity that is dependable (that is, available 9. Financial operations by contractual savings institutions are differentto be called on in accordance with the power purchase agreement). from the financial operations of commercial banks in the sense thatCapacity payments generate sufficient revenue to cover the project's the money multiplier effect of their operations is small and thus doescapital and fixed costs and investor returns, which include costs that not fuel inflation.the project would incur even if the purchaser did not "dispatch" theplant and purchase electrical energy. 10. The impact of life insurance, funded pension schemes, and provi-

dent funds on the quantity of savings is the subject of considerable3. For instance, the annual conversion requirements for servicing the debate, but the impact on the quality of savings, especially throughforeign financing component of an independent power producer the increase in long-term financial savings, is evident.financed with $300 million of debt at 10 percent annual interest andwith a fifteen-year maturity and $100 million in equity with a net 11. In 1994 about 56 percent of Malaysian government securities, withannual rate of return of 20 percent would be equivalent to about $60 a face value of $14.5 billion, had maturities above fifteen years; 21 per-million a year. Therefore a $2 billion annual investment inflow over a cent ($5.6 billion) had maturities of eleven to fifteen years; 19 percentfive-year period on terms similar to those described above could ($4.9 billion) had maturities of six to ten years; and 3 percent ($840potentially lock in an average annual outflow of close to $2 billion miUion) had maturities of four to five years. The government stoppedover a ten- to twenty-year period. issuing securities with maturities of two to three years in 1990.

4. The government's share, held by the Electricity Generating Author- 12. The old pension system was a "pay-as-you-go" system that wasity of Thailand, was reduced from 100 percent to 48 percent. The facing immense financial pressures.remaining 2 percent is held by the Thai Crown Property Bureau.

13. Financial institutions include securities companies ($350 million),5. In addition to the Korean Investors Service there are two other rat- major pension funds ($4.1 billion), life insurance companies ($8.5 bil-ing agencies in Korea: National Information and Credit Evaluation, lion), other insurance companies ($1.6 billion), and the three largestInc., set up by the commercial banking industry, and the Korea Man- trust companies ($35.7 billion).agement and Credit Rating Corporation, established by the KoreaDevelopment Bank.

54

5Conclusion

The need for transition to private financing of should be used to leverage private invest-infrastructure has been generally acknowledged, ment; this paper suggests various methodsbut sustained progress across all sectors requires for doing so. Depending on country condi-that developing countries pursue nationwide, tions, infrastructure financing vehicles of theeconomywide strategies. This process will re- type proposed in the Philippines andquire that careful attention be paid to the follow- Colombia may be promising options.

ing guidelines and caveats: * Restructuring, corporatization, and privati-

zation of public agencies providing infra-* The transition strategy should be publicly structure services is an essential element of

elaborated by senior political figures, in leg- the transition process.

islation if necessary. * External management expertise and finan-

• The transition strategy may have to be imple- cial resources are critical to an expeditiousmented by special-purpose institutions with transition, but political and economic trade-broad representation from the executive and offs have to be carefully weighed.legislative branches of the government. Pri- .Provision of sovereign guarantees is essen-vate representation can also be beneficial. tial for many developing countries embark-Such an approach helps governments con- ing on the transition, but guarantees mustcentrate their capacity-building efforts in the be rovided under accurate and activepublic administration to promote private provia undemend.infrastructure and facilitate the sharing of financial management.lessons and information about activities in * Apart from political uncertainties, elementsdifferent sectors. A common institutional of the regulatory environment or the pres-approach not only helps minimize ineffi- ence of nonviable public agencies in thecient competition among different line agen- market can give rise to market imperfectionscies, it also generates comprehensive policy that can only be mitigated through sover-feedback for identifying and prioritizing eign guarantees. These imperfections shouldreforms. Technical assistance from bilateral be identified and eliminated using carefullyand multilateral sources could be better tar- designed reforms.geted for designing strategies and building Local capital markets development shouldskills for special-purpose institutions. be given top priority within the national

* To the extent possible, incremental capital strategy, with clear targets for achieving aexpenditures that come from fiscal resources tangible shift from foreign to local financing.

55

Main areas of focus should include stimulat- salient features of experiences to date ining demand for long-term assets through terms of financing and contracting mecha-pension reforms and developing institu- nisms. It is important that officials from alltional and regulatory infrastructure for relevant agencies, no matter how special-bond markets. ized, be provided with a broad perspective

on the issues that are involved in the transi-As far as the World Bank and other multilateral tion. This will help ensure cooperationand bilateral development agencies are con- among various agencies in the implementa-cerned, the following are among the useful con- tion process.tributions they can make to the process: * They should design and deploy flexible and

* They should work with governments in adaptable financing instruments that reflectdevising a strategy for transition to private their ability to leverage maximum marketfinancing of infrastructure and provide funding. In this connection they shouldfinancial assistance in the context of a mutu- develop local currency financing and be pre-ally agreed strategy. Various direct and con- pared to manage foreign exchange risks totingent funding instruments can be linked to help expedite the development of localperformance criteria that reflect progress on financial capacities. Another market supportcomponents of the transition strategy. instrument would be to underwrite sover-

eign risk insurance in order to encourage* They should design special seminars and long-term debt financing. These institutions'

courses targeting very high and lowe:r level ongoing dialogue with member countriesofficials from a wide set of agencies in devel- gives them a responsibility for and infor-oping countries. These courses should pro- mational advantage in managing suchvide the rationale for private infrastracture risks. Moreover, they could draw on theand the potential consequences of slow or benefits of diversification of relevant risksno progress within the particular country across large numbers of borrowing memberenvironment. They also should present the countries.

56

Bibliography

Buffie, Edward. 1992. "The Short and Long Run Germidis, Dimitri, Denis Kessler, and RachelEffects of Fiscal Policy." World Bank Economic Meghir. 1991. "Financial Systems and Devel-Review 6(2): 331-51. opment: What Role for the Formal and Infor-

mal Financial Sectors." Organization for Eco-Caprio, Gerard, Izak Atiyas, and James Hanson. nomic Cooperation and Development,

1993. Financial Reforms: Lessons and Strategies. Development Center, Paris.World Bank, Economic Development Institute,Washington, D.C. Habeck, Odo. 1994. "Backstop Lending for Capi-

tal Market Development." FPD Note 29. WorldChrisney, Martin D. 1995. Innovations in Infra- Bank, Vice Presidency for Finance and Private

structure Finance in Latin America: Past, Present, Sector Development, Washington, D.C.and Future Trends. Inter-American Develop-ment Bank, Washington, D.C. Hamilton, Rabinovitz, and Alschuler Inc. 1995.

Mobilizing Private Finance for Latin AmericanCPIAI (Commission to Promote Investment in Infrastructure: Analysis and Suggested Initiatives.

America's Infrastructure). 1993. Financing theFuture. Washington, D.C. Hulten, C., and R. Schwab. 1991. Is There Too Lit-

tle Public Capital? Infrastructure And EconomicEasterly, William, Carlos A. Rodriguez, and Growth.

Klaus Schmidt-Hebbel. 1994. Public Sector Defi-cits and Macroeconomic Performance. New York: IFC (International Finance Corporation). 1994.Oxford University Press for the World Bank. Financing Private Infrastructure Projects. IFC

Discussion Paper 23. Washington, D.C.Economist Intelligence Unit. 1994. "Infrastruc-

ture Development in Latin America." London. . 1995a. Emerging Stock Markets Factbook.Washington, D.C.

Ewing, Andrew, and Susan Goldmark. 1994."Privatization by Capitalization." FPD Note . 1995b. Trends in Private Investment in31. World Bank, Vice Presidency for Finance Developing Countries. IFC Discussion Paper 25.and Private Sector Development, Washington, Washington, D.C.D.C.

1996 "Experience in Promoting Emerg-Forrester, J. Paul, and S. Raymond Tlhlett. 1996. ing Market Investment Funds, 1977-95."

"Debt Finance for Infrastructure Projects." The Working Paper. Washington, D.C.Financier (February).

57

IMF (International Monetary Fund). 1995a. Pri- 92." Policy Research Working Paper 1202.vate Market Financing for Developing Countries. World Bank, Washington, D.C.Washington, D.C.

Santos, Everett J. 1994. "Roundtable Discussion:1995b. World Economic Outlook. Washing- Critical Issues in Infrastructure in Developing

ton, D.C. Countries." In Michael Bruno and Boris Plesk-ovic, eds., Proceedings of the World Bank Annual

Ingram, Gregory. 1995. "It Takes More Than Conference on Development Economics 1993.Money." Project and Trade Finance (January). Washington, D.C.

Kohli, Harinder. 1995. "Infrastructure Develop- Smith, Warrick, and Michael Klein. 1994. "Infra-ment in East Asia and the Pacific: Toward a structure Regulation: Issues and Options forNew Public-Private Partnership." World Bank, East Asia." World Bank, Private Sector Devel-Office of the Regional Vice President for East opment Department, Washington, D.C.Asia and the Pacific, Washington, D.C.

So, Jae, and Ben Shin. 1995. "The Private Infra-Little, Ian, Richard N. Cooper, W. Max Corden, structure Industry: A Global Market of $60 Bil-

and Sarath Rajapatirana. 1994. Boom, Crisis and lion A Year." FPD Note 45. World Bank, ViceAdjustment: The Macroeconomic Experience of Presidency for Finance and Private SectorDeveloping Countries. New York: Oxford Uni- Development, Washington, D.C.versity Press for the World Bank.

Standard & Poor's. 1994. Creditreview.McNaughton, Diana. 1992. Banking Institutions in

Developing Markets. World Bank, Financial Sec- Stewart-Smith, Martin C. 1995. "Industry Struc-tor Development Department, Washington, ture and Regulation." Policy Research Work-D.C. ing Paper 1419. World Bank, Washington, D.C.

Moody's Investors Service. 1992. Global Credit Vittas, Dimitri. 1995a. "Argentina's New Inte-Analysis. grated Pension System: First Year Assess-

ment." World Bank, Financial Sector Develop-Mozes, Dan. 1995. "Long-Term Finance in Infla- ment Department, Washington, D.C.

tionary Economies." FPD Note 35. WorldBank, Vice Presidency for Finance and Private . 1995b. "Pension Funds and Capital Mar-Sector Development, Washington, D.C. kets." World Bank, Financial Sector Develop-

ment Department, Washington, D.C.Muir, Russell, and Joseph Saba. 1995. "State-

Owned Enterprise Restructuring." FPD Note . 1995c. "Strengths and Weaknesses of the57. World Bank, Vice Presidency for Finance Chilean Pension Reform." World Bank, Finan-and Private Sector Development, Washington, cial Sector Development Department, Wash-D.C. ington, D.C.

Naim, Moises. 1995. "Latin America the Morning Vittas, Dimitri, and Y. J. Cho. 1995. "Credit Poli-After." Foreign Affairs (July/August). cies: Lessons for East Asia." Policy Research

Working Paper 1458. World Bank, Washington,Price Waterhouse. 1995. Sources of Finance of BOT D.C.

Power Projects. Report commissioned by thePhilippine BOT Center. Winston, C. 1993. "Economic Deregulation: Days

of Reckoning for Microeconomists." Journal ofRupert, Bruce. 1996. "India vs. Pakistan." Infra- Economic Literature 31.

structure Finance (February/March).World Bank. 1990. Review of Electricity Tariffs in

Sader, Frank. 1993. "Privatization and Foreign Developing Countries during the 1980s. IndustryInvestment in the Developing World, 1988- and Energy Department, Washington, D.C.

58

. 1992. Infrastructure Maintenance in Latin . 1995e. The Emerging Asian Bond Market:America and the Caribbean: The Cost of Neglect The Philippines. Washington, D.C.and Options for Improvement. World BankRegional Studies Program Report 17. Washing- . 1995f. The Emerging Asian Bond Market:ton, D.C. Thailand. Washington, D.C.

. 1994a. "Argentina Capital MarketsStudy." Washington, D.C. . 1995g. "Korea Transport Sector: Resource

Mobilization Challenges and Opportunities."1994b. "Philippines: Private Sector Washington, D.C.

Assessment." Washington, D.C.. 1995h. "Philippines: An Agenda for the

1994c. World Development Report 1994: Reform of the Social Security Institutions."Infrastructure for Development. New York: Washington, D.C.Oxford University Press.

1995i. "Thailand: Increasing Private Sec-. 1995a. Bureaucrats in Business: The Eco- tor Participation and Improving Efficiency In

nomics and Politics of Government Ownership. A State Enterprises." Washington, D.C.World Bank Policy Research Report. NewYork: Oxford University Press. .1995j. Trends in Developing Economies.

Washington, D.C.1995b. The Emerging Asian Bond Market.

Washington, D.C. . 1995k. World Tables. Baltimore, Md.:Johns Hopkins University Press.

1995c. The Emerging Asian Bond Market:Korea. Washington, D.C. . 1996. Financial Flows and the Developing

Countries 3(2). Washington, D.C.1995d. The Emerging Asian Bond Market:

Malaysia. Washington, D.C.

59

Distributors f CHiNA GnEECE ITALY NORWAY SINGAPORE, TAIWAN, VanDienmenEddiosaTectxriqDiti ut r of Ctrina Finaneeial 8 Economic Papasolitiou S A t.icnsa Commissiinana Sansoni SPA tiaesnvG Infommalion Center MYANMuAR BllUNEI Ch de LaCuet 41W orld Bank PstrdshingsHouse 35, Stounara Sir Via DucaDiCalibna. Ill Book Deparment Aaligare Puklsmhng Asia CH1807Blonay.) ~~~~~~~~~~~~8, Da Fo SiDong J,e 106 82 Aheaq Casell'aPtslaleS5S2 PO Box fi`125 Ellrshid Pn-ifil Ple Ltt Td1 j02t)9432673Publications Be(ing Tel. (1) 364-1826 50125 Firenze 1N 0602 slos6 41 F.allarrgPuddng nvad 004 03 Fax (021)9433605

Pi ices anid criedil tr/ ins f IY r / Tel- (1) 333 8257 Fax: (1) 364-8254 Tel. (55) 645-415 Tel. (22) 57-3300 Golden Wheel Buaidingfirriii colnt-lf il tro ountry. Fax (1) 401-7365 Fax: (55) 641-257 Fax: (22) 68-1901 Singapore 349316 TANZANIACeraryuittilI loca distIr IbuoON.OG,MCDTel. (65) 741-5166 0deond University PressCzalisitit yoir IJ cal disli libuffr COLOMBIA Asia 2000 Ltd JAMAICA PAKISTAN Fax (65) 742 9350 Makdaba Slreellbefore placing an order. lnloenlace Llda Sales A Circulalion Department lan andlie Publishers Ltd. M,rza Book Agency e-rail ashraileeasianconneci com PO Box 5299

ApariadoAereno 34270 Seabird House, nlf 1101-02 206 Old Hope Road 65, Shahlrahre Quaid eAzam Dares SalaamBogoIbtDE 22-28 Wyirdhma Slreet. Certrml Kingslon6 PO Box lo 729 SLOVAK REPUBLIC Tel (51)29209

ALBANIA Tel: (t)285-2798 Hong Kog Tel. 809-927-2085 Lahore 54000 Slovad 0 TG Lid Fax (51)46822Adrion Ud Fax (I)2852798 Tel. 852 2530-1409 Fax: 809-977-0243 Tel (42) 7353601 - Krupinsa 4Pedal exepi Sir Fax: 852 2526-':07 Fax- (42) 7585283 PO Box 152 THAILANDPal 9. Shk. I, Ap 4 COTE D IVOIRE ULIL lhip-/Ar salesftasia2000 cornm k JAPAN 852 99 Bratixlava 5 Central Books Dis5ibulionTrrana Cenlre dEdiion el de Diesiorn Eastern Book Service Oxoed Lniversity Pross Tel (7) 839472 306 Sdom RoadTel (42) 274 19; 221 72 Alrticines (CEDA) HUNGARY Hongo 3-Chome, 5 Bangalore Tows Fax- (7) 839485 BangkokFax. (42) 274 19 04 8 P. 541 Foendalion lor Market Bunkys-klo 113 Sharae fraisal Tel 12) 235-5400

Aridan 04 Plaleau Economy Tokyo PO Box 13033 SOUTH AFRICA, BOTSWANA Fax: (2) 237-8321ARGENTINA Tel: 225-24-6510 Dombovad Ul 17-19 Tel (03) 381 0801 Karachi-75350 For single i,tlesOkcinadelLirorherreacional Fax: 22525-0567 H-I1?7 Budaresi Fax (03)3818 0064 Tel Q21) 446307 OdxordrUniversily Press TRINIDAD & TOBAGO, JAM,Ax. Cordoba 1877 Tel: 3612042951 or UlL- itpJAwwbekkroame.orjpysvt-ebs Fax (21)454-7640 SoulhernAirree Systematics SludiesUnit1120 BuenosAires CYPRUS 361204 2948 PO Box 1141 i9WafsStrelTel. (1) 815-8156 CenterofAgpliedfResearch Fax: 3612042953 KENYA PERU Cape To 8000 CurepeFax: (1) 81598354 Cyprous Coege AlAca Book Serice (E A ) Ltd Editorial Desarrollo SA Tel (21) 45 7266 Trinidad. West Indies

6, Diogenes SIlre, Engorni INDIA rairan ilouse, Mlarigano Slreel Apadado 3624 Fax. (21) 45 7265 Tel- 80-662-5654AUSTRAUA, FIJI, PAPUA NEW GUINEA, P.o. Box 2006 Allied Pubrlisiers Lid PO. Box 45245 Lima 1 Fax: 809-662-5654SOLOMON ISLANDS. VANUATU, AND Nicosia 751 Mourntarld Nairobi Tel (14) 285380 For suoscnptrnnordersWESTERN SAMOA Tel. 244-1730 Madras-600 002 Tel (2)23641 Fax (14)208628 Inlereotonal Surscriprion Serce UGANDAD.A fInalo fnServices Fax: 246-2051 Tel (44) 852-3938 Far (2) 330272 PO Box 41095 Gusho Ltd648 Whxiehorse Roed Far: (44) 952-0'i49 PHILIPPIIIES Craighall Marhvaari BuildmgMilcham 3132 CZECH REPUBUC KOREA, REPUBLIC OF International Rooksounce Ceilter Inc. Johannesbtsrg 2n24 PO Box 8997V`c'ria Nalional Inlormation Cenler INDONESIA Dapeon Trading Co Ltd Suile 720, Cilyland IO Tr. (11) e8O 1448 Plot 1614 Jinja RdTel. (61)392107777 p nxde)x Komrliska 5 PI Indira Limiled PO Box 34 CondornMiumiTower2 Fax. (11) 880-6248 KampalaFaa (61)392107788 CS - 11357 Prague I Jaln Borobudur 20 Yervida H dela Costa. comer TelFax: (41)254763UML kopitxrrdadrec'nrom er Tel. (2)2422-5433 PO BOo 19B Senul Valero SI SPAIN

Far (2) 2422-1484 Jakada 10320 Tel (2) 785 1831/4 Makar, Metro Manila Mlrndr 'ruensa Libros, S A UNITED KINGDOMAUSTRIA URL: hilpJlwwe.niscz/ Tel: (2t) 39004290 Fax. (2)784-0315 Tel (2)817 9076 Crslello37 Mcrnoino Lid.Geroldand Co. Fax: (21) 421-4289 Faxa (2) 817-1741 2501M aMrrd PO Box 3Graben 31 DENMARK MALAYSIA Tel I1) 431-3399 AlloS . Hampshire GU34 2PGo-tt1 Wren Sans4stoeroiatur 16114 Lniverooy Di IMalaysa Oorpprasie POTLAND Tar (1)575 395 EnglandTel: (1)533-50-14-0 nlenoemsoAll Sn Kowkab Publishers Bookshop, Limited lternatronal Pubiltshing Service hRp-ttweww saii stmprensa Tel (1420)86849Fax: (1)512-47-31-29 DK-1970 Frededksberg C PO. Box 19575-511 PO Fox 1127 Ul PreNlia 31137 Fax (1420)89889

Tel: (31) 351942 Tehran Jalan Paniai Brnu 00-577 Warrawa Mrndi r-s'Tsa BarcelonaBANGLADESH Far (31)-357822 Tel- (21) 258-3723 59700 Kuala Lumpur Tel: (2) 628-6089 Conselt dx Conl. 391 ZAMBIAMimn Ilrustries Deve"xetope Fax 98)21i)58-3723 Tel (3) 756-5000 Fax- (2) 621-7255 91939 Barcelona Universily iokshoop

Assislarrce Sociaet (MIDAS) ErYPT, ARAB REPUBLIC OF Fax (3) 755-4424 Tel (3) 498 3009 Great East Road CampusHouse 5, Road 16 Al Ahram Kelab Sara Co Puiblishers PORTUGAL Fax (3) 487 7659 PO Box 32379Dlhaxmrrrd lvArpa Al Galaa Street Khaled Eslambobf Axe. MEXtICO Livrana Portirgil LusakaDhaka 1209 Cairo 611h Street IIrOTEC flua Do Camlo 70-74 SRI LANKA. TIlE MALDIVES Tel (1) 213221 Exl. 482Tel: (2) 326427 Tel: (2) 578-6583 Kushehr Delalreoz H4 8 Apadado PoIal 22-860 1200 Listrn Lake Iiouise BookhokoFax (2)811188 Fax (2)578-6833 Telrran 14060 Tlalpan, Tel (1)3474982 PO Boa 244 ZIMBABWE

Tel. 8717819 or 8716104 MexicopF. Fax- (1)3470264 180, Sir ChitlamprlamrA. LongmanZimbabwe (Pte)LtdBELGIUM The Middle East Obsenver Fax: 8862479 Tel: (5) 6806-0011 Gadiner Mawalha Tourle Road, ArdbennieJeaoDeLannoy 41, Sheri Street Fax: (5)606-0386 ROMANIA Colror 2 PO. Box ST125Ao.durRri202 Cavo IRELAND CxrorparDe braralaEcvkieoiSN P. Tei- (1)321%5 StrluileAcmn1060BBrssels Tel: t2)393-9732 GovemmenlStepptesAgency NETHERLANDS Srr. Lipscarrino 26, seNor3 Fax: (1)432104 HarareTel. (2) 53-5169 Fax- (2) 393-9732 Oi6ig an tsolIrilrir De LindeboomrlnOr-Publkilies Bucharesl Tel (4)6216617Fax: (2) 538 0841 4-5 Harcourl Road PO. Box 202 Tel: (1) 6139645 SWEDEN Fax: (4) 621670

FINLAND Dutbn 2 7480 AE laaksbergen Fax: (1) 312 4000 Frares Crisloner ServiceBRAZE1 Aratleenen Kir)oaknsppa Tel (1)i461-311 Tel: (53)574-0004 Regedngalon 12Publicacoes Tecnkcas Inleroacionais PO. Box 23 Fax (1) 475-2670 Fax: (53) 572-9296 RUSSIAN FEDERATION S-106 47 Slockholm

Lida. FI0,-0371 Hesinki Isdaleilvo *Ves Mir, Tel: (8) 680 90 90Rua PeixaoGoGaride, 209 Tel: (0)12141 ISRAEL NEW ZEALAND 9a. Kolpachniy Pereuhlk Fax. (8) 2147 7701409 Sao Pauto SP. Fax- (0) 121-4441 Yoznrol L4eralure LId. EBSCO NZ Lid Moscow 101831Tel (11) 259-6644 UML: hip-ilbeooknel.crtalrellkal PO. Box 56055 PrNiate Maol Bag 99914 Tel (95) 9i17 8 49 Wennergren-Willinurs SBFPx: (11)258-6990 TelAAviv 61560 New Markel Fax. (95)9179259 PD. Box 136'i

FRANCE Tel- (3) 5285-397 Auckland S-171 25 SolnaCANADA World Bank Puobricalions Fax: (3) 5285-39Q7 Tel (9) 524-8119 SAUDI ARABIA, QATAR Tel (8) 705-97-50Renoul Pubiishing Co Lld. 66, avenue diena Fax (9) 524-8067 Jarir Book Slore Fax (8)27-00-711294 Argroma Road 75116Pads RONY nterooAonal PO. Box 31'9Ot)awa OntadoKIB 3W8 Tel: (1) 40-69-30056/57 POB13ox 13056 NIGERIA Aiyndh 11471 SWIzERLANDTel. 613-741-4333 Fax: (1) 40-89-30-68 Tel Aviv 611310 Uriversity Press Limrled Tel. (1) 477-3140 Libraine PayonFax: 813-741-5439 Tel: (3) 5461423 Three Cromns Building Jencho Fax: (1) 477 2940 Service Inslilulionnel

GERMANY Fax. (3) 5461442 Privale Mail Bag 5095 COies-de-MonIIWnon 30UNO-Vedag talorn 1002 LeonanePoppelsdorfer AeeS 55 Paleslinian AuthorlylMidle East Tel. (22)41-1356 Tel: (021)341-322953115 Bonn Index inlormalion Senrices Fax: (22) 41-2056 Fax: (021)-341-3235Tel- (228)212940 PO B 19502 JerusalemFax: (2281 217492 Tel: 12) 271219

I~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

Recent World Bank Discussion Papers (continued)

No. 309 Public and Private Secondary Education in Developing Countries: A Comparative Study. Emmanuel Jimenez andMarlaine E. Lockheed with contributions by Donald Cox, Eduardo Luna, Vicente Paqueo, M. L. de Vera, andNongnuch Wattanawaha

No. 310 Practical Lessonsfor Africafrom East Asia in Industrial and Trade Policies. Peter Harrold, Malathi Jayawickrama,and Deepak Bhattasali

No. 311 The Impact of the Uruguay Round on Africa. Peter Harrold

No. 312 Procurement and Disbursement Manualfor Projects with Community Participation. Gita Gopal

No. 313 Harnessing Informationfor Development: A Proposalfor a World Bank Group Strategy. Eduardo Talero and PhilipGaudette

No. 314 Colombia's Pension Reform: Fiscal and Macroeconomic Effects. Klaus Schmidt-Hebbel

No. 315 Land Quality Indicators. Christian Pieri, Julian Dumanski, Ann Hamblin, and Anthony Young

No. 316 Sustainability of a Government Targeted Credit Program: Evidencefrom Bangladesh. Shahidur R. Khandker, ZahedKhan, and Baqui Khalily

No. 317 Selected Social Safety Net Programs in the Philippines: Targeting, Cost-Effectiveness, and Optionsfor Reform.Kalanidhi Subbarao, Akhter U. Ahmed, and Tesfaye Teklu

No. 318 Private Sector Development During Transition: The Visegrad Countries. Michael S. Borish and Michel Noel

No. 319 Education Achievements and School Efficiency in Rural Bangladesh. Shahidur R. Khandker

No. 320 Household and Intrahousehold Impacts of the Grameen Bank and Similar Targeted Credit Programs in Bangladesh.Mark M. Pitt and Shahidur R. Khandker

No. 321 Clearance and Settlement Systemsfor Securities: Critical Design Choices in Emerging Market Economies. Jeff Stehm.

No. 322 Selecting Development Projectsfor the World Bank. Jean Baneth

No. 323 Evaluating Public Spending: A Frameworkfor Public Expenditure Reviews. Sanjay Pradhan

No. 324 The Bangladesh Rural Advancement Committee's Credit Programs: Performance and Sustainability. Shahidur R.Khandker and Baqui Khalily

No. 325 Institutional and Entrepreneurial Leadership in the Brazilian Science and Technology Sector:Setting a New Agenda.Edited by Lauritz Holm-Nielsen, Michael Crawford, and Alcyone Saliba

No. 326 The East Asian Miracle and Information Technology: Strategic Management of Technological Learning. Nagy Hanna,Sandor Boyson, and Shakuntala Gunaratne

No. 327 Agricultural Reform in Russia: A Viewfrom the Farm Level. Karen Brooks, Elmira Krylatykh, Zvi Lerrnan,Aleksandr Petrikov, and Vasilii Uzun

No. 328 Insuring Sovereign Debt Against Default. David F. Babbel

No. 329 Managing Transboundary Stocks of Small Pelagic Fish: Problems and Options. Max Aguero and Exequiel Gonzalez

No. 330 China: Issues and Options in Greenhouse Gas Emissions Control. Edited by Todd M. Johnson, Junfeng Li,Zhongxiao Jiang, and Robert P. Taylor

No. 331 Case Studies in War-to-Peace Transition: The Demobilization and Reintegration of Ex-Combatants in Ethiopia, Namibia,and Uganda. Nat J. Colletta, Markus Kostner, Ingo Wiederhofer, with the assistance of Ernilio Mondo, TairniSitari, and Tadesse A. Woldu

No. 332 Power Supply in Developing Countries: Will Reform Work? Edited by John E. Besant-Jones

No. 333 Participation in Practice: The Experience of the World Bank and Other Stakeholders. Edited by Jennifer Rietbergen-McCracken

No. 334 Managing Price Risk in the Pakistan Wheat Market. Rashid Faruqee and Jonathan R. Coleman

No. 335 Policy Optionsfor Reform of Chinese State-Owned Enterprises. Edited by Harry G. Broadman

No. 336 Targeted Credit Programs and Rural Poverty in Bangladesh. Shahidur Khandker and Osman H. Chowdhury

No. 337 The Role of Family Planning and Targeted Credit Programs in Demographic Change in Bangladesh. Shahidur R.Khandker and M. Abdul Latif

No. 338 Cost Sharing in the Social Sectors of Sub-Saharan Africa: Impact on the Poor. Arvil Van Adams and TeresaHartnett

No. 339 Public rnd Private Roles in Health: Theory and Financing Patterns. Philip Musgrove

No. 340 Developing the Nonfarm Sector in Bangladesh :Lessonsfrom Other Asian Countries. Shahid Yusuf and PraveenKumar

No. 341 Beyond Privatization: The Second Wave of Telecommunications Reforms in Mexico. Bjbrn Wellenius and GregoryStaple

No. 342 Economic Integration and Trade Liberalization in Southern Africa: Is There a Rolefor South Africa? Merle Holden

THE WORLD BANK

Ihmm\1 A partner in strengthening economies and explandngmarkets to improve the quality of lIe fur peopleeverywhere, especially the poorest

HEADQUARTERS

1818 H Street, N.W.Washington, D.C. 20433 USA

Telephone: 202.477.1234

Facsimile: 202.477.6391

Telex: MCI 64145 WORLDBANKMCI 248423 WORLDBANK

Cable Address: INTBAFRADWASHINGTONDC

World Wide Web: http://www.worldbankorg

E-mail: books@worldbank.org

EUROPEAN OFFICE

66, avenue d'1ina75116 Paris, France

Telephone: 1.40.69.30.00

Facsimile: 1.40.69.30.66

Telex: 640651

TOKYO OFFICE

Kokusai Building1-1, Marunouchi 3-chomeChiyoda-ku, Tokyo 100,Japan

Telephone: 3. 3214.5001

Facsimile: 3.3214.3657

Telex: 26838

1 . _ _ _ _ _j

IM8 : W&474