Analysts at rating agency Fitch Ratings warned recently against relying solely on historical credit default or ratings data to decipher the level of risk involved in a particular CDO tranche. Instead, using market-based methodologies provides a more accurate picture of the amount of risk inherent in a prospective investment.

"Basically we observed that the market-based approaches are superior to the other options in assessing asset correlations in a consistent way across different industries and different regions," said Ahmet Kocagil, a managing director at Fitch, on a conference call last week.

That means portfolio default probabilities based on equity movements are going to provide more accurate results than using other indicators alone, he said, adding that modeling based on the credit default swap market, while cluttered with less noise than the equity markets, is still too premature.

The primary downfalls of using historical rating agency credit default information or historical rating transitions to estimate a portfolio's default correlation is that - although the information is easily accessible - the rarity of defaults do not provide enough applicable detail, and most of the information is U.S.-centric, according to Fitch.

An increase in the average default probability shifts a portfolio's loss distribution to the right or left - depending on the direction of the move - whereas a shift in correlation between individual portfolio assets either fattens or thins out the tail of the loss distribution by transferring risk from the mean toward the tail, according to Fitch.

Because of an increasing level of asset correlation in CDO deals, the behavior of the so-called tail of the default probability curve is carrying increasing importance to investors, as risk will be moved to the tail, Kocagil said.

So far, players in the market are taking varied approaches to determine levels of asset correlation and default probability, Kocagil said, but "pretty much these institutions go with their assumptions without having a benchmark paper comparing methodologies."

Not comparing methodologies could lead to a skewed perception of asset correlation, either too high or too low, depending on the method used.

For example, relying solely on equity-based measures could result in inflated levels of asset correlation. "To the extent equity price movements tend to reflect some non-credit related information, correlations based on equity-based methods, especially those measuring inter-industry relationships are likely to be overestimated," Fitch stated in a recent report. But, this method still trumps the others because the information comes from liquid markets, and has a wider coverage in terms of industry and geography.

Some investors wonder whether the credit default swap market could provide a more accurate alternative to equities when modeling default behavior. While Fitch is currently conducting research on that matter, Kocagil said the rating agency's stance so far is that "CDS-based, as well as bond spread-based measured don't constitute viable alternatives to equity-based measures."

Kocagil said coverage issues are still a "major obstacle for the adoption of CDS spreads as an asset correlation modeling tool at present."

(c) 2005 Asset Securitization Report and SourceMedia, Inc. All Rights Reserved.

Subscribe Now

Access to a full range of industry content, analysis and expert commentary.

30-Day Free Trial

No credit card required. Access coverage of the securitization marketplace, including breaking news updated throughout the day.