RMC Discussion Paper Series, Number 118 Financing Pakistan's Hub Power Project A Review of Experience for Future Projects Michael Gerrard Project Finance and Guarantees Department Resource Mobilization and Cofinancing Vice Presidency August 1997 121 The World Bank CFS/RMC DISCUSSION PAPERS 101 - Privatization in Tunisia, Jamal Saghir, 1993. 102 - Export Credits: Review and Prospects, Waman S. Tambe, Ning S. Zhu, 1993. 103 - Argentina's Privatization Program, Myrna Alexander, Carlos Corti, 1993. 104 - Eastern European Experience with Small-Scale Privatization: A Collaborative Study with the Central European University Privatization Project, 1994. 105 - Japans Main Bank System and the Role of the Banking System in TSEs, Satoshi Sunumura, 1994. 106 - Selling State Companies to Strategic Investors: Trade Sale Privatizations in Poland, Hungary, the Czech Republic, and the Slovak Republic, Volumes 1 and 2, Susan L. Rutledge, 1995. 107 - Japanese National Railways Privatization Study II: InstitutionalizingMajor Policy Change and Examining Economic Implications, Koichiro Fukui, Kiyoshi Nakamara, Tsutomu Ozaki, Hiroshi Sakmaki, Fumitoshi Mizutani, 1994. 108 - Management Contracts: A Review ofInternational Experience, Hafeez Shaikh, Maziar Minovi, 1995. 109 - Commercial Real Estate Market Development in Russia, April L. Harding, 1995. 1 10 - Exploiting New Market Opportunities in Telecommunications: Lessons for Developing Countries, Veronique Bishop, Ashoka Mody, Mark Schankerman, 1995. 111 - Best Methods of Railway Restructuring and Privatization, Ron Kopicki, Louis S. Thompson, 1995. 112 - Employee Stock Ownership Plans (ESOPs), Objectives, Design Options and International Experience, Jeffrey R. Gates, Jamal Saghir, 1995. 113 - AdvancedInfrastructurefor TimeManagement, The Competitive Edge in EastAsia, Ashoka Mody, William Reinfeld, 1995. 114 - Small Scale Privatization in Kazakhstan, Aldo Baietti, 1995. 115 - Airport Infrastructure: The Emerging Role of the Private Sector, Recent Experiences Based on Ten Case Studies, Ellis J. Juan, 1995. 116 - Methods ofLoan Guarantee Valuation andAccounting, Ashoka Mody, Dilip Patro, 1995. 117 - Private Financing of Toll Roads, Gregory Fishbein, Suman Babbar, 1996. JOINT DISCUSSION PAPERS Privatization in the Republics of the Former Soviet Union: Framework and Initial Results, Soo J. Im, Robert Jalali, Jamal Saghir; PSD Group, Legal Department and PSD and Privatization Group, CFS - Joint Staff Discussion Paper, 1993. MobilizingPrivate Capitalfor the Power Sector. Experience in Asia andLatin America, David Baughman, Matthew Buresch; Joint World Bank-USAID Discussion Paper, 1994. OTHER CFS PUBLICATIONS Japanese National Railways Privatization Study, World Bank Discussion Paper, Number 172, 1992. Nippon Telephone and Telegraph Privatization Study, World Bank Discussion Paper, Number 179, 1993. Beyond Syndicated Loans, World Bank Technical Paper, Number 163, 1992. CFS Link, Quarterly Newsletter. RMC Information Center, phone: 202-473-7594, fax: 202-477-3045 Copyright (© 1997 The World Bank 1818 H Street, N.W. Washington, D.C. 20433, U.S.A. All rights reserved Manufactured and printed in the United States of America The findings, interpretations, and conclusions expressed herein are entirely those of the authors and should not be attributed in any manner to CFS, the World Bank, or to members of the Board of Executive Directors or the countries they represent. The World Bank does not guarantee the accuracy of the data included in this publication, and accepts no responsibility whatsoever for any consequence of their use. The paper and any part thereof may not be cited or quoted without the author's expressed written consent. RMC DISCUSSION PAPER SERIES I I 8 Financing Pakistan's Hub Power Project A Review of Experience for Future Projects Michael Gerrard Contents Foreword v Acknowledgments vii Abbreviations and Definitions ix Executive Summary 1 Finance Plan 1 Issues Affecting the World Bank 3 Significant Project "Firsts" 6 Project Development 6 Mobilizing Sufficient and Least-Cost Finance 9 Shortening the Development Period 9 Conclusion 10 Finance Plan 11 Overview I 1 Development of the Plan 14 Banking Strategy 19 Structural Issues 23 Guidelines for Future Projects 27 Project Development 31 Market Experience 32 Timetable Analysis 34 Sponsor Group Formation 35 Development Costs and Risks 37 World Bank Perspectives 40 Guidelines for Future Projects 41 Institutional Development 43 The PSEDF and the ECO Guarantee Facility 43 Institutions in Pakistan 44 Postscript 47 . . Appendixes Appendix 1 Offshore commercial bank facilities syndicate profiles 49 Appendix 2 Project chronology 50 Appendix 3 Principal project agreements 52 Appendix 4 Project participants 53 Appendix 5 Overview of insurance arrangements during construction and operation 55 Appendix 6 Expanded Cofinancing Operation (ECO): Summary of terms of the guarantee 58 Boxes Box 1 Profile of private sector risk 16 Box 2 Role of global depository receipts in Hub Power Project financing 18 Box 3 Saving time through privatization 33 Box 4 Effects of publicity 35 Tables Executive sunmmary table 1 Summary of the base finance plan 1 Executive summary table 2 Facilities funded by offshore commercial banks 3 Executive summary table 3 Reasons for success in syndication 4 Table 1 Base finance plan 11 Table 2 Composition of the Private Sector Energy Development Fund 12 Table 3 Mobilization finance 13 Table 4 Facilities funded by offshore commercial banks 20 Table 5 Reasons for success in syndication 20 Table 6 Bilateral funding and procurement 26 Table 7 Key events and activities in project development 31 Table 8 Principal activities subject to delay 34 Table 9 Timetable of project activities 35 Table 10 Breakdown of development costs 38 Table 11 Personnel deployments 38 iv Financing Pakistan's Hub Power Project Foreword T he Hub Power Project is a 1,292 megawatt oil-fired power station located in the Pakistan province of Balochistan, some 40 kilometers northwest of Karachi. It is a private sector project undertaken by the Hub Power Company Limited (Hubco), a special single-purpose company listed on the Karachi and Luxembourg stock exchanges that has a concession to build, own, and operate the power station. The project has become a landmark in the field of infrastructure finance. From project conception in 1987 through financial close in January 1995, it has evolved in line with the international market for private sector infrastructure. By demonstrating the viability of private finance for a major infrastructure project in a developing country, the project has set an important precedent, particularly in the energy sector. Commercial power was first exported from the station in June 1996, and construction was completed in early March 1997, ahead of schedule. What was foreseen by the government of Pakistan and the World Bank in 1985 as the only way to relieve a serious financial bottleneck in the development of Pakistan's energy sector has since been embraced as economic orthodoxy by governments around the globe and by all types of financial institutions. Moreover, this transformation has occurred not only in the energy sector but also in transport and, increasingly, other infrastructure sectors such as water supply The catalytic effect of the World Bank's participation in the Hub Power Project-relative to financing of the project as well as development of the energy sector and local institutions in Pakistan and of the international market for private sector infrastructure-has been profound. The skepticism that pervaded the market in the early 1980s has been replaced by a "can do" attitude that is due in no small measure to the successful financing of the project. As might be expected during a period of rapid market advancement, a new constraint has emerged-namely, the speed at which these transactions can be implemented. In highlighting this and other market constraints the project has been, to some extent, an exercise in financial research and development for the Bank. The project absorbed about 225 person-years of resources, 25 of them from within the Bank; required more than 200 individual project documents; and claims no fewer than 16 "firsts" in the field of project finance. That so many people within the government of Pakistan, the Bank, the sponsors, and other financiers of the Hub Power Project prevailed over the many setbacks and frustrations inherent in the creation of such novel financial "technology" is to their collective and lasting credit. Because of its pioneering nature, during implementation the project encountered most of the problems likely to be faced by a private infrastructure project in a developing country. Thus the solutions developed and lessons learned have wider applications. This report highlights the solutions and lessons that generally apply to similar private infrastructure projects, with an emphasis on the financial instruments used by the Bank and its management of the project development process. In addition to its significant physical and financial dimensions, the Hub Power Project has sizable institutional dimen- sions: agencies of six governments collaborated with the Bank and the project's sponsors in developing the project, and more than forty international commercial banks are direct lenders to the project. As a result knowledge of and experi- ence with the project have been widely distributed. It is hoped that through this report it will be even more widely shared, for the benefit of future project development. v Acknowledgments his report was prepared under the direction of Suman Babbar, Project Finance and Guarantees Department, with T the assistance of Nina Shapiro, Thomas A. Duval III, and Pankaj Gupta. The report was written by Michael Gerrard, an independent consultant and former project finance manager of HRPG Limited, the Hub Power Project's development company. Other contributors include Maurice Fitz Gerald, an independent consultant and former adviser to HRPG Limited, who helped analyze cofinancier perspectives and reviewed the report; Per Ljung, division chief, Energy and Project Finance Division, Country Department 1, South Asia Region, who contributed widely to the report and in particular to the section on institutional development; and Richard Barton, an independent consultant and adviser to the commercial lenders to the project, who wrote appendix 5. The author is also grateful to the following for their time, recollections, and insights: Dr. Ibrahim Elwan, former divi- sion chief at the Bank and key functionary responsible for the project at the Bank; Adolfo Menendez of K&M Engineering & Consulting Corporation and former managing director of HRPG Limited; Philip Smith and Tracy Graham of National Power; James Chapman of Xenel Industries; Badar Khan of Commonwealth Development Corporation; Jean Fouillaud of COFACE; Dott. Carlo Cafaggi of SACE, Wasif Mustafa Khan of National Development Finance Corporation; Michael Woodroffe of Hubco; Patrick Crawford of Deutsche Morgan Grenfell; Terry Sheat of Linklaters & Paines; Ned Yescombe and Sian Dunstan of Bank of Tokyo-Mitsubishi; Saadia Khaid and Robert Welford of Citibank; Roger Warby and Maarten van Kempen of Credit Lyonnais; Victor Smith (formerly of Sakura Bank); and Bill Morris of NatWest. The report was edited by Paul Holtz and laid out by Glenn McGrath, both with American Writing Corporation. vii Abbreviations and Definitions Arrangers Arrangergroup: Citibank, CreditLyonnais, NatWest, Sakura Bank, and Bank of Tokyo- Mitsubishi; as well as MedioCredito Centrale (with respect to the SACE insured facility) Bridge finance The facilities provided by Al Rajhi (Istisna) and the National Development Finance Corporation (rupee term facility), forming part of mobilization finance. CDC Commonwealth Development Corporation of the United Kingdom COFACE Compagnie Francaise d'Assurance pourle Commerce Exterieur (France's export credit agency) ECO Expanded Cofinancing Operation ECU European currency unit Export credit agency COFACE, MM, SACE, and others, as the context implies Finance plan Sources and arrangements for providing $1,545,000,000 (equivalent) of base finance and $221,000,000 (equivalent) of standby finance for the Hub Power Project Financial close First date (in January 1995) on which all conditions precedent to first disbursement of the ECO guarantee facility and other offshore commercial bank facilities were either satisfied or waived Fuel supply agreement Between Pakistan State Oil and Hubco GDR Global depository receipt HRPG HRPG Limited, the project development company for the Hub Power Project Hubco Hub Power Company Limited IBRD International Bank for Reconstruction and Development Implementation agreement Between the government of Pakistan and Hubco IFC International Finance Corporation Istisna Islamic form of finance JEXIM Export-Import Bank of Japan K&M K&M Engineering & Consulting Corporation of the United States LTCB Long Term Credit Bank of Japan MlTI Ministry of International Trade and Industry of Japan Mitsui Mitsui & Co., Ltd. of Japan (construction consortium leader) Mobilization finance Finance made available to the Hub Power Project prior to financial close to mobilize the turnkey contractor National Power National Power PLC of the United Kingdom (lead sponsor) NIC National Insurance Corporation of Pakistan NDFC National Development Finance Corporation of Pakistan ix Operations and Between National Power and Hubco maintenance agreement PIC Pakistan Insurance Corporation Power purchase agreement Between the Water and Power Development Authority and Hubco Project coordination agreement Between Hubco and its various lenders (intercreditor agreement) PSEDF Private Sector Energy Development Fund, divided into two facilities: PSEDF I and PSEDF II PSO Pakistan State Oil SACE Sezione Speciale per I'Assicurazione del Credito all'Esportazione (Italy's export credit agency) Shariah law Islamic law as applied in Pakistan Sponsors National Power, Xenel, Mitsui, IHI, and K&M USAID U.S. Agency for International Development WAPDA Water and Power Development Authority World Bank International Bank for Reconstruction and Development Xenel Xenel Industries Limited of Saudi Arabia (lead sponsor) x Financing Pakistan's Hub Power Project Executive Summary Finance Plan Overall, the Hub Power Project is financed 24 percent from equity and 76 percent from debt (executive summary The World Bank participates in the Hub Power Project table 1). Debt comprises seven (senior) term loan facilities finance plan in two capacities: as a lender, contributing up (48 percent) and eight PSEDF (subordinated) facilities (28 to $225 million through the Private Sector Energy percent). Equity is contributed by the project's sponsors- Development Fund (PSEDF) administered by Pakistan's National Power, Xenel, Mitsui, Ishikawajima-Harima Heavy National Development Finance Corporation (NDFC), and Industries, and K&M Engineering & Consulting (10 per- as a guarantor of political risks for a syndicate of thirty- cent of overall funding)-as well as by other offshore four commercial banks lending up to $240 million to the investors (12 percent) and onshore investors (2 percent). project through an Expanded Cofinancing Operation All figures and exchange rates used in this report are taken (ECO) guarantee facility. from the information memorandum prepared for Bank syn- Other contributors to the PSEDF include the Export- dication of the Hub Power Project in May 1994, as sup- Import Bank of Japan (JEX[M), the governments of Italy plemented in September 1994. and France, and the U.S. Agency for International The participation of fifteen debt facilities makes the pro- Development (USAID). Forty-three international com- ject the most complicated private project financing under- mercial banks lend directly to the project through the ECO taken to date. A secondary dimension of the finance plan's guarantee facility, a similar facility guaranteed by JEXIM, complexity is the number of currencies (seven) in which and three export credit agency-insured facilities supported sources and applications of finance are denominated. by France's Compagnie Fran,aise d'Assurance pour le Commerce Exterieur (COFACE), Japan's Ministry of Private Sector Energy Development Fund (PSEDF) International Trade and Industry (MM), and Italy's Sezione Speciale per l'Assicurazione del Credito all' Espartazione The PSEDF provides long-term and mostly fixed-rate (SACE). Direct loans are also made to the project by the finance. In so doing it anticipates the two advantages nor- United Kingdom's Commonwealth Development mally associated with a refinancing after completion of con- Corporation (CDC) andasyndicateofninePakistanibanks. struction. No other independent power producer in EXECUTIVE SUMMARY TABLE I Summary of the base finance plan Amount Contribution Type of financing Source (US$ million equivalent) (percent) Equity 371 24 Debt 1,174 76 Senior debt ECO guarantee facility and six other facilities 738a 48 Subordinated debt PSEDF comprising eight facilities 436 28 Total 1,545 100 a. Does not include $60 million in standby debt provided by ECO ($40 million) and JEXIM ($20 million). 1 developing markets has entered the construction phase with on such maturities to Pakistan, the ECO and JEXIM guar- such long-term (twenty-three years) fixed-rate finance. In antees create an environment in which private sources of this context the PSEDF can be seen as not only forward- finance can perform the functions to which they are best looking but also as genuinely innovative. suited-namely, the identification, evaluation, management, The availability of such long-term finance compensates and acceptance of commercial risks. for the fact that the balance of the debt finance (the senior One important role of the ECO and JEXIM guarantee debt) available to the project comes with shorter maturi- facilities in the finance plan is as a provider of untied and ties (twelve years) than commercial banks provide in more entirely flexible finance-that is, as a "sweeper" of costs that developed markets (fifteen to twenty years or more). The cannot otherwise be funded because of strict procurement PSEDF increases the weighted average overall maturity of constraints. One common characteristic of cofinancier con- debt finance for the project to sixteen years. The subordi- tributions to the PSEDF and finance insured by the export nation of PSEDF facilities, in ranking of security interest credit agencies is the rigid conditions that apply to their and by virtue of their longer maturity, creates significant use. Expenditures funded by the ECO guarantee facility, additional risk cover for the providers of senior term debt. by contrast, need only satisfy the more subjective criteria of Loan drawings from the PSEDF are made by the pro- economy and efficiency The Commonwealth Development ject in the currencies of the underlying cofinancier loans to Corporation is similarly flexible. the government of Pakistan and in accordance with cofi- In project finance there are significant costs, such as nancier conditions relating to eligibility of procurement and interest during construction and funding of reserve the like. The project's debt service obligations to the PSEDF accounts, that satisfy neither tied nor Bank international are, however, standardized and become rupee denominated competitive bidding procurement criteria, and yet are just as and when each respective drawing is made, thereby pro- as much capital costs of the project as the power station's tecting the project from exchange rate risks between the equipment. The ratio of these "soft costs" to the "hard rupee and currencies of drawing. costs" of construction generally lies in the range 1:4 to The government's obligations to make debt service pay- 1:1, depending on the type of project and the length of ments to the Bank and other contributors to the PSEDF the construction period. For the Hub Power Project the are not conditional on the project's ability to meet its debt ratio is 1:1.6. The term debt finance in the project finance service payments to the NDFC (which administers the plan is 35 percent flexible and 65 percent inflexible. Equity PSEDF on behalf of the government of Pakistan). That is, finance is 100 percent flexible. the government, not the cofinanciers, accepts limited The problems of funding soft costs are even greater rel- recourse exposure to the project. ative to standby finance (sometimes referred to as contin- gent finance), which is held in reserve to fund unforeseen Expanded Cofinancing Operation (ECO) guarantee costs incurred in completing construction that cannot be facility met from other sources. This contingent reserve typically is equal to 10-15 percent of the base finance (in the Hub The ECO guarantee facility and a similar facility guaran- Power Project the reserve is equivalent to 14 percent, or teed byJEXIM together support $360 million of senior term $221 million), where the construction contract has been debt (including $60 million in standby debt) for the pro- let on a strictly fixed-price turnkey basis. The ECO and ject. The cover provided under the ECO and JEXIM guar- JEXIM guarantee facilities together provide 27 percent of antees is partial, pertaining to certain defined political the standby finance arranged for the project. risks such as nationalization and foreign currency avail- Exchange rate risks between the rupee-denominated ability. The guarantees cover 100 percent of loan principal. tariff revenues of the project and the debt service on the They do not cover interest, which is secured through cash ECO guarantee and other senior term facilities are insured collateral accounts. By removing the political risks that by the State Bank of Pakistan through its foreign exchange would otherwise prevent commercial banks from lending risk insurance scheme. 2 Financing Pakistan's Hub Power Project Commercial bank syndication ance in the overall commercial bank syndicate of several banks that previously had not participated in financing of Offshore commercial banks provided $686 million in lim- private infrastructure. The decision by a group of Italian ited recourse senior term debt for the project, allocated banks, led by Mediocredito Centrale, to participate in the between the ECO and JEXLM guarantee facilities and the first such financing supported by SACE also led other banks three facilities insured by export credit agencies (executive to participate in private infrastructure for the first time. summary table 2). Syndication of these five facilities was carried out simul- Issues Affecting the World Bank taneously by a five-bank arranger group comprising Bank of Tokyo-Mitsubishi, Citibank, Credit Lyonnais, NatWest, Additionality and Sakura Bank, supported byMediocredito Centrale with respect to the SACE insured facility. The arranger group The unanimous and unequivocal view of the government was chosen to reflect the national combination of suppli- of Pakistan, the arranger banks, sponsors, and cofinanciers ers within the construction consortium and associated export of the Hub Power Project is that they would not have credit agency-insured facilities and to include banks with accepted the costs, risks, and diversion of resources required a proven reputation in large limited recourse projects. to complete the development of the project had it not Syndication was launched in May 1994 and closed four been for the participation and support of the World Bank. months later, more than 100 percent oversubscribed. The The Bank's participation was also fundamental for banks reasons for such a successful syndication ranged from the joining the project in syndication. participation of the World Bank and the novelty and scope The Bank provides 39 percent of the PSEDF's funding of the ECO guarantee to the extended export credit agency and, through the ECO partial risk guarantee, supports 29 per- insurance cover (executive summary table 3). cent of senior term finance. Thus, of the $1,766 milion mobi- From the Bank's perspective, syndicationwas anunqual- lized for the project (base and standby finance combined), ified success. It achieved its objective of mobilizing the nec- $465 miDlion (26 percent) was provided or partly guaranteed essary foreign currency commercial bank debt for the project by the Bank. This represents an overall mobilization ratio of with an overall maturity of twelve years and in so doing Bank to all sources of finance on the order of 1 to 4. created an opportunity for thirty-four commercial banks Furthermore, the project is the first time so many cofi- from eight countries to participate in the first ECO- nanciers have participated in a private project. If the agen- guaranteed limited recourse project financing. The pres- cies of the French, Italian, and Japanese governments that ence of the Bank and JEXIM and the novelty of the contributed to the PSEDF are considered distinct from guarantee program are also partly responsible for the appear- COFACE, SACE, and MITI, the project had nine EXECUTIVE SUMMARY TABLE 2 Facilities funded by offshore commercial banks Amount Number of Average participation Facility (US$ million equivalent) banks (US$ million equivalent) Facility agenta ECO guaranteed 240b 34 7.1 Bank of Tokyo-Mitsubishi JEXIM guaranteed 120c 19 6.3 Sakura Bank COFACE insured 45 7 6.4 Credit Lyonnais MITI insured 86 17 5.1 Sakura Bank SACE insured 195 18 10.8 Mediocredito Centrale Total 686 43d 6.0e a. Citibank is the intercreditor agent. b. Includes $40 million of standby debt. c. Includes $20 million of standby debt. d. Most banks participated in more than one facility. e. Average overall participation per bank. Executive Summary 3 EXECUTIVE SUMMARY TABLE 3 not require that all the equipment procured be the cheap- Reasons for success in syndication est available. (percent) In principle the Bank's rules may be applied to a pro- Indicative ject's output costs, such as a power tariff-as in the 60 Rank Reason weighting megawatt Rockfort power project inJamaica (financial close I Presence of the World Bank and novelty 20 October 1994)-which opens the door for all of a pro- of the ECO guarantee 2 Scope of the ECO guarantee 20 ject's capital costs (hard or soft and however procured) to 3 Market bming 15 become eligible for Bank funding. But it may not always 4 Financial and commercial structure and quality IS be possible to adopt this approach, particularly for larger of sponsors 5 Arranger group and market presentation 10 projects in which the tenders for the output costs are con- 6 Italian market underwriting of the SACE insured facilityI 10 ditional upon sources of finance other than the Bank being 7 Pricing 5 8 Extended export credit agency insurance cover S available on appropriate terms. (for commercial risks) after completion of construction In the absence of this or alternative approaches, the Total 100 Bank's participation in a project through a vehide such as Source: Survey of arrangers. the PSEDF will be limited to the extent of procurement that satisfies Bank rules. About 40 percent of the Hub Power cofinanciers. The Bank played a central catalytic role in Project's turnkey construction contract costs satisfied Bank introducing the cofinanciers, sustaining their interest, and procurement rules. The procurement of goods and services bringing them together within the context of the inter- under intemational competitive bidding has the potential creditor agreement. to introduce delays into a project timetable. Most of the offshore banks that participated in ECO, JEXIM, and the three facilities supported by export credit Finance plan fragmentation agencies had never lent to Pakistan before, and many had never participated in a Bank cofinancing before. It has become increasingly apparent that the financing of major private infrastructure projects in developing mar- lntemational competitive bidding kets requires the participation of several cofinanciers and multiple loan facilities. For example, the finance plans for The philosophy underlying fixed-price turnkey contract- the Paiton I power project in Indonesia, the Sual power ing-in which the contractor absorbs cost overruns but ben- project in the Philippines, and the Izmit Su water project efits from savings if it can procure goods or services more in Turkey each comprise at least four foreign currency senior cheaply than the original estimate that lies behind the con- loan facilities supported by various export credit agencies. tract price-is not necessarily consistent with the philoso- The Hub PowerProject finance plan differs from these only phy of international competitive bidding, in which there in the extent of its fragmentation. can be a transparent relationship between the cost of pro- These finance plans are fragmented for two reasons. curement and the price charged to the client. Nonetheless, First, the contractor may divide procurement responsibil- fixed-price turnkey contracting is ubiquitous in private infra- ities into a series of discrete national packages. Second, structure projects with limited recourse financing (such as there is generally a tight ceiling on the political risk cover build-own-operate projects). available for a single country or project from individual For project sponsors the competitiveness of the overall export credit agencies or other cofinanciers. Other cofi- tariff offered to the client (host government) from the infra- nanciers also may be responsible for the contractor opting structure project and the project's deliverability in terms for discrete national packages of procurement. If the Bank of available sources of finance are more important than is involved in the project, the objective of cofinancier mobi- the capital costs of procurement. Thus a project sponsor lization, as well as the Bank's role as lender or guarantor will want to select contractors according to criteria that may of last resort, further encourage fragmentation. 4 Financing Pakistan's Hub Power Project The problem is compounded when export credit agency the process of evolutionary change may be accelerated by cover for a country is limited or when a project wishes to the use of competition to "test the envelope." access a significant portion of the cover available from a These processes have been evident in subsequent inde- single export credit agency. The duty of export credit agen- pendent power projects in Pakistan, which have progres- cies will remain to their exporters, and they may reduce sively fewer government undertakings within their their commitment to a single project rather than see it ster- implementation agreements, lower levels of political risk ilized (that is, committed but undrawn) by project novelty coverage, and less commercial risk coverage under export or complexity, which cause a project to take years to bring credit agency insurance policies, as well as lower debt cov- to financial close. erage ratios. Since 1995 there have been increasing exam- The ECO and JEXIM guarantee facilities use an iden- ples of international banks being prepared to commit tical form of guarantee, whereas each export credit agency uncovered term finance for projects in Pakistan-in con- uses its traditional form of insurance policy. As a result trast to the market position in the early 1990s. four forms of documentation for political risk cover were deployed on the Hub Power Project. Mobilization finance Given that the number of export credit agencies par- ticipating in the project is not abnormally high compared The turnkey contractor for the Hub Power Project was a with other developing country independent power projects, consortium comprisingAnsaldo of Italy, Campenon Bernard the PSEDF's structure is the most significant source of frag- of France, and Ishikawajima-Harima Heavy Industries mentation and consequent complexity. The difficulties of and Mitsui of Japan, with Mitsui as overall consortium coordinating multiparty cofinancings may be partly reduced leader. The contractor was mobilized in December 1992, by bilateral collaboration arrangements and internal reor- some two years prior to financial close. As a result power ganizations, which some cofinanciers-notably export credit from the project was first exported less than eighteen months agencies-have since initiated. after financial close, inJune 1996. Mobilization finance was provided by the Bank and the Risk transfer governments of France and Italy in the form of term loans through the PSEDF; by Al Rajhi and NDFC in the form For developing countries one of the central achievements of bridge loans; by the Commonwealth Development of the Hub Power Project is the creation of a contractual Corporation through a term loan; and by the project spon- framework in which the commercial risks normally associ- sors through equity contributions. Some $423 million in ated with an independent power project are transferred to mobilization finance was advanced from these sources- private lenders and project investors. This reflects one of in the case of debt facilities, largely with the support of the the government's and the Bank's earliest and most funda- government of Pakistan. mental objectives of the private initiative in Pakistan- Mobilization finance played a crucial role in maintain- that commercial bank finance and equity capital be mobilized ing the project's timetable for delivering power to the without direct government guarantees of debt service or Pakistan grid and in avoiding the potential cost increases return. As a result the risk matrix used in the project's that may follow from the expiration of a turnkey contract bank information memorandum is essentially the same as price offer. It also represented a "crossing of the Rubicon" for an independent power project in an industrial country for all concerned-the government of Pakistan, the Bank, (except, of course, with respect to political risks). and sponsors-in terms of creating an unstoppable momen- The envelope of risk that lies with the private sector in tum to the project development effort. In circumstances infrastructure projects maybe expanded in two ways. First, where institutional development forms an integral part of because of increased experience and the confidence that the overall project development effort (as in the case of flows from it, the private sector will be progressively more the Hub Power Project) or where other factors give rise to willing to accept risk, to price it more finely, or both. Second, unacceptable delays in the start of construction, mobiliza- Executive Summary 5 tion finance is one of the few mitigating measures avail- and the first limited recourse financing. For the Bank, it able to the host government and the Bank. was the first private infrastructure project, Bank-financed In arranging mobilization finance that did not disturb infrastructure fund (the PSEDF) to support private pro- the overall permanent finance plan, the project encoun- jects, partial risk guarantee under the ECO program, ECO tered an important limitation on the ability of some facili- guarantee with another institution (JEXIM), and use of ties to refinance construction expenditures that had already the ECO program to support a private project. been funded, but that were otherwise entirely eligible. For the financial markets, it was the first major private This crowding out of some of the permanent facilities would infrastructure project in a sub-investment grade develop- have become a serious problem for the project had finan- ing country to be financed by international commercial cial close been further delayed. banks on a limited recourse basis, international equity offer- ing (global depository receipt) and underwriting for a devel- Cofinancier leadership oping country infrastructure project under construction, stock market flotation of a single power station under con- At the project's cofinancier meetings, sometimes attended struction, multi-cofinanced private infrastructure project, by as many as forty delegates, it was natural that the Bank COFACE-supported limited recourse financing for an infra- be perceived as the catalyst and assume the role of chair- structure project, JEXM-supported limited recourse financ- man, particularly since government institutions were in ing for an infrastructure project, Mm-supported limited attendance. Moreover, pressure on the Bank to assume a recourse financing for a major infrastructure project, SACE- wider coordinating role on behalf of cofinanciers grew from supported limited recourse financing for an infrastructure this chairmanship. Given that the Bank was already a major project, and markup-based (in accordance with Islamic prin- contributor to the development of the energy sector in ciples) limited recourse infrastructure project financing (the Pakistan, this coordinating role had an inherent rationale. locally arranged rupee term facility). The Bank may have been a reluctant leader in circum- stances such as these, particularly given the intensive demand Project Development on personnel resources that such a role brings. But it has a unique capacity to shoulder these responsibilities, and its Development timetables suitability is universally endorsed by the government of Pakistan and the project's cofinanciers, banks, and spon- The origins of the Hub Power Project can be traced back sors. Most cofinanciers see proactive cofinancier coordi- to 1985, to the announcement by the government of Pakistan nation by the Bank as the best hope of speeding up the of an initiative to encourage private participation in power flow of future projects. generation. Work on developing the project began in the The Bank was also, quite naturally, cast in the role of summer of 1987 and expanded rapidly from April 1988 confidante to the government of Pakistan. On occasion this onward, when a detailed feasibility study began. role extended to "honest broker" to help resolve issues Construction of the station began in December 1992, on between the government and the project's sponsors. Both the basis of mobilization finance, and the total finance pack- roles are invaluable when the project under development age for the project became irrevocably committed in is pioneering and the sectoral initiative is new for the host September 1994. Financial close was achieved in January country. 1995. The elapsed time taken to develop the project is generally regarded to be eight years (1987-95). Significant Project "Firsts" If, however, allowances are made for time lost through unforeseen events such as the Gulf war (1990-91), a Pakistan The Hub Power Project occasioned many "firsts" for court ruling on the applicability of Shariah Law to the pay- Pakistan, the Bank, and the international financial markets. ment of interest (1991-92), the disruption caused by hav- For Pakistan, it was the first private infrastructure project ing to reconstitute the construction consortium following 6 Financing Pakistan's Hub Power Project the departure of two members (1990), and the occasional does not easily lend itself to classical project management losses of continuity that attended several changes of gov- techniques that assume a level of control over events by the ernment in Pakistan, the active period of project develop- project management team. There are several common causes ment is closer to five years. of timetable slippage that are rooted in the number of inde- The project's history is not simply about the develop- pendent parties required to contribute to a project, often ment of a project, but about giving life to an entire policy with very different priorities and levels of relevant experi- initiative by the government of Pakistan and the Bank. This ence. These difficulties are compounded by project novelty, endeavor necessarily entailed the creation of new institu- complexity, and the disruptive effects of force majeure. tions in Pakistan as well as a large body of documentation, The preparation of timetables is an important activity all of which are of a general (rather than project-specific) in project development-not because they give accurate application. The speed with which several further inde- predictions of when financial close will occur (which they pendent power projects reached financial close in Pakistan, do not) but because they help define component develop- subsequent to the Hub Power Project, illustrates this point. ment activities, identify potential critical paths, milestones, In developed markets independent power projects usu- and decision points, determine the level of resources ally take one to two years to bring to financial close. In devel- required, and establish priorities. oping markets a pattem is emerging of such projects requiring For parties unfamiliar with private sector project devel- an average of two to three years, although the distribution opment, having only partial control over the project timetable of development periods is skewed toward longer than aver- is unnerving as well as frustrating. It can often lead to age periods. In taking eight years to reach financial dose, sponsors losing patience and confidence and pulling out the Hub Power Project is at the upper end of lengthy devel- of projects as they find that financial close is forecast to be, opment periods-but it is not unique. If the active period say, nine months away-year after year. of project development is considered (five years), the pro- Few companies in the private infrastructure market have ject is in the third quartile of project development periods. sufficient financial strength and resources to undertake Projects that achieve the shortest development periods, major capital projects alone. Moreover, most are relative from the point of view of private sponsors, are usually well newcomers with limited experience working in developing defined technically and commercially by the host govern- countries. As a result the pattem that is emerging interna- ment or authority before the procedure for award of con- tionally is of infrastructure projects being undertaken as cession begins (whether award is by a competitive or joint venture-sponsored or limited recourse-financed negotiated process). Independent power projects are par- enterprises-or, commonly, both. ticularly vulnerable to timnetable overrun because of the The market for private infrastructure projects, and for number of parties and contracts normally required to bring independent power projects in particular, has changed dra- them to fruition. matically since the inception of the Hub Power Project. When One important feature of developing private rather than Xenel sought an electric utilitypartner for the project in 1988 public infrastructure projects is the largely sequential nature and issued an intemational invitation, only four companies of the key technical, commercial, and financial tasks that responded. Of these, only National Power had extensive have to be performed. For example, negotiation of the experience working in developing markets. If such an invi- core commercial contracts supporting the project cannot tation were issued today, between 50 and 100 electric utility take place until the project is defined technically, and finance companies would likely respond. Still, few would have much cannot be arranged until the project is defined commer- operational experience outside their domestic markets. cially. This sequential pattem of activities is a key reason The development of the project has spanned the trans- development periods on private infrastructure projects are formation of the private infrastructure market, from being measured in years. a rare curiosity outside a handful of industrial countries (in An important lesson that can be drawn about project devel- 1987) to mainstream policy for many countries, both indus- opment timetables is that the process of project development trial and developing (in 1997). During this period trends Executive Summary 7 of privatization and utility deregulation have created a mar- ing on the sponsors to complete the assignment success- ket for 'global" utility companies. fully and as rapidly as possible. The stakes were made higher For its project development structure, the project chose for the government by the decision to make the project its a joint venture composed of parties with complementary sole prototype independent power producer, so that the full interests. The core group comprised five companies: Xenel, attention of the government, its agencies, and the Bank National Power, Mitsui, Ishikawajima-Harima Heavy could be concentrated with that of the sponsors on mak- Industries, and K&M. Collectively, they formed a project ing the project a success. Finally, the government's stakes development company (HRPG Limited) that, by virtue of were raised higher still by its support for mobilization finance. being independently resourced and managed, was able to Given the project's size relative to the resources avail- mitigate much of the inertia and dilution of management able within the government and the Bank, no other approach resources that comes from multiparty project sponsorship. was viable. The decision to form a queue of other projects behind the Hub Power Project already appears to have been Costs and risks vindicated by the speed with which subsequent projects have reached financial close. The development costs for projects that are not well defined The frustration of the government and others in having technically or commercially at the outset typically are 2 to to wait until 1995 to see so much effort come to fruition may 4 percent of the project's funding requirement. This rule be partly offset by the fact thatbetweenApril 1988 (the date of thumb appears to hold for most private infrastructure of issue of the first letter of intent for an independent power projects (power, transport, water) and, surprisingly, for pro- project in Pakistan) and March 1996 some 2,700 megawatts jects of all sizes. Thus there do not appear to be economies of independent power producer capacity were brought to of scale for development costs in sponsoring larger projects. financial close. This represents about 330 megawatts a year In fact, there are diseconomies of scale in that development in new capacity-an impressive achievement. costs for larger projects tend to be near the higher end of As the prototype project for Pakistan's independent that range-that is, 3 to 4 percent rather than 2 to 3 percent. power producer program, the Hub Power Project inevitably The Hub Power Project figure of 3.6 percent, while at bore a large measure of onetime development costs and the upper end of the range, is not exceptional. Given the risks that subsequent projects have been able to avoid. project's gestation period and novelty, it is surprising that Although no mechanism exists to share such costs and the figure is not higher. What is inevitably eye-catchingwith risks among project sponsors generally, sponsors have been all large projects is the absolute sum that the sponsors willing to accept them because of the distinction attached must put at risk in order to create a project. Few corpora- to being "the first," whether in Pakistan or any other coun- tions can afford to take gambles of this magnitude on their try. Such is the degree of competition among sponsors in own, particularly since the probability of success in project the market for private infrastructure projects. development can be low. Given the considerable resources that had to be mobi- The personnel required to develop the Hub Power Project lized to complete the development of the Hub Power Project, and bring it to financial close is estimated at 225 person- as well as the risks borne and costs incurred, it is remark- years split broadly as follows: the government of Pakistan able that fatigue did not set in among the government, the and its agencies and advisers, 25 percent; sponsors and their sponsors, the Bank, and the cofinanciers. Herein lies the advisers, 50 percent; the Bank, 10 percent; and other cofi- essential reason the project reached financial close-namely, nanciers, leading banks, and their advisers 15 percent. This none of the key participants wavered in their determina- excludes the resources deployed by the turnkey contractor. tion to see it through. The importance to the government The sponsors were not alone in bearing development and to the project sponsors of the Bank's commitment in risks; the govemment was also exposed to substantial risks. this regard should not be underestimated. All the project's In granting the project an exclusivity period in which to participants should be commended for their stamina and develop and finance the project, the government was rely- patience. 8 Financing Pakistan's Hub Power Project Mobilizing Sufficient and Least-Cost Shortening the Development Period Finance Development planning In most cases the shortage of debt financing is a greater constraint on the development of private infrastructure pro- In its predevelopment preparation for a project, and prior jects than is a shortage of equity. Thus the ECO guarantee to issuing the invitation for private sector proposals, time program is of crucial importance to the power sector. Since spent by a host government in defining the legislative envi- debt is cheaper than equity, maximum gearing is required ronment, the project's technical and commercial require- to achieve the lowest overall cost of funding for a project ments, and policy objectives within the context of what and, in turn, the lowest tariffs for customers. can be financed by the private sector will be more than saved Lower tariffs for customers may also be brought about by by the time financial close is reached. arranging debt finance with extended average and overall It is also advisable to postpone as many nonessential maturities. The main constraint on stretching commercial activities as possible until after financial close. This may bank loan maturities is the reluctance on the part of political require dividing the project into two or more distinct phases, risk guarantors to accept the longer exposure. This reluctance each with a separate finance plan and financial close. Any prevents developing countries from accessing the kinds of consequent reductions in the complexity of a given phase's loan maturities that are available for similar projects in devel- finance plan should yield time savings in reaching the ini- oped markets (that is, fifteen to twenty years or more). The tial financial close, albeit at the probable cost of a tempo- PSEDF, with loan maturities of twenty-three years (nearly rary loss of economies of scale in output tariff. twice that of the commercial bank finance for the Hub pro- Debt syndication is the most obvious financing activity ject), has a significant impact on available loan maturities. that can usually be deferred until after financial close. Still, export credit agencies and other cofinanciers should con- Generally, financial close should be sought on the basis of sider offering maturities of fifteen years or more for infra- an unconditional underwriting against which the project structure projects, to dose the gap between the terms of finance may begin loan drawdowns. that are available in developing and developed markets. Competition is a third route through which the cost of Focused resources funding projects may be minimized. The large number of potential sources of commercial term debt for private infra- The fewer project participants there are the better, be they structure projects indicates what may be possible in this contract counterparties, contractors, cofinanciers, or pro- area. However, the scope for competition is less broad for ject sponsors. Moreover, each participant must dedicate larger projects. The limiting factor on the supply of finance appropriate complementary resources. For sponsors, that is generally political risk cover for term loan facilities. means a strong, dedicated management team whose pri- Any finance plan that stretches to the limit the available mary responsibility is bringing the project to fruition. A sep- finance and associated political risk cover for a project or arate, self-contained, full-time project team is best. The country will face a further constraint in the form of inade- Bank also requires dedicated personnel; this will require quate flexibility of the finance plan, relative to soft costs that resources be freed up from other responsibilities to and the provision of adequate standby finance. This prob- concentrate on the project at hand. lem will usually be worsened by increased fragmentation in the finance plan. The most important feature of the ECO Avoidable complexity guarantee facility in this context is its complete flexibility with regard to the provision of both base and standby finance. Significant complexity arises from a fragmented finance This is a vital point of dlifferentiation between an ECO- plan, which may be unavoidable. However, within any and an export credit agency-supported facility that, in effect, finance plan much time, effort, and cost can be saved in makes the facilities complementary. implementation if the plan has been developed, in at least Executive Summary 9 a rudimentary form, before construction and equipment prior to financial close. Whether the Bank and the gov- supply contracts are tendered. For example, this may be the emrnent of Pakistan would have remained so unwavering only opportunity to achieve a match between the currency in their support of a smaller project remains unknown. of funding and the currency (or currencies) of capital expen- The Hub Power Project is a private project embedded diture or to solve problems that may arise if supplier export in a public environment. The project's fuel supplier (Pakistan credit agencies offer banks less than 100 percent political State Oil) and power purchaser (Water and Power risk coverage. Development Authority) are both owned by the state, as The negative effects of finance plan fragmentation may are several local financial institutions that provided term be partly mitigated by the adoption of harmonized or more loan finance for the project. This pattern of partnership standard forms of documentation. between private and public sectors is likely to be repeated in future project financings that introduce private capital Timetable paradox into a nation's infrastructure development plans. The Bank's role as facilitator of this process was crucial and welcomed To achieve the shortest development timetable a project by all the parties involved in the project. should, paradoxically, be configured to survive a long haul There are many other ways in which the Bank can, and to financial close, whether caused by force majeure or oth- in the case of the Hub project did, contribute to a project's erwise. Particular measures that should be taken include: success. Three measures deserve priority in the future * Determining the level of cross-party political support because of their impact on the flow of private infrastruc- within the host country ture projects in developing markets and on the financial * Testing the sponsors' commitment to contribute to pro- convergence of these with equivalent projects in developed ject financing markets: * Planning for continuity of resources despite sporadic Assumption by the Bank of a leading role with regard progress to coordinating cofinanciers-both generally, in their * Preparing commercial contracts so that they can sur- approach to such projects, and specifically, in relation vive unforeseen timetable slippage without requiring to individual projects that the Bank agrees to support renegotiation. (this is particularlyimportantwhen the project is a "first" Planners should avoid the temptation to believe in "fast for the host country). track" timetables. * Allocation of resources by the Bank on a project-specific and usually dedicated basis for the duration of the pro- Conclusion ject development period (for resource planning purposes, the equivalent of at least two Bank staff per project should A question often asked about the Hub Power Project is be assumed). whether, had it been smaller, it would have been devel- Deployment of ECO guarantees and PSEDF-style facil- oped more rapidly. The best answer that can be given is ities to extend the maturity of available term finance to probably-so long as there was an associated reduction in twenty or more years and to ensure that projects have the complexity of the finance plan. Such a reduction in com- sufficient flexible finance to overcome constraints that plexity might have been achieved, for example, by having may attach to finance that is available from other a single PSEDF cofinancier facility, a single export credit cofinanciers. agency facility, and no rupee facility. The fourth unit of the Hub Project was completed three Much of the project's novelty was independent of its weeks ahead of schedule on 7 March 1997, and formal com- size, as was the time taken to develop the project agree- pletion of the entire station complex was announced, on ments and to establish new institutions in Pakistan. Even schedule, on 31 March 1997. The actual capital cost of com- much smaller independent power projects can have more pleting the projectwas slightly below budget-infact, a sav- than 100 project agreements and documents to complete ings of about 1 percent of the budget were realized. 10 Financing Pakistan's Hub Power Project Finance Plan Overview its finance plan. Under the terms of Hubco's agreements with Pakistan State Oil and the Water and Power The capital cost of the Hub Power Project was budgeted Development Authority, the completion of those facilities at $1,608 million. This total indudes payments for the design, was the responsibility of Pakistan State Oil (as oil suppler) fabrication, erection, and commissioning of the station ($989 and the Water and Power Development Authority (as power million) as well as the associated finance, management, pro- purchaser). ject development, and insurance costs ($619 million). During its commissioning the station was expected to generate $63 Private Sector Energy Development Fund (PSEDF) million in net income from power sales, which lowered the forecast capital funding required during construction to The PSEDF is a funding structure developed by the World $1,545 million. This figure formed the basis for the finance Bank to direct multilateral and bilateral finance into pri- plan. vate energy enterprises in Pakistan. Hub Power was the first project to qualify. The Bank developed a similar vehicle, Capital structure the Private Sector Energy Fund, for the 60 megawatt Rockfort Private Power Project in Jamaica (financial close The finance plan comprises three basic sources of capital: occurred in October 1994). The Bank and other contrib- equity, which accounts for 24 percent of funding ($371 mil- lion); senior debt, which accounts for 48 percent ($738 mil- TABLE I lion); and subordinated debt, which accounts for 28 percent Base finance plan ($436 million; table 1). Equity is in the form of ordinary Amounta Contribution share capital in the Hub Power Company Limited (Hubco), Source (US$ million equivalent) (percent) the company that owns the project. Senior debt is in the Equity 371 24 form of seven individual term loans-some multitranche ECO guaranteed 200 and all with an overall maturity of twelve years-one of JEXIM guaranteed 100 COFACE insured 45 which is guaranteed by the Expanded Cofinancing MITlinsured 86 Operation (ECO). Subordinated debt is in the form of SACE insured 195 CPC 37 two multitranche term loans from the Private Sector Energy Rupee facility 75 Development Fund (PSEDF) with an overall maturity of Subtotal 738 48 twenty-three years. Subordinated debt PSEDF I 322 Construction of a 78 kilometer pipeline to connect the PSEDF 11 114 power station to the fuel oil storage terminal at Pipri (near Subtotal 436 28 Port Qasim) and of a 200 kilometer transmission line to Total 1,545 100 connect the power station to the grid atJamshoro were out- Note: Figures are in end-1993 dollar exchange rates. In practice sources and applica- tions of finance are in mnuftiple currencies. side the scope of the project and thus did not form part of a. Does not include standby debt. 11 utors to the PSEDF make loan advances to the government Expanded Cofinancing Operation (ECO) and Export- of Pakistan which, through the state-owned National Import Bank ofJapan JEXIM) guarantee facilities Development Finance Corporation (NDFC), onlends the money to the Hub Power Project on standardized terms. The Expanded Cofinancing Operation (ECO) guarantee For the first PSEDF facility (PSEDF I) the interest rate facility was developed under the World Bank's ECO pro- paid by the project is fixed at 14 percent a year for the gram. The Hub Power Project's ECO guarantee facility is duration of the loan. For the second facility (PSEDF II) a $240 million senior term loan provided by an international the rate is fixed at 14 percent a year until the start of com- syndicate of thirty-four commercial banks. The loan has mercial operations, then becomes a LIBOR-linked float- limited recourse to the project except for defined political ing rate. Repayments for both facilities begin on the eighth risks, for which lenders are secured by an ECO guarantee anniversary of financial close and are made on a straight- issued by the Bank-that is, a partial risk guarantee. The line basis over the succeeding fifteen years. The average ECO guarantee covers 100 percent of loan principal but loan maturity for the PSEDF is 15.5 years. The govern- no interest, and is secured for up to thirteen months and ment's obligations to make debt service payments to the seven days (the maximum call period of the ECO guaran- Bank and other contributors to the PSEDF are not condi- tee) by a cash collateral account (guaranteed reserve account) tional on the project's ability to meet its debt service oblig- funded by the project. ations to the NDFC. That is, the government of Pakistan The ECO guarantee facility has four currency base (through the NDFC), not the cofinanciers of the PSEDF, tranches (denominated in U.S. dollars, Japanese yen, accepts limited recourse exposure to the project. European currency units, and French francs) totaling $200 The cofinanciers that agreed to provide advances to the million and a single standby tranche of $40 million (denom- Hub project through the PSEDF are shown in table 2. inated in U.S. dollars). Interest on each tranche is a LIBOR- The Bank participates in both PSEDF I and PSEDF H. linked floating rate.A separate guarantee fee of 0.25 percent The Bank's contribution through the PSEDF is $225 mil- a year is payable by the project to the Bank on the facility lion, or 39 percent of the project's subordinated debt amount. The conditions of loan availability follow prece- facilities. dents for commercial bank finance for independent power Although many of the terms on which the NDFC onlends projects in industrial countries and do not impose interna- the PSEDF's component facilities to the Hub Power Project tional competitive bidding or other procurement restric- are standardized, certain important conditions are not- for example, the currency of disbursement and definition CABLE 2 of eligible expenditure, which are controlled directly by Development Fund the respective cofinancier. (US$ million equivalent) Although drawings from the PSEDF are made by the Base Standby project in the currencies of the underlying cofinancier loans Source facilities facilities Total to the government of Pakistan, project debt service oblig- PSEDF I ations to the NDFC become denominated in rupees when World Bank 145 n.a. 145 . . . . . ~~~~ ~ ~ ~~ ~ ~ ~~JEXIM 94 53 147 each respective drawing is made. This approach protects France 27 n.a. 27 the project from exchange rate risks that may arise from a Italy 46 n.a. 46 mismatch between the currency of tariff payment under the USAID 10 n.a. 10 Subtotal 322 53 375 power purchase-agreement (rupee) and the currency of cofi- PSEDF 11 nancier debt service. However, such protection makes tar- World Bank 35 45 80 ., ~~~~~JEXIM 7 1 38 109 iffs under the power purchase agreement significantly more France 8 n.a. 8 complex. The interest rate paid by the project on PSEDF Subtotal 114 83 197 drawings includes a margin to account for this exchange Total 436 136 572 risk insurance. n.a. is not available. 12 Financing Pakistan's Hub Power Project tions on the project. Repayments begin on the fourth anniver- in July 1991, when the contract was awarded, would have sary of financial dose and are made on a straight-line basis lapsed. Since this fixed price lay behind the tariff agreed in over the succeeding eight years. Thus the average loan matu- August 1992 between the Hub Power Project and the gov- rity is eight years. ernment of Pakistan, a contract price revision or rebid could Debt service is paid by the Hub Power Project directly have had serious consequences for the project's timetable to the facility agent for the ECO guarantee facility in the and might have necessitated renegotiating tariffs (see the appropriate currencies. The project's rupee-denominated section on project development, below). By the time of tariff revenues are converted into these currencies by the financial dose about $423 milion of finance had been drawn State Bank of Pakistan, which assures the project of fixed down to fund project development and construction costs exchange rates under its foreign exchange risk insurance (table 3). scheme. The two bridge finance facilities shown in table 3 were An Export-Import Bank of Japan JEXIM) guarantee repaid out of the proceeds ofthe permanent financing.Thus facility comprising $100 milion of base and $20 milion of they are included in the finance plan figure of $1,545 mil- standby finance is provided on terms broadly similar to lion and are not shown separately in table 1. The four term those that apply to the ECO guarantee facility (although finance facilities shown in table 3 rolled over at financial the JEXIM guarantee facility is denominated in only a sin- dose and became part of the permanent finance plan. Bridge gle currency, yen) and is funded by a separate international finance was structured as short-term (for example, twelve syndicate of nineteen banks. The guarantee instruments months) facilities secured by the government. As a result used by the ECO and JEXIM facilities are the same and of mobilization finance the first power exports from the sta- were prepared byJEXIM's legal counsel. tion were achieved on schedule less than eighteen months after financial close, in June 1996. Other senior term loans World Bank perspectives In addition to the ECO and JEXIM facilities, the Hub Power Project has five senior term loans, including three Additionality. The unanimous and unequivocal view of base facilities insured by COFACE, Mlll, and SACE that arranger banks, sponsors, and cofinanciers of the Hub Power are funded by separate international syndicates of com- Projectis thattheywouldnothave accepted the costs, risks, mercial banks (see appendix 1). and diversion of resources needed to complete project devel- The sixth senior facility is a term loan from the Gommonwealth Development Corporation. The seventh TABLE 3 Mobilization finance is a rupee term facility structured on the Islamic principles of "markup" rather than interest. The rupee facility is funded Source (US$ million equivalent) availability by a syndicate of nine Pakistani banks and institutions (see Equity appendix 4). Sponsors 65 January 1987 The seven senior term facilities all have the same over- Others 36 September 1994 Subtotal 101 all maturity (twelve years), repayment profile (straight-line), Bridge finance and average maturity (eight years). Al Rajhi Istisna 92 December 1992 Rupee facility I0 December 1993 Subtotal 102 Mobilization finance. Term finance Bank (PSEDF I) 110 April 1993 France (PSEDF 1) 27 June 1994 The turnkey contractor appointed to construct the power Italy (PSEDF 1) 46 June 1994 station was mobilized in December 1992, about two years CDC 37 September 1994 before financial close (January 1995). Had mobilization Subtotal 220 been delayed further, the construction price that was fixed Total 423 Finance Plan 13 opment had it not been for the World Bank's participation political risk guarantees modeled on the ECO guarantee and support. The government of Pakistan also welcomed deployed on the project. the Bank's participation as a cofinancier, adviser, and coor- dinator, not least because of its long-standing support for Mobilization finance. Mobilization finance played a cru- Pakistan's energy sector. The Bank's participation was also cial role in maintaining the project's timetable for delivery essential to the decisions of banks and investors to join the of power to the Pakistan grid, which was agreed in November project at financial close. 1992 when a mobilization agreement was signed. It also Of the $572 million held by the PSEDF, 39 percent is represented a "crossing of the Rubicon" for all concerned- provided by the Bank. Of the $823 million of senior term the government of Pakistan, the Bank, and sponsors-in finance, 29 percent is partly guaranteed for risk by the Bank. terms of creating momentum for the project development Thus of the $1,766 million in financing mobilized for the effort (see the section on project development, below). project (base and standby finance combined), the Bank The short-term nature of part of the mobilization finance provided or partly guaranteed $465 million (26 percent). (namely, the bridge finance facilities) and the rate at which This represents an overall mobilization ratio of the Bank construction expenditures are generally incurred once a to all sources of finance of about 1 to 4. The project is the contractor has been mobilized placed considerable pres- first time so many cofinanciers have participated in a pri- sure on all concerned. In hindsight this pressure can be vate project. If the French, Italian, and Japanese govern- recognized as beneficial for the project, although poten- ment agencies that contributed to the PSEDF are considered tiallyweakening to the negotiating position of parties already separately from COFACE, SACE, and Mm, there are nine involved in the project and for whom commercial issues cofinanciers. were still outstanding. Such issues had been few and rela- In all, forty-three international commercial banks par- tively minor on the project by this time. But if commercial ticipated in the facilities supported by ECO, JEXIM, and issues had been outstanding at the time of mobilization, the three export credit agencies. Most had never lent to the additional pressure could easily have been destructive Pakistan before, and many had never participated in a Bank rather than constructive. cofinancing. For some banks the Hub Power Project was In arranging mobilization finance that interfaced effi- their first participation in private infrastructure project ciently with the overall permanent finance plan, the pro- financing. ject encountered an important limitation on the ability of The Hub project was the first time that COFACE and some facilities (such as Bank and JEXIM contributions to SACE had supported a limited recourse infrastructure the PSEDF) to refinance construction expenditures that project financing. The Bank clearly played a catalytic role had already been funded but that were otherwise entirely in introducing them, sustaining their interest, and bring- eligible. This crowding out of some of the permanent finance ing them together to share certain common views-for facilities would have become a serious problem for the example, the project coordination agreement between project had financial close been delayed further. Hubco and its various lenders (the intercreditor agree- ment). The rupee term facility was Pakistan's first domes- Development of the Plan tically arranged limited recourse financing for an infrastructure project. Origins The project also has stimulated initiatives among cofi- nanciers that go beyond their direct participation in the pro- The basic structure of the finance plan was established early ject and even among some cofinanciers that did not in the development of the project and underwent rela- participate in the project. For example, COFACE has since tively few changes before final implementation. The pri- established a project financing group, and its Export Credit mary objective, as with all finance plans for infrastructure Guarantee Department and the Asian Development Bank projects, was to mobilize the longest maturity and cheap- are considering participating in project financings through est finance terms possible in order to achieve an accept- 14 Financing Pakistan's Hub Power Project able tariff. The primary constraint on the finance plan was introducing considerable extra novelty to the project. In the limited or, in the case of longer maturities, nonexistent 1991 JEXIM agreed, in principle, to provide a sister guar- availability of foreign currency commercial bank term loans antee facility. One of the most striking features of the Hub for private enterprises in Pakistan, other than with the project finance plan is its complexity and associated level support of cofinanciers. of innovation. There are seven senior loan facilities and, The PSEDF was created in 1988, and its availability to through the PSEDF, eight subordinated loan facilities. the Hub Power Project formed a fundamental part of the The objective of securing the cheapest available term of finance plan from that time. The PSEDF was designed to finance was pursued in three ways. First, the PSEDF was finance up to 30 percent of the capital costs of qualifying structured to offer a fixed (14 percent a year) rupee inter- projects at maturities of up to twenty-three years, subject est cost to borrowers that was (when the PSEDF was estab- to it not providing more than 50 percent of the project's lished) and remains an attractive rate for twenty-three-year foreign exchange costs. subordinated term finance in Pakistan. Second, senior debt The government set the minimum level of equity con- interest costs were based on a combination of LIBOR-linked tribution to the finance plan at 20 percent. (This lower limit floating rates, export credit agency subsidized rates, and was also subsequently adopted as a loan condition by the fixed rates. All were settled on an open-book basis and, PSEDF.) The government briefly adopted a 25 percent limit where competition among finance providers was not pos- on equity contributions in 1988-90, but ultimately it set- sible (which was generally the case), the project was obliged tled back at 20 percent (notwithstanding a target of 25 to demonstrate that reasonable going-market rates had been percent if market conditions permitted, which they ulti- achieved. Third, in the case of equity, which is usually a mately did). more expensive form of capital than debt, the government The fixed components of the finance plan became equity specified a ceiling of 18 percent a year on base-case rates (20-25 percent) and the PSEDF (30 percent), which left of return on equity for the project. Adherence to a regu- a balance of 45-50 percent to be found in mostly foreign lated, base-case rate of return was achieved by using a currency term loans. Identifying sources for these loans computer-based model of the power tariff and finance became the project's main fundraising task. Originally it plan that was maintained by the government, the Water and had been hoped that facilities insured by export credit agen- Power Development Authority, and the project on an open- cies and term loans from multilateral and bilateral agen- book basis. cies (other than the Bank) would meet the shortfall. By It should be stressed, however, that the actual rate of 1990, however, it had become clear that because of the return to investors is not assured by the government or any limited and cydical availability of export credit agency insur- other party. Whether the projections of future shareholder ance cover for Pakistan and such agencies' nervousness (and lender) cash flow contained in the base-case computer about limited recourse finance structures, these sources model are realized depends on the project achieving timely were unlikely to be sufficient. and on-budget completion of construction and subsequent In 1988 the Commonwealth Development Corporation reliable and efficient operation (box 1). had indicated its interest in providing term finance and equity for the Hub Power Project. It was, however, the The PSEDF's role only multilateral or bilateral agency to do so apart from the Bank, although others were invited (including the Asian The PSEDF provides long-term finance, and in so doing, Development Bank, International Finance Corporation, anticipates one of the key advantages of long maturities that and Islamic Development Bank). As a result the Bank's would normally be associated with a refinancing (which timely announcement of the ECO program (in 1990) and would typically be implemented by an independent power its subsequent approval (in 1991) made the ECO guaran- producer once construction was complete). No other inde- tee facility part of the finance plan to make up for the short- pendent power producer in a developing market has entered fall in senior term finance-although at the expense of the construction phase with such long-term finance (twenty- Finance Plan 15 involved in the infrastructure business at that time. Since Box I Profile of private sector risk the Hub project, several smaller independent power pro- Investors in and lenders to the Hub Power Project-including jects have reached financial close in Pakistan without the commercial banks participating in the ECO and JEXIM guar- PSEDF support, relying solely on different combinations antee facilities-are exposed to the same commercial risks that of export credit agency-, IFC-, and JEXIM-supported they would face in a limited recourse financing of an indepen- debt finance, as well as equity and locally arranged term dent power project in an industrial country-that is, during con- finance struction, risks of the plant being built on time and within budget and of its achieving intended performance characteristics; and during operation, risks of the plant performing reliably consum- ECO guarantee facility's role ing fuel, and requiring maintenance in line with expectations. An adverse outcome with respect to one or more of these The ECO guarantee facility and its sister JEXIM guaran- risks resuks in reduced or no retums for shareholders and could, tee facility together provide about 40 percent of base senior in severe cases, lead to Hubco having insufficient cash flow to metdetsevcepyensfalngde term debt-that is, $300 millon of $728 million-fulfill- meet__debt_service __payments__falling__due. ing their primary role of supplementing the $326 million of insurance cover provided by export credit agencies. By three years) in place. In this context the PSEDF can be seen limiting the cover provided under the ECO and JEXIM as not only forward-looking but as genuinely innovative. A guarantees to sovereign contractual obligations (including second important feature of PSEDF I and, to a lesser extent, nationalization, foreign currency availability, and other polft- PSEDF II is the fixed interest rate that it offers the Hub ical risks), the guarantees create an environment in which Power Project. private sources of finance can perform the functions to Such long-term finance compensates for the fact that which they are best suited-namely, identifying, evaluat- the balance of debt finance (that is, the senior term debt) ing, managing, and accepting commercial risks. for the project is available at shorter maturities than com- Another important role of the ECO and JEXIM guar- mercial banks can provide in industrial countries. antee facilities is the provision of untied and entirely flex- Independent power producers in industrial countries can ible finance. One of the common characteristics of finance usually secure construction finance from commercial banks insured by multilateral and bilateral agencies and export with fifteen to eighteen or more years' maturity. For the credit agencies (though not the Commonwealth Hub project the maturity is twelve years. However, with the Development Corporation) is the conditions that apply to PSEDF facilities the project's weighted average overall its use. Project finance involves significant costs, such as maturity of debt finance, at sixteen years ($738 million, or interest during construction and project management, that 63 percent, at twelve years and $436 million, or 37 per- satisfy neither tied nor intemational competitive bidding cent, at twenty-three years), is comparable with those in procurement criteria and yet are just as much capital costs industrial countries. The subordination of the PSEDF facil- as equipment. The ratio of these "soft" costs to the "hard" ities, in ranking of security interest and by virtue of their costs of construction generally are in the range 1:4 to 1: 1, longer maturity, creates significant additional risk cover depending on the type of project and the length of the for the providers of senior term debt. construction period. For the Hub Power Project this ratio The availability of the PSEDF became the driving force is 1:1.6. behind the government's private sector initiative in power In addition to interest during construction, there are sig- generation. Publicity materials for this initiative were nificant soft costs associated with funding reserve accounts issued by the Ministry of Water and Power's Private Power to compensate for the less than 100 percent insurance of Cell in 1989, 1991, and later years. These materials loan principal for political risks under the SACE insured included descriptions of the PSEDF, since without the facility, for the lack of interest cover under the ECO and long-term finance provided by the PSEDF private pro- JEXIM guarantee facilities, and to provide a "markup" jects were not considered feasible in Pakistan by anyone reserve account for the rupee facility. 16 Financing Pakistan's Hub Power Project The problems of funding soft costs are all the greater in extent by onshore investors, in each case a mix of corpo- relation to standby finance, which, if used, is usually needed rations, institutions, and private investors. to meet the costs of a delay in completing construction. The government of Pakistan had always favored a com- Little if any additional procurement of equipment is usu- bination of offshore and onshore investors, and initial plans ally required in such circumstances, and rarely anything that had called for a more even split between the two. However, would lend itself to formal international competitive bid- the difficulty in underwriting such a large issue onshore ding procedures. As a result the ECO and JEXIM guaran- meant that, at least initially, most of the equity finance had tee facilities are crucial to the provision of adequate standby to be mobilized offshore. Ownership of the project is finance for the project. The ratio of standby to base finance expected to migrate slowly and progressively onshore over is typically 10-15 percent, with standby finance covering the life of the station-particularly given the interest shown budgetary contingencies and the construction contract let in the local equity market, where the initial Rs 300 million on a fixed-price turnkey basis-that is, with no contract share offer was four times oversubscribed. This experi- price reopeners (for the Hub project the ratio was 14 per- ence illustrates how difficult it can be to estimate local cent, equivalent to $221 million). This ratio should be suf- market appetite in such circumstances. The portion of shares ficient to sustain a delay in project completion of up to held onshore, about 5 percent, is still low, though it is widely twelve months caused by circumstances for which there is distributed. Hubco now has about 100,000 shareholders no alternative source of funding-such as insurance or liq- in Pakistan. uidated damages from the contractor. The ECO andJEXIM The project had fortunate timing in its launching of the guarantee facilities together provide 27 percent of the international and domestic equity offering (October 1994). standby finance available to the project. The level of interest in the project meant that the target The role of the ECO guarantee facility as a "sweeper" of 25 percent equity contribution to the finance plan was of costs that cannot otherwise be funded because of pro- attainable (in fact, 24 percent). This achievement had curement constraints also extends to its role as a multicur- the indirect benefit of creating a finance plan in which rency facility that sweeps up remaining mismatches between the equity contribution was still greater than 20 percent currencies of expenditure and currencies of funding. The even if all available standby finance were drawn. Of course, only other "sweeper" term facility in the finance plan was by itself fortunate timing is no more sufficient in equity provided by the Commonwealth Development Corporation. markets than in debt markets, and much credit should be Equity and revenues generated during commissioning are given to the thorough and effective way the project was also entirely flexible in their application. presented to the market in financial centers around the The term base finance in the project's finance plan is 35 world. percent flexible and 65 percent inflexible. If equity and rev- The flotation of Hubco on the Karachi stock exchange enues during commissioning are included, these shares are and the global depository receipt issue listed on the more evenly split. Luxembourg stock exchange (box 2), which formed the basis of the successful international offering, were firsts Equity finance for a private power station still under construction. The success of these offerings was greatly assisted by the highly The share of equity in the finance plan remained constant successful senior debt underwriting and syndication that at 20 to 25 percent throughout the development of the Hub preceded them. For international investors that subscribed Power Project. The government, the sponsors, and the to the global depository receipt issue, the central role in financiers preferred that the share of equity remain at the the project of National Power, a major international power higher end of this range, if possible. The sponsor equity utility, was a more important investment consideration than contribution of $149 million represents about 40 percent the World Bank's presence. The shares and global depos- of the project's equity capital. The balance was provided itory receipts have performed well since flotation; shares mainly by other offshore investors and to a much lesser in Hubco have been one of the most actively traded on Finance Plan 17 have implications for World Bank participation in future Box 2 Role of global depository receipts in Hub projects. Power Project financing The shares in Hubco are rupee denominated (10 rupees per International competitive bidding. The first issue is the share par value) and are listed on the Karachi stock exchange. balance that has to be struck between the philosophy of For many offshore investors these features create two issues that fixed price tumkey contracting, which is ubiquitous in pri- may affect their investment decision: investment valuation in a currency not freely convertible or hedgable into dollars and set- vancinfratc reaprec thatorce e l roecose tlement arrangements in an unfamiliar stock exchange environ- financing (for example, build-own-operate projects), and ment. Global depository receipts (GDRs) were created specifically procurement under intemational competitive bidding. In to address these issues: first, they are dollar-denominated: and fixed price turnkey contracting the contractor absorbs cost second, they are listed on the Luxembourg stock exchange. overruns or benefits from savings if the costs of procure- AGDRrepresentstwenty-five shares in Hubco and gives the ment deviate from the original estimate behind the con- holder the right to receive these shares, if desired. The shares . . . that are represented by the GDRs are registered in the name tractprice pUner i ertional com ti bidd ent of a depository (the Bank of New York, New York) and are held is usually transparency between the costs of procurement by a custodian (ANZ Grindlays Bank, Karachi). GDRs are traded and the price charged to the client. through an intemational book-entry settlement system to which A project sponsor will select a turnkey contractor based GDR market makers in London, New York, and other major on many criteria, which may include a willingness to share financial centers have access, through systems such as Euroclear project development costs. Deciding factors may not, how- The total number of Hubco GDRs available may change over time, within certain limits, in response to market demand for ever, cud a reqie nttat indiiduahpiees ofae GDRs relative to the underlying shares. The gradual onshore ment procured by the contractor be the cheapest available migration of Hubco share ownership should mean thatthe num- (in capital cost terms). What matters to the sponsor is the ber of GDRs will eventually decline. More than 13 million Hubco ability to offer the client (host government) the most com- GDRs with atotal subscription price of around $145 million were petitive tariff from the power station (or toll bridge, water sold in October 1994. These GDRs represent about 30 per- treatment works, or other project). Many factors enter cent of the equity in Hubco. into this equation, not least of which are the terms of finance Dividends paid by Hubco on the underlying shares are con- verted into dollars bythe depository, priorto distribution to GDR or even the availability of sufficient finance that the con- holders, through foreign exchange operations in the Pakistan tractor may be able mobilize for the project. domestic market (for which there are certain govemment assur- In principle, the World Bank's rules may be applied to ances in the implementation agreement). GDRowners are enti- a project's output costs, such as a power tariff-as in the tled to instruct the depository on the exercise of the voting case of the Rockfort power project injamaica-which opens rights attached to the underlying shares. The DR isue as cordiatedby Dutsce Mogan the door for all the project's capital costs (hard and soft The GDR issue was coordinated by Deutsche Morgan cs,hoerpouedtobomelglefr Bakfud Grenfell and was fully underwritten by Deutsche Bank: such an costs, however procured) to become eligible for Bank fund- underwriting was itself an important precedent. ing. However, such an approach requires the project to be largely defined technically, commercially, and financially prior to tender, which becomes more difficult as the pro- the Karachi stock exchange, and Hubco has enjoyed much ject's size increases. In these circumstances the tenders for favorable comment from international stock brokering the output cost are heavily conditional on sources of finance firms. other than the Bank being available on acceptable terms. In the absence of this alternative approach, the Bank's World Bank perspectives participation in a project through a vehicle such as the PSEDF will be limited by the extent of procurement that Two key issues-international competitive bidding and satisfies its rules. In the Hub Power Project about 40 per- finance plan fragmentation-arose in the development and cent of the turnkey construction contract cost satisfied the implementation of the Hub Power Project finance plan that Bank's procurement rules. The procurement of goods and 18 Financing Pakistan's Hub Power Project services under international competitive bidding can delay ther pursue these initiatives. This form of collaboration is a project timetable unless the relationship between a pro- greatly facilitated by the creation of specialized project ject's procurement and funding is defined at the outset, finance groups within existing organizations, a development which is seldom possible except on smaller projects with a that has been significantly stimulated for some cofinanciers less fragmented finance plan. by their experience with the Hub project. One positive aspect of finance plan fragmentation, at Finance planfragmentation. The second issue is the frag- least from the perspective of cofinanciers, is the rationale mented nature of the finance plan. A pattern is now emerg- of security in numbers. This rationale carries most weight ing in developing markets of multiple cofinancier if the project in question is novel. Consequently, its impor- participations in these kinds of projects. For example, the tance to cofinanciers is expected to diminish as more large finance plans for the Paiton I power project in Indonesia, and complex limited recourse projects are financed. The the Sual power project in the Philippines, and the Izmit Su conversely negative point about fragmentation is a phe- water project in Turkey each included at least four foreign nomenon that occurs during negotiation, whereby the con- currency senior loan facilities. sensus always reflects the position of the least flexible There are two main reasons for the fragmented nature participant (that is, "the lowest common denominator"), of these finance plans. First, procurement may be divided which generally acts to the disadvantage of the borrower up by the turnkey or the engineering, procurement, and and host government. construction contractor into a series of discrete national In response to the potential difficulties created by finance packages. Second, there is generally a tight limitation on plan fragmentation, in 1993 the arrangers developed a the available political risk cover for a single country or pro- project coordination agreement that is, in effect, a master ject from any individual export credit agency or other cofi- loan agreement and intercreditor agreement combined; the nancier. The latter may, of course, also be responsible for project coordination agreement was used for all senior and the contractor opting for discrete national packages of subordinated facilities on the Hub Power Project. This form procurement. If the Bank is involved, there is the added of documentation has subsequently been adopted on an objective of cofinancier mobilization ("catalysis") and the increasing number of project finance transactions around Bank's role as lender or guarantor of last resort, both of the world. which encourage fragmentation. Given that the number of export credit agencies par- Banking Strategy ticipating in the Hub Power Project is not abnormally high compared with other developing market independent power Syndication producers, the structure of the PSEDF can be identified as the most significant source of fragmentation. The reso- An international syndicate of forty-three commercial banks lution of these and other issues has required the Hub pro- provided $686 million of limited recourse senior term debt ject to develop relatively extensive financial "technology" facilities for the Hub Power Project, allocated between the (particularly in terms of documentation)-certainly higher ECO and JEXIM guarantee facilities and the three export than was expected when the project began. credit agency-insured facilities (table 4). (Syndicate pro- It is becoming increasingly apparent to all cofinanciers files are provided in appendix 1). involved in large private infrastructure projects (including Syndication of these five facilities was carried out simul- the Hub Power Project) that coordination is essential if taneously by a five-bank arranger group comprising Bank finance plan fragmentation is not to impose a heavy bur- of Tokyo-Mitsubishi, Citibank, Credit Lyonnais, NatWest, den of delay in reaching financial close. Thus several cofi- and Sakura Bank, supported by Mediocredito Centrale with nanciers have initiated bilateral collaboration agreements respect to the SACE insured facility. The arranger group was with fellow cofinanciers (particularly export credit agen- chosen to reflect the national combination of suppliers within cies), and there is a strong desire among cofinanciers to fur- the construction consortium and associated export credit Finance Plan 19 TABLE 4 Facilities funded by offshore commercial banks Amount Number of Average participation Facility (US$ million equivalent) banks (US$ million equivalent) Facility agene ECO guaranteed 240b 34 7.1 Bank of Tokyo-Mitsubishi JEXIM guaranteed 120' 19 6.3 Sakura Bank COFACE insured 45 7 6.4 Credit Lyonnais MITI insured 86 17 5.1 Sakura Bank SACE insured 195 18 10.8 Mediocredito Centrale Total 686 43d 16.0e a. Citibank is the intercreditor agent. b. Includes $40 million of standby debt. c. Includes $20 million of standby debt. d. Most banks participated in more than one 6cility. e. Average overall partcipation per bank. agency-insured facilities and to include banks with a proven in the first ECO-guaranteed limited recourse project financ- and leading reputation in large limited recourse projects. ing. The presence of the Bank and JEXIM and the novelty Syndication was launched in May 1994 and closed four of the guarantee program are also partly responsible for months later. During this period some eighty banks were the appearance in the commercial bank syndicate of sev- approached. In the end syndication was more than 100 eral banks that previously had not participated in financ- percent oversubscribed. The reasons for such a successful ings of private infrastructure. The decision by a group of syndication ranged from the participation of the World Bank Italian banks, led by Mediocredito Centrale, to participate and the novelty and scope of the ECO guarantee to the in the first such financing supported by SACE was also extended export credit insurance cover (table 5). responsible for other banks being introduced to the pri- From the World Bank's perspective, syndication was an vate infrastructure market. unqualified success. Syndication achieved the Bank's objec- Though ranked fifth among the reasons for success in tive of mobilizing the necessary foreign currency commer- syndication, the quality of market presentation that was cial bank debt for the project with an overall maturity of achieved on the project-in terms of information memo- twelve years. In so doing, it created an opportunity for thirty- randums and road shows-is widely acknowledged to have four commercial banks from eight countries to participate set new standards for quality and completeness in the field of project finance and to have transformed the project's TABLE 5 image in many eyes. The only negative aspect of such a suc- Reasons for success in syndication cessful syndication is that, in the desire to disappoint no one, (percent) the resulting syndicates are large and fragmented (average Indicative participation per facility is $7.1 million). Such fragmenta- Rank Reason weighting tion gives rise to increased administrative overheads for the I Presence of the World Bank and novelty 20 facility agents and greater difficulties in managing the vot- of the ECO guarantee 2 Scope ofthe ECO guarantee 20 ing procedures under the project coordination agreement. 3 Market timing I S 4 Financial and commercial structure and quality 15 Role of underwriting of sponsors 5 Arranger group and market presentation 10 6 Italian market underwriting ofthe SACE insured facility 10 The dilemma faced by the Hub Power Project of whether 7 Pricing 5 8 Extended export credit agency insurance cover S to underwrite the commercial bank facilities and then syn- (for commercial risks) after completion of construcon dicate or move straight to syndication is one faced by all Total 100 projects. Reasons that support underwriting prior to syn- Source: Survey of arrangers. dication are: 20 Financing Pakistan's Hub Power Project * The early reassurance given to the borrower and cofi- The most significant benefit of underwriting the com- nanciers that there is sufficient capacity in the market mercial bank facilities for the project was that it saved time to complete the financing. on the underwriting and launch of the international and * The possibility of delaying the tasks associated with domestic equity offers by creating a higher level of confi- syndication until after financial close, so that they do not dence in an early financial close. Most participants in the divert the arrangerbanks' and borrower's resources away project, particularly cofinanciers, in hindsight expressed a from the task of achieving financial close. preference for the senior debt to have been underwritten Reasons that favor moving straight to syndication are: much sooner, although they recognized that its value in * Keeping the arranger bank group small and therefore advance of PSEDF I and II being irrevocably committed more easily managed from a logistical point of view, and and the equity fully underwritten would have been largely less prone to the phenomenon of bowing to the lowest symbolic. common denominator during negotiation of term sheets and loan documentation. Commercial bank perspectives * The costs saved in not paying underwriting fees (par- ticularly if the project timetable is liable to slip). Risk tranfer. One of the central achievements of the Hub However, the sheer complexity of the Hub Power Project Power Project is the creation of a contractual framework favored a small arranger group. Still, even a five-bank group in which the commercial risks normally associated with an at times became unwieldy, and the strategy of keeping a independent power project are transferred to private lenders small group that did not underwrite was maintained almost and investors. Thus the project achieved one of the gov- to the end. The need to reassure the government, the spon- ernment's and the Bank's earliest and most fundamental sors, the Bank, and the cofinanciers that sufficient market objectives for the private sector initiative: that commercial capacity existed was satisfied in the summer of 1993 through bank finance and equity capital be mobilized without direct pre-marketing, on the basis of a short information memo- government guarantees of debt service or return. As a result randum. This market sounding showed a level of support the risk matrix used in the Hub Power Project's bank infor- for the transaction that accurately predicted the eventual mation memorandum was substantially the same as for an level of oversubscription in 1994. independent power project in an industrial country (except, In project finance the term underwriting does not always of course, with regard to measures included in the project carry the same meaning it may carry in a corporate finance structure to mitigate against political risks in Pakistan). or capital markets context. From a borrower's perspective, In some instances investors and limited recourse an underwriting is only of value if funds can be disbursed financiers of the project are taking greater commercial risks earlier than would have been the case had the facility been than would usually apply in an independent power project syndicated. The conditions precedent to disbursement, in an industrial country. For example, the fact that the pro- which are usually lengthy in project finance, are generally ject is located in a relatively remote region of a developing the same whether the facility is underwritten or syndi- country makes the commercial job of managing, resourc- cated. ing, building, supplying, and operating the power station In response to a request from the Bank, the Hub Power much greater. No cover is provided to lenders for these Project was underwritten conditionally on 14 September risks under the ECO guarantee. Conversely, by virtue of 1994 by a sixteen-bank group comprising the five arrangers, the ECO guarantee and the World Bank's credit rating, Fuji Bank, the Long Term Credit Bank of Japan, Dai-Ichi commercial bank lenders to the project enjoy an imputed Kangyo Bank, Deutsche Bank, Mediocredito Centrale, and AAA credit rating for the payment obligations for the power six other Italian banks. Syndication closed with the same off-taker (Water and Power Development Authority), which conditionality some two weeks later. These conditions prece- would be unusual (though not unknown) for an indepen- dent were not finally and fully satisfied until 23 January dent power project in a developed market. The same might 1995, at financial close. also be said of regulatory and environmental legislation risks Finance Plan 21 against which, paradoxically, lenders enjoy better protec- International commercial banks' appetite for project tion (through the implementation agreement between finance assets continues to be strong, so the possibility of Hubco and the government of Pakistan) than they might individual banks reaching their notional country or sector expect in a developed market because of the rudimentary limits should not be a constraint on the overall financing legislation and lack of sufficient case law in Pakistan in these of private infrastructure projects in Pakistan or other coun- areas. tries where such partial risk coverage is made available. Because it provides subordinated debt, the PSEDF Whether participating in ECO guarantee facilities would enhances substantially the loan cover ratios on the senior cause banks to reach their limits for projects in a particu- debt and, in so doing, mitigates the effects of any increased lar country more slowly than participating in an export credit commercial risks that arise from the project's location. More agency-insured facility or an IFC B loan, for example, has recent and smaller independent power projects in Pakistan not yet been fully tested. Overall, however, export credit have been successfully underwritten and syndicated in the agency insurance and other forms of political risk cover commercial bank market without a tranche of PSEDF- are likely to be exhausted for private infrastructure projects subordinated debt and so with lower cover ratios. This in a country before commercial bank market capacity for process of subsequent projects progressively pushing out the underlying commercial risks. the envelope of risk transfer to the private sector can also be seen in project agreements, induding the implementa- Strategy formulation. The formulation of a banking tion agreement, in the trend toward gradually reducing strategy-including the appointment of arrangers and deci- levels of political risk coverage being required by banks, sions on the timing of underwriting and syndication and and in the trend for export credit agencies to reduce the their methods of implementation-is, properly, the respon- extent of commercial risk coverage they provide after com- sibility of the private sponsors of the project. Nonetheless, pletion of construction. In the early 1990s Pakistan was these decisions must necessarily be made in consultation unable to access any uncovered finance from international with the respective cofinanciers, including the World Bank, banks other than short-term trade finance; since 1995 there that provide guarantee or insurance support for the com- have been several examples of international banks being mercial bank facilities. prepared to commit uncovered longer-term facilities for Since this was the first time that it had participated, as projects in Pakistan. guarantor, in a limited recourse project financing, the Bank wanted to be more closely involved in formulating and imple- Assetallocation. Banks participating in the project's ECO menting the banking strategy than would normally be the guarantee facility have, as expected, not booked their par- case. The few instances when the Hub project sponsors or, ticipation against Pakistan (in the sense of direct exposure particularly, the Bank acted unilaterally in approaching to Pakistan sovereign risk), but have booked it against the potential lenders and underwriters served only to highlight World Bank. However, the cover provided by the ECO the crucial importance of all those involved adhering to an guarantee does not entirely eliminate lenders' concerns agreed strategy. about the project's location (see the section on risk trans- fer, above). In practice, banks are recording a notional coun- Counter-indemnities. Despite the extensive security pack- try exposure against each loan that enjoys partial risk age that was created for lenders' and cofinanciers' pro- coverage (that is, political risk coverage only) whether the tection-particularly with respect to government obligations partial risk coverage is provided by an ECO guarantee or and the World Bank and JEXIM's subrogation rights as an export credit agency insurance policy. There are, for guarantors of the ECO and JEXIM guarantee facilities- example, some banks that, having participated in more than Bank policy obliged the government to enter into a counter- one independent power project in Pakistan, have ruled indemnity with the Bank. The principle of out participating in any more for the time being, with or nondiscrimination between cofinanciers meant that par- without the benefit of an ECO guarantee. allel counter-indemnities (for political risks) were also given 22 Financing Pakistan's Hub Power Project to JEXIM, the three export credit agencies, and the The diverse currencies of denomination of the PSEDF Commonwealth Development Corporation. Several cofi- (and some senior) facilities caused particular problems. nanciers would not have required a counter-indemnity had In an extreme case, one of the PSEDF facilities was sized the Bank not received one; and, in fact, have not gener- in dollars, drawn in yen, and applied to meet ECU expen- ally required them on subsequent projects supported by diture. The hedging of these exposures was an additional an ECO guarantee. cost burden for the project. International competitive bidding or tied procurement Structural Issues rules generally do not take sufficient account of associ- ated costs inherent in a project financing-for exam- The PSEDF ple, interest during construction-that need to be funded. The government of Pakistan, not the underlying cofi- Assessments of the PSEDF by the parties involved in the nancier contributors to the PSEDF, faces limited recourse Hub Power Project found that its best features were: exposure to the Hub Power Project. * Longer maturity finance than can be secured from com- * A PSEDF administrator that is state-owned (such as mercial sources. At twenty-three years, the PSEDF's the NDFC) may suffer from potential conflicts of inter- maturity is nearly twice that of the ECO guarantee and est when negotiating with a borrower, if the borrower is other commercial bank facilities (twelve years). relying on the state for measures to protect its assets or * The National Development Finance Corporation's revenue stream from financial loss. (NDFC) ability to onlend on a limited recourse basis. * The introduction of variable rate funding after comple- * The PSEDF's ability to fund up to 30 percent of pro- tion of construction of PSEDF H is an unwelcome dilu- ject costs. tion of the rationale for the PSEDF (namely, offering * Available funding, for the most part, at a fixed interest long-term fixed-rate funding not otherwise available to rate. private enterprises in developing markets). * Subordination creating additional risk cover for providers Whereas an increase in the size of the ECO or JEXLM of senior term debt. guarantee facility could have substituted for some of the * Institutional development (see the section on institu- PSEDF in the finance plan, the reverse was not the case tional development, below). because the ratio of inflexible finance (that is, the PSEDF * The PSEDF II standby facility, provided by the Bank and export credit agencies) to flexible finance (all other and JEXIM, is an untied and flexible form of finance. sources) was close to its practical limit, given the pattern The worst features were cited as: of procurement. From the point of view of overall funding * Facilities are complicated to establish because of the flexibility, the Hub Power Project is slightly heavy on the three-step lending route needed to advance loans from PSEDF and light on the ECO and JEXLM guarantee facil- the Bank (and other cofinanciers) first to the govern- ities. Because of the tradeoff between an increased PSEDF ment of Pakistan, then to the NDFC, and finally from contribution to the finance plan (extending the average loan the NDFC to the Hub project. Responsibility for man- maturity) and a decreased ECO guarantee facility (reduc- aging this process and associated documentation was ing the flexibility of disbursement of the available finance) diffused among many parties. the Hub Power Project is close to its maximum use of the * Administration of the eight facilities within the PSEDF PSEDF. is complicated and resource-intensive for lender, admin- An alternative deployment of the PSEDF facility and istrator (the NDFC), and borrower. Implementation of ECO (or JEXM) guarantee facility in a finance plan, and draw-downs was in some instances subject to delay and one that exploits their respective strengths, would be to time-consuming documentation. use the ECO guarantee facility (in parallel with export credit * Inflexible currency of denomination: loans often did not agency and other term debt) preferentially to fund con- match the currency of expenditure that was to be funded. struction expenditure and the PSEDF to refinance the pro- Finance Plan 23 ject after construction was completed. Such a finance plan Some eighteen months were spent negotiating and was developed forJamaica's Rockfort Private Power Project, amending the implementation agreement and a further six albeit without the deployment of an ECO guarantee. This months negotiating the ECO guarantee document on the approach is suited to much smaller projects than the Hub project. To the extent that adoption of a standardized start- Power Project where the constraint on availability of term ing point for these documents is possible, significant sav- construction finance is not so pressing. ings should be possible on the time required to implement The World Bank is also considering a reverse structure future projects. The experience of the Lal Pir power pro- in which the PSEDF finances construction expenditure and ject in Pakistan, which reached financial close within about an ECO guarantee commercial term facility (bank or bond four months of the Hub project, supports this observa- issue) takes out the PSEDF at completion of construction, tion. The Lal Pir project used the implementation agree- in a refinancing. Both structures highlight ways in which ment and JEXIM guarantee documents from the Hub the PSEDF has potential flexibility. project with only minor amendments. ECO andJEXIM guarantee facilities Export credit agency-insured facilities Assessments of the ECO and JEXIM guarantee facilities The basic export credit agency insurance policies used on by parties involved in the Hub Power Project found that the Hub Power Project have been in place for some time, their best features were: although thiswas the first time theywere used by COFACE * 100 percent coverage of loan principal. and SACE on a limited recourse infrastructure project. The * The World Bank and JEXIM's involvement as guaran- export credit agencies' cautious approach to participating tors. in a pioneering limited recourse project was evident in the * Absence of rigid conditions on disbursement relating to fact that, despite contractor mobilization in December 1992, procurement of goods and services. the agencies did not make firm commitments until the end * Availability in multicurrency tranches (not for the JEXIM of 1993 and early 1994. As with the banks, the export credit guarantee facility). agencies are unequivocal in their admission that the Bank's * Flexibility on first loan repayment date and interest peri- participation was instrumental in their decision to partici- ods (in contrast to export credit agency facilities). pate in the Hub project, given its size and complexity. * Guarantee fees paid annually. The export credit agencies' acceptance of a portion of The worst features were cited as: commercial risk after completion of construction is an impor- * Lack of interest cover (this cover will be available in sub- tant difference between these and the ECO and JEXIM sequent IBRD guarantees). guarantees. Otherwise, the guarantees seek to achieve the * Call period of up to thirteen months and seven days. same objective, political risk cover, although in slightly dif- * Poor readability of the guarantee document. ferent ways. The ECO and JEXIM guarantees list the indi- * Requirement for counter-indemnity. vidual circumstances under which the guarantee may be The absence of interest cover under the ECO andJEXIM called. By contrast, the export credit agency documenta- guarantees increased the funding requirement of the pro- tion, being by tradition more of an insurance policy, describes ject by roughly $20 million, reflecting the cash allocated to the insured risks. The somewhat vaguer wording of the the guarantee reserve account and consequential increases export credit agency insurance policy is said to favor the in net interest during construction. A further $15 million insured party (the lenders) where precedent can be invoked. was added to the funding requirement by virtue of an anal- The provision by COFACE and SACE of extended insur- ogous account, the rupee markup account, being required ance cover for commercial risks post-construction was cited by the rupee facility banks. Interest coverage has since been by the arrangers (see table 5) and Hub project participants added to the scope of the ECO guarantee program and as one of the reasons for the highly successful syndication will be available for future projects. of the commercial bank facilities. In hindsight, however, it 24 Financing Pakistan's Hub Power Project is likely that syndication would still have been oversubscribed To the extent that payment of export credit agency insur- (although to a lesser degree) had the export credit agency ance premiums is front-ended (rather than amortized over cover been restricted to political risks. The decision to request the term of the loan)-which is often the case and was on extended insurance cover from the export credit agencies the Hub Power Project-an unwelcome consequence is reflected the inevitable nervousness about banking market an increase in the project's funding requirement, further reaction that accompanies any pioneering and, particularly, magnified by the need to fund consequent interest during large financing. On subsequent projects export credit agen- construction. Guarantee fees on the ECO and JEXIM guar- cies have, on the whole, expected the private sector to bear antee facilities are paid annually in advance over the term progressively more and ideally all the commercial risks-a of the loan and do not give rise to an increased funding policy strongly supported by the Commonwealth requirement. Export credit agency insurance premiums Development Corporation, COFACE, and other Hub Power included in the funding requirement of the Hub project are Project cofinanciers. Some export credit agencies that did about $40 million; ECO and JEX[M guarantee fees paid not participate in the Hub project are more willing to accept during construction are, in aggregate, about $6 million. commercial risks in addition to political risks. Although the export credit agencies provide Hub pro- A second important difference between the ECO and ject lenders with partial cover for commercial risks after JEXIM guarantees and export credit agency cover is the completion of construction, their counter-indemnities with less than 100 percent cover on loan principal provided by the government of Pakistan are similar to the ECO and the agencies' insurance policies. There are at least three JEXIM guarantee facility counter-indemnities that cover ways the portion of loan principal that is not covered for political risks only. For these purposes, the export credit political risks can be addressed on a project. First, the agency indemnities are the same as the ECO and JEXIM bank that has the benefit of the export credit agency insur- guarantee facilities and Commonwealth Development ance policy can accept this exposure and make balance sheet Corporation indemnities. provisions accordingly. Second, the risk can be shifted to One consequence of the multiple and occasionally com- the supplier whose equipment is being financed by the peting objectives of the PSEDF, the ECO guarantee pro- export credit agency-insured facility. Third, the borrower gram, and the export credit agencies was that the sources can arrange supplementary security, typically cash collat- of bilateral funding for the project did not match particu- eral. In the case of the Hub Power Project, supplementary larly closely the procurement supply sources under the con- security was used for the SACE insured facility and at least struction contract (table 6). one of the other mechanisms was used for each of the As a direct result of their participation in the Hub pro- other export credit agency facilities. ject, COFACE and SACE have developed specialized pro- The need to cover a 10 percent uncovered portion of ject finance groups to act as a focus and resource base for the SACE insured facility increased the project's funding future transactions. The experience of the project and of requirement by about $20 million, reflecting the cash allo- other recent major private infrastructure project financ- cated to the relevant collateral account and consequential ings has also increased collaboration between export credit increases in net interest during construction. The higher agencies, including sharing specialist advice to minimize financing costs that this incurred were, in fact, largely off- costs. The effects of greater confidence through these efforts set by the subsidies provided by Mediocredito Centrale to will be increasingly apparent in coming years. the interest rate payable by the Hub project on the SACE insured facility Under the export credit agency insurance Political risk cover policies, interest is covered to the same extent as loan prin- cipal, so no further reserve accounts were required, as was Where lender aversion to host-country political risk is a the case with the guarantee reserve account for the ECO key issue underlying the finance plan, as in the case of the and JE=M guarantee facilities (and the rupee markup Hub Power Project, the main obstacle to achieving finan- account for the rupee facility). cial close is the limited market capacity for this risk and, as Finance Plan 25 TABLE 6 Bilateral funding and procurement (US$ million equivalent) Supplier-mobilized finance Procurement Export JEXIM Country supply credit agency PSEDF guaranteed Total France 194 45 35 - 80 Italy 382 195 45 - 240 Japan 377 86 256 120 462 Total 953a 326 336b 1200 782 a. Total turnkey contract price including charges for customs duties, ncome taxes, and various minor adjustments is $989 million. b. ncludes base and standby facilities. a result, the availability of sufficient finance. This simple Average loan maturities may also be extended within the fact lay behind the creation of the PSEDF and the appli- limit of a fixed overall maturity by adopting annuity or other cation of the ECO and JEXIM guarantees to the project. back-ended repayment profiles. In private infrastructure Two mitigating measures are available to projects to projects, which generally favor leveled tariff profiles, annu- reduce the required funding: phasing construction so that ities are best suited and are commonly available to projects early generation of revenues can partly fund later con- in developed markets. However, this would require guar- struction costs (in the extreme case, splitting the project antor and insurer (that is, the World Bank, JEXIM, and into two or more phases) and drawing equity finance ahead export credit agencies) as well as lender acceptance of of debt to reduce interest during construction. The Hub extended average maturities. Power Project used both measures. With regard to the draw- If export credit agencies are reluctant to extend cover ing of equity finance ahead of debt, a more usual arrange- beyond twelve years, this may open the way for the World ment is for equity and debt financing to be drawn down pro Bank (under its guarantee program) and other cofinanciers rata. to cover the later maturities beyond those covered by the A second constraint on the availability of sufficient polit- agencies or an IFC B loan. Such cover could be within the ical risk cover is the term of the cover. When the commer- context of a single commercial bank facility. However, this cial contracts (Hubco's implementation agreement with the arrangement would not substitute for an ECO guarantee government, power purchase agreement with the Water and facility that covered the period during which loan dis- Power Development Authority, and fuel supply agreement bursements are made, to the extent that the finance plan with Pakistan State Oil) underlying a project are of a much still required a flexible untied "sweeper" facility to com- longer term than the available finance-thirty years in the pensate for the rigidities of export credit agency and PSEDF case of the Hub project-the term of the political risk cover modes of finance. Such an arrangement might, through a dictates the term of finance unless some other limit is reached parallel cofinancing, also be used to encourage export credit in the banking market (such as the term over which banks agencies to extend maturities of cover for infrastructure can access funds). Maturities of fifteen to eighteen years projects. and beyond for independent power projects in developed Despite the work that went into structuring the imple- markets have been secured from commercial banks, and it mentation agreement and ECO and JEX1M guarantees, is likely that had the ECO andJEXIM guarantees and export some banks declined to participate in the syndication of the credit agency insurance policies had terms of that order, the Hub Power Project because they felt that it did not offer a banking market would have responded with longer than a political risk environment comparable to one enjoyed by an twelve-year term for the Hub project (although perhaps independent power producer in, for example, Europe or with less oversubscription). However, OECD consensus North America. Furthermore, a number of well-known pro- rules ultimately limit maturities for export credit agencies. ject finance banks declined to participate simply by virtue 26 Financing Pakistan's Hub Power Project of the project's location, notwithstanding the ECO guar- might be very difficult to achieve, however. Export credit antee, or because all arranging roles were already allo- agencies are known to be highly skeptical of the feasibility cated. Many top project finance banks did participate, of general harmonization, although some are prepared to however. explore it further through the Berne Union and other The nervous reaction by some banks may be partly channels. explained by the perception that there is a class of com- Whichever way it is attempted, the key issues to be mercial risks inherent in a project, such as the Hub pro- addressed in harmonizing documentation will be the need ject, simply by virtue of its location (see the section on for: bank perspectives, above). Moreover, these banks may * Export credit agencies to move to 100 percent cover of believe that such risks cannot be addressed by an ECO or principal and interest. JEXIM guarantee or export credit agency insurance policy * The Bank to move to cover interest as well as principal because they are not, in essence, political risks. The best for political risks (already implemented for subsequent mitigation for these risks is to have strong, experienced, projects). and well-regarded project sponsors and operators. * Export credit agencies to accept annual payment of guar- antee fees and insurance premiums. Harmonization of documentation * Export credit agencies to accept a greater element of the soft costs associated with procurement as eligible costs. The ECO and JEXIM guarantee facilities use an identical * Export credit agencies to adopt the Bank's flexible posi- form of guarantee on the Hub Power Project. Each export tion on issues such as first repayment date and interest credit agency uses its traditional form of insurance policy. periods. Thus four forms of documentation for political risk cover * Uniform policies regarding call mechanisms and tim- were deployed on the project. The first step in any move ing, partial callability, and accelerability. toward harmonization of documentation ought logically to If there are insurmountable difficulties in harmonizing be among the export credit agencies. The need for such documentation, an alternative approach would be for the coordination has become a perennial topic of discussion Bank to be the primary guarantor and for export credit at emerging market and project finance conferences, and agencies to participate on a risk-sharing basis to the extent there are signs that a consensus in favor of harmonization of their exporting contractor's requirements. A borrower is being achieved, although initially on bilateral bases in would then require only a single guarantee for all its com- which the agencies agree on common policies between mercial bank facilities that enjoy political risk cover, and it themselves. would be up to the Bank to enter into risk-sharing agree- Between the export credit agencies and the World Bank, ments with the relevant agencies. A common guarantee there is clearly a need for the Bank to try to increase accep- involving the Bank could not provide commercial risk cov- tance of its political risk guarantee. A suitable occasion for erage for borrowers, which would be seen as an advantage attempting this might be a commercial bank facility in which by most export credit agencies. an agency insures the early maturities and the Bank the later maturities. An analogous arrangement has been success- Guidelines for Future Projects fully implemented through ECO guarantees applied to financings for the Yangzhou Thermal Power Project in China Finance plan objectives and the Leyte-Luzon Geothermal Power Project in the Philippines. The goal of any finance plan for a private infrastructure pro- A more ambitious route would be to attempt general ject is to achieve the lowest tariff to the customer. There harmonization of export credit agency and Bank political are two main routes by which this objective may be achieved: riskguarantees.Theprizeisconsiderableintermsofreduced maximizing gearing and ensuring access to the longest cost and time in arranging cofinancings. Such an outcome possible maturity finance. Market or policy constraints ulti- Finance Plan 27 mately limit how far a project may go along either route. It given the project's size, location and novelty, and timing con- will generally also be the objective of the host government straints. The large number of potential sources of commer- to maximize risk transfer to the private sector. cial term debt finance indicates what may be possible in this area, although political risk cover is generally the limit- Gearing. Since debt is cheaper than equity, maximizing ing factor on the overall availability of finance. gearing will lower the overall cost of funding for a project. A project must be highly developed contractually and Independent power projects are typically financed with gear- particularly from the point of view of risk allocation before ings ranging from 75:25 to as high as 95:5, depending on it can be tendered meaningfully among competing sources the project's contractual structure and the nature of risks of commercial term debt finance. The tradeoff between the that fall inside the project "fence." Globally, the average is increased cost of arranging an underwriting, the timeframe about 80:20. over which the underwriting is expected to remain valid, The Hub Power Project lies at the lower end of the range and the requirement for evidence that sufficient finance for two reasons. First, minimum levels of equity of 20 to will be available, will always be specific to a project's cir- 25 percent were specified by the government and the Bank. cumstances. The value of an underwriting to a borrower is Second, the availability of sufficient debt finance, or at least directly linked to the extent that the term sheet and agree- political risk insurance cover, proved more of a constraint ments underlying the project (all of which support the under- on the finance plan than the availability of equity. The expe- writing) are developed in detail. rience of a greater shortage of debt than equity is becom- ing increasingly common among project sponsors in Risk transfer The envelope of risk that lies with the pri- developing markets and is not unique to the Hub project. vate sector in infrastructure projects may be expanded in This underscores the crucial importance of the ECO guar- two ways. First, as a result of increased experience and the antee program in assisting private sector projects. market confidence that flows from it, the private sector will be more willing to accept risk or to price it more finely Loan maturity. Through a combination of senior debt Second, the process of evolutionary change may be accel- and the PSEDF, the Hub project achieved a weighted aver- erated by use of competition to "test the envelope." The age overall maturity (sixteen years) comparable to that of recent emergence in the international commercial bank independent power projects in developed markets. As far market of a limited willingness to lend to private infra- as average loan maturity is concerned, the project is below structure projects in Pakistan without political risk cover norms for developed markets because of its straight-line from bilateral or multilateral sources (whether from a B repayment profiles. The principal constraint on stretching loan, guarantee, or insurance policy) is an indication of how maturities, overall and average-for example, by adopting quickly the market envelope can move. annuity repayment profiles-is reluctance on the part of political risk guarantors and insurers to accept the longer Avoidance offragmentation. A secondary objective is avoid- exposure. Export credit agencies and the World Bank are ance of finance plan fragmentation, with the predictable largely responsible for deciding whether developing mar- qualification that for larger projects there may be few alter- kets should be able to access annuity and other longer natives to the assembly of a multi-cofinancier package. Even average maturity debt repayment profiles available to sim- splitting a project into several phases may not reduce frag- ilar projects in developed markets. mentation and complexity, but should be considered. The World Bank may be responsible for fragmentation Competition. A third general way to minimize costs in pro- because of its desire to limit its exposure, demonstrate addi- ject finance is the same as elsewhere-namely, to procure tionality, and mobilize multi-cofinancier-supported facili- under competitive conditions wherever possible. However, ties. The implications of fragmentation lie more in the impact the extent to which this principle maybe applied wil inevitably on development costs and the time taken to reach finan- depend on the degree of potential oversupply of finance, cial close than in the long-term cost of finance for the project. 28 Financing Pakistan's Hub Power Project The market performs best under competition, but mean- be tested further in subsequent projects, particularly in rela- ingful competition between equipment suppliers that takes tion to the average and overall maturity of loans. account of the true cost of their respective export credit agency facilities (that is, their impact on the eventual tar- The PSEDF iff) may not be possible at the time procurement decisions are made. Moreover, project sponsors may select suppliers The PSEDF's main role is as a source of long-term, low- based on criteria that do not maximize associated export cost, and, ideally, fixed-rate funding; and, to the extent credit agency cover. that it is available to a project, its use should be maxi- The problem is further compounded when export credit mized. In the form deployed on the Hub Power Project, it agency cover for a particular country is limited in any case, introduces a number of restrictions to the finance plan or when a project wishes to access a significant portion of that do not sit comfortably with either the philosophy of the cover available from a particular agency. The duty of fixed price turnkey contracting or build-own-operate pro- export credit agencies will remain to their exporters, and jects. However, these problems can be (and in the Hub pro- theymayreducetheircommitmenttoasingleprojectrather ject were) overcome, albeit with some difficulty and than see it sterilized (that is, committed but undrawn) by consequential delay. It is strongly advised that the effects project novelty or complexity that causes a project to take of these restrictions be anticipated before the procure- years to bring to financial close. ment plan or the finance plan for a project is finalized. The main restrictions relate to procurement. There are ECO guarantee program two ways that these restrictions can be overcome. First, it is important that the Bank be able to refinance eligible pro- The main roles of an ECO guarantee in the finance plan curement that has already been funded (perhaps under are to: mobilization finance), thus avoiding any loss of available * Supplement available political risk cover in terms of finance for the project through delays in reaching finan- amount and maturity. cial close. Second, project output costs (rather than input * Provide a flexible form of finance that compensates for costs), such as the electricity tariff charged by the power the restrictions that apply to the export credit station, may be subject to international competitive bid- agency-supported and direct cofinancier (for example, ding (rather than the underlying procurement), which opens the PSEDF) forms of finance in terms of funding soft the way for Bank financing of all costs (soft and hard) costs (such as interest during construction), providing during construction or for refinancing after completion of standby finance, and currency of denomination. construction (as in the case of Jamaica's Rockfort Power In fulfilling the first of these roles, the Bank has a diffi- Project). cult balancing act to perform between encouraging export The lack of use of the PSEDF in most of the energy credit agencies and other cofinanciers to participate in the projects that have reached financial close in Pakistan sub- financing and, as always, avoiding crowding them out. Apart sequent to the Hub Power Project is in large part due to from striving for the best balance between ECO guaran- the procurement restrictions that apply to its availability. tee and export credit agency participation in a project, there The optimal mix of the PSEDF and the ECO guarantee are clear benefits to devising structures in which these facil- facility is one that achieves the lowest output tariff from ities complement one another (for example, in terms of the project (that is, the lowest cost of finance) while retain- maturities, risk coverage, and documentation simplifica- ing sufficient flexibility within the matrix of sources and tion). The objective should be to achieve terms for com- applications of finance such that the overall project finance mercial bank finance that more or less match those available plan is workable, including in relation to standby finance. from the same sources in developed markets for similar pri- Deployment of these modes of finance in alternative and vate infrastructure projects. This market envelope should equally complementary structures is also possible. Finance Plan 29 Project Development T he sponsors made their first significant investment The implementation of a general private power initia- in developing the Hub Power Project in April 1988, tive by the government also required that, in parallel with when they initiated a detailed feasibility study. The the project's development, institutions be established in origins of the project, however, can be traced back as far as Pakistan that would have roles reaching far beyond those 1985, to the announcement by the government of Pakistan of simply supporting implementation of the Hub project. of an initiative to encourage private participation in power These induded private power cells within the Ministry of generation. Construction of the station began in December Water and Power, the Water and Power Development 1992 on the basis of mobilization finance. The project's Authority, and the National Development Finance finance package became irrevocably committed in September Corporation (or NDFC; see the section on institutional 1994, by which time the project had suffered two events causing significant slippage in its timetable: the Gulf war TABLE7 (1990-91) and a Pakistan court ruling on the applicability Key events and activities in project development of Shariah Law to the payment of interest (1991-92). Year Extent Working from these key dates and allowing for the time 1987 Sponsors submit proposal Initial site is selected lost through these and other intervening events, the period 1988 Govemment issues letter of intent to project sponsors of active development of the project is estimated at five Detailed feasibility study is completed years. In elapsed time, however-from first involvement Bank approves the PSEDF of the sponsors inl early 1987 until financial close in January Ministry of Water and Power's Private Power Cell is established 1995-it is nearer eight years. The key events in the devel- 1989 I pmplementaron agreement between the govemment and Hubco and power purchase agreement between WAPDA and Hubco opment of the project are summarized in table 7. Appendix are negotiated and initaled 2 contains a more detailed chronology National Development Finance Corporabon's Private Energy The Hub Power Project's history is not simply about Dision is established the development of aroet1990 Fuel supply agreement between Pakistan State Oil and Hubco is the development of a project, but about giving life to an negooiated entire policy initiative by the government of Pakistan and Re-tender of turbine island is effected the World Bank. In most other countries that have attempted 1991 Construction contract is signed such a fundamental reform-inviting large-scale private Arranger banks are mandated investment in what had hitherto been an exclusively pub- ECO program is approved by the Bank lic industry-the reform had been preceded by the enact- 1992 Mobilization finance is arranged Construction is started ment of enabling legislation. In the case of the Hub project 1993 Commercial bank and ECO guarantee term sheets are signed the implementation agreement assumed the role of this leg- Commercial bank due diligence is completed islation, so the project sponsors and the Bank became 1994 Commercial bank debt is underwritten and syndicated inextricably involved in the complex discussions that such Equity is underwritten and placed structural reform involves. The implementation agreement Bank and JEXIM approve guarantee faciliies took more than eighteen months to develop. 1995 Rnancial close 31 development, below). Given that the fuel supplier (Pakistan cause an even greater delay than this rule would suggest, State Oil), power purchaser (Water and Power Development particularly project novelty and force majeure. Such was Authority), and administrator of the PSEDF (the NDFC) the case with the Hub Power Project. are all public entities, the project was and is essentially a Independent power projects in developed markets may partnership between the public and private sectors. take one to two years to bring to financial close; in devel- The speed with which subsequent independent power oping markets a pattern is emerging of these projects requir- projects have reached financial dose in Pakistan owes much ing at least two to three years, although the distribution of to their use of documentation and institutions created for development periods is skewed in favor of longer than aver- the Hub project and the knowledge and experience gained age periods. In taking eight years in elapsed time to reach by those who worked on it. financial close, the Hub Power Project is at the upper end of lengthy development periods-but it is not unique. If Market Experience only the active period of project development on the Hub project is considered (that is, five years), the project lies in Despite the considerable thought and effort that may go the third quartile. into their preparation, projects seldom if ever progress in Recent examples of independent power projects taking accordance with their original (or even revised) timeta- longer than average to reach financial close can be found bles. A typical project development office will have many inAustralia, Indonesia, and the United Kingdom (over four out-of-date timetables filed away, induding some that had years in each case), and in Turkey. In Turkey the elapsed useful lives of perhaps only a few weeks. There are several time for project development for the Birecik project was reasons project development timetables depart from their similar to that of the Hub project. Toll roads appear to original schedule: have the greatest potential for marathon development peri- * Insufficient allowance in the original timetable for the ods, with examples of projects taking five to ten years to lack of timely availability of independent parties, for reach financial close found in the Australia, the United whom the project will be just one of many responsibili- Kingdom, the United States, and a number of developing ties. countries. Projects are as vulnerable to timetable slippage * Underestimation of the committed time required, exdu- in industrial countries as in developing countries. sive of other responsibilities. The projects that achieve the shortest development peri- * Decisionmaking processes and timeframes of indepen- ods, from the point of view of private sponsors, are usually dent parties being poorly understood or defined, or both, those that have been well defined technically and com- at the time the timetable was constructed. mercially by the host government or authority before the * Inexperience of the parties involved with respect to the procedure for awarding a concession begins (whether award principles and practices of private infrastructure devel- is by a competitive or negotiated process). The particular opment, with the project often having to move at the vulnerability of independent power projects to slippage is speed of the slowest. due to the number of parties and contracts normally required * Project novelty and complexity, particularly if the pro- to bring them to fruition. ject requires new legislation, institutions, or forms of The Hub Power Project, for example, requires five core documentation. commercial contracts: the implementation agreement, power * Changes in project scope. purchase agreement, fuel supply agreement, turnkey con- * Force majeure. struction contract, and operations and maintenance agree- All projects suffer from these problems to some extent. ment. Independent power projects seldom require fewer For people working in project development, a useful rule and often require more than five core contracts. Toll bridges of thumb for determining their likely impact on an out- and roads, by contrast, usually require only three core con- turn project development timetable is often to double the tracts (a concession agreement, turnkey construction con- original time estimate. In fact, some of these factors can tract, and operations and maintenance agreement); their 32 Financing Pakistan's Hub Power Project tendency to delay often owes more to choice of routing pared habitually on a "fast track" basis, particularly as the and issues of rights of way. The Hub project ultimately had project's development costs begin to mount. The basic flaw more than 200 project agreements and documents. Most, in all fast-track timetables is that they fail to take into account however, have only minor commercial significance, and a realistic expectation of delays caused by the many exter- are due to the limited recourse nature of the financing, nal influences referred to above. A minimal provision of which requires all aspects of a project to be defined with float means that out-turn events are more likely to diverge legal clarity prior to disbursement of funds. Independent from a timetable, with the consequent risk of demotivat- power projects carried out in industrial countries and of a ing all concemed unless a degree of project development size comparable to the Hub project would have fewer (say, pragmatism, if not resilience, is shown by the project's 100 to 150) final agreements and documents. The addi- sponsors. tional documentation required for the Hub project derives Despite the limited value of project development timeta- mainly from its more complicated finance plan. A list of bles as predictors of when financial close will occur, their the more than forty principal project agreements and doc- preparation and revision remains an important discipline uments supporting the project appears in appendix 3. that helps define component project development activi- One important feature of developing private, as opposed ties and identify potential critical paths, milestones, and to public, infrastructure projects is the largely sequential decision points. It also helps sponsors determine the level nature of the key technical, commercial, and financial tasks of resources required and establish priorities. that have to be performed. For example, negotiation of the core commercial contracts supporting the project (notably, power purchase and fuel supply agreements and the turnkey construction contract) cannot occur until the Infrastructure assets that are financed and built in the public sec- project is defined technically, and finance cannot be tor and subsequently privatized, either individually or collectively, arranged until the project is defined commercially This avoid the problem of commercial arrangements and project- sequential pattem ofactivitiesisaeyreasondevebased (that is, limited recourse) finance arrangements deiaying sequential pattern of activities is a key reason develop- the start of construction. ment periods for private infrastructure projects are mea- Although this route may not be suitable or available for sured in years. larger projects and for hostgovernmentsthat have limited finan- An important lesson that can be drawn about project cial resources, there are obvious attractions in completing nego- development timetables is that project development does tiations of implementation agreements, power purchase not easily lend itself to classical project management tech- agreements, and the like and in arranging project-based finance in parallel with project construction. In these circumstances the nques, which assume that the project management team technicail definition ofthe project, tendering of procurement, and has a level of control over events that is normally not avail- mobilization of public sector finance lie on the critical path to able. This accounts for an aphorism of project development: the start of construction. "it will take as long as it takes." The disadvantages of this route lie in the poor track record For people and institutions unfamiliar with private sec- of public sector construction projects relative to their private sec- torproje,the experience of having only par- tor alternatives in terms of capital costs and the time to com- tiar pontroject develothe pm jent, timetableisunnervingand plete, and the loss of linkage between the responsibility to build and the responsibility to operate the facility. The use of govern- frustrating. It can often lead to sponsors who experience it ment-secured mobilization finance to hasten the start of con- for the first time losing patience and confidence and pulling struction, as in the case of the Hub Power Project, is really a out of projects when they find that financial close is fore- special, hybrid approach based partly on project finance and cast to be, say, nine months away-year after year. (See partly on privatization. However, a key issue underlying such a box 3 for a brief discussion of public sector development project may still be the availability of sufficient public financial projects,that are later privatized.) resources, even if the requirement for mobilization finance is modest and represents only a small portion of the total cost of Paradoxically, this frustration with timetable slippage completing the project. leads to revised project development timetables being pre- Project Development 33 Timetable Analysis meet to resolve outstanding project issues took months of planning, and such meetings were inevitably subject to A timetable prepared for the Hub Power Project in 1987 rescheduling, often on short notice. showed about eighteen months to financial close. If The principal events of force majeure that affected the another project of similar size and complexity were to be project were the Gulf war (1990-91) and a Pakistan court developed in Pakistan today, in the wake of the Hub's ruling on the applicability of Shariah Law to the payment success, it could probably achieve this speed of imple- of interest (1991-92). Some loss of continuity was also mentation. However, the significant delays that the Hub caused by the need to reconstitute the construction con- project experienced relative to this idealized timetable sortium in 1990 after two members resigned; and by peri- are attributable entirely to the factors described in the odic changes in government in Pakistan, of which there previous section-in particular, project novelty, com- were several between 1987 and 1995. plexity, and force majeure. To some extent, each component activity in the Hub The project's novelty is well illustrated by the more than Power Project found itself at one time or another on the sixteen "firsts" that the project achieved (see the executive critical path for the project's development. Nonetheless, summary) and by the need to develop core contractual there are undoubtedly some activities that, in retrospect, agreements from first principles. took too long to complete-even allowing for the factors With a complex finance plan came many parties and referred to above (table 8). attendant logistical difficulties. A good example of these Finally, when considering the Hub Power Project were the problems the project faced in scheduling cofi- timetable account must be taken of a minor though impor- nancier meetings, which required the attendance of no fewer tant factor-namely, its multinational nature. Project oper- than forty representatives of governments, their agencies, ations were conducted in five regions: contractors, sponsors, banks, and advisers. Scheduling sev- * In Pakistan, where project business was carried out in eral consecutive working days during which all parties could Islamabad (location of the headquarters for the TABLE 8 Principal activities subject to delay Activity Comment Implementaton agreement Novelty of the document for all concemed (lack of precedents) Intermittent availability of advisers to the govemment of Pakistan Tumkey constructon contract price Consortium reconstituted following expiration of validity of original bid price Power purchase agreement Inexperience of host authorities, intermittent availability of advisers to host authorities, and resource overhead for all parties caused by transparency of tariff process Complex indexation methodology Private Sector Energy Development Fund (PSEDF) PSEDF I took five years (in elapsed time) to achieve its $375 million target because of diversity of contrib- utors and complexity of documentabon. PSEDF II, by contrast, took one year to reach $197 million Implementation of international competitive bidding procedures Foreign currency senior debt Five of the senior loan faclities required political risk guarantees (ECO, JEXIM, and the three export credit agency facilities). At one bme or another each found itself on the project's critical path since approval procedures for each cofinancier were different, unsynchronized, and generally untested in relation to a limited recourse infrastructure financing Periodic lapsing of indicatons of support from cofinanciers compounded the problems Cofinancier documentation Required three cofinanciers meeting over a nine-month period to agree on intercreditor principles Final export credit agency documentation was, in some cases, not available until financial close Conditions precedent Difficulty in procuring relevant documents in a satisfactory form from so many parties in Pakistan and abroad Local finance Took three years (in elapsed time) Land title acquisition Diversity of land ownership and unavailability of reliable land records 34 Financing Pakistan's Hub Power Project government of Pakistan), Lahore (the Water and Power the subsequent five years, when they had become available, Development Authority), Karachi (Pakistan State Oil was significant in terms of the efficiency of project and the financial center of the country), and Quetta (the development. government of Balochistan). As discussed, the various constraints and force majeure * In Washington, D.C., location of the World Bank and events prolonged the time taken to complete each of the the principal project management office. project's three main activity groups (technical, commercial, * In Tokyo, Japan, home of the construction consortium and financial). Moreover, the sequential pattern of devel- leader. opment activities in the project clearly extended the timetable * In London, United Kingdom, home of one of the main of principal project activities beyond expectations (table 9). sponsors of the project and the center from which most Beyond these factors, the publicity attendant on such a of the finance was arranged. novel project approach likely lengthened the public's per- * In Jeddah, Saudi Arabia, home of the other main pro- ception of the project's timeframe (box 4). ject sponsor. Furthermore, the construction consortium included com- Sponsor Group Formation panies from France and Italy. The project's finance plan includes contributions from all seven locations. Two classic dilemmas arise during project development: A complex and novel project that is also multiparty and what is the optimum size of a project, and what is the opti- geographically spread out creates special difficulties in terms mum number of project sponsors? There are no universal of coordinating activities-even with project offices that answers to these questions. Still, the issues that lie behind are effectively staffed twenty-four hours a day, as was often them are worth discussing because of their relevance to the case on the Hub project. the Hub Power Project and all private sector projects. The key operational issues that arise with any multina- tional endeavor are travel times, nonoverlapping time zones, nonsynchronous working weeks and public holidays, and Box 4 Effects of publicity language difficulties. If it can be said that project finance One of the reasons the Hub Power Project is widely perceived has been made possible on a wide scale by the advent of to have taken a long time to develop is that it attracted publicity PC-based spreadsheet programs, then it can be argued that and an associated level of expectation, particularly in Pakistan, from multinational projects such as the Hub Power Project have an early stage. Moreover, as one year merged into another, the timeframe lengthened in people's minds. By the time the project reached financial close, it was far from being a new story-and between 1987-89, the initial development period for the for many observers there was a strong sense of deja vu. Hub project-during which fax machines were generally The publicity surrounding the Hub project was inevitable not available or were, for a time, illegal in Pakistan-and given its pioneering nature, the World Bank's participation, and the project's size and location. However, much of the publicity TABLE 9 resulted from the numerous conference speeches given and Timetable of project activities press announcements issued by all those involved (the govern- ment, its agencies, the sponsors, and the Bank) as each mile- stone in project development was reached 1987 l Predevelopment work Still, press exposure prior to a project's financial close has 1988 l Feasibility study certain important benefits. One is that the risk offailure becomes 1 989-9 i 3 Commercial contract negotiaions and public, and the potential loss of reputation of those involved consortium formation Insitutional development (including the PSEDF) increases their determination to see the enterprise through to and realization of ECO guarantee program a successful conclusion. Another potential plus is that progres- 1992 i Arrangement of mobilization finance and start sive and positive publicity can stimulate lender and investor inter- of constuction est. Afterfinancial close, the Hub projectwas unanimously voted 1993-94 2 Arrangement of permanent finance "Deal of the Year" by leading specialist financial journals. Total 8 Project Development 35 Project size cially in circumstances where availability of finance is a key project constraint. Arguments that favor larger projects include: A host government may prefer a large sponsoring group to * Economies of scale relative to delivered unit output price. dilute project ownership, avoiding an important infra- * Increased commitment of resources and streamlined structure asset being owned by a single foreign party. decisionmaking by the host government, reflecting the Arguments that favor small project sponsor groups project's macroeconomic significance. inmcude: * Need for new capacity in one increment. * Simpler chain of command in project management. Arguments that favor smaller projects include: * Less potential for overlapping or conflicting interests. * Lower development costs (in absolute terms). * Time savedbynothavingto seekconsensus among share- * Fewer counterparties, reduced complexity, and poten- holders before decisions can be made. tially greater speed. Relatively few companies have the financial strength, resources, * Smaller funding requirement and correspondingly greater and experience to undertake major capital projects alone. flexibility. Oil companies are the best example of the kind of enterprise Many of the benefits of both large and small projects that can pursue such ventures. Within the private infrastruc- are, in principle, accessible to large projects comprising sev- ture market there are even fewer examples of companies of eral phases, each separately financed and constructed. sufficient size to do this, and most are relative newcomers Whether a large project is better split into two or more with limited experience in developing countries. As a result phases is often determined by economic and technical the pattern that is emerging internationally is of infrastruc- considerations. Still, the advantages of smaller and simpler ture projects being undertaken as joint venture-sponsored (phased) finance plans should be recognized. or limited recourse-financed enterprises or, commonly, both. In this context it is worth recapping how dramatically Project sponsor group the market for private infrastructure projects, and for inde- pendent power projects in particular, has changed since the Project sponsors may favor large groups for several inception of the Hub Power Project. When Xenel sought reasons: an electric utility partner for the Hub project in 1988 and * Assembly of complementary skills that may not be avail- issued an international invitation, there were only four able in a single sponsoring company (for example, the respondents. Of these, only National Power (then in the classic combination of a utility company, contractor, form of British Electricity International) had extensive expe- and local partner). rience working in developing countries. Were such an invi- * Reduced downside risk exposure for individual spon- tation to be issued today, between 50 and 100 electric utility sors through the sharing of development costs. companies would respond, although it would still be the * Increased probability of overall corporate success by case that few would have much operational experience out- spreadinglimited resources among several projects rather side their own domestic markets. than concentrating all on one. The development of the Hub project has spanned the * Opportunities for off-balance sheet financing for share- transformation of the private infrastructure market from a holders with equity interests below certain percentage rare curiosity outside a handful of industrial countries such thresholds (depending on jurisdiction). as the United States and the United Kingdom (in 1987), * Shared ownership among the various parties with a com- to mainstream policy for many countries both industrial and mercial interest in the project (for example, fuel sup- developing (in 1997). During this period trends toward pri- plier, contractors, off-takers, and so on). vatization and utility deregulation have created a market * Psychological security from collective corporate action. for "global" utility companies. * Greater ease in arranging finance as a result of an inter- The core Hub Power Project sponsor group comprised nationally diverse group of sponsors. This applies espe- five companies: Xenel, National Power, Mitsui, Ishikawajima- 36 Financing Pakistan's Hub Power Project Harima Heavy Industries, and K&M. The choice of so development costs for larger projects tend to be more in large a group reflected, to varying degrees, the influence on the 3 to 4 percent range than the 2 to 3 percent range. the sponsors' thinking of the above rationales for larger Exceptions to this rule arise where the concession for a pro- groups. Collectively, the project sponsors formed a project ject is tendered by the host government or authority on the development company (HRPG) that, byvirtue of beinginde- basis of a well-defined technical and commercial specifi- pendently resourced and managed, was able to mitigate cation, in which case lower figures (around the 1 to 2 per- much of the inertia and dilution of management resources cent range) would be expected. that comes from multiparty project sponsorship. Only in the Development costs are defined as costs that are borne latter stages of the project's development, following its pri- by the private sponsors of a project and that are sunk in vatization and increased commitment to overseas invest- trying to reach financial close (that is, they will be incurred ment, did it become possible for National Power to play a whether the project reaches financial close or not). They leading (rather than supporting) commercial role in the do not include the costs borne by other parties such as the project and its financing. host government or the Bank, nor do they encompass As the market for international independent power pro- pre-financial dose construction costs where these are funded jects matures and host governments, financiers, and pro- by separately secured mobilization finance. ject sponsors become more experienced, it will become The Hub Power Project development cost figure of 3.6 increasingly possible for smaller groups or even individual percent, while toward the upper end of the market range, companies to bear the costs and risks of developing pro- is not exceptional. Given the gestation period of the pro- jects as large and complex as the Hub Power Project. Still, ject and its high degree of novelty, it is surprising that the the issue of eventual ownership of the project will remain figure is not higher. What is inevitably eye-catching with all dependent on the many factors referred to above, such as large projects is the absolute sum that the sponsors must the desire of sponsors to keep the project's debt off their put at risk in order to create a project. balance sheets. Few corporations in the world can afford to take gam- The creation of a dedicated project development com- bles of this magnitude on their own, particularly when his- pany focused solely on the development of the Hub pro- tory shows that the probability of success in project ject, and with an identity distinct from those of the sponsors development can be low. The primary incentive for a pro- and contractors, was found by the cofinanciers and banks ject sponsor to shoulder front-end risk is the possibility of to be a practical arrangement that greatly assisted their com- creating a much larger investment opportunity at fnancial munications with the project. Despite the protracted devel- close-typically six to ten times the equity capital invest- opment phase, with its intermittent periods of low activity, ment in developing the project. A second incentive is, gen- the sponsors managed through HRPG to achieve a sur- erally, the possibility of earning a higher return on the. prisingly high degree of continuity of personnel as, indeed, development period investment if the project is successful. did the World Bank. This continuity was crucial to the pro- Project development costs can generally be broken down ject's success. into the standard categories, although allocation among these categories can vary significantly by project. The devel- Development Costs and Risks opment costs for the Hub Power Project are shown in table 10. The largest element of the project development Project development costs are usually 2 to 4 percent of a budget was legal costs, which is typical of a project-financed project's funding requirement. This rule appears to hold independent power project. Moreover, given that almost for most private infrastructure projects (power, transport, all 200 project agreements and documents had to be nego- water) and, surprisingly, for projects of all sizes. Thus, as tiated and drafted from scratch, this came as no surprise. far as development costs are concerned, there do not appear Hub project sponsors ended up paying for the services of to be any economies of scale in sponsoring larger projects. fourteen legal firms, most of which were acting for banks In fact, if anything, there are diseconomies of scale in that and cofinanciers rather than for Hubco. Project Development 37 The question of resource mobilization is central to the ers. The total personnel deployed in developing the Hub setting of a development budget. Resource mobilization is Power Project, based on estimates made by the relevant determined by a project's management in light of four participants, is shown in table 11. considerations: The total estimated personnel deployed in reaching finan- * The probability of the project proceeding successfully cial close (225 person-years) can be summarized in broad to financial close, as compared with other projects that categories as follows: the government and its agencies and the sponsors may be pursuing at the time. advisers (25 percent), sponsors and their advisers (50 per- * The minimum critical mass of expertise needed to ensure cent), the World Bank (10 percent), and other cofinanciers, that each project agreement is properly structured and leading banks (arrangers), and their advisers (15 percent). that issues that are interactive between different agree- These figures and those provided in table 11 exdude the ments are properly handled. personnel deployed by the contractor in bidding, negoti- * A pattern of workload that may vary dramatically and ating, and assisting with mobilizing export credit agency unpredictably throughout the development period. support for the project and all activities related to con- * The tradeoffs between the flexibility of having advisers struction. who can be mobilized and demobilized according to Several tasks cut across the above headings and are worth workload, the cost of such consultants, and the risks of highlighting separately because of their relevance to future having knowledge of critical agreements held outside projects. These were development of the implementation the sponsor management group. agreement (ten person-years), development of the power The compromise often struck between these consider- purchase agreement (fifteen person-years), and creation ations is to assemble a core management team comprising, of the PSEDF (twenty person-years). say, six to twelve experts dedicated or seconded from the An equally intensive though often underestimated activ- sponsor organizations and a supporting cast of legal, tech- ity is the building and running of spreadsheet financial mod- nical, and financial advisers. Good advisers may not be easy els to support the project development effort. This activity to retain, however, so individual advisers must be reserved consumed about twelve person-years of resources spread for a project even though they may be required to work only across the sponsors, financial advisers, arrangers, the NDFC, intermittently. This was the approach adopted by the Hub and the Water and Power Development Authority. The Hub Power Project. Power Project models are among the largest and most com- The costs and resources of the project's sponsors are plicated ever built for a project financing, and are capable only part of the picture. Each counterparty to an agreement TABLE II has to staff negotiations and allocate management time. In Personnel deployments fact, the management of advisers is as much an issue for (person-years) counterparties as it is for project sponsors, and it was not Source Length always possible for the government of Pakistan and its agen- Ministry of Water and Power (including Private Power Cell) 10 cies to rely on the timely availability and continuity of advis- Water and Power Development Authority 1 5 Pakistan State Oil 5 TABLE 10 National Development Finance Corporation I 0 Other govemment agenciesa 5 Breakdown of development costs World Bank 25 (percent) Arrangers 25 Cost category Share Cofinanciers 10 Project sponsors and project managers 75 Legal 35 Lawyers 25 Financial 25 Financial advisers 10 Technical IS Technical advisers 5 Project management is Other advisers 5 Cofinancier advisers and due diligence 10 Total 225 Total 1 00 Total_____________________________________________________________________ a. State Bank of Pakistan, Central Board of Revenue, Ministry of Finance, and so on. 38 Financing Pakistan's Hub Power Project of handling fifteen loan facilities and expenditures in seven An important financial risk that the government had to currencies. Most banks that participated in the project and address during the development period was of potential vari- that wanted to run the model had to buy new, high-pow- ations to the power tariff payable under the power pur- ered computers to do so. chase agreement. To cover this risk, the letter of intent and The sponsors were not alone in bearing development the subsequently expanded tariff agreement (1989) con- risks; the government also bore substantial risks. In grant- tained two fundamental protections for the government. ing the Hub Power Project an exclusivity period in which First, the permitted base-case rate of return on equity for to develop and finance the project, the government was the project was limited to an 18 percent annual dollar real relying on the sponsors to complete the assignment suc- internal rate of return. Second, that cost elements of the tar- cessfully and as rapidly as possible. The stakes were made iff could only increase in line with agreed indices and exchange higher by the government's decision to make the project rate movements, such that the tariff's real cost to the gov- its sole prototype independent power producer, so that all emient of Pakistan remained essentially unchanged. of the government's attention and that of its agencies and Furthermore, an open-book approach was agreed between of the World Bank could be focused (along with the spon- the government, the sponsors, and the Bank such that the sors' efforts) on making the project a success. The stakes entire cost structure of the project was transparent. became higher still because of the government's support This strategy could be sustained only to the extent that for mobilization finance. the project's sponsors were able to hold the cost base of Given the size of the project relative to the resources the project constant in real terms. Relative to the turnkey available within the government and the Bank, no other construction contract, this proved difficult over such an approach was viable. The decision to form a queue of pro- extended development period, with two consequences: jects behind the Hub project already appears to have been * The original consortium that had bid a price in 1988 vindicated by the speed with which some of these projects- was unable to maintain it beyond December 1989, at which collectively add up to an additional 1,400 megawatts which time financial dose was still forecast to be at least of private power capacity-followed the Hub project to nine months away. As a result two of the consortium financial close during 1995 and the first quarter of 1996. members (Toshiba and Kumagai Gumi) departed, and The frustrations felt by the government and others with their responsibilities had to be rebid. having to wait until 1995 to see so much effort come to * Having reconstituted the construction consortium in fruition may be partly offset by the thought that between 1990-91, even the revalidated price had only a limited April 1988 (the date of issue of the letter of intent to the validity period, as is normal in contracting. This validity Hub project) and March 1996 some 2,700 megawatts of period expired in December 1992, at which time finan- independentpowerproducercapacitywasbrought to finan- cial close was still forecast to be at least nine months cial close, representing an average of about 330 megawatts away. Rather than face an additional rebid, the decision (net) a year-an impressive achievement. was made by the government, the sponsors, and the Bank As the prototype project for Pakistan's independent to mobilize the contractor based on a limited down- power producer program, the Hub Power Project inevitably payment in December 1992. bore a large measure of one-time development costs and In hindsight, the expectation that so much front-end risks that subsequent projects have avoided and will con- development work on agreements, institutions, and modes tinue to be able to avoid. Although no mechanism exists of finance could be completed and financial close achieved whereby these one-time costs and risks can be shared among within the typical validity period of a turnkey price offer subsequent projects, it generally has not deterred sponsors was overly optimistic. In the end the only way so many par- from trying to achieve the distinction of being "the first," allel and interrelated activities could be brought to a suc- whether in Pakistan or any other country. This sponsor cessful conclusion was for mobilization finance to be used enthusiasm reflects the degree of competition among spon- to remove the issue of turnkey price validity (and therefore, sors in the market for private infrastructure projects. that of the power purchase agreement tariff) from the critical Project Development 39 path to financial close (although mobilization finance had chairmanship of these meetings and increased with the num- its own maturity deadline that maintained pressure on all ber of cofinanciers participating in the project. Moreover, parties to reach financial dose). The support for mobiliza- given that the Bank was already a major contributor to the tion finance represented a second important financial risk development of the energy sector in Pakistan, this role had that had to be addressed by the government (see the dis- an inherent rationale. cussion on mobilization finance, above). The Bank was also, quite naturally, cast in the role of The adoption of an open-book approach to the develop- confidante to the government of Pakistan. On occasion this ment of the project cost structure and tariff inevitably intro- role extended to being an "honest broker" to help resolve duced a significant overhead for the project development issues between the government and the project's sponsors effort, particularlyin terms of the resources required to ensure and banks. Both roles were invaluable given that the pro- that all parties involved were apprised of the project's detailed ject under development was pioneering and the whole sec- economics and finance plan at each step along the way to toral initiative was new for the host government. Without financial dose. The alternative and more common approach the Bank's contribution as unofficial leader of the cofi- of contracting on the basis of a single indusive tariff figure, nanciers and confidante of the government, development with no transparency offered to the purchaser, is preferred. of the project could have presented insurmountable prob- This, in fact, is the basis on which subsequent independent lems to the government and the sponsors. power projects in Pakistan have contracted. Such roles are, however, resource intensive and rapidly The more one considers the resources that had to be become full-time-which presents a problem given the other mobilized to complete the development of the Hub Power duties that managers in the relevant Bank regional depart- Project-the risks borne and costs incurred-the more ment may have. The figure for Bank resources deployed remarkable it seems that fatigue did not set in among the on the Hub project-twenty-five person-years-makes the government, the sponsors, the Bank, and the cofinanciers. point forcefully This is equivalent to five people working And herein lies the essential reason that the project reached full-time for the duration of the active development period. financial close: none of the key participants wavered in their Without additional resources that can be dedicated to a determination to see it through. The importance to the gov- project until it reaches financial close, it is possible that ernment of Pakistan and the sponsors of the World Bank's attempts to balance the management demands of a project commitment in this regard should not be underestimated. withwider sectoraldevelopmentwill compromise both. Thus All the project's participants should be commended for their the Bank's leadership role would best be formalized, prob- stamina and patience. ably through its appointment as chairman of a cofinanciers steering committee. Although the Bank may be reluctant to World Bank Perspectives assume leadership of the cofinancier and government liai- son aspects of a project, it has a unique capacity to shoul- Leadership role der these responsibilities. Moreover, the Bank's suitability for this role is universally endorsed by the government of It was natural at the Hub Power Project's cofinanciers meet- Pakistan, the cofinanciers, the banks, and the sponsors of ings that the Bank should assume the role of chairman. the Hub Power Project. In situations where the Bank is Cofinanciers are accustomed to the Bank showing leader- proactive, the development time of projects will generally ship in such circumstances, particularly where (as in the be shorter and, therefore, the flow of projects greater. case of the Hub project) the venture was a "first" for the country concerned. Indeed, for cofinancier meetings held Project ownership early in the development period, the invitations to attend had to be issued by the Bank. Though described as a private sector project, the Hub Power The pressure on the Bank to assume a coordinating Project is to a large extent a partnership between the pri- role on behalf of cofinanciers grew naturally from its vate and public sectors. After all, the principal input and 40 Financing Pakistan's Hub Power Project output contracts (the fuel supply and power purchase agree- financing, and implementation of an independent power ments) are with public entities whose performance is guar- project in a developing country. Finally, the project has anteed by the govemment through the implementation demonstrated that, complex as they are, such deals can be agreement. Moreover, all forms of debt finance for the done. project are either provided by international or Pakistani Furthermore, in light of the development experience of public agencies or partly guaranteed by such. A similar this and other independent power projects, it is possible to pattern can be expected for such "private" infrastructure identify a number of specific recommendations that should projects in many other countries with which the Bank has help reduce the risk of future projects encountering delays an active program. The Bank and others involved in such or failing: projects share a similar dilemma: "Whose project is this?" * Time spent by a host government prior to inviting pri- Host governments, project sponsors, and the Bankwill each vate sector proposals in defining the legislative envi- legitimately feel ownership over the enterprise, although ronment, clarifying what is required from the project joint ownership may reassure lenders and investors. technically and commercially, and framing its objectives within the context of what can be financed in the pri- Mobilization finance vate sector will be more than saved by the time finan- cial close is reached. If the power purchaser is a public One of the Bank's most important contributions to short- entity then, at a minimum, the host government's respon- ening the timetable to the start of construction was its sup- sibilities should include specifying available sites, plant port for and participation (through the PSEDF) in size, fuel type, mode of plant operation, and provision mobilization finance. In circumstances where institutional of grid interconnection facilities. The more such para- development forms an integral part of the project devel- meters are defined prior to invitation, the greater is the opment effort, as in the Hub Power Project, or where potential for meaningful competition between sponsor other factors give rise to unacceptable delays in the start of groups and the shorter should be subsequent project construction, mobilization finance is one of the few miti- implementation periods. gating measures available to the host government and the * Efforts should be made to postpone as many nonessen- Bank. One constraint, however, is that in such circumstances tial activities as possible until after financial close. The important commercial issues must be settled prior to deploy- logical extension of this point is to split the project into ment of mobilization finance unless the host government two or more phases, each with a separate finance plan is prepared to complete construction of the project in the and its own financial close. Any consequential reduc- public sector, if necessary. In the case of the Hub project, tions in the complexity of a given phase's finance plan all major commercial agreements were signed prior to mobi- should yield time savings in reaching the initial financial lization of the contractor. close, albeit at the probable cost of a temporary loss of economies of scale in output tariff. Guidelines for Future Projects * Debt syndication is the most obvious financing activity that can usually be deferred until after financial close. The successful financial close of the Hub Power Project will Generally, financial close should be sought on the basis help shorten the time needed to develop future indepen- of an unconditional underwriting against which the pro- dent power projects in which the Bank is involved for three ject may begin loan draw-downs. reasons. First, much of the commercial, financial, and legal * Sponsor resources should be matched to the nature experience embodied in the project documentation is of a and scale of the project. Moreover, these resources must general, not project-specific, nature. Second, a large group be focused on the project at hand by assembling a ded- of individuals spread across numerous institutions inside icated team whose primary responsibility is to bring the and outside Pakistan now have first-hand knowledge of the project to fruition. A dedicated team supported by strong principles and practices of project development, project project sponsors is best of all. Project Development 41 * The need to allocate dedicated personnel applies equally 100 percent of the loan principal and interest), and are to the Bank, which must be prepared to free up or retain expected to require the least fragmented finance plan. resources for this purpose. For planning purposes, the The chosen single currency of bid should be one in which Bank should expect to assign the equivalent of two peo- debt and equity funding for projects is commonly denom- ple more or less full-time to a project in which it pro- inated and in which interest rates are readily hedged. poses to participate, and the commitment should be for * The power purchase agreement tariff should ideally be the duration of the development period. denominated or determined in the same currency as * The fewer project participants the better, which applies the project funding, or in a currency that may be hedged as much to contract counterparties, contractors, and cofi- against the currency of funding. nanciers as to members of the sponsor group. * A suitable host government institutional framework * The temptation to believe in "fast track" timetables should be in place and sufficient administrative resources should be resisted. mobilized to expedite needed project agreements and * To achieve the shortest development timetable, a pro- approvals, prior to commencement of substantial pro- ject should, paradoxically, be configured to survive a long ject development work. haul to financial close, whether delays are caused by force * All contract counterparties should strive for continuity majeure or other factors. In particular, a project should in the advisers who provide specialized expertise, as determine the level of cross-party political support within and when required. This issue is particularly important the host country; test sponsors' commitment to con- if the project enters the long haul. tribute to project financing; strive to achieve continuity - Projects with fragmented finance plans would benefit in the key individuals responsible for developing the prin- from the creation of a formal cofinanciers steering com- cipal project agreements within the host government, mittee during their development. The committee would sponsors, and Bank, and their respective advisers; and be chaired by the Bank. prepare commercial contracts so that they can survive * Logically, smaller projects should have simpler finance unforeseen timetable slippages without requiring rene- plans, although this is not always the case. Where they gotiation. do and an order of magnitude reduction in complexity e Invitations issued to turnkey contractors should favor is achievable (for example, through having a single bids that are from smaller rather than multiparty con- PSEDF cofinancing facility or export credit agency), sig- sortiums, are denominated in a single currency, include nificant timetable savings should be realized. However, export credit agency cover in line with the requirements the number of agreements involved in such projects does of the commercial banks that will fund the respective not fall as rapidly with scale of financing as one would facilities (including, for example, political risk cover for hope. 42 Financing Pakistan's Hub Power Project Institutional Development T he institutional framework to support private par- account the feedback received from private investors and ticipation in Pakistan's energy sector evolved as a the international financial community. Refinement of the major component of the work carried out for the framework of incentives was also needed to make Pakistan development of the Hub Power Project. In 1988 the gov- internationally competitive in attracting financial resources erinent of Pakistan initiated a program to address macro- and to integrate with these measures the actions taken by economic constraints and liberalize the economy. The governments to deregulate the economy and increase program, among other actions, required rationalization of reliance on the private sector. A new policy for private energy public expenditures. Budgetary allocations for revenue- that integrated all these measures, amendments, and refine- earning public entities were reduced. Investment financ- ments was promulgated in March 1994. The policy incor- ing for these enterprises was, in the future, to be mobilized porates the original policies introduced in 1988 together through internally generated funds and borrowings from with the amendments made since. the domestic and international capital markets. This pro- The policies for the creation of an enabling environment gram had a profound impact on the energy sector, which for the private sector have four components: criteria for the historically had accounted for more than 33 percent of consideration of private sector projects by the govern- annual development allocations, since internally gener- ment, fiscal and financial treatment of private energy pro- ated funds and projected borrowings fell short of the invest- jects, incentives for investors and lenders, and a structure ments needed to meet the forecast demand. In response, for the agreements, provisions, and facilities needed to safe- the government outlined a strategy for rationalizing the con- guard the interests of the government, lenders, and investors. sumption of energy and increasing the role of the private sector in energy to fill the gap left by the reduction in pub- The PSEDF and the ECO Guarantee lic investments. Facility The government identified three key constraints to mobi- lizing private capital: the absence of a policy framework The PSEDF was established in 1988 to provide financing for the private sector-that is, incentives, fiscal treatment, support to long-gestation energy projects. The PSEDF repatriation of profits and capital, availability of foreign was to provide partial, long-term subordinated debt financ- exchange, pricing, and so on; the lack of long-term finance ing for energy investments. The NDFC manages the PSEDF for projects with relatively long gestation periods and eco- for the government of Pakistan. The PSEDF was structured nomic life; and the inadequacy of the institutional struc- to allow for its possible detachment from the NDFC when ture for the review, negotiation, and approval of private its staff are fully trained and when projects that it has already sector projects. financed provide sufficient revenue to cover the cost of its An initial framework of incentives to attract private invest- operations. As the development of projects (including the ment into the energy sector was put in place in 1988. During Hub Power Project) progressed, it became apparent that the negotiations for the Hub project the government rec- subordinated financing would not be adequate to mobi- ognized the need to fine-tune the framework to take into lize the balance of loan finance required, since lenders 43 generally were unwilling to provide sufficient long-term was assigned responsibility for negotiating and adminis- financing without suitable protection from the political risks tering power purchase agreements and integrating the oper- that they perceived to be inherent in the transactions. ations of private power plants within its system. Each of In 1990 the ECO guarantee program was considered these agencies set up new units specifically charged with by the Bank's Board of Directors. The program built on supporting private energy investments. The Ministry of the guarantee aspects of the Bank's B loan program and Water and Power set up the Private Power Cell, the NDFC aimed at using the Bank's partial guarantees to mobilize created the Private Energy Division, and the Water and private finance for public or public-private projects in the Power Development Authority established the Private Bank's lending program. Many private infrastructure pro- Power Organization. jects are being developed and financed on a project finance The Private Power Cell was created in 1988 to provide (that is, limited recourse) basis, whereby a private investor an institution that would evaluate and negotiate private mobilizes debt financing secured primarily or solely on the power projects on behalf of the government. In an effort revenues generated by the project and a legal charge over to streamline the institutional structure for private power, the project's assets. Project revenues, in turn, are often the revised energy policy of 1994 called for the Private Power based on long-term sales contracts. This concept underlies Cell to be designated as the Private Power and Infrastructure most build-own-operate (BOO) and build-own-operate- Board to function as a single-stop investment window and transfer (BOOT) projects. Based on the emerging trends assume full responsibility, on behalf of the government, in private infrastructure projects, in 1991 the Bank broad- for negotiating all private power investments. The board ened the concept to enable guarantees to support private was so designated in June 1994. commercial financing for private sector projects. The Hub Under PSEDF I the NDFC was designated as the admin- Power Project was the first application of the guarantee to istrator of the PSEDF and created the Private Energy a private infrastructure project. Division specifically for this purpose. The division is struc- The ECO guarantee program was reviewed in 1990 tured to allow for its future detachment from the NDFC. and 1992, with certain procedural modifications introduced The division now has a suitable complement of staff com- in the second review based primarily on experience gained prising young professionals who have undergone extensive with the Hub project. In 1994 the Bank replaced the ECO and specialized training, some of it abroad. The division program with a more generalized use of guarantees as a was set up as a self-contained unit responsible for its own mainstream instrument in Bank operations. The main- decisions on the technical and financial viability of pro- streaming of guarantees built on the successful features of jects presented to it for financing, after they have been the ECO program, but at the same time proposed modifi- approved by the Private Power Cell. The Private Energy cations to some governing rules no longer considered nec- Division has successfully undertaken all the pre-construc- essary, primarily to facilitate a broader use of guarantees for tion activities in connection with the appraisal of the Hub limited recourse-financed projects. project. The Water and Power Development Authority organized Institutions in Pakistan the Private Power Organization under a managing direc- tor supported by a general manager, four directors, and six The third set of measures for the promotion of private assistant directors, and with additional specialized support energy involved the creation of institutions for the review, from consultants. The organization's original function was negotiation, and approval of private project proposals. to coordinate all the Water and Power Development Under PSEDF I the Ministry of Water and Power was Authority units involved in load dispatch, system operation, assigned responsibility for reviewing and approving power finance, system planning, and so on, to handle the com- proposals and the Ministry of Petroleum and Natural plex issues of negotiating tariffs and power purchase agree- Resources for reviewing oil and gas and coal development ments, and to outline contractual and technical aspects for proposals. The Water and Power Development Authority the integration of private sector supplies into the Water and 44 Financing Pakistan's Hub Power Project Power Development Authority's system. The new policy organization now plays a major role in the implementation has transferred these functions from the Private Power of the recently approved Power Sector Development Project, Organization to the Private Power and Infrastructure Board, particularly with regard to the reorganization and corpora- which now has sole responsibility for them. tization of the Water and Power Development Authority In 1993 the Water and Power Development Authority into a holding company with decentralized generation, trans- assigned the Private Power Organization the additional mission, and distribution subsidiaries operating as discrete responsibility of planning the privatization of its assets. The autonomous profit centers. Institutional Development 45 Postscript T he fourth unit of the Hub Power Project was com- Currency movements aside, all base loan facilities have pleted three weeks ahead of schedule on 7 March been drawn substantially as envisaged in the finance plan, 1997, and formal completion of the entire station although considerable difficulty was experienced in fully complex was announced, on schedule, on 31 March 1997. using the World Bank's contribution to the PSEDF due to The output capacities and efficiencies of all the units has restrictions that apply to the application of drawings (see proven to be at or above guaranteed levels, and through- the section of project development, above). The standby out their first winter period of operation (1996/97) the com- loan facilities have not been drawn, by virtue of the project missioned units operated at almost full output. The station's being completed under budget. contribution to the reduction of load shedding in Pakistan The most significant underusage of a senior loan facil- has already been widely and publicly acknowledged. ity has been the base ECO guarantee facility which, being The pipeline connecting the station to the Pipri oil ter- a flexible "sweeper" of costs and currency mismatches, has minal was commissioned in October 1996 and has an ended up being only 75 percent drawn (compared with a installed capacity sufficient for a 2,000 megawatt power sta- forecast usage of 90 percent in the finance plan). The export tion. The first of two export power lines was commissioned credit-insured and other senior facilities have been fully in December 1995, the second in September 1996. drawn. Revenues generated from plant operations prior to The actual capital cost of completing the project has completion have been ahead of the forecasts made in the been slightly below budget-in fact, savings of about 1 finance plan, to the extent that some twenty-seven addi- percent on budget has been achieved, equivalent to $13 tional unit-weeks of generation were made available by the million. At the same time, cross-currency movements have early completion of the second, third, and fourth units. caused the dollar value of all loan facilities available to the The broadly smooth experience of managing monthly project to increase by about $50 million relative to the drawdowns of the fifteen loan facilities that comprise the original finance plan. Much of this apparent increase in loan Hub project financing has been greatly assisted by the invest- availability has been within the PSEDF's JEXIM facilities ment that Hubco and the intercreditor agent (Citibank) and gives rise to a reported underusage of the PSEDF: 85 made in assigning qualified financial analysts and managers, percent drawn compared with 100 percent in the finance at the outset, to design and implement a computer-based plan. loan management and reporting system. 47 Appendixes Appendix I Offshore commercial bank facilities syndicate profiles Facility ECO JEXIM COFACE MITI SACE ABN AMRO Bank NV / / / ANZ Grindlays Bank pic / / Banca Nazionale del Lavoro Spa-London Branch V Banco di Napoli International SA (Luxembourg)a / Banco di Napoli Spa-Hong Kong Branch I Bank of Scotland / / Bank of Tokyo, Ltd. / / Banque Francaise du Commerce Exterieur / / Bayerische Landesbank Girozentrale / Centrobanca-Banca Centrale di Credito Popolare Spa I C'tibank NA / / I Cooperatieve Centrale Raiffeisen-Boerenleenbank BA (Rabobank) / / / Crediop-Credito per le imprese e le opere pubbliche Spa / Credit Foncier de France and Auxilaire du Credit Foncier de France (jointly and severally)a / / Credit Lyonnais SA / / Dai-lchi Kangyo Bank, Limited / / / Daiwa Bank, Limited / / Deutsche Bank I / . Dresdner Bank AG / / / Fuji Bank, Limited / / / GiroCredit Bank Aktiengesellschaft der Sparkassen / I Industrial Bank of Japan, Limited / / / Internatonale Nederlanden Bank NV / / Kredietbank NV / Long Term Credit Bank of Japan, Limited I / / Mediocredito Centrale Spa I / Mediocredito Toscano Spa / MeesPierson NV / / Mitsubishi Bank, Limited / 1 / Mitsubishi Trust and Banking Corporation / / / Mitsui Trust and Banking Co., Ltd. / / / National Westminster Bank PLC / / / Norinchukin Bank / / / Orix Corporation / Royal Bank of Scotland pic / I Sakura Bank Limited / / / Sanwa Bank, Limited / / / SB General Leasing (UK) Limited / Standard Chartered Bank / / Sumitomo Bank Limited / / / Tokai Bank, Limited / / Yasuda Trust & Banking Co. Ltd. / I Westdeutsche Landesbank Girozentrale / / Banks per facility 34 19 7 17 18 a. Appears as a single bank on the tombstone for the facilities. 49 Appendix 2 Project chronology 1987 1988 1989 1990 Govemment of Pakistan July: April: January-December: Initial proposal submitted Letter of intent issued Implementation agreement to govemment of Pakistan negotafons December: Implementation agreement initialed Commercial contracts May: May-August: January-December: January-December: and corporate development Initial invitation to bid for Tender and appointment Power purchase agreement Fuel supply agreement turnkey consortium leader of tumkey consortium negotiafons negotations Tender and appointment of plant operator December: Power purchase agreement May-June: initialed Project offices opened in Islamabad, Washington, December: and London Interim tariff agreement signed Engineering and construction October: May-November: May-October: June-December: Inital site selecton Feasibility study Construction contract Internatonal competitive negotations bidding for turbine island December: December Tumkey pnce validity expires New tumkey consortium Institutonal development November: April: Ministry of Power and Water's National Development Finance Private Purchase Cell formed Corporaton's Private Energy Water and Power Develop- Division foamed ment Authority (WAPDA) private power organizaton formed Other August: Gulf war starts Finance April: November: March: March: Initial financial review First cofinanciers meeting Second cofinanciers meeting Mobilization finance review Preliminary informaton World Bank considers memorandum issued January-December: Expanded Cofinancing to cofinanciers Discussions with cofinanciers Operation (ECO) program December: Worid Bank approves the Private Sector Energy Development Fund (PSEDF) 50 Financing Pakistan's Hub Power Project 1991 1992 1993 1994 Apnl-November: April-August: January-September: Implementation agreement Implementabon agreement negofatons Bank due diligence on development implementation agreement August: Implementation agreement signed November: Supplementary implementation November: agreement signed Mobilization agreement signed April-November February-September: January-September: September: Power purchase agreement development Operations and maintenance Bank technical and commercial Reference tariff agreement signed agreement negotiations contract due diligence May: Fuel supply agreement initialed April-August: October: Power purchase agreement and fuel Indicative tariff agreement signed May: supply agreement negotiations Revised tariff agreement signed November: August: Supplementary contracts signed October: Power purchase agreement and fuel Project site acquired supply agreement signed Shareholders agreement signed June: Offices in London and Washington September: enlarged Operations and maintenance agreement signed August: Hubco formed with technical listing on Karachi stock exchange Januory-June: September: September: Construction contract negotiations Groundbreaking Site clearance July: December: November: Construction contract signed Contractor mobilization Foundation laying commences June: Private Power and Infrastructure Board formed Februory: December 1991 -May: Gulf war ends Appeals lodged against Federal Shariat court ruling November: Federal ShaHat court ruling March: September-December: January-June: February: First application to the PSEDF Mobilization finance arranged Mobilization information Fifth cofinanciers meeting memorandum prepared ECO guarantee finalized May: July: SACE commitment Export credit agency meeting Arrangers remandated February: Senior debt term sheet signed May: August: December: Information memorandum issued Arranger banks mandated Mobilization finance-Istisna facility April: agreement signed and drawn down PSEDF loan agreement signed; July: December: first disbursement COFACE and SACE loans signed ECO program approved by the Bank May: September: ECO guarantee term sheet agreed Senior debt and equity underwritten Rupee term loan signed July: ECO, JEXIM, and MITI loans signed Third cofinanciers meeting Project coordination agreement signed Brefing informaton memorandum issued Syndication closed October October: Project coordination agreement finalized Global depository receipts and share issues launched November: Fourth cofinanciers meeting November: Word Bank and JEXIM approve guarantee December: facilities and further PSEDF facilities Commonwealth Development Corporaion loan agreement signed December: Rupee mobilization finance loan Restated PSEDF loan agreement signed agreement signed COFACE and MITI commitments January 1995: Financial close Appendixes 51 Appendix 3 Principal project agreements I. Project coordination agreement 2. Definitions agreement 3 ECO facility agreement 4. JEXIM facility agreement S. COFACE facility agreement and amending letter 6. SACE facility agreement and amending letter 7. SACE reserve account agreement and SACE reserve account letter and related letters 8. MITI facility agreement 9. CDC amending agreement and amended and restated CDC facility agreement 10. Senior rupee facility agreement I . PSEDF facility agreement 12. ECO guarantee agreement 13. JEXIM guarantee agreement 14. Guarantee coordination agreement IS. Security trust deed 16. Composite security agreement 17. Assignment of insurances 18. Mortgage deed 19. Escrow agreement 20. Implementation agreement 2 I. Government of Pakistan guarantee 22. Government of Pakistan approval letter 23. Construction contract 24. Fuel supply agreement 25. Power purchase agreement 26. Operations and maintenance agreement 27. National Power guarantee 28. Consultancy services agreement and supplemental agreement 29. Raytheon guarantee 30. Consultant's agreement of warranty and undertaking 3 i. Support services agreement 32. Appointment letters a. Technical advisers b. Insurance advisers 33. Exchange risk insurance letters from State Bank of Paklstan: a. confirmation letter b. in relation to the ECO facility c. in relation to the JEXIM farility d. in relation to the MITI facility e. in relation to the CDC facility f. in relation to the COFACE facility g. in relation to the SACE facility 34. Govemment of Balochistan agreement 3 5. Ansaldo hedging agreement 36. Shareholders agreement (I1992) and novation agreement 37. Shareholders agreement (I1994) 38. Sponsors equity agreement 39. Memorandum and articles of association of the Hub Power Company Limited 52 Financing Pakistan's Hub Power Project Appendix 4 Project participants Pakistan ministries and agencies Ministry of Water and Power Ministry of Finance Water and Power Development Authority Pakistan State Oil State Bank of Pakistan Government of Balochistan National Development Finance Corporation Private Power Infrastructure Board Sponsors National Power Xenel Industries Mitsui Ishikawajima-Harirna Heavy Industries K&M Engineering & Consulting Corp. Cofinanciers Worid Bank JEXIM CDC COFACE MITI SACE U.S. Agency for intemational Development Government of France Govemment of Italy Construction consortium Mitsui Ishikawajima-Harima Heavy Industries Ansaldo Campenon Bemard Plant operator National Power Principal international banks Citibank Arranger Credit Lyonnais Arranger Bank of Tokyo Arranger NatWest Arranger Sakura Bank Arranger Mediocredito Centrale SACE joint coordinator Local banks National Development Finance Corporaion Habib Bank Limited Muslim Commercial Bank Limited National Bank of Pakistan Industrial Development Bank of Pakistan Allied Bank Limited Pakistan Industrial Credit & Investment Company Saudi Pak Industrial & Agricultural Investment Company (pvt) Umited United Bank Limited Appendixes 53 Arrangers of mobilization finance Al Rajhi Islamic Investment Company of the Gulf ANZ Grindlays Equity brokers Deutsche Morgan Grenfell Global coordinator of GDR and share issue Deutsche Bank GDR underwriter West Merchant Bank Crosby Securities Jardine Fleming Bear, Steams Intemational Limited Bear Steams Jahangir Siddiqui Ltd. Consultants to local equity offer Advisers to government of Pakistan and public bodies Latham & Watkins Legal advisers to govemment of Pakistan, Water and Power Development Authority, Pakistan State Oil, and National Development Finance Corporation Price Waterhouse Accounting advisers to govemment of Pakistan Fatehali W. Vellani & Co. Legal advisers to National Development Finance Corporation Advisers to sponsors Morgan Grenfell & Co. Limited Financial advisers Linklaters & Paines Legal advisers in United Kingdom Hogg Insurance Brokers Insurance brokers K&M Engineering & Consulting Corp. Technical advisers K&M/Ebasco Owner's engineer during construction Ford, Rhodes, Robson, Morrow Auditors to Hubco Arthur Andersen Accountants in United Kingdom Rawlence & Brown Accountants in United Kingdom (for HRPG) Surridge & Beecheno Legal advisers in Pakistan Ughi e Nunzianti Legal advisers in Italy Slaughter and May U.K. legal advisers to GDR issue Cleary, Gottlieb, Steen & Hamilton U.S. legal advisers to GDR issue Elvinger, Hos & Prussen Legal advisers to Hubco and arrangers in Luxembourg Advisers to arrangers Clifford Chance Legal advisers in United Kingdom Coopers & Lybrand Accounting advisers (financial model) in United Kingdom Stone & Webster Technical advisers Richard Barton Consulting Limited Insurance advisers Arthur D. Little Hazop consultants Rizvi, Isa & Company Legal advisers in Pakstan Chiomenti e Associati Legal advisers in Italy Advisers to cofinanciers Mitsui, Yasuda, Wani & Maeda Legal advisers to MITI Adlen & Overy Legal advisers to JEXIM 54 Financing Pakistan's Hub Power Project Appendix 5 Overview of insurance of the start-up date of the project being deferred on account arrangements during construction and of loss of or damage to imported equipment during the operation transit to the project site, provided such loss or damage is covered (or would have been covered but for the exis- In limited recourse transactions the insurance cover put in tence of a deductible) by the Marine Cargo insurance. place by the project is an important part of the security taken * Construction "All Risks" insurance covers the works by the project lenders over the physical assets of the trans- under construction on a replacement cost basis until han- action and over the revenue stream generated by those dover to the Project Company, and any temporary works assets under various project agreements. Accordingly, and and, if the parties so agree, the contractor's plant and in order to control this part of the collateral in priority to equipment. Cover continues after handover for the con- any other project party, the lenders take a security interest struction contractor until the end of the defects liabil- in the insurance cover that the project has put in place on ity period specified in the construction contract. behalf of itself and all other transaction parties. * Construction "All Risks" Delay in Start-up insurance pro- Since lenders in limited recourse transactions rely on vides cover for the Project Company and the lenders for the revenue generated by the transaction assets to service the loss of revenue that the Project Company may incur their debt, and to cover the project's fixed and variable oper- as a result of the start-up date of the Project being deferred ating costs and its profit, it follows that the lenders will require on account of loss of or damage to the works under con- wide-ranging insurance cover to be a source of cash that is struction, provided such loss or damage is covered (or external to the transaction in the event that there is an insured would have been covered but for the existence of a incident which results in a loss of or damage to the trans- deductible) by the Construction "All Risks" insurance. action assets, a reduction in or deferment of their revenue- * Third Party Liability insurance covers the liability of the earning ability, or the generation of a liability to third parties. project parties to third parties with respect to loss of or damage to the property of third parties and injury to or Limited Recourse-Financed Projects death of third parties. During the operation phase of a limited recourse onshore In the construction phase of a limited recourse onshore transaction the principal insurances effected annually by transaction the principal insurances effected by the project- the Project Company on behalf of itself and the relevant owning company ("the Project Company") on behalf of operation phase parties generally comprise physical dam- itself and the relevant construction phase project parties age insurance, physical damage business interruption generally comprise: insurance, machinery breakdown insurance, machinery * Marine Cargo insurance breakdown business interruption insurance, third party * Marine Cargo Delay in Start-up insurance liability insurance, and insurances required by law. * Construction 'All Risks" insurance * Physical Damage insurance provides cover on a replace- * Construction 'All Risks" Delay in Start-up insurance ment cost basis for loss of or damage to the project assets * Third Party Liability insurance that have been handed over to the Project Company. * Insurances required by law * Physical Damage Business Interruption insurance provides The cover provided by these insurances can be outlined cover for the Project Company's loss of revenue follow- as follows: ing an incident covered by the Physical Damage insurance. * Marine Cargo insurance customarily covers the imported * Machinery Breakdown insurance provides cover on a equipment on an "All Risks" replacement cost basis from replacement cost basis for the Project Company for loss or place of manufacture to the project site. damage to machinery, plant, boilers, and the like resulting * Marine Cargo Delay in Start-up insurance provides cover from mechanical and electrical breakdown or derangement. for the Project Company and the lenders for the loss of * Machinery Breakdown Business Interruption insurance revenue that the Project Company may incur as a result provides cover for the Project Company's loss of rev- Appendixes 55 enue following an incident covered under the Machinery ances were placed with an insurer in the country or coun- Breakdown insurance. tries providing the external loans or aid. Section 13 of the The construction and operation phase insurances are con- NIC Act requires the NIC to reinsure with the Pakistan tracts of utmost good faith; they are expressed for specific Insurance Corporation (PIC) the proportion of insurance limits of liability and contain self-insurance provisions that it cannot retain for its own account. (deductibles); and their terms and conditions define the To help minimize the lenders'exposure to Pakistan coun- nature and scope of the cover that they provide. try risk, the government of Pakistan agreed to provisions in the implementation agreement whereby Hubco would Public Sector-Financed Projects be exempted from certain of the regulatory requirements referable to the construction phase insurances. Thus Hubco When state entities (or major private corporations) finance was given permission: their capital investments "On Balance Sheet," they frequendy * to insure the Marine Cargo and Marine Cargo Delay in make different and less wide-ranging insurance arrange- Start-up insurances outside Pakistan. ments for such investments. For example, the Construction * to procure reinsurance in the international markets for 'All Risks" and Third Party Liability insurances are often 100 percent of the Construction Risk, Construction Risk arranged by the construction contractor, and the Delay in Delay in Start-up, and Third Party Liability insurances, Start-up and Business Interruption risks are often self-insured. less only the small participations underwritten by the The reasons given by state entities and major private cor- NIC and the PIC by agreement with the lenders. porations for adopting, or continuing, such a fundamen- * for direct contractual arrangements, giving Hubco (and tally different insurance strategy vary considerably, but there the lenders by means of the provisions of an assignment are increasing signs that state entities have discerned merit of insurances agreement) rights to: in insuring more of their risks with commercial insurers than * invest the control of claims under the insurances with with their taxpayers, whereas major corporations are self- the offshore reinsurers. insuring more risk because they are improving their ability * pay in currency directly to the NIC, the PIC, and the to identify and manage the risks to which their substantial offshore reinsurers their respective proportionate balance sheets are exposed. shares of the premiums referable to the insurances. In the Hub Power Project the Project Company (Hubco) * receive directly in an offshore insurance account the effected, on behalf of itself and all project parties, the full respective proportionate shares of claims paid by range of construction phase insurances that are customar- the NIC, the PIC, and the offshore reinsurers. ily required in a limited recourse independent power pro- * resolve any disputes under the insurances, or the rein- ducer transaction. The Project Company will be effecting surances, under a third-party law (English). comparable operation phase insurance cover in due course. The combined effect of the consents given by the gov- ernment of Pakistan and the agreements with the NIC Structure of the Principal Construction and the PIC gave the lenders a direct contractual relation- Phase Insurances ship with the offshore reinsurers. Thus the lenders were put in substantially the same position as they would have been Under the National Insurance Corporation Act No. XXII if the insurances had been placed directly offshore, and of 1976 (the NIC Act), Hubco was required to insure with Hubco was able to comply with Pakistan law and practice. the National Insurance Corporation (NIC) because the Hub Power Project is defined as "public property," being Special Temporary Funding a project financed out of external loans and aid. Section 10 of the NIC Act did, however, give Hubco the ability to One of the responsibilities of the government of Pakistan effect the Marine Cargo and the Marine Cargo Delay in pursuant to the terms of the implementation agreement is Start-up insurances directly offshore, provided these insur- the provision of special temporary funding under a speci- 56 Financing Pakistan's Hub Power Project fied range of force majeure events that may be grouped it was in immediately prior to the relevant event, if feasi- broadly as follows: ble; and second, to meet the fixed costs of the project * Pakistan political events-for example, war, terrorism, that fall due during restoration of the project or during civil commotion, national strikes, changes in law, expro- an interruption to the project that is caused by a relevant priation, changes in consents, and so on. event. * Political events not directly connected with Pakistan but Where special temporary funding is provided, the gov- affecting, among other things, oil deliveries. ernment of Pakistan will cause the Water and Power * Natural events causing damage to the plant that cannot Development Authority to pay a supplemental tariff suf- be covered by insurance. ficient to cover interest and principal payments, due on * Natural events causing damage away from the plant, but the special temporary funding, which is treated as a loan affecting the plant indirectly. from the government to Hubco for those purposes. Even * Acts or omissions of the Water and Power Development where the government disputes a claim for special tem- Authority or Pakistan State Oil that affect the construction porary funding, it must still pay the debt service and fixed contract price or the completion date. operating costs elements of the tariff payable under the Two types of special temporary funding are provided: power purchase agreement while arbitration is taking first, to meet the cost of restoring the project to the state place. Appendixes 57 Appendix 6 Expanded Cofinancing PSEDF I (Ln. 2982-PAK) and the replenishment loan under Operation (ECO): Summary of terms of the proposed PSEDF II loan to be presented to the the guarantee Executive Directors in concurrence with this ECO. The amounts allocated to Hub would be US$377 million and Borrower US$195 million, respectively. The Hub Power Company Limited (Hubco). Procurement Guarantor The equipment, materials, and services under the Turnkey International Bank for Reconstruction and Development Construction Contract for the power plant which have been (World Bank). or will be awarded through a competitive process meeting the requirements of economy and efficiency. Arrangers The Bank of Tokyo, Ltd., The Sakura Bank, Citibank Drawdown International Plc., Credit Lyonnais, S.A., National Disbursements under the Loan will be for eligible expen- Westminster Bank Plc. ditures for the power plant. Hubco will maintain special disbursement accounts with Citibank N.A. or The Bank Lenders of Tokyo, Ltd., in Pakistan, denominated in the curren- A syndicate of commercial banks. cies of the Loan, into which the proceeds of the Loan (other than disbursements to be paid directly to third parties and Amount disbursements needed to maintain the required balance US$240 million equivalent, consisting of two facilities: on reserve accounts) will be deposited monthly, based on a. ECO-guaranteed base credit facility of US$200 million the projected expenditure requirement. Hubco will pro- equivalent; and vide the World Bank with periodic reports on the use of b. ECO-guaranteed standby credit facility of US$40 mil- funds for expenditures for the power plant, including yearly lion. audited financial statements, accompanied by a separate opinion of auditors on whether the expenditure state- Currencies ment supports the withdrawals from the special disburse- ECO base credit facility: U.S. Dollars, Japanese Yen, French ment account. Francs, and ECU, of which: a. ECU 29 million; Availability Period b. French Francs 157 million; The earlier of March 31, 1998, or six months after the pro- c. Japanese Yen 2,000 million; and ject completion date. d. U.S. Dollars 123 million. ECO standby credit facility: U.S. Dollars Repayment The Loan will be repayable in sixteen equal semiannual Use of Proceeds installments falling due on January 10 and July 10 in each To finance part of the overall cost of the power plant, includ- year, commencing on the first interest payment date after ing design and construction costs, financing costs, and inter- the end of the Availability Period, provided that the final est during construction. The balance of the costs will be maturity date will be no later than January 10, 2006. financed by four syndicates of commercial banks covered by the Export-Import Bank of Japan and the Export Credit Interest Rate Agencies of Italy (SACE), France (COFACE), and Japan ECO base credit facility (MITI); other senior debt; and subordinated equity. a. Six-monthLondonInterbankOvernightLendingRate Subordinated debt is to be provided by the Fund under (LIBOR) for the relevant currency plus 2% per annum 58 Financing Pakistan's Hub Power Project until the eighth anniversary of the date of the loan agree- the ECO Guarantee is covered by the ECO Guarantee ment; and thereafter. Reserve Accounts (summarized below). b. Six-month LIBOR plus 2.25%. c. The ECO Lenders' Agent Bank may (and, if so instructed by the Majority Lenders, shall) call the ECO ECO standbyfacility Guarantee upon the occurrence of the following spec- a. Six-month LIBOR plus 2.5% until the eighth anniver- ified events: sary of the date of the loan agreement, and thereafter * Hubco has failed to make a principal payment or a six-month LIBOR plus 2.75%. portion thereof when due and payable to the ECO b. The interest rate per annum on the ECO Facilities will Lenders either on a scheduled payment date or on be reduced by 0.25% when Hubco's annual debt ser- acceleration or on voluntary or mandatory prepay- vice coverage ratio is 2:1 or greater. ment; and Hubco has failed to make such payments for rea- Conditions Precedent to Effectiveness sons which include (directly or indirectly) a Guarantee The ECO Facility Agreement and the ECO Guarantee Event (as described below). will be effective only when, inter alia: a. the other debt financing for the Plant is underwritten Guarantee Events and all conditions precedent for the utilization thereof A claim may be made under the ECO Guarantee in the have been fulfilled; following circumstances, if they result in a failure by Hubco b. the equity standby financing for the Plant has been fully to make a payment of principal under the Loan: subscribed and paid for (or where standby equity is to a. the Government has failed to pay Hubco any amount be subscribed or paid for subsequently, letters of credit due under the Implementation Agreement or under the or guarantees satisfactory to the Lenders are in place Government's guarantee issued thereunder of the oblig- to assure that such equity will be subscribed for when ations of Pakistan State Oil (PSO) under the Fuel Supply required); Agreement, Water and Power Development Authority c. all agreements required in connection with the con- (WAPDA) under the Power Purchase Agreement, and struction and operation of the Plant (including but not the State Bank of Pakistan under the Foreign Exchange limited to the Implementation Agreement, the Power Risk Insurance Scheme (other than any amount payable Purchase Agreement, the Fuel Supply Agreement, the by the Government as the result of the failure of PSO Construction Contract, the Operations and Maintenance to deliver fuel oil to Hubco due to certain oil-related Agreement, Security Trust Deed, the Government of foreign political events). Such payment obligations are Pakistan Guarantee, and the Exchange Risk Insurance summarized below; Letters) have been executed and are effective; and b. the Government through the State Bank of Pakistan d. Hubco has been capitalized and staffed in a manner has failed to. make available the necessary foreign satisfactory to the Lenders. exchange in accordance with its agreement under the Foreign Exchange Risk Insurance Scheme; Guarantee Provisions c. any restriction has been imposed under the laws of Scope Pakistan on the free transfer of foreign currency funds a. The ECO Guarantee will be provided by the World held by, or on behalf of, Hubco out of Pakistan; Bank in parallel with the JEXIM Guarantee byJEXIM d. any restriction has been imposed under the laws of in the initial ratio of 2:1. The ECO Guarantee and the Pakistan on the free transfer of rupees held by, or on JEXIM Guarantee will be callable on a pari passu basis. behalf of, Hubco from a project account for the pur- b. The ECO Guarantee will cover the principal payments poses of the Foreign Exchange Risk Insurance Scheme; due under the Loan and remaining unpaid. Interest due e. there has been a change in Shariah law (or such a change to the Lenders on the occurrence of risks covered by is due to occur within six months) which causes the Appendixes 59 following to become unlawful, unenforceable, or invalid: nation amount, the level of which depends on the reason Hubco's obligation to make a payment or perform an for termination but which in all cases would include the obligation under any of the project or financing agree- Loan amount due the Lenders. The Government is oblig- ments; Hubco's enjoyment of its rights under any such ated to pay a termination amount where the Implementation agreement, or enforcement, or rights to enforcement Agreement is terminated because of: in Pakistan, by Hubco or the Lenders of any obligation * a Government default thereunder; of the parties to such agreements; provided that such * the occurrence of a Shariah Event; unlawfulness, unenforceability, or invalidity affects a * the occurrence of a natural force majeure event which payment obligation in excess of $100,000 in respect of affects the WAPDA grid or the PSO system (but not interest or fees or in excess of $1 million in respect of the Plant) where Hubco and the Government agree to principal or any security or right constituted by the secu- terminate or an expert determines that the Project is rity which is material to the Lenders or Hubco (a "Shariah no longer viable; Event").The ECO Guarantee specifically exdudes from * the occurrence of a Pakistan political force majeure coverage the existing Federal Shariah court decision event where Hubco and the Government agree to ter- on interest, which has been appealed to the Supreme minate or an expert determines the Project is no longer Court of Pakistan. If such a decision is upheld by the viable; or Supreme Court, such a decision would be a Shariah * the occurrence of an uninsured natural event causing Event from that time; or damage to the Plant. f. the performance by the Government of any of its oblig- ations, or the exercise by Hubco of any of its rights, Claim on ECO Guarantee Where There is a Commercial under the Implementation Agreement, or the perfor- Default and a Guaranteed Default mance by Hubco of any of its obligations, or the exer- Whether or not the occurrence of the Guarantee Event was cise by the Lenders or the Lenders'Agent of any of their the sole cause of Hubco's failure to pay such principal rights, under the ECO Facility Agreement has become amount will not affect the ability of the ECO Lenders' Agent void, illegal, invalid, or unenforceable as a result of any to make a daim under the ECO Guarantee. However, where change in Pakistan law or any court decision which notwithstanding the occurrence of the Guarantee Event, renders inaccurate the legal opinions obtained by the Hubco would still have been unable to perform in full its Lenders at financial closing (an "Indemnified Event"). payment obligations because of a commercial default, the amount payable under the ECO Guarantee will be reduced Payment Obligations of the Government under the Implementation by that portion of the payment not paid by Hubco because Agreement and the Government's Guarantee which are Covered of the commercial default. by the ECO Guarantee Under the Government's guarantee, the Government is Claim on ECO Guarantee Where There is a Commercial responsible for guaranteeing the payment obligations of PSO Default Followed by a Guaranteed Default under the Fuel Supply Agreement and WAPDA under the The ECO Lenders' Agent may claim under the ECO Power Purchase Agreement and for guaranteeing the oblig- Guarantee in the case where the Loan has been accelerated ations of the State Bank of Pakistan under the Foreign solely because of a commercial default and the ECO Lenders Exchange Risk Insurance Scheme. are attempting to enforce their security but are unable to do so because of a change in Pakistan law, expropriation Termination Amounts or certain force majeure events. In such circumstances, The ImplementationAgreement provides that either Hubco the amount payable under the ECO Guarantee will in or the Government may terminate the Implementation essence be the lesser of (i) the amount of the Loan remain- Agreement in certain specified events. In certain of these ing unpaid and (ii) the difference between the amount the cases the Government is obligated to pay Hubco a termi- ECO Lenders would have recovered had they been able 60 Financing Pakistan's Hub Power Project to freely enforce their security and the amounts they actu- ECO Guarantee must be made within 60 days of the lat- ally recovered upon enforcement. ter of (i) the due date in respect of which a failure to pay has occurred and (ii) the expiry of any applicable grace or Limitations on Enforcement cure period relating to the relevant Guarantee Event. If Hubco fails to make a principal payment on the Loan as Calls on the Guarantees will be permitted for up to 60 a result of a Shariah Event or an Indemnified Event, or days after the scheduled final maturity date of the Facilities. because the Government has failed to make Capacity However, the Lenders have the option to extend the guar- Purchase Price or Special Temporary Funding payments antee, on an annual basis, for a further five years upon under the Implementation Agreement solely as a result of payment of an annual guarantee fee. The purpose of the a dispute as to whether such amounts are payable, the extension is to provide coverage in respect of the Lenders may call on the ECO Guarantee for any unpaid Government's continuing obligations under the scheduled installment of principal but may not accelerate ImplementationAgreement following a commercial default the Loan and take enforcement action unless the default during the time when the Lenders are seeking to enforce has not been remedied by certain specified dates, thus giv- their security interests as described above. ing the Government the opportunity to remedy the situa- tion before a claim maybe made under the ECO Guarantee Exclusion of Liability for the full unpaid principal amount of the Loan. The cure The World Bank's liability under the ECO Guarantee will periods are six months for disputes regarding Capacity be reduced to the extent a claim arises out of or as a result Purchase Price and Special Temporary Funding payments, of: (a) Hubco, the Lenders, the Lenders' Agent, the secu- nine months for a Shariah Event, and twelve months for rity trustee, or the intercreditor agent voluntarily entering an Indemnified Event. In all cases, however, the cure period into an agreement with the Government regarding such will be reduced, if necessary, so as not to exceed the period claim which materially reduces the claim against the of time for which interest payments on the Loan are cov- Goverrment; (b) the insolvency (except where contributed ered by the amount in the Guarantee Reserve Account. to by the action of the Government) of the Lenders' Agent, any of the banks holding the project accounts, or any of Guarantee Reserve Account the banks responsible for remitting funds to the State As the ECO Guarantee does not guarantee interest pay- Bank of Pakistan for the Foreign Exchange Risk Insurance ments on the Loan, an offshore Guarantee Reserve Account Scheme; or (c) a failure by such an account or remittance will be established to provide the Lenders with interest cov- bank to transmit necessary funds (except where such fail- erage during the cure periods when the Lenders are pre- ure is due to a Pakistan political event). vented from accelerating the Loan and claiming under the The ECO Guarantee will lapse if the Lenders amend or ECO Guarantee for the full principal amount in the cir- waive certain material provisions of the ECO Facility cumstances described above. The Guarantee Reserve Agreement or the intercreditor agreement without the World Account is required to be funded in an amount equal to 13 Bank's consent or if the Lenders' Agent fails to request the months' and 7 days' interest on the Loan. The Lenders may intercreditor agent to issue a drawstop notice following draw on the Guarantee Reserve Account for interest pay- the World Bank's demand to do so (as described below.) ments on the Loan in those cases where Hubco has failed to make a scheduled principal payment installment and Guarantee Fees the Lenders are entitled to call on the ECO Guarantee for The guarantee fee will be payable in the respective currency such payment. guaranteed by Hubco to the World Bank in installments. The first installment of approximately two years' guaran- Claims tee fee will be payable on the date of the first drawdown Unless the ECO Lenders have notified the World Bank of under the Loan. Thereafter, an annual guarantee fee will a payment default within 60 days thereof, claims under the be payable, commencing on the first anniversary of the Loan Appendixes 61 and on each subsequent anniversary date (until but exclud- Commitment Fee ing the anniversary date preceding the final maturity of the Three quarters of one percent per annum on undrawn com - Loan). The fee will equal one quarter of one percent of (a) mitments. the Loan commitments during the Availability Period, and (b) the maximum Loan Amount thereafter. The ECO Pre-payment Guarantee will lapse if the guarantee fees are not paid when Hubco may voluntarily prepay the Loan in full or in part due. (in agreed minimum amount and multiples thereof) on any repayment date. On project completion, any surplus Drawstop Notices funds must be used to prepay the ECO/JEXIM Facilities. During the Availability Period, if Hubco has issued a pre- After project completion, 45% of surplus operating rev- liminary termination notice under the Implementation enues must be use to prepay the senior debt pro rata, after Agreement due to a default by the Government, the World permitted shareholder distributions. Bank has the right to require the Lenders' Agent to request the intercreditor agent to issue a drawstop notice to pre- Indemnity by Pakistan vent Hubco from making any further drawings under the The Islamic Republic of Pakistan (Pakistan) will enter ECO Facility Agreement. At any time while such a draw- into an Indemnity Agreement with the World Bank in stop notice is in effect, the Lenders may claim under the respect of its Guarantee. Under such Indemnity Agreement, ECO Guarantee for all amounts of principal which Hubco Pakistan would undertake to reimburse and indemnify the fails to pay for any reason whatsoever. World Bank on demand, or as the World Bank may oth- erwise determine, for any payment made by the World Bank Subrogation under the ECO Guarantee and for all liabilities and Following a payment by the World Bank under the ECO expenses incurred by the Bank with respect to the ECO Guarantee, the World Bank will be immediately subrogated Guarantee. to the rights of the Lenders (except for any such rights which relate to amounts in the Guarantee Reserve Account or in Governing Law and Jurisdiction respect of any amounts due from the Government under The ECO Facility and the ECO Guarantee will be gov- the Implementation Agreement or in respect of amounts emed by English law and provide for the non-exclusive juris- attributable to commercial default). However, until the diction against the World Bank of the English courts. The Lenders have been repaid in full, the World Bank will not Indemnity Agreement will follow the legal regime, and be subrogatedto anyvoting entitlements held bythe Lenders include dispute settlement provisions, which are custom- or be entitled to direct the Lenders as to how to exercise ary in agreements between member countries and the World such rights. Under the terms of the ECO Guarantee, the Bank. World Bank has the right to waive its subrogation rights and rely exdusively on its rights against the Government under Agent the Indemnity Agreement. Bank of Tokyo International Limited. 62 Financing Pakistan's Hub Power Project The World Bank Headquarters 1818 H Street, N.W. Washington, D.C. 20433, U.S.A. Telephone: (202) 477-1234 Fax: (202) 477-6391 Telex: MCI 64145 WORLDBANK MCI 248423 WORLDBANK Cable Address: INTBAFRAD WASHINGTONDC European Office 66, avenue d'1ena 75116 Paris, France Telephone: (1) 40.69.30.00 Facsimile: (1) 40.69.30.66 Telex: 640651 Tokyo Office Kokusai Building 1-1, Marunouchi 3-chome Chiyoda-ku, Tokyo 100, Japan Telephone: (3) 3214-5001 Facsimile: (3) 3214-3657 Telex: 26838