Rating Agency Liability Wouldn't Have Prevented the Crisis

Over the past six months, there have been a lot of changes on Wall Street due to the new regulation thrust on the market by the Dodd-Frank bill. One industry that Washington sought to reform was the credit rating agencies. In particular, Washington sought to hold them liable for their mistakes. Yet, the new liability standard, as it should be applied, probably wouldn't have prevented the bad ratings that investors are so angry about. But that's okay.

There seems to be some confusion on this issue, however. In Bloomberg Markets Magazine this week, John Lippert appears to assert that raters can now be held liable for forming an incorrect opinion:

Largely overshadowed in the Dodd-Frank overhaul, 18 of the 848 pages strike at defenses that have allowed Moody's, Standard & Poor's and Fitch Ratings to defeat every ratings lawsuit they've faced. The bill says ratings aren't protected by the First Amendment as free speech or by securities laws that say opinions can't be wrong.

While that's sort of true, in practice it's more complicated. There are two different liability standards applied to the agencies by Dodd-Frank. One of them sought to expose the agencies to an expert liability standard, from which they were previously exempt. If their ratings are included in official deal documentation, now they could be held liable if they are wrong.

The Securities and Exchange Commission effectively nullified this standard, however, as explained here. When the bill passed, the agencies said that their ratings could no longer be included in official deal documentation, opting out of the expert standard to avoid that new liability. The market nearly seized as a result, as the bonds needed ratings to be sold. So the SEC saw that it needed to accommodate the agencies' decision and relaxed its requirement that ratings be included in deal documents. The SEC's rule is in place indefinitely, but it is not permanent. Eventually the regulator will decide the best way to solve this problem.

The other standard, still in effect, would allow an investor to sue an agency for knowingly or recklessly failing to conduct "reasonable" analysis. This is a more liberal rule than a strict fraud standard, which the agencies were always subject to. Time will tell how courts decide what a reasonable investigation or verification consists of, but presumably there's some wiggle room here.

How much should there be? It might help to go to the Dodd-Frank language to capture its spirit:

Because credit rating agencies perform evaluative and analytical services on behalf of clients, much as other financial ''gatekeepers'' do, the activities of credit rating agencies are fundamentally commercial in character and should be subject to the same standards of liability and oversight as apply to auditors, securities analysts, and investment bankers

It says the raters should be liable just like auditors, analysts, and investment bankers. But there's a key difference between the agencies and these other parties: the agencies must predict the future, and the others don't.

For example, auditors have a number of important functions. To name a few, they check the accuracy of documents, verify financial statements, and even compare financial assumptions to market standards. But that's all based on the past; they don't try to predict how a company or security will perform in the future. Some equity analysts provide price guidance, but if Goldman Sachs has a "Buy" rating for Apple and the stock declines, you won't get very far suing Goldman for bad advice. The information these parties are liable for is all based on factual data that helps to evaluate the present or past performance of a company or security.

That's why it's fundamentally bizarre to try to hold the agencies liable for their ratings. When an agency says that a bond is rated AAA, for example, it predicts a very low likelihood of its default. The only way liability could come into play in a meaningful way here is if a rater made identifiable mistakes in its analysis that a reasonable person shouldn't have made. That's different from if a big economic shock hits the market or if a bubble goes unseen by nearly everyone, and a bond defaults that wasn't expected to.

As a result, this liability standard wouldn't have conceivably prevented flawed assumptions about the housing market that raters made prior to the financial crisis. If they miscalculated statistics that led to bad ratings, however, then that's another story. Then, maybe they should be held liable. But if we're talking about guessing wrong about the macroeconomic future of the U.S. economy, then it's just crazy to try to hold them responsible. Most "reasonable" market participants did, in fact, make the same mistake.

Frankly, it's also better for the market if the agencies' opinions are treated as just that: opinions. Investors shouldn't be overly reliant on what the raters predict; instead, they should develop their own models and based on their own assumptions. They should treat the an agency's view as an opinion to consider, not infallible gospel. If there had been a little more diversity of ideas in the mid 2000s, then perhaps the housing bubble never would have expanded as dramatically and popped so disastrously.