United States: Debt Finance For Infrastructure Projects

According to the World Bank, developing countries currently spend about $200 billion a year on infrastructure development. ¹ Although primarily financed by governments to date, this level of government spending places an enormous burden on public finances and is probably unsustainable over the long run, especially in Asia where infrastructure investment is expected to rise from 4% to 7% of gross domestic product by the year 2000. The Asian Development Bank has recently projected over $1 trillion of infrastructure investment in Asia by 2004. The scope of this projected new infrastructure investment is awesome. For example, China wants to install 8 million telephone lines a year after 1995. This in itself is an impressive goal, until one notes that in 1992 China only had a total of 18 million installed lines. ² In addition, a common problem with existing infrastructure investments in developing countries is inferior performance when compared with comparable investments in developed countries. Until recently, most infrastructure has been financed by the public sector through budget allocations and, as often as not, by sovereign borrowings. Often the results of such public spending have been quite unsatisfactory with clear evidence of corruption, inefficiency and substantial waste. On the other hand, infrastructure investments can deliver major benefits in economic growth. ³ Moreover, failure to make necessary investments in infrastructure can impair or derail a country's economic growth and development.

Private infrastructure investment (as opposed to public investment) is growing rapidly throughout the world, but especially in the developing countries and emerging markets in Asia, Latin America and the former Eastern bloc. According to one recent survey, private investment in infrastructure outside the United States doubled from 1993 to 1994 and the survey reported over 900 projects amounting to almost $680 billion that have been constructed or are under development in 72 countries. ⁴ This survey noted the importance but limited availability of commercial bank financing for these projects and that obtaining long-term institutional debt and equity financing in sufficient amounts remained a challenge even for financially sound, well-structured projects. ⁵ This survey also noted that the large construction companies increasingly dominate infrastructure development, but are funding this development on their own balance sheets. Another recent survey of project sponsors echoed this concern and also complained of a lack of creativity and sophistication among project advisors and lenders. ⁶ These and other similar surveys provide clear evidence of a need for an exchange of information and education in this area.

Financing for infrastructure is available from several traditional sources: government funding (consisting of grants, loans and credit enhancement) and direct investment, suppliers (with respect to construction financing), multi- and bi-lateral agency funding, credit facilities provided by commercial banks or other financial institutions (provided in reliance upon the credit of the corporate sponsor of the particular infrastructure project (whether contractor, owner or operator) or sovereign sponsor of the infrastructure project), capital markets financing based on such credit (in the form of U.S. public offerings, municipal bond offerings, Eurobond offerings and private placements), project financing (dependent solely or principally on the assets and cash flow of the project in question) and securitization (consisting of the transfer of the right to receive the revenue from the project, often under a government-granted concession, to a special purpose vehicle, and the contemporaneous issuance of securities that are payable from revenues generated by the project).

Financing which is based upon the credit of a private sector or sovereign project sponsor is limited in availability, and is also subject to changes in the perception of the creditworthiness of the sponsor, local market considerations and developments unrelated to any particular project that may adversely influence the financial condition or business of the sponsor. In addition, private sponsors of international infrastructure projects generally seek to avoid incurrence of indebtedness that permits creditors to have recourse to the sponsor. In fact, limitation of on-balance sheet indebtedness or recourse (they are not necessarily the same thing) is one of the principal objectives of most project sponsors.

By contrast, project financing is structured so as to be reliant only (or principally) upon a specific project, and the revenues anticipated from an identified source (such as tariffs, tolls, take-or-pay contracts or power purchase agreements) under contractual arrangements negotiated as part of the financing. Thus, the risks arising in this type of financing are limited primarily to those associated with the particular project, and the industry (or segment of the economy) served by the project. Provided the relevant industry (or segment of the economy) performs as anticipated, repayment of the financing should be isolated from most factors (other than currency devaluation) which might otherwise adversely affect a more traditional cross-border, credit-based financing.

Project finance has been employed in almost all capital intensive industries and particularly in transportation (aircraft, rail and shipping), in mineral (including oil and gas) and other natural resource exploration and development and, most recently, in power generation. Project finance has also commonly been used in countries whose capital markets are small relative to their project finance requirements (e.g., Australia and Canada) or are relatively immature (e.g., certain developing countries). Project finance allows the project assets to be separated from the sponsor and to be financed upon the cash flow from the project assets. It allows a sponsor to undertake a project with greater risk than that for which the sponsor is willing to accept full recourse. The typical project financing is structured to provide lenders with only limited recourse to the sponsor, which relieves the sponsor of a general contractual obligation to repay the infrastructure financing. Under this approach, individual projects or select groups of projects can be financed separately, to minimize the effect of the financial condition and profitability of the sponsor on the availability of its financial resources, and to limit the risk to the sponsor from the financial problems of a particular project. Accordingly, through reliance upon project finance techniques, infrastructure can be financed through limited recourse securities which, if structured correctly, carry manageable (and therefore marketable) risks.

Moreover, many have noted that project finance, by unbundling and encapsulating the risks comprised in international infrastructure projects⁷, allows project participants to assume those risks that they can best control and, at least in theory, to leverage multi- and bi-lateral agencies' participation into a larger number of total supported projects. ⁸

Since World War II, public international organizations have assisted economic development in the emerging markets, particularly the World Bank (or, as it is officially known, the International Bank for Reconstruction and Development or IBRD). One of the principal problems to be addressed after the war was the need for investment capital to rebuild war-torn economies and for economic development in less developed countries. As Lord Keynes stated at the Bretton Woods Conference (at which it was determined to form the World Bank):

. . . only a few of the member countries will be in possession of an investable surplus available for overseas loans on a large scale, especially in the years immediately following the war. It is in the nature of the case that the bulk of the lending can only come from a small group of member countries, and mainly from the United States. How then can the other member countries play their proper part and make their appropriate contribution to the common purpose? ⁹

This challenge still remains today.

Infrastructure projects represented more than 25% of the $2.9 billion in financing approved by the International Finance Corporation (the private sector member of the World Bank group, "IFC") in 1995, which was the fourth consecutive year of more than 10% growth in financing volume. The total value of all projects supported by the IFC in 1995 was $19.5 billion, a record for the IFC. ¹⁰

Arranging cross-border infrastructure financing requires that the project participants assume certain risks, in addition to those common to infrastructure projects, including:

Currency risk with respect to an international infrastructure financing has three components: the risk arising from incurring liabilities to fund a project denominated in a currency (typically U.S. dollars) different from the currency of project revenues and the risk of inconvertibility of, or inability to repatriate, currency arising from exchange controls or other currency restrictions. Investors in financings for foreign borrowers must consider and evaluate the risks associated with foreign currencies, including the possibilities of devaluations, exchange requirements and repatriation restrictions.

Any devaluation of the local currency in which project revenue is denominated will require a corresponding increase in the project's cash flow to assure that economic returns are maintained, and to service the project's debt (especially if denominated in another currency). Any significant devaluation of a local currency, without offsetting factors, will directly impair the project's likelihood of repaying its lenders and other investors.

Additionally, exchange controls could result in international investors not being able to receive payments in specified currencies on a timely basis. If only a limited supply of a specified currency is available, or if new approvals or extra fees are required to effect conversion to the currency of payment, the adequacy of project cash flow could be impaired, or investors could have to settle for payment in another currency. Repatriation restrictions could prohibit or limit the ability of lenders to move payments received, or of investors to repatriate earnings, out of the country in which the project is located.

Loans and other investments dependent on revenue streams from foreign projects are subject to risks specific to the country involved, causing investors to examine differences in law and politics from their own nation. Country-specific risk must be taken into account, including (i) the political stability of the applicable governmental units (central and local), (ii) the attitude (historical and current) of the government towards foreign investment, in particular regarding currency exchange, repatriation of earnings, taxation, privatization and the need for infrastructure development, and (iii) the degree of involvement of the government in the economy and industry segment served by the project.

Furthermore, investors must also take into consideration the risks of unanticipated developments that might adversely affect the project, or the industry segment served by the project. These risks include the risk that underlying contracts will be changed or impaired, the risk of expropriation or confiscation of project assets (both outright and creeping), and risks of war and civil disturbance. Investors must also consider the likelihood, and probable effect, of any change in political regime, and whether policies and attitudes toward the particular project, towards international financing in general or towards the sponsors in particular will change. ¹¹

Beyond the impact all government policies and attitudes may have on an economy, an industry segment or a project, tax policies may have an important effect upon a project and therefore warrant special consideration. Current and expected income, assets or property, operational, stamp, mortgage, withholding and other revenue and financing-related taxes need to be considered carefully by investors in evaluating and structuring infrastructure financings. Significant increases in such taxes or cancellation, modification or termination of favorable tax treatment of the project entity, the project or the sponsors could have an adverse effect upon the availability of cash flow from an infrastructure project, thereby diminishing the revenues available for the repayment of the project's debt.

In addition, local withholding laws, and the existence of international tax treaties applicable to foreign investors, will often have an important effect upon the availability and amount of project revenues for debt service. The termination of a favorable tax treaty or the implementation of new withholding and other taxes will generally require the issuer to make "gross-up" payments to lenders and other investors, thereby further diminishing cash available for debt service.

Aside from these cross-border risks, infrastructure financing involves the following project risks:

Insulation from the general credit risk of the private or sovereign sponsor is one of the benefits of infrastructure project financing. In addition, infrastructure involves basic products and services needed or required by the public. However, such financing is also insulated from (and deprived of the benefit of) any diversity in activities of the sponsor which may cushion an economic downturn affecting the project. Consequently, a financing which is wholly dependent on the demand for a particular product or service from a particular facility (such as a single power plant with a single power purchaser, or a toll road with only a single destination) is directly subject to the economy of the country in general, and to the specific industry segment of the economy served by the project. Historical use, projected trends and economic expectations are therefore very important considerations to all investors and other project participants.

Project finance involves limited recourse to sponsors and other non-project assets. Investors must recognize that project financing of infrastructure is subject to additional restrictions in the exercise of traditional remedies. For example, foreclosure on a power plant with a sole power purchaser (such as a state-owned power company) is of little use if the problem is an inability of the power purchaser to make the required payments. If there are no other significant purchasers of power available in the relevant market, the power plant, to the extent physically and legally removable, may have only salvage value to lenders.

Similarly, where the financing is structured as a securitization of concession rights (which, unlike the related cash flows, are generally non-transferable), upon the occurrence of an event of default there may be no asset on which to foreclose (as a practical matter, toll roads and bridges cannot be moved).

Since there is often incomplete data or in some cases no reliable data concerning project revenues prior to completion, lending for construction of a project is properly regarded as a relatively high-risk proposition. The ability to finance any project, particularly in the capital markets, is therefore directly tied to its stage of development and the availability of operational data. A project that is complete and operating is considerably easier to finance than a project subject to construction risk, completion risk, technology risk or performance risk. A project with an operating history is easier to finance than a project subject to unquantified operating risk. Historical and pro forma operating data, and a careful analysis thereof by an independent expert, are essential.

Project financing is usually secured by the cash flows (including, in the case of financing a concessioned project, the right to receive revenues under the concession), the underlying contracts and the project assets themselves. In evaluating any limited recourse financing, secured creditors always want to know the likelihood and cost of being able to foreclose upon, and realize the benefit from, the assets which secure repayment of their financing. The ability to perfect a security interest and, more importantly, to be able to transfer the asset at the time of foreclosure to reduce or repay in full the outstanding indebtedness are highly important considerations to lenders, particularly in the context of project finance. In considering the transferability of project assets, project lenders focus on the following issues:

• Are the licenses, permits and concessions transferable to someone who can finish the project and operate it?
• Are hard assets transferable or removable?
• Is there a fully effective and legally enforceable assignment of all rights to future cash flow?
• Are the underlying operational contracts transferable for performance by someone else?

At the risk of oversimplification, the current universe of debt instruments issued to finance infrastructure comprises or is provided by the following:

1. Commercial Banks
2. Private Placements
3. Rule 144A/Public Offerings
4. ECA/RDB/IBRD Financing
5. Municipal Finance

Most private infrastructure developments (not yet projects) will be financed on a sponsor's balance sheet (although the amount to be so financed will usually be relatively small). Construction companies will also finance retainage, warranty liability and any residual interest in the project on their balance sheets. Sponsors of private infrastructure projects will normally seek to obtain financing at the project company level because of the ability to obtain better ratings, and thus better financing terms. However, a number of independent power companies (e.g., AES, Cogentrix, Kenetech and Mission Energy) have obtained financing through public offerings. These offerings are a hybrid of corporate/project finance and the indentures governing such offerings provide significant restrictions on the issuer's ability to incur indebtedness at the project level because of the "seniority" of such indebtedness over the issuer's indebtedness.

Commercial banks have always had an active role in project finance transactions. ¹² In fact, project finance is generally thought to have begun in the 1930s when a Dallas bank made a non-recourse loan to develop an oil and gas property and to have "come of age" in the 1970s and 80s with the successful project financing of the North Sea oil and gas, Australia's Northwest Shelf gas, independent non-utility power generation in the United States, and similar substantial projects.

Commercial banks have successfully provided project financing because of their demonstrated ability to evaluate complex project financing transactions and to assess and assume the construction and performance risks usually involved in such financings. However, principally due to the short-term nature of a commercial bank's liabilities (i.e., its deposits), commercial banks usually limit in amount and otherwise closely monitor and control their project finance exposure.

The project's sponsor will normally request commitments from its commercial banks for both construction financing and, following completion, the permanent, long-term financing of its project. Typically, the commitment for construction financing will be for 2 to 3 years and, following completion and demonstration of operational performance, for permanent financing will be from 10 to 15 years, although in rare cases commercial banks have provided permanent financing commitments of 20 years or more. Most permanent financing commitments by commercial banks will include specified increases in the applicable interest rate ("step-ups" in the applicable margin or spread over the commercial bank's cost of funds) in an effort to create escalating incentives for the commercial bank financing to be refinanced before its scheduled maturity.

The successful commercial bank syndicate for a project financing will usually seek to "sell down" its underwritten commitments in a further coordinated syndication to a larger bank group. This subsequent syndication may occur before financial closing (i.e., the execution and delivery of definitive financing documents) or afterwards, depending upon the confidence of the original underwriting banks in the "marketability" of their transaction in the bank project finance markets (and their willingness to assume the risk of adverse change in such markets), the timing constraints of the project, the project sponsor's preferences in this regard or the original underwriting banks' desire to reduce their level of commitments or all or any combination thereof. The project's construction financing, which will normally bear interest at a floating rate, will usually require interest rate risk to be hedged through an interest rate swap, cap or collar (although, if such hedging is under swaps and collars, because payments thereunder may be due from the project and thus the swap or collar providers may become creditors of the project, project collateral will have to be shared with such providers). Even though the commercial banks provide a permanent, long-term financing commitment, upon completion of construction and demonstration of the project's acceptable performance, most sponsors will seek to refinance the project with permanent, long-term and fixed rate financing. This refinancing will usually be on terms that allow the project more operational flexibility because construction risk has been eliminated from the project and because obtaining waivers from institutional holders is more difficult.

This traditional model has proven very successful over a considerable period of time and in a wide variety of industries and specific applications. It has provided and will continue to provide substantial capital to qualifying projects throughout the world.

Private placements have also played a prominent role in infrastructure finance. This market has demonstrated strong tolerance of "story credit" and transaction complexity, lacks registration requirements and offers less maturity sensitivity than commercial bank financing (because of the longer-term nature of an insurance company's assets and the reduced asset/liability mismatch). Project finance is a key element of recent and dramatic growth in the private placement market. However, insurance companies (the private placement market's most active buyers) face recently imposed risk-based capital requirements and there is a demonstrable "flight to quality" in this market which has made it much more difficult to obtain financing (or at least more expensive to obtain financing) for a marginal project. While an explanation of the National Association of Insurance Commissioners' ("NAIC") rating system is beyond the scope of this paper, the market "break" seems to be at NAIC 2 (the equivalent of S&P's "BBB" rating), below which there are significantly fewer available investors.

Notably, both Standard & Poor's Rating Agency Group ("S&P") and Duff & Phelps Credit Rating Co. ("D&P") have announced private placement rating systems which are more lenient than their respective public ratings (insofar as the private placement ratings express views about ultimate payment and unlike public ratings do not place the same stress on timeliness of payment).

Approximately $4 trillion of the U.S. capital markets is comprised of public and corporate pension assets, which usually seek high-quality, long-term investments. The U.S. capital markets are therefore naturally attractive to infrastructure projects which seek long-term financing. ¹³

The attraction of the U.S. capital markets for infrastructure financing has been enhanced considerably by the adoption, in 1990, by the Securities and Exchange Commission (the "SEC") of Rule 144A ("Rule 144A") under the Securities Act of 1933, as amended (the "Securities Act"). As discussed below, Rule 144A established a non-exclusive exemption from the registration requirements of the Securities Act for resales by investors to eligible institutions of privately placed securities, subject to certain limitations. By facilitating resales of securities issued in private placements, Rule 144A has created liquidity in the secondary market for privately placed securities, causing the U.S. private placement market to become more attractive to foreign issuers, including those offering debt securities to finance infrastructure projects.

With the adoption of Rule 144A, a broader market for private financing in the U.S. capital markets has become available, especially for foreign issuers, characterized by greater liquidity than existed prior to adoption of the Rule and ease-of-entry. Prior to the adoption of Rule 144A, privately placed securities were subject to significant restrictions upon resale under applicable U.S. securities rules, rendering such securities more illiquid.

Rule 144A provides a non-exclusive safe harbor from the registration requirements of the Securities Act for resales of certain se4curities to "qualified institutional buyers" ("QIBs")¹⁴. The expressed intention of the SEC in adopting Rule 144A was to increase the liquidity of privately placed securities by allowing unrestricted resales of such securities among QIBs, and to increase access to the U.S. capital markets by foreign issuers.

In a Rule 144A placement, an issuer will sell, in a traditional private placement (in reliance upon an exemption from the registration requirements¹⁵), its securities to one or more investment banking firms which will then resell, in reliance upon Rule 144A, the securities to a larger number of QIBs. From an issuer's standpoint, the way in which Rule 144A placements are conducted is very similar to traditional underwritten public offerings, especially in that securities are offered by the investment banking firm to potential investors on a "take it or leave it" basis. Instead of direct negotiation with investors, issuers rely upon the bank's perception of the marketability of the issuer's securities.

A sale in compliance with Rule 144A must meet four basic criteria not discussed in detail in this paper: (1) the securities must be offered and sold only to QIBs; (2) the securities must not, when issued, be of the same class as securities listed on a U.S. securities exchange or quoted in a U.S. automated interdealer quotation system (such as the NASDAQ System); (3) the seller and the prospective purchaser must have the right to obtain certain information about the issuer if such information is not publicly available¹⁶; and (4) the seller must ensure that the prospective purchaser knows that the seller may rely on Rule 144A.

Since 1990, in response to the adoption of Rule 144A, a large number of foreign and domestic companies have issued securities in the U.S. private placement market in reliance upon the resale exemption afforded by Rule 144A and, as expected, there has developed a liquid secondary market for these securities among QIBs. The creation of this market and the growing appetite among U.S. institutional investors for high-yield securities (which, by definition, includes those issued by private or public sector issuers in the emerging markets) have made it possible to structure a capital markets financing for infrastructure projects targeted to the U.S. private placement market¹⁷.

Infrastructure financing using Rule 144A or public offerings seemed to offer the best of both worlds -- access to the deepest segment of the U.S. capital markets (because of its liquidity and required minimum credit quality) and availability of long-term, fixed rate debt capital. The latter is especially important to infrastructure projects because shortening the maturity of financing can dramatically increase debt service requirements which will drive infrastructure tariffs to higher levels. For this reason, the long-lived assets which characterize infrastructure investments ideally should be financed with comparable debt maturities. But, this market generally requires an investment grade rating. See "Investment grade" discussion below. The use of a U.S. capital market financing for an infrastructure project is a "major development of the 1990s." ¹⁸While S&P believes that some European and Asian issuers will obtain an investment grade rating, this will depend significantly on the transaction's structure and, in particular, on how currency and political risks are handled.

One of the most interesting recent developments in infrastructure finance is the growth in export credit agency ("ECA") and regional development bank ("RDB") activity, as well as significant growth in infrastructure finance activity at the World Bank and International Finance Corporation ("IFC"). ¹⁹ This growth is evident in the "reengineered" Export-Import Bank of the United States ("US ExIm"), which has formed a Project Finance unit and has been very proactive in educating potential users about its services and in streamlining its application requirements and approval procedures. A similar proactivity has been exhibited by the Overseas Private Investment Corporation ("OPIC"), which has likewise established a special Project Finance unit, increased its maximum project loan amounts and has both sponsored and made significant investments in several country-specific funds. ²⁰

IFC has created and expanded its infrastructure department, sponsored and invested in several infrastructure funds and increased its cofinancing (the so-called "B loan") program. ²¹In a laudable transaction this year, the IFC Latin America and Asia Loan Trust 1995-A (some 2 years in gestation) sold interests in a pool of 73 LIBOR-indexed, dollar-denominated loans to non-governmental borrowers in Argentina, Brazil, Colombia, Chile, Mexico, Venezuela, China, India, the Philippines and Thailand. ²² This transaction introduces the potential benefits of secondary liquidity and diversification to investors. ²³ One can only hope that this transaction will provide the same stimulus to the capital markets and create the same secondary liquidity for similar instruments as did the groundbreaking RTC transactions for mortgage-backed securities.

More controversially, the IFC has sought and obtained several significant financial advisory engagements and syndicated many of its B loans.

The World Bank has attracted a lot of recent criticism and calls for reform. ²⁴Most of this criticism concerns the delay in obtaining World Bank financing, conceded by the World Bank to be over 2 years ²⁵ (although, in the case of the Hub River project in Pakistan, it was 6 years), and inflexible rules that limit its guarantees to less than 5% of its lending volume and prevent guarantees from being extended to low-income countries that qualify for its concessioned loans. ²⁶

Similar criticism has been made about the ECAs and RDBs, although several ECAs have implemented reforms aimed at expediting approvals increasing the use of guarantees and other credit enhancements to increase the total value of supported projects. ²⁷ US ExIm has admitted that it subsidizes its insurance premiums (although it claims to do so only to the same extent as other ECAs). ²⁸The OECD arrangement or consensus, to which US ExIm and most other ECAs subscribe, provides only general rules and guidelines which result in substantial differences between ECAs.

While the participation of multi- and bi-lateral institutions in infrastructure finance is necessary at least in the near term, it inherently distorts free competition to provide capital to infrastructure projects and may significantly influence whether and, if so, which projects will be financed.

The municipal bond market is the capital formation market used by American states, cities, counties and other units of local government to finance necessary governmental infrastructure improvements. The municipal bond market owes its separate existence not so much to the fact that it raises capital for units of government, but rather because interest on municipal bonds and notes is generally exempt from taxation by the United States federal government and (usually, but not always) from taxes levied by state and local governments. The tax-exempt nature of municipal bonds permits American local governments to secure financing at lower borrowing costs than can be obtained by corporations of comparable credit quality. Because the municipal bond market is based upon the exemption of interest payable on its bonds and notes from federal income taxation, it is extremely sensitive to changes in United States federal tax policy. A reduction in the maximum marginal United States federal income tax rate, for example, makes tax-exempt income less valuable to investors in municipal bonds and, accordingly, can drive up the interest rates at which state and local governments may borrow money. Legislation enacted by the United States Congress can expand or contract the ability of state and local governments to issue tax-exempt bonds. The Tax Reform Act of 1986, which made sweeping changes in the Internal Revenue Code, including the enactment of numerous restrictions on the right of municipalities to issue tax-exempt debt securities, is the most recent and perhaps the most dramatic example of this. Today, the municipal bond market once again finds itself roiled by the prospects of a federal "flat" income tax.

Nevertheless, while somewhat tax sensitive and relatively unique to the U.S., municipal finance has clearly demonstrated its value in providing municipalities with flexible funding for infrastructure investments. Municipal finance has proven its value over time with its innovative instruments and solutions to inevitable political changes.

The U.S. municipal bond market is characterized by a wide variety of public bodies that issue debt in that market. Large states and cities, such as the states of Michigan and Ohio or the cities of New York and Chicago, are frequent issuers of municipal bonds, but rural villages with only a few hundred inhabitants also raise money through the municipal bond market. In addition, a series of bonds issued by a municipality can have very different security provisions from other series of bonds issued by the same municipality. The traditional "general obligation" bond or note is supported by the municipality's pledge of its "full faith and credit," including a levy of taxes for the specific purpose of repaying the bonds or notes. However, the same municipality could also issue bonds supported solely by revenues of the municipality's water or electric utility system, bonds payable strictly from assessments charged to those property owners specifically benefited by the improvements being financed, bonds supported only by the municipality's promise or "moral obligation" to repay, or industrial revenue bonds ("IRBs"), the proceeds of which are loaned to a private corporation and, accordingly, are payable only from payments made by the private corporation.

The complexity of the municipal bond market is increased by the fact that the market, at this time, is largely unregulated. The SEC has, from time to time, made efforts to improve the quality of disclosure in offerings of municipal securities and, more recently, to compel municipal issuers to make some continuing disclosure about themselves. In sharp contrast to private companies that publicly sell their securities in the United States, municipal issuers are generally not required to register their securities with the SEC or to observe the financial and other information reporting requirements required by the United States federal securities laws. In addition, again in sharp contrast to the corporate market, there is no uniform standard of accounting principles or rules applicable to the financial statements of municipal issuers. Prospectuses (usually called "official statements" by market participants) are commonly prepared and used in the original distribution of municipal bonds and notes, but the quality of disclosure can vary widely depending upon the size of the financing and the sophistication of the participants.

The rating agencies play an important role in setting standards and bringing order to this unregulated market. Securing a rating is an important part in the marketing of most municipal bond issues. In order to issue a rating for a municipality's bonds, the rating agencies require the municipality to submit certain information about its financial condition, operations and future prospects. Once a rating is issued, the municipality is supposed to regularly update the information originally submitted to the rating agency. Ratings are, in fact, raised and lowered and, on occasion, withdrawn, but the system by which municipal bond ratings are issued and maintained cannot be compared to the process by which a continual stream of financial information is publicly disclosed to the U.S. corporate equity and debt credit markets.

Last, but not least, the municipal bond market is well supported by an extensive and sophisticated financial infrastructure. Municipal bonds are underwritten and placed not only by Wall Street, but by regional investment banking firms and money-center, regional and local commercial banks as well. This financial infrastructure did not spring up overnight, but, in a sense, its development was inevitable since only United States-based taxpayers have any appetite for investments that generate United States federal tax-exempt income.

In 1812, the City of New York issued the first bonds secured by a property tax. Since that time, general obligation or "G.O." bonds have been the workhorse of American municipal finance. In a G.O. bond financing, the state or local government pledges its full faith and credit to repay the bond. Except for tax revenue streams specifically dedicated by state statute or other bond financing ordinances to the repayment of other obligations, the state or local government's pledge obligates all of its financial resources and taxing powers. If the issuer of the G.O. bond is a city, county or other local government, this pledge usually includes the contractually enforceable promise to levy an unlimited ad valorem property tax. Community residents (who are also citizens and voters!) are naturally concerned about the ability of their local city council or other local legislative body to continually increase property taxes. As a result, most state laws require that once property taxes reach a certain level the local government must seek the approval for an issue of G.O. bonds by a referendum vote of the local citizens. State governments, which typically have no right to levy ad valorem property taxes, would pledge a significant unrestricted revenue stream such as income or sales taxes to support their G.O. bonds. Although state income and sales taxes are usually not subject to a referendum process, the decision of a state governor to seek, and of a state legislature to enact, income or sales tax increases is anything but a "pure" financial decision and is bound to be influenced by the political process.

Because property, income and sales taxes are relatively constant and predictable revenue sources, general obligation bonds are well regarded by investors. Nevertheless, there are clear differences in credit quality within the universe of G.O. bonds. The capacity of a particular state or local government to repay its G.O. debt is analyzed by examining the local economy, the financial performance and flexibility of the issuing state or local government, the other outstanding indebtedness and debt burdens of the issuing state or local government and the quality of the state or local government's administration. Of these four factors, the evaluation of the local economic base is by far the most critical.

As noted above, G.O. bonds are the mainstay of U.S. municipal finance. However, over the last two decades the dominance of G.O. bond financings relative to other municipal finance structures has shifted. The reasons for this shift are significant because they highlight the increasingly politicized nature of municipal finance. The attractiveness of financing capital projects through user fees rather than broad-based taxes, legal limitations on the issuance of G.O. debt by an angry electorate (e.g., "Proposition 13" in California), the practical capacity of even large units of local government to incur and service ever larger debt burdens and continuing market innovations are the reasons most frequently cited for this shift.

Revenue bonds are typically secured by and repayable from a user fee or particular dedicated tax rather than the full faith and credit and taxing power of the public body. Bonds issued by a city to finance improvements to its municipally-owned and operated water system where the source of repayment and security is limited solely to a pledge of the fees and charges paid by residents for the use of the water is a common example of a revenue bond. An issue of bonds to finance improvements to highways or roadways secured by a dedicated fuel tax (a much narrower tax than a state income or sales tax) is a second example. Revenue bonds have their origin in the development, at the end of the 19th century and during the early part of this century, of municipally owned and operated electric, water and sewage systems. Over time, their use has expanded to finance a wide variety of municipal facilities (e.g., airports, toll roads and convention facilities) where the facility has the natural capacity to generate fees and revenues. This financing technique has gained in popularity in an era where taxpayers (and especially local property taxpayers) are dissatisfied with the "tax and spend" practices of government at all levels. It is often easier, from a political standpoint, for a local government to authorize an increase in user fees than present the general community with a new or higher tax.

In evaluating the revenue bond, it is useful to contrast it with the general obligation bond. The first and most obvious difference is the limited source of repayment of revenue bonds. Situations in which the revenue bonds of a municipally run utility are actually rated higher than that municipality's G.O. bonds indeed exist, but are few and far between. In a revenue bond financing, investors focus on the dedicated source of repayment and structure the financing accordingly to guard against the risk of impairment of the revenue source. A second characteristic that sets revenue bonds apart from G.O. bonds is that competition between providers of the same service is more likely to be an investment consideration in a revenue bond financing. Toll roads and airports are examples of facilities that both compete with other public and private entities and are well-suited to a revenue bond financing structure. Third, revenue bonds typically are not subject to voter authorization and, more importantly, the decision to increase rates or charges is typically not subject to the regulatory process before public utility commissions and the like required in the United States of investor-owned utilities.

These credit considerations shape the structure of revenue bond financings. The revenues generated by the facility are dedicated to the repayment of the revenue bonds with the bondholders standing behind only the costs of operating and maintaining the facility in the "waterfall" of revenues. Other legal safeguards typically include required debt service coverage ratios, the creation and funding of debt service reserve funds and facility repair and maintenance funds, the covenant of the issuing local government to charge rates sufficient to maintain the target debt service coverage and the right of bondholders to require the issuer to retain experts and consultants to assist and advise in the management of the facility if target debt service coverage ratios are not achieved. It is important to note that S&P believes revenue bonds are inherently superior to project finance for comparable infrastructure projects, primarily because of the ability to adjust tariffs to accommodate change of circumstances. ²⁹

Although general obligation and revenue bonds are the predominant U.S. municipal obligations, they by no means exhaust the list of the financing structures now at work in the municipal capital market. Among the most significant of these alternative municipal finance structures are lease ³⁰ appropriation backed obligations, moral obligation bonds, financings using a so-called "state aid intercept" structure, tax increment financing bonds and state revolving funds. Each of these financing techniques has in common the fact that it was developed to respond to a particular problem or situation that could not be addressed by traditional municipal finance techniques. In addition, with the exception of the "state aid intercept" structure and the state revolving fund, each financing technique involves a significantly higher risk of default to investors than G.O. and revenue bonds. As a result, these alternative financing techniques are best suited to specialized situations. However, above and beyond being interesting financing techniques, these new ideas are indicative of the flexible and creative nature of the municipal bond market. ³¹

Under a bond financing supported by a "moral obligation," bondholders are asked to rely on a "best efforts" promise of a public body (generally a state) to seek appropriations as needed to pay the related debt. Moral obligation bonds have been most frequently used to support state housing finance agencies and their bond financing programs to fund pools of mortgage loans for first-time home buyers and multi-family residential rental apartment projects for low and moderate income tenants. Moral obligation bonds are generally rated one full rating category below the issuer's general obligation bond rating. In order to obtain a rating on moral obligation bonds, the issuer must covenant to maintain a debt service reserve fund dedicated to the payment of the bonds equal to the bonds' maximum annual debt service requirement. If the amount of the debt service reserve fund falls below the required level, a notification process resulting in the request (usually by the state's governor) for an appropriation from the legislative body to fully restore the debt service reserve fund is required.

As noted earlier, the size and financial capacities of public bodies that issue bonds in the U.S. municipal bond market vary greatly. Small public bodies that lack name recognition in the municipal credit market and public bodies of all sizes with poor or limited credit quality and resources suffer when they bring their bonds to market. "State aid intercept" financing techniques have been developed as a way to permit these local governments to issue debt based upon not their own credit, but rather that of their state. This financing technique is based upon the fact that most local governments receive a significant amount of aid payments from their state government. These state aid payments are "intercepted" and pledged to secure payment of the debt of the local government. The marketplace carefully studies the state laws by which the "state aid intercept" operates. However, local governments are increasingly utilizing this structure as a means to lower interest costs and increase market acceptance of their bonds.

Tax increment financing (commonly called "TIF") is an increasingly popular form of financing which works hand in hand with efforts of local governments to develop unimproved land or redevelop a blighted or deteriorating downtown or business district. In a TIF financing, the local government creates a defined tax increment financing district and pledges any increase in ad valorem property taxes within the district from the "base year" to the repayment of the TIF bonds. The TIF bonds are used to finance infrastructure improvements (e.g., improved roads, sidewalks, sewers, waterlines and other infrastructure improvements) that will presumably make the TIF district attractive to businesses, retailers and developers. TIF financing projects are generally undertaken either in conjunction with the development of the TIF site as a major economic development project (e.g., an out of state company that is relocating a major office center or plant) or in connection with the efforts to redevelop a blighted downtown or business or industrial area. Particularly in the latter situation, it is common for the local government to team up with a real estate developer who will be responsible for building and renovating the buildings and securing the businesses that will populate the revitalized TIF district. Investors in TIF bonds essentially assume the risks and rewards of the upside increase in real property taxes collected on the formerly vacant or blighted land generated by the new real estate development. Because those risks are often quite substantial, local governments frequently issue so called "double-barrelled" bonds, i.e., bonds backed not only by the TIF pledge of future increased property taxes, but the municipality's full faith and credit as well.

The United States federal government has, in recent years, sent increasingly strong signals that it will no longer undertake to provide funding for a wide variety of state and local problems. The need of state and local governments for infusions of capital to pay for badly needed infrastructure maintenance and improvements has not been spared this trend. Starting with the "New Federalism" of the Reagan Presidency, the federal government has announced the end of various types of federal grants and aid for local infrastructure improvements. This announcement has frequently been accompanied by a large lump-sum payment to the states. As an alternative to simply spending the lump-sum federal payment on a specific clean water or highway project, a number of states have created state revolving funds that seek to leverage the federal contribution. Bonds are issued that are secured by the lump-sum payment. In so doing, a much larger pool of money, a revolving fund, is created from the bond proceeds. The state then loans revolving fund moneys at attractive rates to local governments to finance various infrastructure improvements. The repayments received from the local governments will eventually replenish and even increase the revolving fund.

Finally, mention should be made of structures that permit several governmental bodies to cooperate to finance certain infrastructure improvements. Increasingly, many large infrastructure improvements, such as convention centers or large water treatment and distribution systems, have a regional character and require inter-governmental cooperation and pooling of resources. Quite naturally, particularly where the participants are political bodies, issues of trust and control arise. Without the cooperation of all parties, the project will not be financed, but why should one government take a back seat to another? The creation of special purpose authorities or districts has been an effective solution to this problem. Through agreements set forth in the statute creating the authority or district, the participating governments are given some amount of control of the governance, funding and operations of the authority or district. The special purpose authority or special district is an infrastructure financing vehicle that both acknowledges the regional nature (that transcends political boundaries) of large infrastructure projects and provides a realistic solution to the bickering and suspicion that too often undermines efforts of independent political bodies to work together.

As noted, the availability of an investment grade rating can provide significant benefits in an infrastructure financing.

What is "investment grade"? Technically, of course, it is a long-term debt rating of BBB- or better (S&P and D&P) or Baa3 or better (Moody's Investors Service). What is required to obtain an investment grade rating? This question is more difficult to answer, for the nationally recognized statistical rating organizations only provide us with tantalizing glimpses of what it means or, more accurately, what it means in the context of a particular transaction. See S&P's CreditReview "Global Project Finance" March 1995, especially "Infrastructure Finance Adopted for Global Markets" contained therein. What constitutes "investment grade" may, like beauty, lie in the eye of the beholder. Too frequently for comfort, the rating agencies see particular transactions quite differently. A recent study ³² noted a surprising divergence between rating agencies when rating foreign sovereign credit as compared with comparable corporate debt. This divergence increases as credit quality is lower. Moreover, there seems to be a notable difference between these rating agencies in rating foreign sovereign credits for domestic currency payment ability. ³³

In any event, the same rating does not mean that issuers will pay the same coupon rate, even for comparable maturities. Comparably rated sovereign debt currently trades (at least since the recent Mexico peso devaluation) at a substantial yield premium to corporate debt.

Caution in the value placed on ratings is historically justified at least. In the 1920s an explosion in sovereign lending ended in a flurry of defaults. At the time of default, Moody's had rated a majority of these sovereign issuers as investment grade. ³⁴

Notable investment grade-rated transactions (and the authors' view of their respective noteworthy elements) are as follows:

A typical example of a "single user" infrastructure project is the independent power project ("IPP") model, where a special purpose project entity (usually a limited partnership, limited liability company or similar pass-through entity for tax efficiency) contracts for the construction of a power generation plant on a fixed, guaranteed, lump sum basis for the long-term supply of fuel for such plant and for the sale of capacity and electrical energy to be produced by such plant to a utility on a long-term basis (with capacity payments thereunder in an amount sufficient to repay the project financing and equity return to the project sponsors and energy payments in an amount sufficient to cover operation, including fuel, and maintenance expenses). By pledging such power purchase agreement, the project entity is able to raise the required project financing. Again for tax reasons, the project sponsors will usually seek to defer any required equity contribution until the project is complete and producing income in order to minimize nonincome-producing assets. However, this will require 100% financing for the project. This model, with slight variations, has been used for oil and gas, mining, power, petrochemical, and pulp and paper industrial projects.

By contrast with the single user type, the "multiple user" infrastructure project looks not to a single project output purchaser, but to many users of the facility. Examples of multiple user projects include toll roads, bridges and tunnels. With minor variations to the multiple user model (usually involving some degree of governmental support), it can be applied to landfills and waste treatment (transfer, recycling, composting and incineration) and similar infrastructure facilities. Financing for both single user and multiple user infrastructure projects will be secured by the facility itself and all revenues (including future revenues) therefrom.

Distinguishing features of infrastructure financing include the size and complexity of these projects and the socially necessary or desirable public services that they provide. Infrastructure projects tend to be very large and complex transactions with long developmental periods, occasionally utilizing new technology or novel applications of old technology, extensive licensing requirements and sophisticated, tax-efficient financing.

Notwithstanding their socially desirable or necessary purposes, infrastructure projects are often the subject of great public interest and occasionally strong opposition. This is particularly true for large power plants and waste treatment facilities, and especially for hazardous or medical waste treatment facilities. Accordingly, protracted delays in obtaining necessary governmental or regulatory approvals are common features.

The foregoing clearly demonstrates the desirability of interim financing for an infrastructure project until the project has been completed and its operating performance sufficiently demonstrated in order that the project may be able to receive an investment grade rating and to obtain long-term capital.

From an investor's perspective, risk inherent in its investment must be reasonably circumscribed and adequately rewarded. From a sponsor's perspective, residual risk must be adequately rewarded and it must be allowed to undertake its transaction (with reasonable flexibility to make changes thereto) without undue interference from other investors. To both, this represents a balancing of risk/reward.

Political risk aside, commercial banks have a competitive advantage in the provision of construction financing for infrastructure projects. The political risk can be encapsulated and shifted to an ECA or RDB for a nominal premium (as noted above, it may be knowingly underpriced). Generally, ECAs and RDBs prefer to avoid construction and technology risk, although some of them are on a "crash-learning" course. The cost to the project sponsor of commercial bank construction financing is intensive involvement by the banks or their designated representative (e.g., an "independent" engineer) and sponsor recourse for cost overruns, unanticipated environmental liability, and the like. The benefits to the project sponsor are floating rate financing, the ability to capitalize interest, the ability to flexibly schedule drawdowns with expenditure requirements and the "comfort" that experienced commercial banks will "do the right thing" if an unexpected problem arises.

Many commentators (including one of the authors) have suggested that the international capital markets, especially the United States Rule 144A market, might provide an alternative source of project financing to commercial banks. ³⁵

Recent events, including the Mexican peso devaluation, have again highlighted the volatility and, from a project sponsor's viewpoint, unreliability of the capital markets as an alternative to project financing by commercial banks. These events have adversely affected projects in emerging markets and have even resulted in the abandonment of capital markets financings for such projects at the eleventh hour. ³⁶In addition, the capital markets are not particularly receptive to project financing during construction of a project for three principal reasons: first, the investors do not competitively price the construction risk that they would assume; second, the project will incur negative arbitrage on funds raised in anticipation of construction costs not yet incurred (which can be a substantial financial penalty depending on the construction timetable and scheduled project draws and, of course, on the applicable spread between interest paid to investors and interest earned on funds invested); and third, as a technical matter, it is difficult to write definitive documentation (particularly the construction draw provisions) that provides appropriate investor protection during the construction period and provides the project with necessary flexibility to deal with unexpected events.

As noted, private placements offer fixed rate, long-term financing and are "story" and transaction-complexity tolerant. However, private placement investors are less flexible regarding construction drawdown scheduling (or, more accurately, changes thereto) but are otherwise (in the author's view) comparable to commercial banks, except that they cannot/do not provide letters of credit or commodity, currency or interest rate swaps. While the private placement market is somewhat credit rating-sensitive, it is less so than the Rule 144A/public markets.

Rule 144A/public offerings offer the most liquidity, the longest tenors and fixed rate financing for infrastructure projects. A Rule 144A offering usually, and a public offering always, requires at least one rating, preferably an investment grade rating. These "public" markets are not as "story" tolerant as the private placement market and generally (although the Sithe/Independence transaction is an exception) do not assume construction or technology risk. Because they usually finance projects that are complete and have demonstrated operational performance, the investors are not as intrusive in the project as commercial banks or private placement investors. It is more difficult to obtain a consent or waiver from public investors than from commercial banks or private placement investors, and, accordingly, documentation for a public offering is designed to obviate any required consent or waiver except in a material situation.

ECA/RDB/IBRD are "gap" fillers, insofar as they provide cover against certain risks which construction lenders will not assume (e.g., political risk), and occasionally "cheap" money, because of the subsidization of their export promotion/development activities.

We offer the following five suggestions for possible innovation in infrastructure finance:

1. Exploit the opportunities of project finance to isolate and encapsulate embedded risks in order to take advantage of the most cost effective solution thereto.
2. Adapt and use U.S. municipal finance models, which have demonstrated flexibility and creativity in balancing political, social and economic goals (all of which are usually present in an infrastructure project).
3. Use equity enhancements to reduce debt service requirements yet maintain an investor's return. These enhancements can be tailored to suit the specific situation and, at least in theory, can be very flexible (e.g., using "shadow" equity and puts/calls to provide floors/ceilings on an investor's return).
4. Reinvent "older" infrastructure finance securities, such as income bonds and other contingent interest securities.
5. Use and adapt securitization techniques, especially for multiuser infrastructure projects, to structure securities designed to maximize their investor appeal and the benefits of portfolio diversification and secondary liquidity.

While somewhat frustrated with the lack of recent innovation in debt finance for infrastructure projects, we are nevertheless optimistic about the prospects for future innovation in this area. Our optimism stems primarily from the need to be innovative in order to raise more capital for the compelling infrastructure requirements around the world.