Under the newly imposed stress test rule required by the Dodd-Frank Wall Street Reform Act (DFA), the Federal Deposit Insurance Corporation (FDIC) stipulates that covered financial institutions, including all U.S. banks with consolidated assets between $10 billion and $50 billion, are required to conduct annual stress tests calculated under a multi-scenario analysis.
A "stress test" is generally described as the process to assess the potential impact on the consolidated earnings, losses, and capital of a covered financial institution over the planning horizon; considering the risks, exposures, strategies, and activities.
The final rule pushed back the compliance date for model submission until October 2013, and additionally delayed the public disclosure provisions until the completion of the 2014 data collection cycle. The delay applies only to those U.S. banks or subsidiaries that did not participate in the Supervisory Capital Assessment Program (SCAP).
The purpose of the stress test rule is to monitor capital adequacy. That is, each covered financial institution must model their capital sufficiency to absorb losses resulting from adverse economic conditions. As a result, covered financial institutions will be liable for systems modeling starting in 2013.
As required the DFA, the FDIC is coordinating the rules, test scenarios, reporting and disclosure with the Federal Reserve Board, the Office of the Comptroller of the Currency, and the Federal Insurance Office to ensure consistency between the Federal regulatory agencies. Although complete specificity regarding the scenario analysis application and subsequent disclosures remain forthcoming; the lack of details will likely not be a mitigating factor for any additional time extension.
The regulatory guidance confirms that capital adequacy is one of the critical factors that will be analyzed when conducting stress modeling supervisory activities. The stress test rule requires a minimum of three variations of model-based economic scenarios including: (i) baseline; (ii) adverse; and (iii) severely adverse. The baseline scenario reflects the macro-economic outlook based on consensus views, government agencies and other public-sector organizations.
The adverse scenario requires historical data input of economic variables consistent with mild or moderate recessions. The severely adverse scenario requires historical data input of economic variables consistent with severe recessions together with real-time correlations. For example, a covered financial institution with significant trading activity must model a global market shock component in both the adverse and severely adverse scenarios and to measure potential market and counterparty risks. The severely adverse model should embed overlay scenarios using real-time or recent data inputs.
Synthesizing the rule with the requirements contemplates that a regulatory-sufficient model will capture inputs that include real-time broad economic data and asset prices to effectively monitor and detect procyclicality. Procyclicality in this context refers to the exacerbation of business cycle tendencies caused directly or indirectly by regulatory requirements.
For instance, real-time measurement of procyclical shifts can mitigate the deterioration of the lending market in times when loans are necessary to keep the economy moving.
Integrated models will need to account for the prescribed Basel III-defined capital conservation buffer and be capable of procyclical analysis. This is true because during periods of accelerated credit expansion, a banking organization would need to hold an additional capital buffer which could only be measured during business-cycle stresses as well as event-specific tail risks.
The FDIC anticipates that the stress test scenarios will be revised annually to ensure that each stress scenario remains relevant under prevailing economic and market conditions. These yearly revisions will enable the scenarios to capture evolving risks and vulnerabilities.
Although Federal regulatory agencies maintain a reservation of authority to require a covered financial institution to use different or additional model methodologies, this could likely be avoided if the deployed models are commensurate with an institution's risk profile and relevant activities.
For instance, in addition to real-time monitoring of procyclicality and connectivity to historical simulation data, a stress model should adequately estimate losses, revenues, and changes in capital positions over the planning period, including securities, trading losses, counterparty exposures, and requisite regulatory capital ratios.
A properly designed and integrated risk system that can comply with the stress test rule will likely avoid model disqualification under the regulators' reservation of authority and will only need to be updated based on the annually prescribed revisions. As a result, with the right risk systems in place, covered firms should be enabled to stress less about stress tests, as it would appear that the control is held by the financial institutions.