51768 noTE no. 48 ­ JunE 2009 GRIDLINES Sharing knowledge, experiences, and innovations in public-private partnerships in infrastructure Another lost decade? Effects of the financial crisis on project finance for infrastructure James Leigland and Henry Russell R apid growth in project finance, driven by as well as developed countries (box 1). In just huge increases in liquidity, helped fuel six years the international project finance market the gains in private participation in infra- grew by four or five times, depending on the mea- structure (PPI) in developing countries in the sure used (Abadie 2008). past decade. But when the financial crisis hit, the excess liquidity began to dry up as lenders Much of the growth of activity in that wider market backed away from practices that had helped was driven by huge increases in liquidity generated generate it. The effects are already apparent-- mostly in developed markets. Changing banking in greater delays in financial closures, more practices accounted for some of this liquidity cancellations, and higher financing costs growth. Project finance banks shifted from "origi- for PPI projects. If full recovery of the proj- nate and hold" to "originate and distribute" strat- ect finance market takes much longer than egies: they would book project loans, then quickly expected, some of the measures that are now distribute them into the market through syndica- being adopted to avoid shutting down project tion, securitization, secondary market sales, and pipelines might have unintended--and very other techniques. The emergence of a secondary negative--consequences. market in loans (in contrast to securities such as bonds) was helped in Western Europe by the rap- After the Asian financial crisis, it took a full decade idly growing use of credit ratings on loans. These for private participation in infrastructure (PPI) to ratings made investors more comfortable about recover: 2007 was the first year in which invest- buying project loans while knowing little about ment commitments to PPI projects in developing the borrowers or the underlying projects. countries exceeded the record level of 1997 (fig- ure 1). Today's global financial crisis poses a new Bond issuance to finance larger projects in West- threat. Investment commitments to PPI projects ern Europe or the United States was facilitated by are estimated to have dropped sharply in 2008 monoline insurance, which passed on to the bond and are expected to drop again in 2009. Govern- issue the strong credit rating (usually triple-A) of ments and development institutions are under- the monoline insurer. Again, investors had little standably concerned about returning as quickly incentive to conduct their own due diligence of as possible to the investment levels of 2007--and the issuer or the underlying project as long as the avoiding another "lost decade" of private invest- bonds carried such strong ratings. These monoline ment in infrastructure. "wraps" were also critical for the sale of all kinds of asset-backed securities, some of which bundled project finance debt for off-loading to secondary PPIAF Approved Logo Usage Structural changes in the market investors. Understanding how to regain the investment peak Logo - Black of 2007 requires an appreciation of what it took to reach that level in the first place. The growth James Leigland is acting program manager of PPIAF. of PPI in the developing world in 2003­07 was PUBLIC-PRIVATE INFRASTRUCTURE ADVISORY FACILITY Henry Russell is manager of the Finance and Guarantees facilitated by a precipitous rise in the use of proj- Unit of the World Bank's Finance, Economics, and Urban ect finance in a variety of sectors, in developing Development Department. Logo - 1-color usage (PMS 2955) Helping to eliminate poverty and achieve sustainable development PUBLIC-PRIVATE INFRASTRUCTURE ADVISORY FAC ILITY through public-private partnerships in infrastructure 2 FIgu RE 1 A lost decade followed the Asian financial crisis Investment commitments to infrastructure projects with private participation in developing countries, 1991­2007 180 160 140 120 2007 US$ billions 100 80 60 40 20 0 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 Source: World Bank and PPIAF, PPI Project Database. These methods of distribution created space in the a few in South Africa, they have little long-term originators' balance sheets for new lending and money available at affordable interest rates for helped banks make money without holding onto infrastructure investment. And what capital mar- loans. The resulting increase in liquidity generated kets exist in the region are shallow. But the huge tremendous competition among lenders, driving amount of international project finance liquidity down the price of loans and giving borrowers the in 2002­07 meant that international lenders had upper hand in negotiations. space in their balance sheets for the occasional riskier loan in emerging markets, and some larger Excess African projects benefited directly. In other cases liquidity Liquidity spillover the spillover was more indirect: local or regional African banks borrowed from these international for project Most of this excess liquidity for project finance was banks or from capital markets to raise financing finance generated in developed markets and helped fuel the large PPI programs in Western Europe. In the needed to make loans to PPI projects, or used credit lines from international banks to back lend- spilled United Kingdom the government's Private Finance ing to large African projects. over into Initiative reached record levels of investment. But not all the liquidity could be absorbed in these developing markets, and some spilled over into developing From syndicates to clubs countries countries. Much of it went to Eastern Europe, where Western European lenders were quick to This project finance liquidity began to dry up take advantage of the need for project finance for when the mechanism for distributing it shut down. public-private partnerships and privatization in The interbank borrowing market was the main infrastructure, reforms driven by EU accession. way in which international banks raised money The surge in project finance liquidity also ben- for project finance, and when that market was dis- efited Latin America, with a long tradition of using rupted as the subprime mortgage crisis began to international credit ratings and monoline insur- raise doubts about the ability of banks to repay ance to access project finance from international their own debts, the distribution of project finance banks and bond markets. came to a stop. But the underlying mechanism for generating this liquidity also collapsed. By early Larger projects in poorer regions like Africa also 2009 the "originate and distribute" model was no benefited. Banks in Africa tend to be less tightly longer working because originators could no lon- integrated into the global banking networks than ger distribute project loans into the market. The those in East Asia or Latin America. Apart from secondary market for this debt collapsed because Another lost decade? 3 stage forecasts of value for money, because "market Box 1 flex" arrangements allow lenders to change prices A note on terminology and other terms at the time of financial closure and sometimes even afterward. Most but not all of the projects included in the World Bank­PPIAF PPI Project Database both involve Competition among lenders has lessened, and project finance and can be classified as a public- bank credit committees have regained control private partnership (PPP). Project finance normally over lending decisions and pricing from origina- But with tion teams. The balance of negotiating power in means that project loans are secured by project the markets has shifted back from borrowers to the onset assets and repaid from project cash flows rather than lenders. Excess liquidity is no longer being gener- of the from the sponsor's corporate balance sheet. PPP ated in developed markets nor spilling over into normally implies a risk sharing relationship between the developing world. Many of the world's leading financial public and private project partners, formalized in a project finance lenders are sharply reducing their crisis, the activity in the project finance market (Royal Bank contract. The database includes projects, particularly of Scotland, Lloyds Banking Group, DEPFA Bank, liquidity in telecommunications, that are corporate-financed Fortis, Dexia, Halifax Bank of Scotland). In addi- began to "merchant" projects without PPP-type risk sharing tion, Basel II banking regulations, which require contracts between public and private partners. It banks to hold more assets in reserve against the dry up also includes divestitures (including partial share risks of longer-term lending, are already leading sales) that do not involve project finance. to a general repricing of project finance credit risk. The additional regulation that is likely to come in response to the crisis will probably push the big international project finance banks even further credit ratings of loans lost credibility and triple- away from PPI projects. A monolines virtually disappeared as a result of credit downgrades. Impacts in developing countries Even normal loan syndication techniques were no longer being used. Previously, one or two lead In developing countries the effects of the crisis are arrangers would reach agreement with a borrower already apparent. PPI projects in most develop- on the size and pricing of a loan, then syndicate ing regions are seeing greater delays in financial parts of the loan to banks or other investors. In the closure, more cancellations, and higher financing new environment most big project finance deals costs (Izaguirre 2009). In Eastern Europe and East began to involve expensive and time-consuming Asia project finance has come to a standstill. In "club" arrangements in which borrowers negoti- Latin America some large deals are still closing ate with a set of potential lenders to put together through "mega-club" arrangements involving nine a total financing package. When such deals could or more banks. In Chile PPI projects in transport be done, borrowers had to tolerate lower "lender are on hold--this in a country where traditionally hold" levels, shorter terms, higher pricing, and about 40 percent of highway concession invest- greater requirements for sponsor equity. ments have been financed by bonds guaranteed by triple-A monolines, most of which have now In addition, the borrower-friendly "covenant light" been downgraded. In India normal loan syndi- lending that accompanied the intense competition cation by private banks has slowed, and so has among banks quickly gave way to the growing reap- the pace of financial closures for the government's pearance of legal boilerplate clauses that protect "ultra-mega" power projects. The government has lenders from market changes but also reduce bor- struggled to keep toll road and energy projects rower certainty about the ultimate cost and terms afloat with lending, guarantees, and refinancing of financing. "Market disruption" clauses allow pledges by public financing institutions that have lenders to increase prices during loan drawdown, been allowed to raise less expensive funding for and "material adverse market change" clauses their clients. allow lenders to walk away from deals after closure on the basis of a determination by the lenders of In Africa club arrangements are evident in a seriously disadvantageous market changes. The variety of project finance sectors. The Nige- syndication that is still possible is complicating rian press has reported that international banks government PPI project reviews that require early- have rejected dozens of requests by local banks 4 for guarantees and credit lines needed to allow looking for ways to streamline their internal pro- lending to local infrastructure projects. Some local cedures for approving PPI projects in a "market banks and investment funds in Africa, particularly flex" environment where firm information on debt in South Africa, still have the liquidity needed pricing is not available (and therefore cannot be to participate in syndicated or club deals. But there evaluated) in advance of financial closure. Gov- are fewer such lenders now, their interest rates ernments are also considering temporarily replac- are increasing, and their hold levels decreas- ing scarce commercial finance with public money ing. Many lenders simply do not want to make from government budgets or donors. But the risks commitments because of near-term market associated with many of these emergency measures uncertainties. are substantial if a "return to normal" means a stabilized project finance market with much less liquidity and much higher prices for debt than in Returning to "normal"? 2007. Governments risk huge contingent liabili- ties on projects that may not have been vetted as Banking practices such as interbank lending and thoroughly as necessary--and risk being trapped loan syndication will eventually return to some as long-term investors in projects that cannot be semblance of normal functioning, and many of refinanced with commercial debt. the banks that now forswear project finance will eventually return to the market. But some of the Many governments, particularly in developing other recent changes in the structure of the market countries, will need help in identifying and man- for project finance are likely to affect liquidity and aging these kinds of risks. They will need to do pricing for a long time to come. The secondary at least three things: carry out the kind of pru- market in project debt that drove liquidity gen- dent due diligence that banks do to identify com- eration at the top of the project finance cascade mercially viable projects, determine the optimal in developed countries will likely take years to mix of public and private money in these projects, rebuild, and that is likely to have long-term down- and avoid accepting so much project risk that pri- stream effects on PPI in developing countries. vate partners no longer have incentives to meet performance expectations. Given the likely long- But some lenders recently quoted in financial pub- term effects of the crisis on infrastructure project lications are not at all sorry to see the generation of finance, this kind of rigorous appraisal, selection, liquidity slow down. They maintain that loan pric- development, structuring, and implementation of ing in mid-2009 better reflects underlying credit projects will be essential to ensure that aggressive fundamentals than it did when excess liquidity was rescue measures do not end up transforming con- fueling lender competition and fire-sale pricing in tingent liabilities associated with poorly prepared the buildup to the record project finance levels of projects into huge financial burdens for govern- 2007. Given these considerations, the recent mar- ments that cannot afford them.1 ket low point in 2002 (or perhaps even the early 1990s), not the peak level of 2007, is beginning Note to look more like the "normal" market level of PPI 1. A companion note (Leigland and Russell 2009) discusses in investment. greater detail the need for PPI-related technical assistance. References If it takes much longer than expected for the mar- Abadie, Richard. 2008. Infrastructure Finance--Surviving the Credit ket for infrastructure project finance to fully Crunch. London: PricewaterhouseCoopers. PPIAF Approved Logo being recover, some of the measures now Usage GRIDLINES Izaguirre, Ada Karina. 2009. "Assessment of the Impact of the adopted by governments and devel- Crisis on New PPI Projects: Update 2." Private Participation in Infrastructure Database, World Bank, Washington, DC. Gridlines share emerging knowledge opment agencies to avoid shutting Leigland, James, and Henry Russell. 2009. "New Needs for Technical on public-private partnership and give an down project development pipe- Logo - Black Assistance: Responding to the Effects of the Financial Crisis on overview of a wide selection of projects from lines might have unintended--and Private Participation in Infrastructure." Gridlines series, no. 49, various regions of the world. Past notes can be very negative--consequences. PPIAF, Washington, DC. found at www.ppiaf.org/gridlines. Gridlines are a publication of PPIAF (Public-Private Infrastructure For example,R Agovernments are IT Y PUBLIC-PRIVAT E IN F S T R U C T U R E A D V IS O RY FA C IL Advisory Facility), a multidonor technical assistance facility. Through technical assistance and knowledge dissemination PPIAF supports the efforts of policy makers, nongovernmental organizations, research institutions, and others in designing and implementing Logo - 1-color usage (PMS 2955) strategies to tap the full potential of private involvement in c/o The World Bank, 1818 H St., N.W., Washington, DC 20433, USA infrastructure. The views are those of the authors and do PhonE (+1) 202 458 5588 FAx (+1) 202 522 7466 not necessarily reflect the views or the policy of PPIAF, gEnERAL EMAIL ppiaf@ppiaf.org wEB www.ppiaf.org PUBLIC-PRIVATE INFRASTRUCTURE ADVISORY FACILITY the World Bank, or any other affiliated organization.