Given the uncertainty as to whether banks can continue to lend vast sums for an increasing number of projects worldwide, the project CLO looks to ease the tightening for capital. According to Guy Cirincione, director of project finance at CSFB ( the first and only bank to securitize project loans to date): ?The primary motivation is to provide capital release. Also there is a certain amount of credit risk transfer. In our institution, the impetus was started by the project finance department and, of course, we needed the sign-off from senior management.?

Moreover, the project CLO allows banks and lenders the opportunity of being able to better control their balance sheets. ?Although I believe banks are great originators of commercial paper, balance sheets have tended to be constrained? says Paul Forrester, partner at Mayer Brown & Platt. ?Banks do not have 20- to 25-year assets against which to lend.?

In simple terms, securitization is based on the law of large numbers; homogenised numbers of underlying assets which can be structured with diversification in order to satisfy the rating agencies' criteria. Standard transactions have tended to take the form of routine types of property: credit card pools, mortgages and leases. However, as the sophisticated form of securitization, the CLO looks set to change this.

Since NatWest closed its $5 billion ROSE Funding No. 1 transaction in 1996, a host of large volume CLOs have hit the market. A year later, NationsBank emerged with a $4.2 billion CLO transaction, one of the largest US deals to date. In 1999 alone, some $60 billion to $70 billion in CDOs (collateralized debt obligations) were closed. At present, the CDO market encompasses loan, bond and mortgage obligations, as well as including secured loans and high-yield bonds.

Between 1998 and 1999, a number of CDOs were downgraded, due to credit problems in the emerging markets. Since 1988, however, out of some $257 billion CDOs rated by Moody's, only $8.8 billion have suffered downgrading; all of which were in emerging markets. Since then rating agencies have treated certain CLOs with a greater degree of scepticism and severity.

However, the incorporation of derivatives has further improved structures, in the ability to pay down pro-rata rather than sequentially. Monoline insurers have also entered these structures through taking the benefit arbitrage and capturing the spread.

In addition, the project CLO structure uses total return swaps, which pay interest and repay capital in the event of bankruptcy and default. In addition, credit linked notes are sold through a special purpose vehicle (SPV), which pays principal in order to raise payment and issues note swaps. The interesting twist to structuring through an SPV is getting the assets into it, without rendering them subject to tax, the problem being that certain jurisdictions can be extremely unfriendly to these transactions.

While the CLO looks to be good news for project financiers, transactions will not add greatly to overall CDO deal volumes. ?CSFB's securitising $1 billion in project finance is a lot. On the other hand, the commercial loans being securitised makes project finance pale,? comments Cirincione.

For many banks, with large portfolios of project loans, the CLO could be an answer to what has been a difficult question: how to liquify an illiquid portfolio? Through restructuring project loans, banks now have the opportunity of being able to reallocate their capital. Both risk-based capital and cash can be freed up in order to be re-deployed into lower-risk weighted and higher yielding assets.

Furthermore, from a banking perspective risk can be passed on freeing up extra capital, which needs to be set aside as a contingency against defaults under central bank regulations. Under BIS (Bank of International Settlements) regulations, banks must retain risk capital worth 8% against loans made.

Given the benefits of the CLO, why are not more project financiers looking at restructuring their project debt? Many banks prefer to hold onto their projects, earning the interest on their substantial portfolios. ?It is a case of perceived risk and reward. Rather take on LIBOR +2 on a project loan, than LIBOR +5 to high quality corporate,? muses Forrester. In addition, it may stem from a rating problem, in that few agencies have come forward to look at these transactions.

Yet, certain bankers are not taken in by the hubbub surrounding the emergence of the structure. ?It is not really a viable business proposition. It is just a good adjunct,? says a US commercial banker.

Good adjunct or not, the project CLO remains as a viable proposition. For a number of institutions, an overriding problem is the reconciliation between long life commitments and the short-term character of their balance sheets. Credit departments have trouble in deciding whether 18- to 20-year commitments will really be paid back within three to four years as their project financiers would have them believe.

Given the number of institutions looking to change direction and emphasise their commercial capabilities, the project CLO can be a way of achieving this. According to a US investment banker, ?If there is a danger of too much concentration, you can sell up and lighten up.?

However advantageous project CLOs may be to a given institution, they can provide ammunition for competing banks. Thus, it has come as no surprise that certain bankers have even gone so far as to query CSFB's motives for its project CLO transactions. ?If Deutsche Bank were to do this type of transaction, it would give a completely different signal,? muses a New York-based investment banker. ?The question is whether you are doing these deals in order to facilitate an exit from the market or recommit your capital.?

In response, CSFB does not rise to the bait and states that any facilitation of an exit from the project finance market is wholly inaccurate.

CSFB is generally acknowledged to have been the first to close a project CLO (there are rumours that CSFB has been pipped by a US bank, though no specific institution has been named). And the quality of both transactions look to withstand any scrutiny, given the handful of new deals expected to hit the market before year-end. ?There have been a lot of enquiries,? states Henry Albulescu, director with Standard & Poor's. ?Some transactions are fairly active and we should see one or more by the end of the year.?

Although rumours abound as to new candidates for project debt CLOs, no bank has as yet owned up. Margin-conscious Morgan Stanley, Merrill Lynch and Goldman Sachs are said to be looking at the structure. However, all three institutions have little in the way of project debt on their books.

The most likely candidates are thought to include Citigroup, Chase Manhattan, Bank of America, Dresdner bank and Deutsche Bank. In addition, there is growing speculation that some of the French banks will look to restructure their debt and lighten their portfolios. However, there remains an air of secrecy not only over prospective deals, but those currently in the market.

Any deal expected to hit market is likely not to be larger than that of CSFB's first transaction, Project Funding Corporation I. According to Cirincione, ?We have been telling them that you should have $500 million, because below that amount you start to stretch. There is also a top limit for marketing reasons. The ideal is between $500 million and $800 million.?

Whether future transactions will comprise one sole institution remains to be seen. Given that CSFB and BHF shared Project Funding Corporation II, 76% and 24% respectively, there is no reason why three banks could not come together on a given transaction. But with any more than four, transactions would become over managed and over complicated.

Aside from the intrinsic benefits of having closed two transactions, CSFB is now selling its expertise adds Guy Cirincione. Detractors state that by acting as an arranger and not concentrating on originating new transactions, CSFB is making a tacit exit from the project finance market. ?That is hardly true. Anyhow, these banks have different people to structure their CLOs,? states a US attorney.

Whatever the reasons behind structuring a project CLO, the benefits look to outweigh the few reservations. The only tangible hurdle being the view that certain rating agencies might take of project transactions. Thus, it seems unusual that the only transactions have been closed by CSFB. Cirincione agrees, ?It begs the question as to why nothing has been done.?

One reason put forward is the fact that other institutions had waited to see to what mix of projects would succeed. Closed in March 1998, the first CSFB $617 million pool of project finance loans, with the exception of a Chilean mining project and a US steel plant, all originated in the US power sector. Thus, any emerging market risk was kept to a minimum.

Although the first transaction was groundbreaking, for its debt class, the very projects chosen were in fact the easiest to securitize. What CSFB had chosen to do was test the CLO market with an easy set of assets. Given a pool largely made up of completed US power deals, denominated in US dollars, institutional investors were hardly going to turn their nose up at such an opportunity.

Two years later, the investment house re-emerged with an altogether more audacious transaction. The $498.6 million CLO, known as ?Project Funding Corporation 2?, secured 40 projects. Only 50% of loans were booked in G7 countries, primarily the US, Germany and the UK, compared with the near total bias of the first transaction. More importantly, the remainder consisted of emerging markets and non-OECD debt, which included Chile, Colombia, Venezuela, Morocco, Turkey, Thailand and Qatar (see box).

Given the diversity of G7 and emerging market projects, the main hurdle that needed to be overcome was rating the range of credit profiles. According to Arthur Simonson, director at Standard & Poor's, ?From an S&P perspective, we need to find out as much as possible about the underlying assets. There is a lot of due diligence.?

Rating a project CLO is a four level analytical process. Firstly, the underlying assets need to be assessed as to their credit quality; secondly, the default and loss severity nature of the pool is analysed; thirdly, cash flows of the projects are forecasted; and fourthly, the structural features of the transaction are examined.

Although the due diligence process of a project CLO may be arduous for superior assets, the fact remains that project debt lends itself rather well to the CLO structure. Amoritizing debt, tailored debt service payments, covenants and cash traps make projects attractive.

Standard & Poor's has stated that refinancing risk, prevalent in corporate CLOs, is mitigated by amortitizing debt. Therefore, this allows the rating agency to appraise the repayment obligations, beyond the tenor of the debt. In addition, amoritized debt allows for a decrease of debt leverage over the given period, making it advantageous from a credit perspective.

Tailored debt payments, inherent in project finance structures, are also favoured by the rating agency. Given that there is no limitation of the how principal is repayed, repayment can be correlated with prospective cash flows.

For the capital markets there is a dislike of two basic tenets of project financing, namely construction risk and the dearth of a law of large numbers. Conversely, high recovery value and quicker recovery periods make projects highly attractive. According to Forrester, ?The experience with projects which have had problems have been worked out consensually by participants rather than through bankruptcy. This culminates in shorter work-out periods.?

The fact that project finance by nature is highly leveraged, debt-expense elements of total expenses tend to be much higher compared to corporate loans. Thus, projects have a greater propensity to default and run into difficulties. Yet, the majority of projects will be able to cope with a high amount of leveraged debt, given that cash flows will tend to outstrip expenses by a sufficiently large margin.

In stress tests, project loans do not perform badly. The loans can be restructured in such a way as to extend the maturity dates, while repayments decrease over the course of this period. ?From stress tests, it has been corporate loans that have not looked that good,? adds Forrester.

More importantly, the transfer of the CLO from corporate loans to project debt has hit difficulties in diversification. The average corporate CDO transaction of between $300 to $500 million maintains a diversity test, requiring 40 to 60 issues across 30 industries. Standard & Poor's uses either a single jurisdictional or multi-jurisdictional default model to assess the probability of default of a pool of loans.

While the single jurisdictional model is used for transactions where there a fewer loans in one sovereign jurisdiction, the multi-jurisdictional model assesses a high concentration of loans. Thus, Standard & Poor's has pronounced that diversification penalties will occur where there is a reduction in an implied rating due to a correlation in assets. A correlation between similar assets will depend on the nature of the asset and its congruity with other assets of a similar vain. In certain circumstances, this may be negligible due to macroeconomic forces creating similar credit deterioration on a number of assets.

The problem with the model arises in particular emerging market jurisdictions. ?What about telecoms, for example? Does Brazil correlate with Argentina?? asks Forrester. Standard & Poor's has bracketed both Chile and Colombia together, which has raised a number of eyebrows. Chile remains AA rated, while Colombia has slumped to BBB. ?The criteria do not make a lot of sense. For international projects, there needs to be a more flexible stress test,? comments a US banker.

However, given that the criteria are based on CSFB's first deal, they should not be seen as intransigent. Standard & Poor's guidelines will look to evolve as more deals come to market and the project CLO becomes more sophisticated. In reality, both rating agency and bank take a close sample of some 40 loans, and when mapping look at the largest.

Although both bank and rating agency complete their own individual mapping, differences tend not to occur. Banks may rate certain projects higher, because of political risk insurance (PRI), which they maintain decreases default. However, according to Standard & Poor's, PRI only enhances any recovery in the event of default.

With a host of ostensible advantages and little in the way of downside ? bar expense and complexity ? it remains unusual that more deals have not hit the market. ?These are very long loans, between 15 and 18 years. Therefore, banks should be motivated to try to move them off balance sheet,? comments Cirincione.

Given that most banks are disdainful of 30-year paper, using the CLO structure for project loans would seem a natural move. However, as one US lawyer states, ?Banks are not very good at looking at things on a risk and reward basis. They are schizophrenic about being in or out of the market?.

This seems to be the case, even when rewarding fellow bankers for their foresight in structuring groundbreaking transactions. Whether CSFB facilitates its exit from the project finance market remains to be seen. If the investment house, with its lighter portfolio, fails to originate a host of new deals, competitors will feel themselves vindicated.

According to Forrester, ?Looking at the impact of CDOs on the leveraged loan market, one would expect project finance CLOs to bring substantial liquidity to the market and potentially become the market drivers in pricing and structure.? With this in mind, there is a feeling that some of these drivers will have been CSFB's detractors.