In the November RMA Journal Digital Edition: Credit Risk Ratings and Noise

The November RMA Journal digital edition features an article on the impact of new credits, renewal plus modification to existing credits, and ongoing account management on credit risk ratings.

The article notes, “Ratings for wholesale credits can be distilled from market information based on equity prices and volatili­ties, credit default swap spreads, or similar market-based information. To the extent that there are multiple contributors to the market estimates reflecting consensus views, noise is likely to be diminished. Where market information is unavailable attempts to relate historical financial informa­tion in complex models to default rates have had mixed success. These began with early efforts by Edward Altman in 1967, where Z-scores for public companies were calculated based on discriminant analysis using five financial factors and their relation­ship to historical default rates. Other approaches using neural networks and other statistical techniques have also been explored. What is missing from these models is the application of judgment from experienced credit risk profession­als. Factors not easily quantified include the nature of industry risk; competitive position; technological advantage; product diversification; and management experience, integ­rity and leadership. One can’t easily regress or incorporate past judgment into these models and account for how much they might have contributed to the ratings. Judgment is particularly difficult to apply when the starting point is a complex model that stretches understanding, and it is unclear as to how much these qualitative factors may have affected default rates or previously assigned ratings.”

To read your Journal or subscribe, follow the link to the Digital Edition, which offers a convenient choice of viewing platforms, or download The RMA Journal App at the Amazon, Apple, or Google app stores.