Editorial: Rating the raters

The US Securities and Exchange Commission's (SEC) report on its 10-month investigation into the manner in which rating agencies assessed structured financial products, and their role in the sub-prime crisis, makes for interesting reading. The review period went as far back as January 2004 and involved scrutinising all manner of internal records, including e-mail between staffers, staffing patterns/the competence of those doing the ratings, fee structures, deviations from known models, and so on "" over 50 staffers of the SEC, the report says, participated in this exercise. Not surprisingly, then, the report has enough documentation of the shoddy, even criminally negligent, work that the rating agencies did; of how they didn't always even have a model or a template that adequately captured the risks; of how even the templates available were frequently tweaked to the benefit of the clients paying for the ratings. Even if you don't understand the fine details, snatches of e-mail are enough to bring the point home. Thus, one e-mail has an analyst saying her firm's model didn't capture "half" of the deal's risk but that "it could be structured by cows and we would rate it" another says "our staffing issues, of course, make it difficult to deliver the value that justifies our fees". In another case, the report says, one rating agency regularly reduced the loss expectations on sub-prime second lien mortgages from what the model showed ... The list goes on.

 

In the event, the SEC has proposed some changes in the manner in which rating is done. One, no rating agency should be allowed to rate a structured product unless all the information is made public "" so that someone else, not receiving a fee, can use this information to do another rating. Then there are to be disclosures on whether different models are used for ratings surveillance as compared to the initial ratings; ensuring there is no link between ratings given and salaries/bonuses; and publishing of performance statistics so that the public knows just how well the raters have performed. As happens in all such reviews, some of the suggestions are meaningful, some aren't worth the paper they're written on "" prohibiting gifts of over $25 from those looking for ratings to those giving them is surely one of the latter, given how impossible it is to police. Similarly, instead of leaving it to raters to develop policies that address the conflict of interest which arises when the firm seeking a rating pays for it, why not ban such ratings and ensure that only stock exchanges on which such paper is traded pay for the rating?

But in order to have real impact on the way raters function, the SEC will have to go well beyond what it has done so far. After all, the reason why rating firms came into being was precisely because the ratings offered an advantage to those using them. So, for instance, the SEC rules allow broker-dealers to maintain less capital if they have higher-rated debt. Various regulations for banks (the Basle requirements) and pension funds, similarly, mandate the use of credit rating agencies. Putting all this power in the hands of the rating agencies was surely tempting fate. And while it is true that good ratings allowed many market players, such as banks and financial institutions, to suspend their judgment (how could anyone think that loans given to sub-prime borrowers could be prime, no matter how they were packaged?), all of them have paid a huge financial price. All except the ratings firms.