In last October's Working Paper on the Internal Ratings Based Approach to Specialized Lending Exposures, the Basel committee suggests that project finance loans ?possess unique loss distribution and risk characteristics? which ?exhibit greater risk volatility? and in times of distress could lead banks to ?both high default rates and high loss rates,? relative to corporate lending. A bad omen for project finance, one might think.

Indeed, this interpretation has met with much criticism from key project financiers on both sides of the Atlantic. Project Finance has held two roundtables to discuss the Basel II accord, in order to stimulate a consensus behind the lobbying of regulators ? not least to mitigate the proposals' potentially punitive effect on project lending. The most recent, held in London in January, served as a forum for key European industry participants to voice their concerns. The resulting discussion focussed on the proposals' possible effects on project lending. And the most obvious conclusion was the need for a common inter-institutional methodology going forward.

According to Mike Wilkins, head of project and infrastructure finance ratings at S&P in London, and co-chair of the London roundtable, ?what we need to do is to present a coherent and united case for the appropriate treatment of project finance loans under the new Basel regulations.? The question is which body is best suited to the task.

Presenting a united front

Citigroup is coordinating a multi-bank study of defaults and recoveries from project finance loans. At present, four lead arranger banks, representing roughly a quarter of the project loan market over the past five years, are combining their data, and, according to sources at Citi, are working with Standard & Poor's to aggregate and analyze the data on a consistent basis. Expectations are that the results will show that project loans perform substantially better than unsecured corporate loans. The first phase of the study should have been completed by the end of March, and will be submitted to the Basel Committee, and to the regulators in each of the countries where the banks in the study are based.

The hope is that additional banks will join in subsequent phases of the study. According to a source at Citi, ?joining in the study will be in the interests of institutions which, in order to be treated as ?sophisticated' banks by their regulators, will need to have data available to back up the regulatory treatment they would like to receive on project loans. It will also be in banks' interests to build an industry data base in project finance, and will be useful for banks' internal discussions with risk managers and others who allocate capital among products and businesses within each bank.?

The Basel committee is attempting to carve out ?a common industry standard for a rigorous, empirical, and risk-sensitive approach to economic capital estimation? of project exposure, albeit informed by an arguably narrow interpretation of project lending risks.

In contrast, for corporate exposures, the committee was ?comfortable that banks had in place, or could develop within the relevant time frame, internal rating systems capable of assessing the quality of the exposure, and quantifying these assessments.?

At issue is the probability of default and the probable amount that a bank will recover from a loan in default. And the consensus among the participants of the London roundtable is that the Basel committee has exaggerated the probability of default on project loans, while overstressing assumptions on losses in the event of a default. The result is the tendency to assign high risk weightings to projects, thus making project lending perilously expensive. This could mean that project finance loans will become less available, and that banks will try to minimize their exposure to project finance lending.

Lower rated loans are to be weighted at more than 100%. Under the draft proposals for specialized (including project and asset) lending, that means a 150% risk weighting coming into effect at the single B and double B level, with triple-B loans remaining 100%. The capital adequacy ratio is to remain at 8%.

One of the more radical features of Basel II is the role which banks themselves will have in determining risk weighting under the internal ratings-based (IRB) approach.

When using the IRB approach, banks have to estimate their own probability of default (PD) numbers to feed into a formula that calculates the risk weighting for an asset. Banks have to prove to their regulators that the way they collect default data is sound, and that the system they use to estimate probability of default is adequate too.

Basel has set minimum requirements for default histories of five years under the foundation IRB approach and seven years for the advanced IRB approach. There is, however, a transition period until 2008 that allows banks to go ahead with three-year default data for the foundation approach after 2005.

But the Basel proposals are at an early stage yet. The October Working Paper is by no means the final word on new capital adequacy provisions. Says WestLB's Wolfgang Nickels, ?Since the October proposals, several elements have reportedly been changed ? especially with regards to the definition of the project finance portfolio. In other words, the fundamental questions are still open.?

Even taking the October paper as a valid basis for discussion, Nickels says, ?I still think there is no clear framework for calculating the impact on lending institutions' portfolios. My understanding is that what we have before us is simply a first attempt by the Basel Committee to come to terms with some very complex issues.? But, he adds, ?the proposals cannot be set into motion, given their present ambiguities. Most of my European and American banking colleagues are facing just this same problem, when it comes to evaluating the possible impact on our portfolios.?

As a preliminary measure, however, most top tier banks have at the very least set up internal task forces, as a first attempt to address the committee's proposals.

Says BES' Nigel Purse, ?the first step is essentially a data gathering exercise. The committee has set forth three methodologies ? basic, foundation, and advanced ? and they're seeking feedback from the industry on all of those. It is the industry's responsibility to go back with the relevant data, tracing back default and loss given default analysis for the appropriate historical time frame.?

S&P, for one, has begun looking over its data on project finance ratings over the last decade. ?We're looking at substantial amounts of data, some of which is also based on work we've done more recently for banks that are looking to do project finance CLOs, where we've had to undertake rating estimates on each of the numerous loans in their portfolios,? says Wilkins.

There are many uses for this data, including helping the industry forward in managing its discussions on Basel, as well as helping refine institutional risk management procedures.

?We have been operating a project finance rating model for over a year now,? says Simon Byrne, ABN Amro's head of project finance for natural resources in the EMEA region. ?We've been aggressively building it out over the last year to ensure that we get the right ratings methodology in place.? This, it seems, is roughly the stage at which most key European institutions find themselves.

Says Nickels, ?we've had systems for project finance transactions up and running over the last three years. But we don't yet have any historical data on the project finance rating system, particularly on LGDs.?

Patrick Blanchard, head of project finance at Dexia Credit Local, believes that even defining default is difficult: ?The definition of default in the October paper is just too harsh for project finance. If implemented, it will have perverse implications, such as banks having more relaxed term sheets in order to avoid a situation of technical default. Now this is clearly counter-productive.?

He adds, ?my own experience of project finance, having also been involved in corporate banking, is that, in terms of default rates, in the long run project finance is less risky than corporate banking. The problem is that the regulators may not yet be on that same wavelength.?

This clearly is one part of the problem ? the relative risks of project versus corporate lending. Another related issue, says Wilkins, is the mis-stated relationship between default and recovery rates: ?The October paper states that there's a direct correlation between default rates and recovery rates in project finance. In our experience this is simply not the case. In fact, you can have quite a high default probability for a project, or a low rating if you want to view it that way, but quite a good recovery, or low loss given default ? and this really reflects the special characteristics of project finance.?

But the Basel committee, assuming a direct correlation, is thereby forced to conclude that the restructuring of a project finance loan is a negative event, whereas restructurings have actually increased the recovery of project finance loans.

Project bankers are clamouring for the proposed assumptions on default and recovery to be altered. Suggests Rene Kassis, then director of project finance at Credit Lyonnais, ?the more covenants you have, and the more stringent they are, then the higher the probability of breaching them. But the fact that you have a default situation by itself is clearly not necessarily a bad thing ? we need to differentiate between the kind of default that occurs as an early warning signal, and the type of default that actually damages a transaction.?

One size doe not fit all

Another point of concern is the lack of analysis by industry. Says Byrne, ?a ?one size fits all' approach is untenable. We need to be able to assess project finance risk by industry and to take account of statistical data by sector, in order to back up the argument that the early warning signals, the security structures and so on, which are particular to project finance, will actually benefit across the board.?

Accordingly, historical data on both probability of default and loss given default, specific to project finance transactions, needs to be analyzed, by sector. Although, as Peter Robinson, National Australia Bank's London Project Finance head, points out, ?there's no point doing it bilaterally among 40 institutions. It makes much more sense to actually pool that information together and go in, together with the ratings agencies, with a forceful argument.?

But the question remains ? if the current Basel proposals are implemented, what will be the impact on project finance banks?

One possibility is that this will lead to more structured financings, CLOs, and securitizations as banks try to manage their exposures while looking for regulatory relief. This tool could prove particularly appealing since capital weightings for banks with such securities on balance sheet will be set at 20% under Basel II.

Says Wilkins, ?I know of some banks already going in that direction. There have only been three closed project finance CLOs but there are hundreds in the corporate finance domain. The question is whether we'll ever get to the stage where project finance CLOs become commodity transactions.? That may depend on who the investor community for the CLOs is, and the extent to which banks adopt the IRB approach.

Several more banks, including four or five European institutions, are understood to be actively considering project finance CLOs. But, says Byrne, ?The cost of doing the CLO both in terms of human effort, cash and time, is huge. I'm really not convinced this is the most attractive option.?

But it is clearly an ever-evolving tool. Says Wilkins, ?we are now much further along the curve, and each new CLO will take less time to do. And given the ultimate size of the project finance world, it won't take that long to get to the point of duplication.?