But within the agencies, among their debt-issuing clients and the ranks of investors who use their ratings, there has been a huge debate and much work analysing the nuts and bolts of what credit ratings really mean.
The main three agencies have recently issued proposals or requests for comment on a variety of enhancements or additions to the traditional ratings scales that run from triple A at the top end to C at the bottom.
At the end of last week, the leading global industry association for capital markets issued its recommendations on what should be done to enhance investors’ understanding of ratings in order to bring confidence – and money – back to the markets for mortgage-backed bonds and other structured debt.
This trade body, the Securities Industry and Financial Markets Association, rejected out of hand one of the few practical suggestions that has received broader attention: the idea of changing the traditional ratings scale, or adding a suffix or identifier to structured finance ratings.
This idea has been championed by Michel Prada, the head of the French financial regulator, along with other international regulatory and supervisory groups. But Sifma’s task force said it would lead to significant unnecessary costs – related to updating computer systems, investment mandates and state and national laws – while potentially triggering more forced sales of structured bonds by investment managers no longer mandated to hold them.
In any case, it added, the changes would add little in terms of transparency, saying they would “at best be a cosmetic solution”.
The agencies are not hot on the idea either. Michel Madelain, Moody’s chief operating officer, told the Financial Times in May that his agency had already decided against such a change due to industry feedback. Standard & Poor’s and Fitch have also said the feedback they have received is not generally favourable.
Most industry focus has been on how to address the volatility of ratings – or the perception that when structured debt deals start to go bad, they get downgraded very quickly rather than weakening over time as companies are perceived to do.
Before the credit crunch, the agencies routinely insisted that structured credit ratings had broadly performed better than those for corporates or governments, although this was based on a limited history of data. But since the crunch, the downgrades have been coming thick and fast, especially for subprime mortgage-related bonds and collateralised debt obligations.
Moody’s is introducing what it calls an “assumption volatility score” to address the risks of uncertainty around the assumptions and modelling that underlie a rating. However, in what is hardly a strong sales pitch, the agency says such scores “could help signal” that some of those deal types “have greater potential ratings volatility than other similarly rated securities”.
The agency is also bringing in a loss-sensitivity measure to address how vulnerable a rating is – or how big a downgrade it would receive – under a higher forecast loss rate for the assets to which a bond is exposed. This is designed to help differentiate between, for example, two slices of the same securitisation, both rated triple A but where one ranks junior to the other in repayment terms.
This second idea is closer to one of Fitch’s three proposals, for which it is assessing feedback from the past month. The first of these, a slightly different kind of loss-severity ratings scale, it believes would add the most value because of wide variation of losses after a default between traditional corporate debt and most structured finance.
Fitch’s other proposals are to increase disclosure of underlying collateral quality and to extend its “ratings outlooks” to give more of a view on medium-term ratings volatility. The agency said last month that meaningful scales of potential ratings transition, or volatility risk, would be difficult to develop separately from traditional debt ratings.
Fitch launched a ratings volatility scale for CDOs of corporate debt more than two years ago, but the initiative was canned in part due to lack of investor interest but also, industry insiders say, because it proved too great a drain on resources. Fitch would not comment further on this on Monday.
However, none of these directly satisfies Sifma’s direct plea in its paper last week that agencies “should ensure that ratings performance of structured products is consistently in line with [that] of other asset classes”.
S&P comes closest to this with its proposal explicitly to recognise stability as an important ratings factor, on which the deadline for comments is Wednesday. This move would act as a “limiting factor” on the rating that could be achieved by deals or structures that are most vulnerable to instability. It would mean that many of the deal types that have proved to be more volatile than expected could never have achieved top-notch ratings.
This will be unpopular with banks whose main business has been finding ways to get the highest level of return for the safest ratings.
But if the benign days when investors where desperate for yield by any means are past, that may not matter a great deal anyway – and S&P’s proposal may not be hugely relevant.