Joint World Bank-USAID Discussion Paper Nov. lqqL1 Mobilizing Private Capital for the Power Sector: Experience in Asia and Latin America David Baughman Matthew Buresch FILF United States Agency for International Development The World Bank November 1994 I I I Joint World Bank-USAID Discussion Paper Mobilizing Private Capital for the Power Sector: Experience in Asia and Latin America A collaborative study between World Bank Cofinancing and Financial Advisory Services Department World Bank Industry and Energy Department USAID Office of Energy, Environment and Technology Authors: David Baughman, World Bank Matthew Buresch, RCG/Hagler Bailly Study Advisory Committee: The World Bank Inder Sud (Chair) Cofinancing and Financial Advisory Services Trevor Byer Industry and Energy Department Rafael Moscote Latin America and Caribbean Region Anil Malhotra Asia Region D.C. Rao Financial Sector Development Department International Finance Corporation Vijay Chaudhry Infrastructure Department USAID David Jhirad Office of Energv, Environment and Technology Acknowledgments This study is the result of a collaborative effort between the World Bank's Cofinancing and Financial Advisory Services Department, and Industry and Energy Department, and USAID's Office of Energy, Environment and Technology. An Advisory Committee was formed in June 1993 to guide the effort and comprised representatives from the World Bank's Cofinancing and Financial Advisory Services, Industry and Energy, and Financial Sector Development Departments, the Technical Departments of the Asia and Latin America and Caribbean Regions, IFC's Infrastructure Department, and USAID's Office of Energy, Environment, and Technology. The analysis was performed primarily by David Baughman of the Project Finance Group in the World Bank's Cofinancing and Financial Advisory Services Department and Matthew Buresch of RCG/Hagler Bailly for USAID. Several others made contributions to this study, including: Suman Babbar, Peter Cordukes and Ranjit Lamech of the World Bank; Jenifer Wishart of IFC; and Alain Streicher, John Armstrong, Suzanne Smith, Amit Dalal, and Ashraf Laidi, of RCG/Hagler Bailly. Antoinette Burnham of the World Bank provided editorial and design assistance. Foreword This study examines the recent experience of independent power projects in developing countries. Governments world-wide are increasingly looking to the private sector to provide infrastructure, after decades of public sector dominance in securing financing as well as ownership and operation. The main objective of the study is to analyze a sample of projects to understand the key financing structures, risk sharing arrangements, and host government support that enabled the project sponsors to mobilize private capital. The study focuses on Asia and Latin America, which have the most widespread experience to date, but the insights gained from the projects examined can apply more generally to other regions as well. The study looks at eight projects, all of which involve significant amounts of private sector financing, in eight countries with varied macroeconomic environments, power sector characteristics and regulatory systems. The countries are: China, Chile, Colombia, Philippines, Jamaica, Pakistan, Guatemala, and Belize. The projects examined have a range of technologies, sizes, financing needs and sponsorship arrangements and thus illustrate the breadth of experience to date with independent power. Nonetheless, the comparative analysis in the study highlights the similarities among the projects as well as evolving trends in this growing industry. Both USAID and the World Bank have important functions as clearinghouses for the experience around the world in the area ofprivate sector development, especially in the rapidly changing infrastructure sectors. This study is a part of that dissemination. Inder Sud David Jhirad Director Director Cofinancing and Financial Advisory Services Office of Energy, Environment and Technology The World Bank USAID Acronyms ADB Asian Development Bank BLT Build, Lease, Transfer BOO Build, Own, Operate BOT Build, Operate, Transfer CDC Commonwealth Development Corporation (U.K. development agency) ECA Export Credit Agency GW Gigawatt (one million kilowatts) IDB Inter-American Development Bank IFC International Finance Corporation IPP Independent Power Producer (or Project) LRMC Long-Run Marginal Cost MW Megawatts (one thousand kilowatts) NPC National Power Corporation of Philippines OECD Organization for Economic Cooperation & Development PPA Power Purchase Agreement ROE Return on Equity USEXIM US Export - Import Bank Glossary of Terms Used in the Study Commercial Risk Generally, those risks under the control of the project owner. This may include project development, construction of the plant within budget and on time, efficient operation of the plant within budget, employment of qualified personnel, etc. Debt Service Cover Ratio The ratio of 1) cash available (i.e., after all operating and other expenses are paid) to pay debt service to 2) the amount of debt service due (principal and interest). Development Period This includes the time it takes to complete all studies, conclude negotia- tions, and arrange the financing required to achieve financial closure. The starting date of this period varies depending on how the project is initiated (e.g., bidding process, issuance of letter of intent). The ending date would be financial closure. Dispatch The process of drawing electricity by the utility from a range of available plants to meet demand. Normally, this is done on a merit-order basis; that is, the plant that is capable of producing electricity at the lowest cost is cho- sen first, followed by the next lowest, etc., so that the plants with the high- est operating costs are only utilized to meet peak demand requirements. Financial Closure The point (normally a specific date) when the legal documentation for all project agreements and financial commitments (debt and equity) are ex- ecuted. In most cases, disbursements are made very soon after financial closure and construction begins. Force Majeure Uncontrollable events, generally defined to include a range of occurrences, including natural events (e.g., lightning, floods) and political events (e.g., revolution, civil unrest). Whether these latter events are controlled by gov- ernments is often debated during negotiations. Grace Period The period of time under a loan agreement that principal amounts are not payable. Investment Grade Indicates that the investment is of good quality in terms of the ability of the obligor to meet debt service obligations. It is normally indicated by an as- signed rating; for Standard and Poor's, the designation is BBB. Least-Cost Planning A planning method to determine the optimal sequence and timing of new plants to meet the future needs of the power system. It is derived from ex- pected demand, and the cost and operational characteristics of candidate plants; it is normally calculated using optimization software. Limited Recourse Generally refers to limitations on lenders' ability to seek payment or other benefits from the equity parties. It can also refer to limitations on a project entity's recourse to the host government for support in certain circum- stances. Political Risk Generally, those risks under the control of the government. This will vary depending on the circumstances and the perceptions of project participants. They may include the performance of government-owned entities (includ- ing the purchasing utility), regulatory environment, issuance of critical per- mits (customs, environment), and foreign exchange convertibility. Private Risk Capital The amount of private sector financing provided for a project that is not directly guaranteed by government. This is divided by the total amount of financing to arrive at a percentage figure. Project Finance A financing approach whereby debt is mobilized on the basis of the perfor- mance of a specific project rather than on the strength of an existing company's balance sheet. It involves a careful assessment of the project's cash flow performance under various scenarios to ensure a high probability of repayment on the project's debt. Sometimes referred to as non- or lim- ited-recourse financing. Senior Indicates an order of payment that has priority over other creditors. Sponsor Party that takes a leading role in structuring the project, negotiating the project agreements (PPA, etc.) arranging the financing, negotiating and securing the construction contract, and other tasks associated with project development. This party also quite often supplies the majority, or a signifi- cant portion, of the equity financing. Subordinate Indicates an order of payment that is junior to other creditors. This may include debt. Payments to equity would always be subordinate to all debt. Term (Loan) The number of years until the final maturity of principal. Table of Contents Acknowledgments ......................................................................i Foreword ..................................................................... ii Acronyms ..................................................................... iii Glossary of Terms Used in this Study ..................................................................... iii-iv Abstract ..................................................................... vii Note on Source Material ...................................................................... viii Section I: Background ..................................................................... 1 Power Sector Financing Needs . ..................................................................... I Trends in Power Sector Financing ........................ .............................................. 2 Key Issues Addressed in the Study ......................................................................3 Linkage to Country and Sector Context ...................................... ................................ 3 Financing Structures and Sources ..................................................................... 3 Public and Private Sector Risk Sharing ......................................................................4 Section II: Project Characteristics ...................................................................... 5 Country Risk and Regulatory Environment ......................................................................7 Project Characteristics ...................................................................... 7 Financing Structures ..................................................................... 10 IPP Pricing Issues ..................................................................... 14 Section III: Linkage to Country and Sector Context .................................................................... 9 Country Risk and Regulatory Environment ......................................................................9 Competitive Bidding/Environment/Least-Cost Planning ........................................................ 12 IPP Pricing Issues ..................................................................... 14 Section IV: Financing Structures and Sources . .................................................................... 16 Project Finance Approach ..................................................................... 16 Extent of Private Risk Capital ..................................................................... 17 Debt Financing ..................................................................... 18 Equity Financing ...................................................................... 20 Foreign vs. Domestic Capital Mobilization ..................................................................... 21 Equity Returns ..................................................................... 22 Section V: Public vs. Private Sector Risk Sharing ..................................................................... 23 Introduction ..................................................................... 23 Approach to Analysis ..................................................................... 23 Completion and Technical Performance ........................................... .......................... 25 Utility Performance ..................................................................... 26 Fuel Supply ..................................................................... 26 Foreign Exchange ..................................................................... 27 Government Financing ..................................................................... 28 Summary Project Comparisons ..................................................................... 28 Section VI: Future Developments ..................................................................... 30 Supply vs. Demand for Capital ..................................................................... 30 IPP Approach and Replicability ..................................................................... 30 Financing Structures and Sources ..................................................................... 31 Power Pricing and Private Sector Development ..................................................................... 31 Need for Credit Enhancement ..................................................................... 31 Tables Table 1: Project Profiles .................................................................... 5 Table 2: Project Sponsors ................................................................... 8 Table 3: Country Risk Profiles .................................................................... 9 Table 4: Bidding/Least-Cost Option/Environmental Standards .................... .......................... 13 Table 5: Private vs. Public Power Tariffs ................................................................... 14 Table 6: Financing Structure ................................................................... 17 Table 7: Financing Structure-Summary Debt Terms ........................................................ .... 18 Table 8: Financing Structure-Equity ................................................................... 20 Table 9: Government Risk Coverage ........................ ........................................... 18 Figures Figure 1: Additions to System Capacity ....................................................................6 Figure 2: Estimated Project Development Time .................................................................... 7 Figure 3: Private Power Policy Framework ........................................... ........................ 10 Figure 4: US Independent Power Framework ................................................................... 13 Figure 5: Country Risk and Government Support: Expected and Actual ............. .................... 24 Figure 6: Extent of Government Risk Assumption .................................................................. 25 Abstract Given the rapidly evolving conditions in the market for private power and continuing privatization efforts, new financing models and approaches to risk allocation are emerging. The study examines pri- vate power projects (often referred to as independent power projects or IPPs) in a range of countries with varying degrees of country risk, access to international capital markets, and regulatory environ- ments. To assess private power initiatives under varying circumstances, three groups of low- to middle- countries were considered: Category 1-countries with high country risk profiles and little or no access to capital markets (Pakistan, Jamaica, Belize, Guatemala); Category 2-countries with moderate coun- try risk profiles and some access to capital markets (Philippines, Colombia); and Category 3-countries with good country risk profiles and access to capital markets, but which may be constrained by amounts and/or terms (Chile, China). Eight project finance transactions were analyzed and compared to identify the similarities in risk shar- ing and the extent to which the country and sector environments required particular levels of govern- ment support. This is important because of the close interface between the public and private sectors that is often present in infrastructure projects and which can pose a significant political risk. The study also looks at power pricing issues, the role of multilateral and bilateral guarantees, financing sources and structures, levels of private risk capital mobilized, and competitive bidding, least-cost planning and envi- ronmental compliance issues. The results suggest that while private power projects alone are not likely to fill the large gap in financ- ing the power sectors in developing countries, they will have an important role in many countries and offer substantial amounts of private capital. Basic structures (e.g., BOO/BOT) can be replicated across countries, but each project will have its own unique set of circumstances that will lead to different solu- tions regarding risk sharing between private and public entities. Projects can proceed in countries with- out fully established private sector power policies, as specific private sector projects can be regulated through contractual arrangements. The sector's long-term development, however, will hinge on the success of power sector reforms. Government guarantees and the involvement of multilateral and bilateral agencies played a critical role in enabling project sponsors to mobilize private debt and equity in countries with higher levels of country risk. Reducing, or eliminating, these guarantees and various forms of credit enhancement will only occur over time, as power sector reforms take hold and utilities improve their creditworthiness in well-functioning regulatory environments. The projects examined involved substantial levels of private risk capital not guaranteed by government, while most of the capital was foreign with generally little or no involvement of domestic private investors. Private power pricing seems to be broadly in line with public alternatives, despite the existence of power tariff subsidies. Additionally, sponsors of privately-owned projects tend to be selected through a competitive process, follow international environmental standards, and were generally compatible with least-cost expansion programs. After a slow start, the IPP market appears to be accelerating, and the complex project models in these cases are being replicated and modified in many countries. Project development, however, is costly and time consuming. In some of the larger countries, the gap between the power sector's demand for capital and the level of investor/lender interest is large and will likely be slow to diminish. Note on Source Material This study drew information from numerous sources. The tables in the text provide some reference citations, but generally, project-specific information was taken from the loan documentation of multilateral financing institutions, including the World Bank, IFC and ADB. U.S. Government sources were also utilized. Supplementing this material, and where these agencies did not participate, is publicly available information from magazine or journal articles. Financial and other project-specific data were also obtained from project sponsors and lenders, and export credit and investment insurance agencies. This information was obtained primarily through personal interviews. Figures 3 and 6 were formed using judgments based on all available information. Figure 6 is also drawn from information provided during personal interviews with project participants. The tariff data in Table 5 were taken from the following sources: * Chile: IFC loan documentation; Chilectra Annual Report. * Colombia: K&M Engineering (project sponsors). * Philippines: IFC and ADB loan documentation. * Pakistan: World Bank data and loan documentation. * Guatemala: IFC loan documentation. * Jamaica. Coopers & Lybrand, Power Sector Regulatory Framework and Privatization Study, 1993. * Belize: IFC Loan documentation; World Bank data. Section I Background This study examines the recent experience of Power Sector Financing Needs eight Asian and Latin American countries in mobi- lizing private financing for power sector invest- The extraordinary financing needs of the Latin ments. The capacity requirements of these two American and Asian power sectors derive from a regions loom large in the rapidly expanding global broader need to expand infrastructure generally in demand for infrastructure. This is particularly true these two regions. As the economies of these re- for electricity service, where, after decades of public gions try to sustain their recent rapid economic sector dominance, many countries in these and growth, the need for adequate supporting infra- other regions are now looking with increasing structure such as roads, telecommunications, and interest to the private sector to provide critical electricity is becoming critical. Typically the de- investments as well as to own and operate power mand for infrastructure services outstrips the growth plants and other power-related facilities. rates of the economy when countries are at low While all the countries examined in this study income levels. The investment requirements in are trying to achieve similar objectives ofexpanding many countries also reflect the need to upgrade or capacity and increasing efficiency, each has taken a replace fully depreciated assets, or facilities that slightly different approach. All have encouraged have not been well-maintained and are being re- entry in the generation market by way of build- tired prematurely. own-operate or transfer (BOO/BOT) projects. At the project level, power sector investments, Many have also pursued far-reaching sectoral re- like infrastructure investments generally, tend to be forms to varying degrees. large, which means large financing packages are The countries examined are: Chile, China, Phil- required. This results from two factors. First, there ippines, Colombia, Jamaica, Guatemala, Belize, are scale economies in both the development of and Pakistan. They were selected primarily to re- these projects and their operation. Second, their flect a range of circumstances with respect to two long lead times require significant capacity addi- criteria: level of country risk and degree of stability tions in order to keep pace with rapid growth in and coherence in power sector regulation. These demand. Financing packages also exhibit some countries also provided a sample of projects involv- scale economies, making larger projects generally ing different technologies, fuels and sizes. They more efficient. thus provide a range ofsettings and project-specific Over the past few decades, most large public characteristics with which to examine the experi- sector projects have been financed by multilateral ence of private investment in the power sector. development banks and bilateral aid, often by They were also chosen because they are among a securing several sources of finance based on the select group of countries with actual experience in government's sovereign credit. Mobilizing similar mobilizing large amounts of private capital for amounts of private financing for private sector independent power projects (IPPs). projects in recent years, however, has been prob- 2 Mobilizing Private Capitalfor the Power Sector: Experience in Asia and Latin America lematic. This stems from the unwillingness of pri- the 1980s, followed more recently by Argentina vate lenders and investors to assume significant and Peru, which are undergoing similar transfor- exposure in any one country, because of perceived mations. Power sector reforms in these countries sovereign or political risks. This is particularly true are aimed at stimulating private sector investments for projects that earn local currency and feature a and encouraging improvements in electricity ser- close connection to regulated public sector entities, vice. These changes have occurred within the con- as in power projects. Development banks, given textofmacroeconomicreforms designed to liberalize their predominance in meeting many countries' economic activity and spur growth. As a result, financing sources, are in a better position to assess Latin America has become a prime market for IPP and control these risks, which private lenders can- investors, from both within and outside the region. not. Given its rapid growth and large populations, Trends in Power Sector Financing Asia holds a large share of the developing world's demand for new power sector investments. The Notwithstanding the tremendous need for power World Bank's Industry and Energy Department sector financing, many governments have found estimated in 1990 that Asia will require about 244 that mobilizing the amounts required is extremely Gigawatts (GW) of new generation capacity by the difficult given the severe fiscal constraints they face. end ofthe decade to meet expected demand. Trans- Traditional sources of funding are not sufficient as lated into dollar terms, this represents almost the needs faroutstrip the abilityofmultilaterals and US$280 billion (constant 1989 dollars) of new bilaterals to meet the requirements, even assuming investments, or about US$28 billion per year. reasonable mobilization of utilities' internal re- Including financing costs, the needs amount to well sources. For the two regions examined in this study, over US$30 billion per year. the major multilateral lenders are the World Bank, If China's rapid economic growth continues, it Asian Development Bank (ADB) and the Inter- alone is expected to account for 36% of this de- American Development Bank (IDB). Power sector mand, or about US$10 billion per year. India's lending from these institutions averaged roughly generation investments will representaneven larger US$4 billion in 1991-1993, which represents only portion of the region's needs; the Bank has esti- a fraction of the US$45 billion per year required in mated India's investment requirements at about generation alone. US$12 billion per year for the decade, not includ- Consequently, a growing number of countries ing financing costs. Although India's needs are are turning to the private sector to provide these large and its market for private power is developing facilities, as well as arrange financing and take on rapidly, it had not undertaken anyprivately funded operation and maintenance responsibilities. This projects at the time of this study, and is thus not comes at an opportune time given the recently examined here.' Because many projects are at an renewed flow of private capital to developing coun- advanced stage of development in India, it will be tries spurred by opening markets and improved worth revisiting once a few projects have reached macroeconomic conditions. The challenge is to financial closure. promote sector reform and project structures that Latin America, while not as large as Asia, is facilitate the channeling of this capital to the power another very dynamic region with favorable eco- sector. nomic prospects and an imposing requirement for Private sector participation in the power sector power sector investments. Recent World Bank represents a broader trend toward unbundling some assessments indicate that this region will require utility services that have traditionally been pro- some 40 GW of new generation capacity for the vided by one vertically integrated entity. Until remainder of the decade, amounting to US$75 recently, these services have been seen as natural billion in new investments, or roughly US$15 monopolies and were heavily regulated and/or con- billion per year. Much of this capacity will be trolled by the state. Unbundling has created new installed to exploit the region's hydro resources. entry points for private investors in these sectors. In Probably more than any other region, Latin the power sector, this has been realized mostly America has experimented with new regulatory regimes and industry structures that encourage I There are several privately-owned electrical utilities in private sector participation. The region is led by India. This study, however, is focused on project finance Chile which privatized its electricity sector during transactions normally undertaken on a BOO/BOT basis. Mobilizing Private Capital for the Power Sector: Experience in Asia and Latin America 3 through new entrants into the generation business, reveal both elements common to BOOs and unique and has resulted from a number of factors such as: aspects that influence the outcome of the negoti- * Governments and utilities can secure large ated risk sharing arrangements and financing pack- amounts of capital and significant capacity ages. additions by contracting with one plant; * The interface between the IPP and power pur- Key Issues Addressed in the Study chaser (utility) is relatively well-defined; and * Theprecdentsestalishd in he UitedDuring the mid and late 1980s, numerous pri- • The precedents established in the Unitedvate developers applied the BOO/BOT model to States and United Kngdom which provide a infrastructure projects, particularly private power pool of investors with experience in develop- projects. Initial attempts in several countries, haw- ing independent power projects. ever, met with long delays or were abandoned, The deregulation of the U.S. and U.K. markets leading many observers to question the appropri- serves as a model for a number of developing ateness of this model. Recently, activity has acceler- countries and inform the debate in many others. ated and after much debate, a consensus is beginning The experiences ofprivate investors, manyofwhich to form on the key questions facing those involved originate from these two countries, have certainly in the emerging independent power market. Most influenced their approach to the power business. of the issues fall into one of three categories: linkage Private sector participation can take manyforms, to country and sector context, financing structures including: privatization (change of ownership); and sources, and public and private sector risk build, own, operate (BOO) schemes; and opera- sharing issues. tions and maintenance contracts. Corporatization The answers to these questions affect the activi- is another approach that, while not necessarily ties of all the main actors in the private power involving private sector participation, can lead to business including developing country govern- better performance and may, in fact, represent a ments, multilateral and bilateral development in- transitional phase and make subsequent privatiza- stitutions, developers, investors, and lenders. By tion easier. analyzing the actual experience of eight projects, In the end, long-run solutions involving sector this study brings together a range of issues that can reform are preferable and will be the most durable. be framed by the following questions: But establishing new regulatory systems and the corporatization of existing entities can take a long Linkage to Country and Sector Context time. Because of this, many governments are opt- ing for two-track strategies. This involves promot- * Country Risk and Regulatory Environment: ing entry in generation as a way to meet pressing How does country risk influence IPPs? And, capacity requirements through private sector par- to what extent is a well-developed policy, ticipation, while at the same time devising regu- legal, and institutional framework supporting latory arrangements to encourage sector-wide privately owned power companies essential to operational and financial improvement as well as the development of individual projects and an new investment to meet future requirements. overall IPP industry? This study examines new entry through BOOs. * Bidding/Environment/Least-Cost Planning: The basic approach to BOO schemes is by now To what extent are IPP sponsors selected well-known. Typically, they involve the govern- through competitive bidding, and to what ment awarding a concession, or right, for a speci- extent do IPPs comply with environmental fied period of time, to a private investor to arrange guidelines, and fit into least-cost plans? financing, and construct, own and operate a facility * IPP Pricing Issues: How do IPP tariffs com- to provide a service, such as electricity. Included in pare to some common benchmarks associated this study are eight projects that match or vary only with public sector provision of electricity? slightly from this approach: Shajiao C in China, What are the key pricing issues? Pangue in Chile, Pagbilao in Philippines, Mamonal in Colombia, Hub in Pakistan, Rockfort in Ja- Financing Structures and Sources maica, Puerto Quetzal in Guatemala, and Macal River Hydro in Belize. These projects represent a * Extent of Private Risk Capital: What level of good cross-section of experience with BOOs and private risk capital is mobilized by successful 4 Mobilizing Private Capitalfor the Power Sector: Experience in Asia and Latin America IPPs? What is an appropriate measure of "risk particular country? What are the necessary capital" in the context of IPPs? conditions for eliminating the requirement for * Foreign versus Domestic Capital Mobiliza- such government guarantees? tion: What proportions of foreign versus do- * IPP Replicability: How replicable are these mestic capital were mobilized and how projects; are they special cases, or the basis for important a role do domestic capital markets more rapid development of an independent play? power market? * Risk versus Equity Return Tradeoff: What Section II summarizes and compares the main range of risk-adjusted equity returns is typi- project characteristics of the eight projects studied. cally called for by IPP projects in developing Section III discusses the relationship of these IPPs countries? to country risk environments and power sector policies. Section IV looks at financing structures Public adPiaeScand sources. Section V analyzes the risk sharing * Government versus Private Risk Assump- arrangements and discusses the similarities and tion: What level of government support and/ differences among the projects. Last, Section VI or guarantees of specific risks are required in provides some observations on the future develop- order to enable the development of IPPs in a ment of the industry. Section 11 Project Characteristics Table 1 presents some key characteristics of the the debt and often extend beyond the maturity of projects studied. They include a range of technolo- project loans to account for any potential interrup- gies, and both hydro and thermal plants using tion in cash flow during the life of the contract (see different fuel sources. The thermal projectswere all discussion on project finance on page 16). land based, with the exception ofPuerto Quetzal in There is also a wide range in plant sizes, which Guatemala which is a barge mounted facility. somewhat depends on the capacity requirements in Typically, IPPs are baseload plants due to the each country. As mentioned above, the arguments desire of purchasing utilities to maximize their for large plant size include economies of scale in utilization in order to realize the benefit from development, operations, and financing. Econo- expected greater operating efficiencies compared to mies of scale in operations will vary, however, existing plants in service. Most projects are also fully dispatchable and therefore compensate the projec entit for te amun of caact avial 2 In dispatchable projects, the utility compensates the project entity for the amount of capacity available project for all fixed costs (debt service, fixe d O&M, and for dispatching during the payment period. equity returns) in a "capacity payment" if it meets agreed For most projects, revenues are secured through technical performance standards (e.g., amount of capac- long-term Power Purchase Agreements (PPAs), ity and availability). In return, the utility controls the ranging from 14 years (Mamonal) to 40 years dispatch of the plant to meet system needs. To theex (Macal River Hydro), with the average being about tent the utility dispatches the plant, it compensates the project for fuel consumption and other variable costs in 20-25 years. This follows traditional "project fi- an "energy payment.' The project is also subject to nance" structures, whereby revenues are secured by agreed efficiency standards, for example, in terms of heat contracts that at least match the term (maturity) of rates and O&M costs. Table 1: Project Profiles Fuel Base or Total Cost PPA Term Project MWIFuel Source Peak Load (US$ millions) (years) China: Shajiao C 1,980/Coal Domestic Base 2,000 20 Chile: Pangue 450/Hydro Domestic Base 437 n.a. Colombia: Mamnonal 100/Gas Domestic Base 70 14 Philippines: Pagbilao 700/Coal Imported Base 933 25 Pakistan: Hub 1,292/Oil Imported Base 1,900 30 Guatemala: Puerto Quetzal 110/Oil Imported Base 92 15 Jamaica: Rockfort 60/Oil Imported Base 130 25 Belize: Macal River 25/Hydro Domestic Base 60 40 n.a. Not Applicable 6 Mobilizing Private Capitalfor the Power Sector: Experience in Asia and Latin America depending on the type of plant (i.e., steam, gas set in terms of risk sharing between the IPP and the turbine, etc.). Thus, the optimum plant size thresh- government as well as financing arrangements in a old may be different for individual projects. Even if particular country, project size could be increased. the threshold is relatively low, economies in devel- Most of the projects examined here, however, opment and financing can reduce the cost of power did not seem to follow this strategy. As Figure 1 when the alternative to meeting expected capacity shows, on average, these eight projects added about needs is developing several IPPs simultaneously. In 18% (weighted by size) to the existing capacity, addition, negotiating several IPPs would, in many indicating the large stakes governments and utili- countries, severely tax the administrative capabili- ties have invested in these ventures to meet pressing ties of government agencies. needs. This may have occurred because project size For some countries it has been pointed out that was dictated more by system requirements, rather small plants (relative to the size of the total system) than considerations of economies in development are a less risky approach than large projects, given and financing. the complexities in raising financing for IPPs in Figure 2 shows that there can be a wide variance many developing countries, and the often arduous in the time from the beginning of project develop- learning and negotiating process that ensues once ment (defined as the start of negotiations or bid- the developer is selected. Where there are difficul- ding) to financial closure and the commencement ties in raising financing for IPPs, the risk is mani- of construction. In general, however, for large fested in project implementation delays and the projects selling to public utilities for public supply resultant unmet power needs. (which includes all the projects with the exception The balancing factors in this trade-off are best of Pangue and Mamonal), the average time to determined based on the specific circumstances in develop IPPs was two and a half to three years. This each country and for each project and the expected includes several months to bid the project and transactions costs incurred in developing and nego- select a developer, one to two years to complete tiating the first one or few projects. It has been negotiationsonaPPA,andseveralmonthstoayear suggested that it may be prudent for the first few to arrange financing. projects to be relatively small, provided the plant It should be noted that, for lack of consistent size is economic in the context of system require- information, the timelines in Figure 2 have some- ments. This reduces the impact on the system in the what different reference points for the starting date. event of delays while at the same time may acceler- It is difficult to pinpoint an exact starting date for ate project development and financing given the a project, whether it is the submission of a proposal smaller financing needs. Once a precedent has been by a developer, the award of a bid, or initial contact Figure 1: Additions to System Capacity (Percent) 60 50 40 30 20 1 0 0 E a) .1 gr CD 0- Notes: China system size figure incorporates only Guangdong Province. Pangue amounts to about 12% of Chile's total capacity, but 23% of Endesa's capacity (Endesa is the parent company of Empresa Pangue, S.A., the project company). Colombia is not included as Mamonal is an enclave project not primarily for public supply. Mobilizing Private Capitalfor the Power Sector: Experience in Asia and Latin America 7 between the sponsor and the government or utility. example, Shajiao B (Shajiao C's predecessor) was Hence, the timelines may not be directly compa- developed over two years, or about the same time as rable in all cases, although they do provide a general Shajiao C, although it should be noted that C was indication of development periods. approximately three times as large as B. Navotas 1, Development time depends on several factors, the first IPP in Philippines, was developed over two including country risk, project size, the complexity and one half years, compared to Pagbilao which of the financing plan and the difficulty in mobiliz- took four years, or about one and a halfyears longer ing private debt sources. With the exception of than Navotas. Again, Pagbilao was a significantly Shajiao C, the large projects tended to have longer larger and more complicated project than Navotas. development times than the small projects. The This suggests that achieving noticeable reductions relation between development time and other fac- in development time may, in fact, require more tors such as extent ofgovernment involvement, and than just one or two projects and that development number of financiers is less clear. Thus, aside from time will depend on the nature of the project. The size, there does not appear to be a strong relation- indications will become clear as more projects are ship between development times and specific coun- completed in China and Philippines, as well as in try, project or sectoral factors. other countries. Many observers expect average development Hub's extended timeline may be a special case to time to diminish over time as subsequent projects some extent, given the number of new initiatives are developed. The experience of two countries in being undertaken by the Government of Pakistan, this study, however, suggest that development time as well as in the financing of the project. Various may not necessarily decrease over time, at least political difficulties in Pakistan also delayed the during the initial phase of IPP involvement. For project at several points. This latter factor effec- Figure 2: Estimated Project Development Time Years Project (Start Date) 0 1 2 3 4 5 6 Colombia: Mamonal Jan. 1992: Bid Issued _ 7 Guatemala: Puerto Quetzal - - $ Mid-1991: Negotiations Belize: Macal River End-1991: Negotiations _ China: Shajiao C Jan. 1991: Negotiations --- r - Jamaica: Rockfort March 1992: Bid Issued _ _ Philippines: Pagbilao Jan. 1989: Bid Issued --------- ---- Chile: Pangue 1989: Negotiations ---------- $-| Pakistan: Hub April 1988: LOI __ _ -- - - - ---- __ PPAIContracts Signed T Construction Begins Financial Closure 8 Mobilizing Private Capitalfor the Power Sector: Experience in Asia and Latin America tively added several months to the development the country where the IPP market is most devel- time. oped. At the other extreme is the Mamonal Project in Various participants in a project (government, Colombia, which involved both a much smaller utility, developer, lenders, multilaterals, etc.) will project and private sector purchasers in a private be involved and/or more active at certain junctures commercial transaction, and no direct government of project development, while others will be en- involvement. It was also undertaken in a far more gaged throughout. For instance, lenders to IPPs, creditworthy country, which facilitated the mobi- including some multilaterals such as IFC, begin lization of financing. This structure may account analyzing the potential for their participation only for the compressed development time indicated. once a PPA has been negotiated by the developer Table 2 shows the sponsoring parties for each and the utility or is close to signature. Developers, project and the origin of the sponsors (i.e., head- on the other hand, as the project promoters, are quarters or primary domestic market). Only two involved in the process for the entire development projects involved domestic sponsors (Pangue and period and often invest large sums in "development Shajiao C), while the rest involved foreign spon- costs" such as legal and other advisory fees, travel, sors. Half of the foreign sponsors are U.S.-based, and their internal costs. Table 2: Project Sponsors Project Sponsors Origin China: Shajiao C Guangdong General Power Corp./Hopewell China/Hong Kong Chile: Pangue Endesa Chile Colombia: Mamonal K&M Engineering U.S. Philippines: Pagbilao Hopewell Hong Kong Pakistan: Hub National Power/Xenel Industries U.K./Saudi Arabia Guatemala: Puerto Quetzal Enron U.S. Jamaica: Rockfort Hydra-Co Enterprises/ U.S. Energy/ U.S. International Energy Finance Belize: Macal River Dominion Resources U.S. Section /I1 Linkage to Country and Sector Context Country Risk and Regulatory Environment The table broadly indicates the degree of coun- try risk, with China and Chile topping the list The first item investors look into when contem- (indicating a lower degree of country risk), Colom- plating a power sector investment is the country bia and Philippines in the middle range, and Paki- risk and regulatory environment they will face. The stan, Jamaica, Belize and Guatemala toward the countries analyzed in this study represent widely lower end. Four of the eight countries have S&P divergent circumstances, which allowed for com- sovereign ratings, with Chile and China having parison of different levels of creditworthiness and investment-grade ratings of BBB. Colombia and sector regulatory development. Philippines have less than investment-grade ratings The eight countries were classified in three cat- of BBB- and BB respectively, but are nonetheless egories from a country risk perspective: low, me- attractive to manyIPP investors from acountryrisk dium and high risk. While somewhat arbitrary, this standpoint. For non-rated countries, IPP invest- classification is useful for comparing experiences. ment opportunities are conditioned on the market Table 3 provides some indicators of country risk, size and the existence of specific policies and pro- gleaned from various market and other sources. grams that encourage private investment in the Table 3: Country Risk Profiles Average Country Standard & Poor's Debt Service to Ranking Sovereign Exports 1992 Country (/0) Rating (%) China 18 BBB 11 Chile 28 BBB+ 21 Colombia 34 BBB- 39 Philippines 52 BB 18 Pakistan 60 .. 21 Guatemala 62 .. 28 Jamaica 63 .. 30 Belize .. .. 8 .. Not rated Note: Average country rankings are based on the average percentile ranking of Euromoney, Institutional Investor, and Bank of America country rankings. Lower number indicates higher average ranking. 10 Mobilizing Private Capitalfor the Power Sector: Experience in Asia and Latin America power sector. Also, countries with higher country place for private sector ownership and operation. risk profiles may offer good diversification poten- Other countries have dealt with increased private tial for developers with a portfolio of IPP invest- sector involvement on an ad hoc basis without ments in several countries. Of course, investors will explicit and uniform criteria or processes. undertake assessments of country risk, regardless The cases examined here indicate that the exist- of whether or not the sovereign credit is rated. ence of a fully developed policy and institutional Even small countries (like Belize, Guatemala framework for the power sector is not essential to and Jamaica) with limited opportunities in scope the development of the first few IPPs, but is critical offer advantages to some IPP investors. They may to the long-run development of a power market appeal to developers who wish to limit their expo- with substantial private sector participation. In sure by virtue of the generally smaller project size, most of the cases studied, governments have estab- or to smaller developers with limited financial lished the broad legal and institutional framework, resources for equity investments. Also, the gener- sometimes with development assistance from do- ally smaller units required in these countries may nor organizations. Policy formulation and project entail less risks in construction and operation, development, however, often proceed simulta- thereby possibly enhancing their appeal. The larger neously, where the policy, legal, and regulatory countries such as China and Pakistan obviously framework influence the market, while experience offer greater potential given in larger market and with projects in turn refines the policy and institu- project sizes, which in general tend to attract larger tional development process. developerswith greater financial strength, although Figure 3 provides an overview of the policy they face the challenge of raising large amounts of framework facing investors in the sample countries. debt financing. Chile offers a third scenario with its As is clear from the figure, most of the countries in sound economic policies, good growth potential, the study have adopted policies relating to IPPs, at and well-structured and established regulatory en- least to some extent. The benefits of these policies, vironment. This combination of factors has appeal if maintained in a consistent manner, are that they to many developers, provided electricity demand is can provide a clear framework for developers to sufficient. pursue projects. In reality, these policies have evolved Adding to the overall macro framework, many overtime (e.g., inPakistan, Philippines, andChina), of these countries adopted specific policies to en- which has resulted in uncertainties for developers as courage private power investments, whether or not to the rules regarding risk sharing and other provi- broader sectoral regulatory arrangements are in sions. In the end, the greatest benefit of IPP policies Figure 3: Private Power Policy Framework IPP Competitive Policies Independent Bidding Country or Law Regulator Framework China O Q O Chile 0* Colombia O O O Philippines (O Pakistan Q Guatemala O ( 0 Jamaica Q O Belize O O Q * Established ( Developing 0 Limited or None ' Chile's power sector is primarily privately owned. Competitive bidding is not mandatory for utilities purchasing power from independent generators, although they can bid on their own accord. Mobilizing Private Capitalfor the Power Sector: Experience in Asia and Latin America 11 may be that they send a strong signal to investors in this report is an important milestone in the about the opening of the power market. development of the sector as it is the largest new The market in the Philippines, which is the most private investment in the sector since privatization. developed, evolved from a number of factors. For- With the exception ofthe Philippines and Chile, eign investmentlaws, which provide favorable treat- the power sectors in most of the sample countries ment to power investments, combined with a severe are regulated at the ministry level with no indepen- power sector crisis brought on by plant closings, dent regulator. While several countries are estab- drought and poor maintenance gave developers an lishing independent regulators, developers tend to opportunity to enter the market. The Philippines look at established practice in structuring their risk launched its private power policy in 1987. While mitigation arrangements, and do not grant much the first private power project, Hopewell's Navotas importance to the intended impact of government 1, began operation in 1991, other private power policies. This means that only regulatory systems proposals failed to progress in large part because of that have solid track records will have credence with poor coordination between government agencies developers and hence allay their fears of political and inadequate regulations. interference or arbitrary changes in procedures or In 1993, in response to continued severe elec- laws that negatively impact their projects. tricity shortages and resultant outages, as well as In all cases except Chile and Colombia, the some difficulty in attracting investors following entities that purchased the power were publicly Navotas, the government issued a decree that estab- owned and operated electric utilities. This is the lished fast-track procedures for screening and nego- case for most power sectors world-wide, and, to the tiating IPPs and allowing for tariff adjustments extent the utility is financially unstable, this pre- which ordinarily would involve a time-consuming sents a risk from the point of view of the developers process. Since then activity has picked up and who will rely on that entity to fulfill the financial several projects have reached financial closure. The and operational commitments made to the IPP. lessons from this experience caused the govern- Often, other providers of inputs (e.g., fuel) and ment to establish the Investment Coordinating other facilities (e.g., transmission interconnection) Committee and adopt new BOT laws and imple- critical to the successful operation of the project are menting regulations. Since these improvements, publicly owned entities as well. For example, fuel the Philippines has seen an increase in the projects supply for the thermal projects was provided by a reaching financial closure, and continued strong public entity in four out of six cases. This public- interest on the part of developers. sector interface complicates investor risk mitiga- In Jamaica, the first IPP (Rockfort) achieved tion strategies to the degree that it heightens the financial closure in October 1994, about two and perceived political risk to the project. a halfyears after the government first called for bids In virtually all the projects studied where the from developers. Based on this experience, the purchasing utility is a publicly-owned entity, gov- Jamaican utility is now negotiating with several ernments "regulated" IPPs through contractual other private sector projects to add more capacity, arrangements. In turn, investors have a clear frame- which the government hopes to bring on line work for doing business, embodied in the con- relatively quickly. In Pakistan, several projects long tracts, and do not have to rely on unproven sector identified by the government as priority private regulatory arrangements. In China, for instance, sector projects are now proceeding after several the Shajiao C project and Shajiao B before it, were years of relative inaction as developers awaited the implemented on the basis of contractual agree- outcome of the Hub Project, which is also expected ments in the absence ofan established policy frame- to close in 1994. work, let alone specific laws and regulations. It has The transformation of Chile's power sector also even been argued that waiting to establish the proceeded gradually. The privatization process be- policy framework may impede IPP development, if gan in the early 1980s with the passage of new investors have to wait through the complications power sector legislation and pension fund reform anddelaysinvolvedinpassinglegislationandbuild- while the actual privatization of the largest state- ing regulatory capacity. In the Philippines and owned utility, Endesa, occurred over several years Pakistan for example, the executive and legislative during the 1 980s. Regulation of the sector has been debates around independent power policies could efficient and transparent, although the framework be regarded as having complicated and delayed continues to evolve. The Pangue project discussed some of the negotiations with developers. 12 Mobilizing Private Capitalfor the Power Sector: Experience in Asia and Latin America Given this situation, a dual track effort that ing and transmission function generally calls for involves simultaneous policy and project develop- careful regulatory oversight. ment can be a viable approach to independent Theevolutionofapolicyandinstitutionalframe- power market development. Contractual provi- work that accommodates private participation of- sions can be designed to cover the gaps in the policy ten takes time to reach a stage that inspires investor and legal framework. The structure of these initial confidence. As shown in Figure 4 the development projects, however, needs to be regarded as transi- of the US IPP industry took several years from the tional. As the market evolves and more utilities are initial passage of the Public Utilities Regulatory privatized, project development will move away Policy Act (PURPA) to withstand legal challenges from special contractual provisions with one pur- and utility/regulator intransigence and to realize chasing entity to more reliance on sector regulatory major growth. provisions. This can be seen in the Pangue project in Chile which relies on the sale of energy that is not Competitive Bidding/Environment/Least- Cost under contract for about 40% of its expected Planning revenue. This kind of arrangement requires that generators have the right to access the grid to allow Table 4 provides information on the use of them to sell to distributions companies and other competitive bidding for the selection of sponsors, large users. each project's adherence to environmental stan- As demonstrated in such countries as the US and dards, and whether the project was the least-cost the UK, there are, of course, some basic provisions alternative. that shouldbe inplace,suchastherightofindepen- Competitive Bidding. Jamaica and Pakistan, in dent generators to enter the generation business. addition to articulating policies toward IPP invest- Theexistenceof an independentregulatorybodyto ments in generation, have also adopted policies to limit the monopoly power that can arise in the promote competitive bidding for the right to de- power sector is useful, but often there are other velop new plants. This may reflect a growing trend means to insure against monopolistic practices. in developing countries to cope up-front with the Long-tern development of sector requires open influx of new proposals for BOO plants and the access to transmission, and close coordination be- desire to achieve competition. tween economic dispatch and the settling of ac- In many cases, the scope for competition among counts among generators with varying commercial generators is limited given the current state of most obligations. In turn, management of the dispatch- regulatory systems. Competition in energy and Figure 4: US Independent Power Framework Annual non-utility capacity Development Timeline additins exceeds utility 50 Further dereg. 1 stlPP not of IPPs 45 - tOlimited (R_ M to renew- :jR 40 ables or 35 ~~~~~~~~~Compet-coe a 30 itive Bidding 25str 20 Legaland 9! 15 Regulatory Disputes E 10 1978 1980 1983 1987 1989 1990 1992 1994 PURPA Law Passes, opening Other Expansion of Environmental Control up market for Important reewbleand2 Trends Integrated Resoulef Planning antioOSM Mobilizing Private Capitalfor the Power Sector: Experience in Asia and Latin America 13 capacity requires sophisticated regulatoryoversight, documents and procedures will vary. Thus, which in many countries is not currently practical. the rigidity of the process, and hence the ob- Financial constraints based on the need to secure a jectivity of the criteria and transparency of the steady cash flow stream to service loans and provide procedures, will depend on the circumstances. a return on equity can also hinder competition in * Power Sector Needs: the more urgent the generation, which may inject too much uncertainty need for new generation, the more likely utili- with regard to revenues, particularlyforIPPs. Hence, ties and governments will streamline bidding, many governments are turning to competitive bids with the line between competitive bids and to capture savings from competition up-front. negotiated deals growing blurred as selection The use of competitive bidding for selection of criteria become more general. sponsors in the sample of projects studied was not Projects funded by most multilaterals normally universal, but nonetheless was relatively common. require competitive bidding for the sourcing of Out of eight projects, five involved some formal equipment, as distinguished from selecting the competitive bidding process. The degree ofcompe- project sponsor through a competitive process, tition and the formality of the procedures utilized although the Jamaica example has provided a pre- undoubtedly varied among the five projects that cedent for some flexibility. Some multilaterals, were selected competitively, although lack ofinfor- such as IFC, do not require competitive bidding for mation makes it difficult to compare the degree of either sponsor selection or equipment sourcing, rigor used for each projects. In Jamaica, the com- although they insist on good business practices. petitive bidding procedure used was approved by Environmental Standards. The experience of the the World Bank and IDB, as they provided a projects analyzed in this study suggests that most portion of the funding for the project. Because of BOO/BOT projects adhere to international (often this, the Banks did not require a subsequent com- World Bank) environmental standards. Those petitive bid for equipment. projects involving IFC or the World Bank did meet In general, the use of competitive bidding for World Bank guidelines. Other projects are re- sponsors will depend on several factors that will be ported to meet these standards as well although this influenced by local conditions and policies. Some could not be verified. In general, however, foreign of the most prominent of these include: private power developers are likely to be as respon- * Legal requirements: some countries have laws sible in their compliance with environmental stan- or regulations that oblige utilities to tender the dards as local governmental entities. In addition, right to develop new plants on a competitive private lenders are increasingly looking to the envi- basis; ronmental standards of multilateral institutions in * Institutional Capabilities: the ability of a util- those circumstances where local standards are not ity or government to develop complex bidding yet developed. Table 4: Bidding/Least Cost Option/Environmental Standards Competitive Bidding Least-Cost Environmental Project for Sponsor Option Standard China: Shajiao C No .. World Bankt Chile: Pangue No Yes World Bank Colombia: Mamonal Yes .. World Bank' Philippines: Pagbilao Yes Yes World Bank Pakistan: Hub No Yes World Bank Guatemala: Puerto Quetzal Yes Yes World Bank Jamaica: Rockfort Yes Yes World Bank Belize: Macal River Yes Yes World Bank Unknown Reported to comply, but existence or content of Environmental Impact Assessment is uncertain. 14 Mobilizing Private Capitalfor the Power Sector: Experience in Asia and Latin America Least-Cost Option. The extent to which each without major disruptions. It was also generally project fits into a least-cost expansion plan in each below the economic long-run marginal costs country's power sector is less clear. While inte- (LRMC) of providing power, although the degree grated resource planning (i.e., least-cost analysis to which the utility's retail tariffs reflect marginal combined with demand-side management) prac- costs varied among countries. tices are evolving and being applied to varying One would expect IPP tariffs to be below both degrees in developing countries, the selection of system LRMC (the figure that is commonly avail- most projects reviewed generally did fit into some able) and the utility's retail tariff. They should be form of formal or informal least-cost planning below system LRMC because it represents the framework. It is important to note that, aside from marginal cost of generation, transmission and dis- the least-cost question per se, limited data on tribution, while IPPs costs are at the generation privately developed projects suggests that they do level.3 Generation generally accounts for about achieve high levels of operational performance, 60% of LRMC. Thus, ideally IPP tariffs should be which enhances the economic and environmental compared to the marginal costs of generation, but performance of the power sector overall. Shajiao B this figure was not available in most cases. IPP in China averaged 95% availability in its first three tariffs should also be below utility retail tariffs and a half years of operation. Navotas 1 in Philip- because the utility must recover the financial costs pines averaged 90% availability in its first two years of providing transmission and distribution service of operation. Availability in the second year was in addition to generation. reduced due to a failure in one unit for which To some extent, it can be argued that IPPs, insurance proceeds were sufficient to repair in a because of their need to fully recover all costs, short period of time. including all financing costs, introduce pressure on utilities to adjust tariffs to realistic levels and there- IPP Pricing Issues fore assist in moving pricing mechanics and levels to more market-based systems. On the other hand, Power pricing is a complicated and much de- the degree to which utilities can impose these tariffs bated subject, and one that has been widely dis- on consumers who are accustomed to subsidized cussed with respect to IPPs as more projects achieve rates is often in question. financial closure and attention is focused on the Table 5 provides some pricing data for both the "price" for introducing private power. IPPs and for the relevant public utility ifapplicable, From the available information, the price of ortheclosestproxytothepriceofpubliclysupplied independent power for most projects studied was electricity. generally far below the average retail tariff at the 3 In this context, it is important to note that LRMC is an time the project came on line, which should allow economic calculation, while IPP pricing reflects financial the utility to pass through these costs to consumers costs. Table 5: Private vs. Public Power Tariffs (Per Kilowatt Hour) Average Tariff Charged Retail by Independent IPP IPP Nominal LRMC' Utility Power Project Levelized Tariff/ Utility Power Pricing Factor per kWh Tariff Nominal Levelized Tariff/LRMC Average Tariff Chile: Pangue .. $.074 .. $.029 .. 39% Colombia: Mamonal .. $.08' $.045 .. .. 56% Philippines: Pagbilao $.08 $.072 $.076 $.06 75% 106% Pakistan: Hub $.066 $.051 . $.058 88% 114% Guatemala: Puerto Quetzal $.062 .. $.054 $.064 103% Jamaica: Rockfort $.12 $.14 $.069 57% 49% Belize: Macal River .. $.19 .. $.11 .. 58% Not applicable or available Long Run Marginal Cost, the weighted average long-run cost of generation, transmission, and distribution. 2 Average tariff in Colombia (1993), i.e., estimate of what the power purchaser would be paying to public supplier of electricity. Mobilizing Private Capitalfor the Power Sector: Experience in Asia and Latin America 15 This table, to the extent possible given data account for risk-adjusted returns on equity limitations, shows four pricing figures: system investment, and in fact they often do not LRMC per kilowatt hour (kWh); the utility's aver- provide any return to the shareholders (gov- age retail tariff per kwh; the tariff charged by the ernment). While the government's "return" is IPP at the time of project start up; and the levelized presumably passed through in the form of low IPP tariffs over the life of the PPA. tariffs, the process is not transparent, making In all of the projects for which there are data, the cost comparisons with IPPs difficult at best. levelized and nominal IPP tariff was approximately * Data Limitations: Gaps in the data limit cost at or below LRMC, ranging from 57% to 103%. As comparisons of projects in many ways. Public explained above, this is expected, although one sector accounting and cost data are often sub- would expect IPP tariffs to be closer to 60% of standard and not current. In addition, indirect LRMC as a general rule. costs (for instance, the use of staff from other In four out of six cases, the IPP tariffs were well ministries or agencies in developing a project below the utility's retail tariff. Again, one would and general overhead expenses) are generally expect the ratio to be close to 60%, considering not accounted for. recovery of transmission and distribution cost. IPP While the cost of private sector capital is higher tariffswere about 50% below retail tariffs on average than public sector capital, better availability of forthe fourprojects that are belowthe utility's retail privately owned plants, greater efficiency in opera- tariff. Where IPP tariffs are greater than retail tariffs tions, and efficient management systems can off-set (Pakistan and Philippines), it indicates a low retail these costs. The extent to which private investors tariff rather than high IPP tariffs, given the fact that can emphasize the latter measures to offset the LRMC is above the retail tariff in both countries. former will determine the competitiveness of pri- These figures are provided only as indicative vate power. While in the US, IPPs have, on balance, measures of IPP tariffs and utility pricing. IPP resulted in a reduction in the cost of power genera- pricing and its relationship to utility pricing in tion, the ability to make these precise calculations developing countries is an area which needs closer indevelopingcountriesisconstrainedbythelackof attention now that several projects have been con- a sufficient track record and the required data. This cluded. Two issues should considered when analyz- points to the need for additional analysis of this ing tariff data for public and private entities, question. Projects could be compared in several particularly in the case of IPPs: ways. One way would be to examine the same cost * Transparency: IPPs often provide a more category for public and private sector projects side- transparent accounting of costs than state- by-side, imputing public sector costs with market owned utilities, since investors (equity and proxies when necessary. In this event, adjustments debt) require returns on all costs. Also, gov- would also have to be made for risk-adjusted fi- ernments and therefore public projects do not nancing terms and costs. Section IV Financing Structures and Sources Project Finance Approach and expertise in pursuing profitable investment opportunities. A second important reason is the BOO/BOT projects are generally undertaken risk sharing which is achieved through project on a "project finance" basis.4 Project finance refers finance structures, where sponsors share project to a range of financing structures whereby lenders risks with lenders. If the project fails, lenders absorb depend on the performance of the project itself for any losses suffered along with the sponsors. Other repayment, rather than the credit of the sponsor. It reasons for project finance include tax benefits for is also sometimes referred to as non-recourse or the sponsors, and favorable accounting treatment limited recourse financing. These terms refer to the resulting from categorizing project finance invest- fact that lenders either have no or limited recourse ments as off balance sheet, thereby limiting a to the sponsors for repayment of loans. company's debt exposure. Project finance normally involves the following Governments, of course, also attempt to limit elements: the recourse of lenders to their credit, except to the * Lenders' reliance on the cash flow of the extent that they mayguarantee the performance the project for repayment without full recourse to performance of a state-owned entity, for example the sponsor; through a sales contract. Nevertheless, they are * Thorough technical and financial evaluation often closely linked to the operation of the project, of the project by lenders, including the source if they are not a direct participant in the project in of revenue stream, construction contractors, some capacity. operating arrangements and other project Table 6 provides an overview of the financing features that are key to maintaining adequate structure for the projects studied. Generally, all the cash flow for debt service; projects were structured as some variation of BOO/ • Complex loan and security documentation, BOT, and were financed on a project finance basis. often involving several lenders and investors; The primary source of capital for most projects was dtaie involving several lnrisk aocatond ionvsto foreign private investors and commercial banks * A detailed process of risk allocation amongst alnwihmtlteladbltrlagcesTe project participants including sponsors, lend- alnwihmtlteladbltrlagcesTe projectq tipant suicludie, sonsors, lend- percentage foreign capital was between 80% and eors, equipmasers,input suppliers, cointo rsoera- 100% in all but two cases. Given that the prime r, p haers, motivation for independent power policies is to address a capital shortage by attracting foreign Project finance structures have appeal primarily investment, this relationship is to be expected. because they allow sponsors to undertake invest- ments that they otherwise would not be able to The general discussion of project finance in this secton is make on the strength of their own balance sheet. In partly drawn from the publication, Project Finance by this way, it allows them to leverage their resources Clifford Chance. Mobilizing Private Capitalfor the Power Sector: Experience in Asia and Latin America 17 They were also highly leveraged transactions, nancing structures that result from specific project- with debt-equity ratios ranging from 80/20 to 70/ related needs, financing amounts and debt service 30, with the exception being the Shajiao C Project, cover ratio considerations. The alternative to pri- and to a lesser extent Pangue, which had debt- vate sector involvement is public sector sponsor- equity ratios closer to 50/50. Foreign capital played ship and financing ofprojects. Traditionally, under a dominant role in funding these projects, prima- these arrangements, the government guarantees rily stemming from the underdeveloped nature of the financing required for a project, which would domestic capital markets in most of the countries then typically be undertaken by a government- examined, and the large foreign exchange require- owned entity. The extent of private risk capital ments for equipment, which is usually imported provided through private sponsorship of power from industrialized countries. projects is then the degree to which government does not directly guarantee repayment of debt or Extent of Private Risk Capital equity return to financiers. Some projects, although they do not involve 100% private risk capital, do Table 6 also shows an estimate of the private risk nonetheless represent a significant step forward in capital provided by these projects. This was mea- terms and amounts, given the country and sector sured as the amount of private debt and equity risk environments. financing provided (in dollar terms) that does not There are two observations on the concept of benefit from direct governmentguarantees taken as private risk capital worth highlighting: a percentage of total (dollar) funding. Measured as * As noted in Table 6, many governments pro- such, the percentage of risk capital ranged from a vided a guarantee of the purchasing utility, as low of 25% to 100% in several cases. in the case of Philippines, Jamaica, Guatemala The level of private capital was substantial in and Pakistan. It is, however, important to most of the projects examined. More than 80% of distinguish between indirect guarantees like the total funding requirements was met by private these which are one step removed, versus di- risk capital in five out of eight cases. Most of the rect guarantees of loan repayments. In the projects involved some form of government guar- latter case, governments take both commercial antees to cover the utility's performance under the and political risks, as the debt service obliga- PPA (where the purchaser was state-owned), the tion must be met regardless of the project's supply and price of fuel, and the availability of performance. A guarantee of utility perfor- foreign exchange. Under these conditions, the pri- mance, on the other hand, can be seen as a vate investor or lender was at least covering most, if political risk guarantee, given that the perfor- not all, of the commercial risk. mance of government-owned entities is often It is important to put this measure in the context controlled in some fashion by the government of the country risk, government policies and fi- itself. In these cases, the government is gener- Table 6: Financing Structure Debt Equity Foreign Capital Private Risk Capital Project Type Ratio (Percent of Total) (Percent of Total)' China: Shajiao C BOT 38/62 44 252 Chile: Pangue BOO 60/40 42 100 Colombia: Mamonal BLT 80/20 100 100 Philippines: Pagbilao BOT 75/25 100 1002 Pakistan: Hub BOO 80/20 85 48 (Political)/67 (Commercial)2 Guatemala: Puerto Quetzal BOT 77/23 100 922 Jamaica: Rockfort BOO 70/30 98 100 (Yrs 1-5)/35 (Yrs 6 17)2 Belize: Macal River BOT 77/233 100 1002 I Financing that has not received direct sovereign guarantees. 2 Government guarantees provided to cover one or more of the following risks: utility performance, fuel supply, and foreign exchange. 17% of debt is subordinated and convertible into equity. Note: BOT-Build, Operate, Transfer; BOO-Build, Own, Operate; BLT-Build, Lease, Transfer 18 Mobilizing Private Capitalfor the Power Sector: Experience in Asia and Latin America ally not obligated to repay the loan in the foreign exchange, involves a greater degree of"in- event the project does not perform to agreed direct" government guarantees. These outcomes availability and operational standards, and were determined based on a specific set of circum- lenders and investors assume these and other stances, and thereby exemplify the range of poten- risks related to the technical and commercial tial structures. It is not possible, however, to precisely management of the project, which can be quantify the risks covered by government guaran- significant. Of course, the more risks the gov- tees in those cases where the government provided ernment guarantees beyond merely the perfor- indirect guarantees. Therefore, the distinction be- mance of government entities, the closer these tween these projects needs to be kept in mind when indirect guarantees are to repayment guaran- making comparisons. tees. As countrycreditworthiness improves and power As discussed below, the extent of risk capital sectors become more attractive for investors, one provided must also be analyzed and judged in can expect even higher proportions of private risk the context of overall project risk sharing. Low capital in the future for most countries. Others may levels do not automatically mean the host be constrained for some time due to sovereign government is "overexposed." It may be the creditworthiness considerations. case that the underlying project agreements pass many risks from the public sector to the Debt Financing private sponsor, notwithstanding levels of less than 100%. This can be judged by analyzing Table 7 provides an overview of the debt terms the sponsor's cmmitmntsgad th pnaltzies achieved in the projects. Most private debt financ- the sponsor's commitments and the penalties ddbcomrilanswheto contained in the PPA. It can also be assessed ing was provided by commercial banks, while twO by analyzing the degree to which other o - projects involved bond issues. The sponsors of tions, like fuel supply, are placed in the hands Rockfort in Jamaica, tapped medium-term bond of sponsors. financing available in Puerto Rico to fund con- struction.5 In Chile, the sponsors of Pangue are The Mamonal and Pagbilao projects illustrate the latter point. While both feature 100% in pri- The bonds were issued by the Caribbean Basin Projects vate capital, Mamonal has virtually no government Financing Authority, a Puerto Rican institution which involvement and support while Pagbilao. by virtue channels investment funds from US corporations operat- involvement and support while Pagbilao, by virtue ing in Puerto Rico to private sector projects in Carib- of a government guarantee of NPC's purchase and bean countries. These bonds are tax-advantaged under fuel supply obligations, as well as provision of the US tax code and offer below LIBOR interest rates. Table 7: Financing Structure-ummary Debt Terms Weighted Avg. Weighted Avg. Percentage of Debt Grace Term Debt Enhanced/ Project Rank (years) (years) Guaranteed' China: Shajiao C Senior 3 7.5 100 Chile: Pangue Senior 5 14 20 Subordinated 14 15 0 Colombia: Mamonal Senior 1 8.5 0 Philippines: Pagbilao Senior 4 13 80 Pakistan: Hub River Senior 4 12 70 Subordinated 8 23 100 Guatemala: Senior 0.5 8.7 0 Puerto Quetzal Subordinated 0.75 10.75 0 Jamaica: Rockfort Senior 5 8.5 100 Subordinated 2.5 10 0 Belize: Macal River Senior 4 12 0 Subordinated 4 14 0 ' This includes: loans made by governments, loans made by bilateral agencies and institutions; and private loans guaranteed by export credit agencies or multilateral institutions. Mobilizing Private Capitalfor the Power Sector: Experience in Asia and Latin America 19 seeking to tap the long-term domestic market role in Pangue, where the purchasing entities are through a bond issue (see page 21). privately owned. The ECAs providing coverage for The maturity of the loans ranged from 5 year commercial banks under these arrangements for the construction financing in Jamaica to 23 years for sample projects include those of France, Italy, Japan, government subordinated debt from the in Paki- the U.S., Sweden, Canada and Germany, which stan. The average term for senior commercial debt provide the bulk of export credit cover globally. was just over 9.5 years, with an average grace period ECAs traditionally have provided political and of 3.5 years. A significant share of this financing commercial risk coverage for public sector projects benefited from partial risk guarantees from export in developing countries. While the need for this credit agencies or multilateral development banks. type of coverage remains, particularly for political Additionally, commercial financing was mobilized risks, the nature of the borrower and financing through IFC's B-Loan program for three projects.6 structures are beginning to change rapidly. Given The terms of this debt tend to be somewhat shorter the importance oftheir participation in many IPPs, than IFC loans, but are normally much longer than ECAs have therefore had to reassess their coverage previous commercial loans to these countries. policies, the pricing of their risk coverages, and the The challenge for developers in structuring fi- extent to which they can rely on host government nancing packages for IPPs is to secure debt with counter-guarantees. For the projects in this study, maturities long enough to accommodate a reason- the extent and nature of these counter-guarantees able tariff profile. This is especially important for was not clear from available information. those projects selling electricity to a public grid, Many ECAs are willing to support project fi- given the difficulties of increasing tariffs on a regu- nance transactions because of the expanding busi- lar basis. Shorter maturities would "front load" nessopportunitiesforelectricalequipmentsuppliers tariffs, making it very difficult for utilities to meet in projects being financed on this basis. As experi- their payment obligations in the early years of a ence accumulates with the risk sharing provisions, project. coverages and documentation involved with these Equity returns are normally constrained in the arrangements, one can expect to see ECAs become earlyyearsbyminimumdebtservicecoverageratios even more active in IPP project financing. The and restrictions on dividend distributions imposed U.S.- Eximbank's recent reorganization and em- by lenders. These are specified in loan agreements phasis on project finance reflect a broader move in and represent threshold levels for lenders. In the many ECAs to adjust policies and practices to earlyyears of a proj ect's operations, debt repayment accommodate this approach. dominates dividend distributions. Over time, loan ECA cover and other forms of enhancement repayments diminish, however, leaving larger cash through multilateral guarantees or programs like balances for distributions. IFC's B-Loan syndication facilitate the extended Generally, the tariff will fluctuate over time maturities necessary for IPP financings. ECA cover accordingtotherevenuerequirementsoftheproject is generally limited to 10 years, but for power for operational expenses plus debt service and eq- projects it can extend maturities for up to 12 years uity returns, and does not remain fixed. Project under the OECD rules that govern ECA policies. sponsors must structure their revenue stream such By comparing the terms of the "enhanced" debt that annual tariffs remain reasonable on a year to financing with what had been achieved in the year basis, as well as over the life of the project (i.e., country without enhancement, one can get an idea levelized). Several of the projects utilized subordi- of the benefit of the comfort provided by the nated debt in the financing plan as a way of provid- guarantee. On average, based on the terms of debt ing a cushion to senior lenders in terms of debt financing achieved for the projects in the study, service coverage. For their part, providers of subor- ECA and multilateral support extended maturities dinated funds sometimes receive options to convert by about 7 years. This unquestionably played an their claim into an equity stake at some point or important role in enabling the IPPs to offer com- higher interest rates to compensate for the greater petitively priced power to the purchasing entity, as risks associated with a subordinate position. well as achieve an acceptable pay-back period for Export credit agency (ECA) participation was their equity investments. an interesting feature of several projects. ECAs 6 Where IFC, in addition to providing a loan for its own played a significant role in Pagbilao and Hub, account, is the lender of record for a syndicated com- where the purchaser is a public utility, and a lesser mercial bank loan. 20 Mobilizing Private Capitalfor the Power Sector: Experience in Asia and Latin America What is becoming clear, and is amply demon- riods to the satisfaction of lenders for the invest- strated in Tables 6 and 7, is that for many of these ment to be viable. projects to go forward, it has been necessary to The reluctance of lenders is due to many factors, involve one or more of the multilateral agencies, or includinglackofconfidence in commitments made bilateral export credit agencies. For those projects by governments and a lack of familiarity with the which show little or no direct government guaran- capabilities and reliability of the key public sector tees (i.e., Pangue, Mamonal, Pagbilao, Puerto agencies involved. The weak credit standing of Quetzal, Macal River), three out of five involved many governments also undermines the attractive- commercial funding mobilized under IFC's de ness ofgovernment commitments. Many countries facto preferred creditor (B Loan) umbrella (Pangue, have implemented new regulatory systems to en- Pagbilao, Puerto Quetzal), which was a significant hance the attractiveness of the power sector to factor in making the investment possible. There investors, but investors look for a track record to have been IPP-type projects that have reached assess the effectiveness and predictability of these financial closure without multilaterals or ECA in- arrangements, which will take time to develop. volvement; the YTL project in Malaysia, which is Multilaterals and bilateral agencies are uniquely being financed from domestic sources, is a good positioned to assist commercial parties to enter example. To that extent, the project sample may be these new and complex markets in the meantime, somewhat biased, although most of the successful given their long presence in these countries as IPPs not covered in this study are either in countries development agencies and their widespread experi- with relatively low country risk environments, or ence in power sector development. Multilaterals involved bilateral and multilateral support. may support IPPs in a number of ways, including Although the demand for IPPs is large and the provision of a portion of the financing, guaran- expanding, private lenders have demonstrated an tees against political risks, and support to public across the board reluctance to enter these markets sector entities designed to improve their creditwor- without some form ofpolitical riskcover, given the thiness (e.g., commercialization initiatives). The domestic orientation of infrastructure projects and financial commitments of these agencies, however, the difficulty of absorbing cross-border risks on arelimitedbytheirownexposureguidelines. Hence, their balance sheets. Commercial lenders, particu- for larger projects it will be necessary for several larly banks, are willing in general to assume com- agencies to participate in order to mobilize enough mercial risks, such as the ability of a contractor to financing to complete the financing plan. provide a technically acceptable power plant and efficient operation of the plant by the owners. They Equity Financing are not, however, well-positioned to assume politi- cal risks, or those risks under the control of the Table 8 provides some details on equity financ- government. These risks need to be addressed for ing. The equity structure is most influenced by the both the pre-completion and post-completion pe- objectives and business strategy of the sponsoring Table 8: Financing Structure-Equity Sponsor's Equity Domestic Equity Estimated Equity Project (% of Total Equity) (% of Total Equity) Return (%IRR) China: Shajiao C 100' 60 >20 Chile: Pangue 98 98 19 Colombia: Mamonal <50 0 >20 Philippines: Pagbilao 87 0 24 Pakistan: Hub 26 26 18 Guatemala: Puerto Quetzal 1002 83 20 Jamaica: Rockfort 35 6 25 Belize: Macal River 100 5 Unknown Project entity is a joint venture between Guangdong General Power Corp. and Hopewell. 2 The sponsor, Enron, will retain 50% of equity as a long-term investment and will bring in suitable partners for the remainder. EEGSA, the power purchaser, will purchase an income note of US$7.25 million from the project company at start-up, repayable between years 8-12 interest free. Mobilizing Private Capitalfor the Power Sector: Experience in Asia and Latin America 21 party, the degree of sovereign risk and the level of This marketmaybeencouraged ifpowerprojects domestic capital market development. From the could be refinanced after construction. This would figures in Table 8, two strategies are apparent. In allow institutional investors, comprising pension the more common strategy, the sponsor's equity (as funds, insurance companies, and mutual funds, to opposed to more passive equityparticipants) played replace commercial banks as the primary long-term a dominant role, accounting for greater than 50% lenders. While this investor group is immeasurably oftotal equity in six out ofeightprojects. Hopewell, larger in developed countries, there are potentially the sponsor of Pagbilao and the foreign joint- large amounts of capital to be tapped in developing venture party in Shajiao C offers a good example of countries through these intermediaries. To enable asponsorthatgenerallyprovidesasignificantequi- this, many countries will need to reform their ty stake and does not leverage its own equity capital markets, and pension fund systems specifi- position. Inthesecondstrategy,thesponsorssought cally, in conjunction with power sector reforms to leverage their participation by seeking passive that establish sound regulatory arrangements to equity partners for the majority of the equity fi- free up such resources for private sector investment. nancing required. The sponsors of the Rockfort, When an IPP is structured as a BOT project, Mamonal and Hub projects followed this strategy, where ownership is ultimately returned to the providing less than half of the total equiry in the power purchasingutility, the prospects ofthe project project. being sold on the secondary market may be limited, and the ability of domestic and institutional inves- Foreign vs. Domestic CapitalMobilization tors to participate is inhibited. Of the eight projects evaluated, four are BOTs, three are BOOs, and one Domestic participation in equity tends to be is a Built, Lease, and Transfer project. Given the small with some exceptions. Most foreign sponsors initial dominance of the BOT model, the existence see local equity participation as potentially helpful of three BOO projects in the cases examined may in terms of familiarity with the business environ- indicate atrend toward the more open-ended BOO ment and regulations, but not absolutely essential. as a preferred approach. Scarcity of risk capital inhibits local participation as In the U.S. the market has evolved to the point well, particularly for a new industry such as power that debt securities issued by at least one project generation. have been rated prior to completion (i.e., the Sithel The Shajiao C project in China was an anomaly Independence Project in New York State). While with the provincial utility, the Guangdong General this type of rating may be a long way off for many Power Corporation, taking a 60% equity stake. In developing countries, it is now conceivable that addition, Pangue in Chile had a large domestic well-structured projects with strong sponsors in stake through the sponsor Endesa, which provided countries with a sovereign rating could be rated in almost 40% of total financing through its equity the future. The problem for many countries today investment. In all the remaining cases, both a is that, without adequate credit enhancement, the majority of the debt and equity came from foreign poor or non-existent sovereign credit rating creates investors and lenders, with bilateral and multilater- a ceiling on the creditworthiness of projects in that alinstitutionsprovidingnotonlyfinancingbutalso country. Rating projects is also made difficult by various forms of credit enhancement in the form of the rating agencies' lack of familiarity with the guarantees and insurance. varied regulatory and operating environments in The limited or non-existent role of domestic most developing countries. capital in these initial projects may present a barrier The Pangue project in Chile provides an inter- to market development should it persist over the esting case of local capital market development and long term. Ultimately, a framework must be estab- the role of domestic rating agencies, of which there lished to channel domestic savings into long-term are several in Chile. The sponsors are currently investment opportunities provided by infrastruc- attempting to tap the domestic bond market for ture projects. The establishment of domestic capi- long-term funding, but need to access the institu- tal markets will no doubt be an evolutionary process tional market. In order to do this, the sponsors have in view of the need for the requisite policy, regula- requested the rating agencies for a rating to allow tory, and institutional frameworks. Nonetheless, them to market the issue in the public markets. The this will greatly facilitate the formation of a more rating agencies, in turn, are grappling with the active market for power generation in the future. difficulties of rating a project finance transaction in 22 Mobilizing Private Capitalfor the Power Sector: Experience in Asia and Latin America which the issuing entity has no track record. If 15% range, due to increased competition and some successful, this could set a precedent for bringing standardization in financing structures and risk together long-term investment opportunities with allocation. For some countries in the sample, the pension funds and other investors with extended premium over U.S. investments inherent in the investment horizons. observed 20-25% range may not seem adequate to compensate for the substantially higher risks devel- Equity Returns opers face in markets which are yet to be developed on a large scale. This is particularly true of those Surprisingly, the expected (exante) rate of return investors who bear the costs of "development risks" for all projects was very similar, in the 20-25% during the negotiations of project agreements and range, with none higher than 25%. This is some- the arrangement of financing. However, increasing what unexpected because of the diversity of risk competitionamongdevelopersindevelopingcoun- environments faced by developers in the countries tries is beginning to exert downward pressure on examined, which, all things equal, would suggest a equity returns. broader range of returns as investors appropriately More fundamentally, quantifying the risks in adjust their returns commensurate with the risk large infrastructure projects is difficult and poses they face. substantial complications in determining a fair There are several caveats to this finding. First, in return. Rather than focus the negotiations on the a few cases, expostreturns may be higher because of tariff, countries such as China, and initially India, efficiency incentives built into the tariff formula in announced caps on the return investors can earn on the PPA (e.g., Pagbilao, Hub). Second, the spon- power projects based on the rationale that these soring investors often earn a premium given their caps serve the public interest by preventing exces- earlierfundingcommitmentfordevelopmentcosts sive profits. There has been speculation that the (and hence assumption of higher risks) compared policy in China could in part be attributed to the with investors who subscribe equity later in the perceived high returns that Hopewell realized on development process, or even after construction the Shajiao B project. Some observers have sug- begins. Therefore, there may be a range of equity gested that the Chinese appeared not to want to returns embedded in Table 8, which shows average establish this project as a model, and thus insisted equity returns for all equity holders. Last, it is on majority ownership and less attractive terms on difficult, due to confidentiality, to uncover pre- the Shajiao C project. India has since adopted a cisely what ROE is expected on some projects, so more flexible approach. the figures for at least half of the projects are rough The argument put forward by investors is that estimates. These estimates are based on discussions governments should focus on negotiating a com- with developers as well as published sources. petitive tariff for power and allow investors to earn The narrow range of equity returns required by an appropriate risk-adjusted return at a tariff the investors (20% - 25%) may be attributed primarily utility can absorb. Although the level of the tariff is to an implicit "cap" due to substantial levels of partially a function of the return on equity, it has implicit and explicit credit enhancement by gov- been suggested that the existing policies oriented ernment, bilateral, and multilateral institutions. toward capping equity returns limit investment There are also constraints on utilities in absorbing and do not give investors the incentive to optimize (i.e., passing through) high tariffs. their performance. In this way, many investors IPP investors in the U.S. market, while earning argue that the regulation of IPP returns rather than very high returns (>30%) in the early years of tariffs is posing a significant barrier to attracting PURPA, are now realizing returns more in the 10- private capital. Section V Public vs. Private Sector Risk Sharing Introduction played a key role in various projects (i.e., Pagbilao, Mamonal, and Hub). This results in large part from Risk sharing, in one form or another, is at the projects' needs to secure long-term funding, which heart of most of the reservations about BOO/BOT is difficult to arrange in today's market, particularly projects generally, and specifically for IPPs, where for non-export oriented projects in high risk coun- some observers have questioned the "real benefits" tries. of this approach. While there is no one way to It is important to note that the analysis of these measure these benefits and the costs of IPPs, some cases represents a snapshot in time in what is the key areas of risk sharing are highlighted here for early stage of an evolutionary process in most of the each project. Risk sharing does not easily lend itself countries examined. The development of the inde- to quantitative analysis; thus much ofthe following pendent power market has brought more experi- discussion is qualitative in nature. Nonetheless, by ence and some standardization across countries. looking in some detail at how risks are mitigated This means that the role of multilateral and bilat- across several projects and in different country and eral institutions will also evolve and is likely to sector environments, one can pinpoint areas of diminish as the commercial framework improves. divergence and similarity to form an overview of the range of experiences. Approach to Analysis In countries with high country risk levels and electric utilities that are seen as uncreditworthy, the This study analyzed the main project-specific targeted role of government, bilateral, and multi- characteristics within the various country and sec- lateral guarantees and credit enhancement is often tor settings, in order to identify key similarities and critical to the successful financing of at least the first differences in project structures. In each case, the projects. This support can either cover country focus is on the particular arrangements that were (political) or project (commercial) risk; govern- necessarytosecureprivatefinancingfortheproject, ments and multi- or bilateral agency guarantees the terms of the financing, and the level of govern- have focused on the former risks, over which inves- ment support required to achieve a satisfactory tors and lenders have little or no control. credit package acceptable to lenders and investors. For instance, in Shajiao C, Hub, Pagbilao, Thelevelofcountryriskcanbecapturedthrough Rockfort, Puerto Quetzal, and Macal River, gov- indicators such as those presented above in Table 3. ernment undertakings were essential to mobilizing Based on these indicators, the projects included in private investments. In these cases (with the excep- the sample were allocated to three country risk tion of Shajiao C), and also Pangue, the role of the categories: IFC, World Bank, ADB, CDC and other develop- Low: Pangue, Shajiao C ment agencies was important in mitigating risks to Medium: Mamonal, Pagbilao commercial investors and lenders. In addition, High: Hub, Rockfort, Macal River, Puerto export credit agency and bilateral financial support Quetzal 24 Mobilizing Private Capitalfor the Power Sector: Experience in Asia and Latin America Depending on circumstances, IPPs may require too high, either in terms of returns demanded by a range of government support to mitigate country investors to absorb the risk, or prohibitively expen- risk or more project-specific risks that confront sive private insurance. Also, in many instances investors. Examples of this support include: private insurance is not available for periods be- * financial guarantees backing payment obliga- yond one year or less. This type of government tions of government-owned entities (particu- support allows countries to realize the benefits of larly where the sole purchaser of output is IPPs in private financing, development and opera- government-owned) tional expertise, while achieving a tariff that can be * provision of fuel accommodated in a utility's revenue stream with- * guarantees of foreign exchange. out a major disruption in tariff levels. Government support in these areas is designed Given this situation, one would expect that in to lessen risks to the private investors and lenders. high-risk countries, investors would look to gov- These risks are heightened in less creditworthy ernment for greater levels ofsupport. The expected countries given the uncertainties in economic poli- scenario is illustrated in the left side of Figure 5, cies, changes in governments, and other factors that which depicts the country risk on one axis and the may cause an interruption in project cash flows. In level ofgovernment support on the other. The right these countries, investors, generally look to the side of Figure 5 shows actual outcomes (see below). sovereign government to cover these risks (e.g. To arrive at the right side of Figure 5, the role of insurance) since private arrangements may not be the government in mitigating several risks was available, or government-owned entities are them- examined. While the number of risks in a project selves not creditworthy. In addition, in most high- financing are numerous given the range of un- risk countries, the government is usually in the best knowns that may adversely affect the project in the position of any party involved to influence or future, the study focused on eight specific risks. correct policies that may be detrimental to the The approach followed identified areas in which project and are thus the logical party to insure the projects may diverge (rather than areas of broad against these events. acceptance) in order to determine the degree of Alternatively, there may be times when a gov- government support and to test the above relation- ernment wishes to provide specific financial and ship. Other areas where specific approaches to risk other commitments to the project. This is espe- sharing have become common in BOO/BOTs cially true if the price for covering a certain risk is were not analyzed. Figure 5: Country Risk and Government Support: Expected andActual Expected Actual Macal R_ i fort High Puerto Quetzal Puerw Hub Rockfort Quwitral 441 Hub Mal Mamonal c Medium Pagbilao Pangue langu Low Shajiao C Low Medium High Low Medium High Extent of Govemment Support Very Somewhat Somewhat Very Underguaranteed Underguaranteed E Overguaranteed Overguaranteed Mobilizing Private Capitalfor the Power Sector: Experience in Asia and Latin America 25 Figure 6 shows the extent of government risk investor and not the government. In the sample of assumption associated with certain identified risks, projects studied this was generally the case. Spon- induding the extent to which governments tookan sors will normally put at risk their portion of the equity stake in the project and provided some fundingwhile construction lenders assume comple- portion of the debt financing for the project. The tion risks for the debt portion. To a large extent, eight areas examined indude: these risks are passed on to a construction * Project Completion contractor(s) who provides a performance bond up * Project Technical Performance to a fixed amount (i.e., percentage of the contract), * Utility Performance to back their commitment to complete the project * Fuel Supply on time and within budget. Lenders often also seek * Foreign Exchange insurance from ECAs or other agencies to cover * Force Majeure political risks during construction. This is the case * Equity Financing for Pagbilao and Hub for a portion of the debt * Debt Financing provided by commercial banks. As shown in Figure 6, there was a wide range of Occasionally, sponsors will back completion experience in terms of the degree to which govern- beyond their equity contribution, but this is the ment took on a major risk for each item identified. exception. Two examples are Pangue and Puerto The extent of overall government support for the Quetzal. In Pangue, the parent company, Endesa projects ranged from Mamonal and Pangue, where provided a completion guarantee that was released there was no government involvement, to Shajiao by securing a certain portion of the project's rev- C where the provincial government played a sig- enues through long-term sales contracts and oper- nificant role across the board from financing, to ating the project at levels agreed with the lenders for operations, and part ownership of the facility,. one year. This guarantee was for the benefit of the Several of these items are highlighted below. lenders. In Puerto Quetzal, the sponsor (Enron) provided a completion guarantee with the lenders Completion and Technical Performance disbursing only at completion of the facility. As in Pangue, the project had to operate for one year at Project completion and technical performance agreed levels for the guarantee to be released. are generally seen as strictly commercial risks, and As shown in Figure 6, governments do on occa- are thus normally borne by the private lenders and sion play a role in these areas for different reasons. Figure 6: Extent of Government Risk Assumption Project Technical Utility Financing Com- Perform- Perform- Fuel Foreign Force Project pletion ance ance1 Supply Exchange Majeure Equity Debt China: Shajiao C O O * * * Chile: Pangue Q Q Q Q Q Q Q Q Colombia: Mamonal Q Q Q Q Q Q Q Q Philippines: Pagbilao 0 0 S 0 0 0 Pakistan: Hub * * * * 0 0 Guatemala: Puerto Quetzal 0 0 S 0 0 0 (O Jamaica: Rockfort O 0 O Q O Q O Belize: Macal River 0 0 O 0 0 0 0 * Major risk to government O Some risk to government 0 Little or no risk to government 'Indudes payment obligations. 26 Mobilizing Private Capitalfor the Power Sector: Experience in Asia and Latin America In the case of Shajiao C, the provincial government long-term contracts with private distribution com- of Guangdong is ajoint venture partner, and in this panies and large industrial consumers, and through capacity is acting as part owner of the facility. As sales in the short-term market to other generators. such, any commercial risk generally assumed by an About 60% of forecast sales will be secured though owner, such as "completion risk," would be borne contracts, which is approximately the proportion by the province given its ownership stake. In Figure of debt in the financing plan. The choice of cus- 6, this is shown as a partial risk to the government tomers in this case was a business decision influ- due to its ownership share. enced by the need for steady cash flows (i.e., InthecaseofHub, thegovernmenthascommit- creditworthy buyers) and the desire of lenders to ted to pay a specified amount when a project secure long-term commitments. It was not dictated agreement is terminated due to a government de- by the existence of a single buyer, which is often the fault. To the extent this may occur before project case when projects produce for public supply. In completion, the government is assuming comple- addition, Chile's regulatory system has been func- tion risks. This is shown as a partial risk due to the tioning well for several years, both thereby obviat- limited nature of government commitment, i.e., ing the need for government intervention both only in the event of a government-caused default, from the point of view of the government and the and only for a specified (capped) amount. investors and lenders. This is a unique situation in developing countries, but one that may become Utility Performance more common as power sector reforms take hold. Mamonal in Colombia benefited from the fact In all six projects which involved government- that it was primarily an enclave operation selling to owned purchasers, the government provided a guar- a consortium of industrial customers. The project antee of the utility's obligations. This is common was developed as an enclave partly as a result of a where there is no clear division between the govern- power crisis in Colombia and the consequent need ment as owner and operator of the utility and there for reliable electricity. The investors were able to is no independent regulator. As such, it is very assess the buyers' creditworthiness and also could difficult for investors to assess the creditworthiness sell power at unregulated prices, thereby reducing of the utility, which may be subject to government their regulatory risks, which governments normally interference in its financial and operational man- would be expected to cover by guaranteeing the agement. In addition, where data is available, quite performance of the utilities. In a sense, Mamonal often investors discover that the utility is in poor was able to target what would be the utility's better financial condition and hence is not creditworthy. customers and hence isolate a better credit risk than In these cases the government's credit is sought in that of the utility, which serves a broader customer place of the utility's. base. If this approach were to be replicated more One might expect that, where the utility's finan- broadly, however, the result would be a fragmented cial condition is reasonably good, there would be industry, possibly with some loss of efficiency and no need for a government guarantee. This was, in flexibility in meeting total system needs. It would fact, the case for Pakistan, Jamaica, Guatemala, and also eventually relegate customers with higher credit China. However, even in these cases, the investors risks to the public utility. sought government guarantees, presumably be- cause of the lack of independent regulation and the Fuel Supply consequent uncertainties regarding the ability of the utility to meet its obligations in the future. In three of the six thermal projects examined, the Until both the utility's financial condition is ac- government assumed a substantial fuel supply risk. ceptable and the regulatory arrangements are inde- Fuel risk is a critical part of a project's risk profile, pendentandwell-established, investorswillcontinue as a reliable supply of fuel of acceptable quality is to require sovereign guarantees of purchasing utili- fundamental in order for the plant to meet its ties. Where the sovereign credit itself is weak, they supply obligation to the purchaser. In many coun- may also seek credit enhancement from third party tries fuel is imported; this adds foreign exchange guarantors. exposure to the project owner, who may have to On the other hand, in Mamonal and Pangue, source fuel from overseas markets. This can be the government did not provide a guarantee. For alleviated by sourcing locally if feasible, but there Pangue, project revenues are provided through remains reliability and quality risks with these Mobilizing Private Capitalfor the Power Sector: Experience in Asia and Latin America 27 arrangements. Where the government controls the the sample of projects in this study, investors uti- fuel supply in a country through regulation or lized three methods to overcome this risk, or at least government-owned entities, investors may seek a minimize it: government guarantee of the supplying entity. * Payment in Hard Currency: The best way to In the case of Shajiao C, the provincial utility is overcome exchange risk, from the investors' responsible for fuel supply, thereby putting the and lenders' points of view, is by arranging for province of Guandong at substantial risk for non- payment in hard currency. This effectively fulfillment of this obligation. In the Philippines, eliminates foreign exchange risk to the project, the government has pursued a policy ofcontracting although there is still a possibility that hard IPPs under an "energy conversion agreement," currency will not be available. This solution where the government, through NPC, provides the was negotiated for the Pagbilao and Mamonal land and fuel, and contracts with private investors projects. For Shajiao C, a substantial portion to "convert" the fuel into electricity. While this of the payments from the provincial utility is arrangement has proven attractive to many IPP payable in hard currency, while the rest is pay- developers, it transfers significant risks from the able in local currency. Given the large portion private sector to the government in terms of meet- of domestic funding for Shajiao C, this ar- ing fuel commitments under the conversion agree- rangement effectively eliminated the exchange ment. In Hub, the government is exposed through risk to the foreign investors (principally the its performance guarantee of PSO, a majority state- sponsor Hopewell and foreign banks). Inves- owned fuel marketingcompanythathascontracted tors in all of these projects, however, are still with the project company to supply fuel. left with the payment risk of the purchasing In the Rockfort project in Jamaica, the fuel entity, although this is no different than if the supplier is a Jamaican government-owned entity payments were made in local currency. (Petrojam refinery), However, the government did Foreign Exchange Risk Insurance: Some in- not provide a guarantee of its performance. The investors are instead relying on a fuel supply con- vestors arranged for foreign exchange risk in- tract under which they will look to enforce their surance from bilateral or multilateral agencies. rights inaccordancewithJamaicanlawifnecessary. The investors in Mamonal, for example, ob- rIghtsn acro prdancwth JPamaicand lawcif necesary. tained this insurance from the Overseas Pri- In hydro projects (Pangue and Macal River) vt netetCr. h ..ivsmn water rights, rather than fuel supply, become im- iane Investmors.inh Rockfovestment portant, and raise a different set of issues. In this surance agency. Investors in Rockfort have case, access to water rights becomes paramount, arranged inconversinilHty Insurance through which has implications regarding monopoly rights e ngesura in this case through to the water, but is not necessarily a risk-sharing scheme offered by thet Bank ofPksa issue. In the case of Pangue, the water rights had wchite ountry's entalBank of Pak.stan, been owned by Pangue's parent company Endesa for many years prior to its privatization. Indexed Local Currency Payments: Where the payment is made in local currency, inves- Foreign Exchange tors negotiated adjustment clauses in the PPAs to compensate them for currency deprecia- None of the projects studied had export sales tion. PPAs for the Hub, Rockfort and Puerto with which to earn hard currency. This results in Quetzal projects were denominated and paid foreign exchange risk to the project's foreign own- in local currency and featured adjustment ers and lenders, given the expectation that the clauses to account for local currency deprecia- revenue stream will be in local currency. There are tion. While payments are normally made on a three aspects to foreign exchange risks. These in- monthly basis, they are made on a weekly clude exposure to depreciation of the local cur- basis for Puerto Quetzal, which is designed to rency, the risk of non-availability, i.e., that there lessen the impact of tariff adjustments on the may not be foreign exchange for the investor to buy utility's cash outlays and therefore increase the in the short run, and the ability to transfer exchange probability of payment. out of the country. Risk sharing differs among these alternatives. IPP investors cover these risks in many ways, The greatest exposure to government is obviously sometimes through a combination of measures. In payment in hard currency, as the obligation to 28 Mobilizing Private Capitalfor the Power Sector: Experience in Asia and Latin America secure foreign exchange rests with the government ing is met with local currency, it alleviates a (either directly or through the utility). The least portion of the foreign exchange constraint exposure is indexation arrangements; while com- most projects face. The larger the proportion pensating the project for depreciation, they do not of funding, the less burdensome the obliga- oblige governments to actually provide foreign tions become, although it could also at the exchange. Local currency costs incurred because of same time diminish the amount of commer- depreciation are quite often recovered through cial funding in certain cases. Where long-term adjustment provisions in customers' monthly bills, commercial debt financing is constrained, which greatly facilitates this solution. The degree to however, it is difficult to say that any financ- which this arrangement is acceptable to all finan- ing is displaced, especially where none has ciers depends on many factors, especially the capac- been provided previously. In the case of multi- ity of the foreign exchange market to meet the lateral and bilateral loans channeled through needs of a large power project, and the availability governments, as in Pakistan and Jamaica, re- of back-up measures employed such as inconvert- payment from the private borrowers to the ibility insurance. government will be in local currency for which the government charges the borrowers a Government Financing premium over its cost of borrowing. * Subordination: Governments may subordi- Some projects (Shajiao C, Hub, and Rockfort) nate r thi fudn to comrillnes to also benefited from partial government funding, as provide pri lenders additional inenterst is depicted in the last two columns of Figure 6. The tofine priorty prs Subordinatiohs rationale for such funding is grounded in the desire th imant preorety projects. Subordinathon has of governments that have little or no access to long- vie covrtagefrati of senir lenders basedo term commercial debt financing to leverage their epce cash raflos f te projer as- own funds to facilitate private financing. Govern- euping cash rikws a, res pofits sordi- ment funding can facilitate private financing in nate p it he rent can charge - three ways. ~~~~~~~~nate position, the government can charge a three ways. premium on its subordinated loan. The *Lteverage: Governments Lending can leverage PSEDF loan to Hub is a subordinated facility. government resources to attract private capi- tal. For instance, a 30% government funding Notwithstanding the rationale for government contribution would be expected to mobilize funding, it is clear that the IPP market is rapidly equity of 25-30% plus additional debt for the evolving, particularly in the financing area. It is balance of 40-45% of total funding needs. reasonable to expect a diminishing role for govern- This is an especially appropriate strategy for ments in IPP financing, resulting from improve- large projects in less creditworthy countries, ments in both country creditworthiness and power where the sheer magnitude of the required fi- sector regulatory environments, as well as the priva- nancing is limited by the exposure consider- tization of key purchasing entities. Barring specific ations of banks and other lenders. Also, policy considerations with regard to IPPs, the need considering the alternative is a public sector for government funding will no doubt be greater in project, 30% government funding can achieve less creditworthy countries with very limited access a significant incremental reduction in public to private financing. Once the first one or few sector financial commitments to the power projects are successfully implemented, it is conceiv- sector. In Hub, the government is providing able that lenders will be more amenable to partici- approximately 35% of the funding through pating. Even if this scenario is realized, however, the Private Sector Energy Development Fund some form of credit enhancement will likely be (PSEDF), which is funded by multilateral and required for the high-and even medium-risk coun- bilateral agencies. In Rockfort, the govern- tries. ment is providing a take-out loan which will replace five year commercial debt raised for Summary Project Comparisons construction. This long-term debt amounts to 65% of project funding requirements. Project comparisons in the right side of Figure 5 * Lessen Foreign Exchange Burden: To the ex- reflect the outcomes examined above, after examin- tent that the repayment of government fund- ing the eight factors shown in Figure 6. Mobilizing Private Capitalfor the Power Sector: Experience in Asia and Latin America 29 Contrary to the expected alignment set out in Mamonal veers from its expected relationship Figure 5, manyofthe projects did notexhibit aclear primarily on account of the nature of the project, relationship between country risk and government which is designed to service a small group of private support. In some cases, government support was industrial customers rather than a broader client beyond what might be expected. These are termed base through a public utility. In this sense, it is by "over-guaranteed" in Figure 5-right, and include definition isolated from some country risks and Shajiao C and to a lesser extent Pagbilao. most sector risks, and therefore it may be somewhat Some projects were "under-guaranteed" (such as of an overstatement to term Mamonal "under- Puerto Quetzal and Macal River, and to a lesser guaranteed." Thus, the projects' placement in Fig- extent Rockfort). The Mamonal project in Colom- ure 5 is derived more from its enclave structure bia is shown as very under-guaranteed based on the rather than risks sharing between government and fact that there is no government exposure in this the private investors, as in the other projects. project. On the other hand, some projects did Forotherprojectsthatareeithersomewhatover- exhibit the expected relationship as outlined in the or under-guaranteed, the deciding factor was fuel, left side of Figure 5; these include the Hub and foreignexchangeorfinancing,thoughtheseprojects Pangue. are close to expected levels of government support. This figure reveals the range of experience to Puerto Quetzal and Macal River had little or no date with IPPs, in terms of project structures re- government funding (8% for Puerto Quetzal), and quired to secure financing in varying circumstances fuel supply risks for the former were assumed by in different countries. It is important to realize that private investors and lenders. As a hydro project, these relationships are determined based on a vari- Macal River had no fuel supply risks. Though the ety of factors and that each divergence shown in government is providing funding to the Rockfort Figure 5 can, however, be accounted for by project- project, it is structured as a take-out loan, replacing specific factors or particular policy factors. commercial loans secured by the project company For instance, the Chinese government's promo- for the first five years of the project from financial tion of joint ventures with foreign private investors closure. In addition, fuel supply risks are solely the for new IPPs, instead of accommodating projects responsibility of the sponsor, and payment from with only private shareholders, partly accounts for the utility is made in local currency. The govern- the significant extent of government involvement ment does not undertake to guarantee foreign in the Shajiao C project. Other factors influencing exchange. the Shajiao C project's "over-guaranteed" categori- Pagbilao was categorized as somewhat over- zation include guarantees of payment in foreign guaranteed primarily based on the government's exchange and provision of fuel by the government obligations for fuel delivery and providing foreign utility. exchange, which are significant risks. Section VI Future Developments Several observations on the future development transmission and distribution resulting from the of private power are summarized below. recent privatizations of many electric utilities, fi- nancing of generation using IPP structures will Supply vs. Demandfor Capital undoubtedly remain a major market for private investment. From the investor's point of view, this The global gap between the supply ofcapital and stems partly from the desire of many IPP sponsors the demand for power is large and cannot be to undertake projects on a non-recourse basis. expected to diminish significantly through IPPs Utilities can also benefit from the potentially sig- alone in the near term despite the recent upsurge in nificant amounts of financing that can be secured IPP financings. Developing countries will need to in one transaction forlarge segments oftheirexpan- tap many sources of public and private finance to sion needs using this approach. fund the expected annual additional generation While transferring the IPP model across coun- requirements of about 40-60 GW. The capacity tries is not as straightforward as might be expected thatdoesgetfinancedwillnodoubtbemetthrough at first glance, the experience with basic project a combination of publicly-owned utilities through structures and contracts has to a large extent been traditional funding sources, newly-privatized utili- transferred from one country to the next. In some ties borrowing as corporates in international and cases, this has been facilitated by the involvement of domestic markets, and IPPs. multilaterals and bilaterals, who bring interna- Currently, no more than 2-4GW of new private tional experience. These institutions also play an generation capacity, through IPPs and cogenera- important role in advising governments on policy tion, is being brought on line each year. While this matters which has been helpful in the transfer of is only a fraction of total demand, it is important to experience. Also, by supporting external advisors remember that India and China comprise a large through technical assistance, multilaterals and share of these needs. Thus, for some smaller coun- bilaterals have a very useful role in helping to lessen tries IPPs may meet a substantial share of their the development time of IPPs by enhancing the future requirements. The case analyses of Guate- efficiency of the negotiation process. A well-in- mala, Chile, Jamaica and Belize demonstrate this formed government will generally prove a more possibility. This may also be the case if one consid- responsive and flexible negotiator than an unin- ers regional requirements at the state or provincial formed one. level in the larger countries. Perhaps more importantly, within a particular country, the establishment of a policy and institu- IPP Approach and Replicability tional framework and/or the development of the first few projects may facilitate a more rapid devel- This study has focused on private capital mobi- opment of follow-on projects. The experiences of lization for the expansion of generation facilities. the Philippines, Pakistan, and Jamaica demon- While opportunities exist for private financing of strate this possibility (see page 11). Mobilizing Private Capitalfor the Power Sector: Experience in Asia and Latin America 31 As the power sector becomes more diversified from IPPs multiply. While it is recognized that over time in many countries, a sound regulatory access to longer-term finance would help assuage framework will enhance the financing possibilities this pressure on retail tariffs, more empirical inves- not only for IPPs, but also for privatized utilities tigation ofthe precise link between financing terms looking to finance rehabilitation of existing assets, and power pricing is required. as well as capacity expansion. Utility financing on Although all of the projects studied achieved balance sheet would generally add flexibility to much longer maturities in general than had been their financing strategies bybroadening their sources realized before in each country, the impact of a and types of financing. project's debt service commitments on tariff levels For the long-term, the implications of integrat- would be further moderated with longer maturi- ing IPPs into more diversified power sectors must ties. This highlights the need to establish new be considered, including the link between commer- financing mechanisms and to tap new sources of cial arrangements (i.e., long-term sales contracts), and funding to stretch out maturities beyond what has dispatch practices, and the potential role of IPPs in been achieved to date. Raising this funding in meeting the intermediate load of a system. An domestic markets would not only have the benefit investigation of the experience in developed coun- of aiding the development of the local capital try markets may provide insight in this area. market, but would also alleviate some ofthe foreign exchange burden imposed by large repayment com- Financing Structures and Sources mitments to foreign banks. While the private debt financing mobilized for Needfor Credit Enhancement the projects examined in this study primarily in- volved commercial bank lending, some sponsors Government guarantees and credit enhance- are beginning to look seriously at the broader ment from bilateral (e.g., ECA) and multilateral capital markets to fund IPPs. Enron's US$105 institutions will continue to play an important role million financing in the U.S. capital market of the in IPP financing during the transitional phase from Subic Bay project in the Philippines may set a state-dominance to a more market-oriented sys- precedent. Funding for U.S. IPPs in the capital tem. Governments generally prefer to minimize market on a corporate or project finance basis is their exposure to political risks only, but the precise well-established, though it is still not common to scope of political risks may often be in question, raise construction financing in this way. More particularly from the point of view of investors and often, commercial banks provide construction fi- lenders. nancing, but are taken out by long-term institu- As a policy matter, the provision of limited tional investors at completion. The reason for this government guarantees in certain critical areas (e.g., is that banks are generally better able to assess and utility performance, fuel supply, access to foreign analyze the risks involved in construction projects exchange) may be the best option in the short run than other providers of debt finance and also are to bring on line new IPP investments, while efforts not eager to tie up their capital for long periods of are made to change the policy and institutional time. environment in the power sector. Assuming public Developing countries pose country risk barriers sector financing options are either not feasible or to tapping these markets. But with the large amount favored, this is especially applicable when new of long-term financing available in the capital mar- capacity needs are pressing and not meeting these ket, this financing avenue will likely be critical to needs will seriously impair future economic growth meeting the large demands for infrastructure fi- prospects. nancing in emerging markets. A transparent and well-functioning regulatory environment and creditworthy power purchasing Power Pricing and Sector Development utilities will greatly enhance the mobilization of IPP funding in the long-term and eventually lessen Perhaps the most important issue for developing the need for government support. This environ- countries is the price they pay for private power. ment will no doubt need to be sustained for many Passing through the full cost of generation is al- years in order to be considered "tested" by private ready difficult in many countries, and this situation investors-only then will the need for government may be exacerbated as commitments to purchase involvement diminish. Note on Source Material ,I r g i,; i; , * E lIi ii ii . iiii