Fitch Ratings expects to change the ratings methodology it uses for collateralized debt obligations to bring it in line with the methodology it uses on n-to-default structures, according to Stefan Bund, senior director in London. The agency published a report on the importance of correlation in rating n-to-default products last week.
Fitch currently uses a Monte Carlo simulation for n-to-default ratings, which allows the agency to incorporate correlation assumptions into its models. For rating CDOs, however, Fitch uses estimates of recovery rates by examining different scenarios. Monte Carlo simulation is a well-established tool for cash flow analysis and should make the recovery assumptions more accurate, according to Alexander Batchvarov, managing director in structured credit research at Merrill Lynch in London.
Alex Reyfman, credit derivatives strategist at Goldman Sachs in New York, said this move should bring the rating agency's methodology more in line with valuation models and that can only be a good thing.
