In February 2008, Fitch issued a consultation paper seeking market feedback on proposed changes to its rating methodology for corporate CDOs. The proposal has two broad aims. First, the proposal seeks to identify and assess portfolio and transaction features historically associated with CDO underperformance. Broadly speaking, these risks relate to industry and asset concentrations, adversely selected assets, and reduced prospects for average recovery upon default of corporate assets. Secondly, the proposal seeks to better align CDO ratings performance going forward with other asset classes.
As Fitch had hoped, market participation during the consultation period has been considerable and productive, focusing on several constructive analytical areas. Fitch currently is undertaking additional analytical work related to the feedback received. As a result, Fitch now expects to issue final updated corporate CDO rating criteria no later than April 30, 2008 (rather than March 31, 2008 as previously announced). At that point, Fitch expects to publish its updated, final ratings methodology, discuss implementation plans, and lift the self-imposed embargo and begin assigning new corporate CDO ratings upon request.
Fitch received substantial feedback on the implications of its proposed methodology for existing CDO ratings. Analytical review of market feedback is continuing. Some of the modifications under consideration are expected to result in more moderate downgrades than detailed in the initial exposure draft. However it is too early to express definitively the magnitude of relief the points under consideration will yield. In addition, a detailed update on Fitch's intentions with respect to assessing existing ratings will be published concurrent with the final criteria.
Transparency and Feedback:
Fitch recognizes the importance of transparency in the ratings process and restoring trust and credibility in a fragile market, and has committed significant time and resources to the consultation process. In return, Fitch appreciates the efforts and contributions of so many market participants in providing feedback. Fitch conducted over 150 one-on-one meetings around the globe, as well as conducted conference calls, website broadcasts, a dedicated feedback e-mailbox, and regional conferences to collect market opinion on Fitch's proposal. The revised timeframe to the end of April will enable Fitch to analyse and incorporate market feedback, as well as provide better clarity on both the revised methodology, as well as its application to existing ratings. The one-on-one meetings in particular provided valuable, constructive input into the proposed criteria changes. Broadly speaking, feedback fell into two broad categories: analytical-based and implementation-based:
Throughout the consultation period, Fitch consistently received feedback in support of the aims of the new approach. That is, there was general support for an approach which uses unadjusted empirical inputs wherever possible, captures the systemic risk posed by industry concentrations, the idiosyncratic risk posed by obligor concentrations, and the risk of early CDO downgrade due to unexpected early migration in the underlying corporate portfolio. There was also broad support for the reduction in assumed recovery rates, reflecting the market's forward view of corporate credit risk. There were, however, specific points raised that Fitch believes warrant further analytical consideration.
Base Calibration and Risk of Industry Concentrations:
To calibrate CDO ratings against historical default cycles, hypothetical CDO portfolios were created that could be expected to exhibit performance characteristics consistent with broad-based, publicly available default statistics. Concerns, however, have been raised as to whether the baseline default data includes some level of industry concentration that was inconsistent with a well-diversified CDO portfolio. Fitch is re-examining default data to determine the level of any industry concentrations implicit in the historical figures, and whether this results in an unnecessarily high default threshold for diverse portfolios. On a related point, there has been varied feedback on the proposed correlation framework, and the way in which industry risk is captured. The final methodology proposal is expected to include an updated calibration, as well as an updated correlation framework to better capture the systemic risk posed by industry concentrations, particularly in high-yield portfolios.
Risk of Obligor Concentrations:
The notion of capturing the idiosyncratic risks posed by individual obligor exposures, particularly for investment-grade portfolios, was widely supported. However, the specific proposal was challenged as being overly punitive to high-yield portfolios, where portfolios are often smaller, and where the greater proportion of the risk is being captured in the base inputs. The final methodology will ensure that CDO liabilities can still withstand a minimum number of obligor defaults, an intuitive result. The proposal for capturing this risk, however, will be modified to balance the idiosyncratic risk measurement with the systemic risk measurement inherent in the base proposal, better distinguishing as needed between high yield and investment-grade portfolios.
Risk of early CDO downgrade (termed 'adverse selection'):
Market participants are generally in agreement that multiple-notch downgrades of highly-rated CDO notes early in the life of the transaction is undesirable. However, the proposed use of Fitch CDS Implied Ratings as a way to 'filter' portfolio credits with more immediate downside risk was met with some level of dissent. Fitch believes that CDS Implied Ratings can be a valuable tool for protecting CDO investors, particularly in the case of static synthetic CDOs. That said, other mechanisms may be equally effective and less problematic for managed transactions and Fitch, therefore, is considering other ways to identify and measure the risk of adversely selected names. Fitch is seeking to balance the identification and mitigation of the risk posed by 'excessive arbitrage', with the ability of a manager to usefully apply their expertise to add value to CDO note holders.
CDOs with Short Remaining Life:
Fitch has recognised that the current methodology proposal results in loss coverage well above targeted levels for portfolios with short remaining term to maturity. It has also been recognised that with a short term remaining, exposure to individual asset default events may identify CDO risk more effectively than statistical analysis. Consequently, Fitch will propose to supplement the model output with an event-risk matrix calling for loss coverage equating to a minimum number of obligors, depending on portfolio size and credit quality. The application of this supplemental criteria is expected to reduce, and in some cases, avoid negative impact on investment-grade corporate CDOs with remaining terms of less than three years.
Fitch's overarching goal is for ratings that are timely, accurate, and prospective, while providing consistency across the many sectors and products to which its ratings are applied. While these goals were generally supported in the feedback received, concerns were raised about the potential impact of the proposed methodology on existing transactions, including the potential for crystallisation of mark-to-market losses. Specifically, many market participants opposed any significant rating downgrades caused solely by model and methodological changes, at a time when underlying corporate performance, particularly investment grade corporates, has largely been sound. Additionally, some market participants felt it would be difficult effecting portfolio changes to align with model and methodological revisions given current market conditions in the proposed 90 day implementation timeline.
Between now and the proposed April 30, 2008 release date, Fitch plans to thoughtfully consider a number of these specific analytical points, develop and test consequent model changes where appropriate, and prepare a detailed implementation plan. All of this will be detailed in the final criteria report to be published at that time.