By Jody Martin
Note: The Dodd-Frank Wall Street Reform and Consumer Protection Act does not introduce any capital planning stress testing requirements on banks with less than $10 billion in assets. This article simply highlights the benefits of forward-looking analysis as outlined in existing supervisory guidance, and does not introduce new regulatory requirements.
Stress testing has been a risk management tool of financial institutions for decades, but supervisors and bankers are placing renewed emphasis on this approach. One of the biggest appeals of stress testing is that it takes a forward-looking view of performance. In contrast, many of the other methods used to assess risk give only a historical, or point-in-time, perspective. While this can certainly be informative, it provides us with an incomplete view of existing exposures.
The impact of the 2007–2009 recession on financial institutions has taught both bankers and supervisors the dangers of assuming that history will always repeat itself. Supervisors and bankers are also concerned with knowing what the performance expectations are going forward. This forward-looking view is an integral part of the stress analysis. As with any effort to predict future performance, uncertainty is introduced in the process that you won’t encounter if your analysis is limited to historical information. Documenting the reasoning, judgment and assumptions that went into the forecast can help the various stakeholders assess the level of uncertainty around the results.
Generally, the Federal Reserve is recommending a quarterly, two-year horizon for stress testing. That is, forward-looking analysis should encompass future performance estimates for each of the next eight quarters. This rule of thumb provides enough of a forward view to be useful in decision-making, without going so far out in the future as to make the estimate unreliable.
In addition to being forward-looking, stress testing allows management and supervisors to evaluate performance under a variety of conditions. As noted above, one of the biggest lessons the recent recession has taught us is: “Don’t assume that economic or market conditions will stay the same.” A rigorous stress testing framework allows for the introduction of scenarios that reflect current conditions as well as changes in conditions. The combined results of all the scenarios assessed will provide management information about potential risks, which should then inform the strategies employed to manage those risks.
Determining which risk drivers to adjust is critical to developing a beneficial stress testing framework. Stakeholders need to examine the exposures, and areas of operations that are being analyzed, deconstruct the risk factors and apply varying assumptions about those factors over a future time horizon. Simply put, the firm must engage in the process of asking, “What might happen, and how will that affect what we do?” Or, “What can go wrong, and how will we manage it?”
As firms begin the process of developing a formal stress testing framework, it may make sense to initially limit the number of risk factors that are being adjusted and, consequently, limit the number of resulting scenarios. Once the firm is comfortable with the stress testing process, the variables stressed and scenarios created can be increased, as is reasonable for the scope of the firm’s activities.
Senior management and the board of directors can be instrumental in determining the risks, exposures and lines of business that will benefit most from forward-looking analysis. It is frequently true that the process of developing the stress testing framework — identifying the areas of focus, defining the scenarios, setting the parameters for the stress variables — provides as much benefit to the bank as the results of the analysis itself. This is because the process of developing a robust stress testing framework will engage staff from across the firm to provide input about each of the analytical dimensions, ensuring a comprehensive view of the risks being analyzed and reinforcing the strategic priorities of the company.
One of the biggest benefits of performing forward-looking analysis is that the same principles can be applied across a wide variety of bank functions. For example, individual credits can be stressed to allow risk managers to assess the effectiveness of underwriting standards; loan portfolios can be stressed to identify risk concentrations; investment portfolios can be stressed to identify market sensitivities and to evaluate investment standards; operational infrastructures can be stressed to assess information technology capabilities; and enterprisewide stress scenarios can be used to inform strategic planning, contingency funding and liquidity planning, as well as capital planning. These activities all contain the same basic elements:
This final step, the review of stress test results, will help stakeholders assess the effectiveness of existing risk management practices and identify potential areas for improvement.
For all the benefits provided by stress testing, there is still resistance to implementing a stress testing framework at many institutions. One of the concerns that supervisors hear most often is that stress testing requires complex models or statistical estimation methods in order to be effective. Supervisors don’t believe this is true. There are a number of approaches that can be easily introduced using nothing more than standard spreadsheet software. This is valid whether performing loan-level analysis or enterprisewide analysis. However, supervisors do expect that the sophistication of a bank’s stress testing framework will be consistent with the level of systemic and inherent risk in the bank’s exposures and business activities in order to ensure that risks are effectively evaluated.
One thing that is necessary for an effective stress testing framework of any level of complexity is access to data. That is, access to accurate and timely data that is relevant to the risk being examined, and data that is available in a manner that allows for analysis. This almost always means data that’s available in electronic files or databases. If a firm’s key risk data is contained in hard copy form within loan files, then the firm must first develop a better framework for managing risk data before trying to build a stress testing process. Again, supervisors don’t believe it’s necessary for smaller firms to make significant new technology investments. Simply change the firm’s business processes so that it becomes standard practice to capture the necessary risk information in electronic form at the time the exposure is created. This information must then be updated as necessary. Introducing these practices is essential to a sound risk management function and not just specific to performing forward-looking analysis. Once the practices are in place and the data is available, the ability to conduct effective stress testing is greatly enhanced.
While conducting stress testing will require bank resources, there are significant benefits to be obtained — benefits that far outweigh the costs. Firms that integrate stress testing within the risk management infrastructure, rather than treating it as a “regulatory exercise,” will be better prepared to weather future volatility within their markets. The process of conducting forward-looking analysis enhances internal communication within the firm about risks and risk management practices, provides opportunities to re-examine and reinforce the board of director’s strategic direction for the firm, and complements existing risk management practices based on the analysis of historical results. Additionally, the stress testing framework can improve the dialogue between bankers and supervisors regarding the firm’s expectations for its markets, exposures and business activities.
Further information can be found in the following guidance:
For firms with $10 billion or more in assets, there is also proposed guidance relating to stress testing.
Jody Martin is a Risk and Policy team leader with the Charlotte branch of the Federal Reserve Bank of Richmond. He can be reached at firstname.lastname@example.org
The analyses and conclusions set forth in this publication are those of the authors and do not necessarily indicate concurrence by the Board of Governors, the Federal Reserve Banks, or the members of their staffs. Although we strive to make the information in this publication as accurate as possible, it is made available for educational and informational purposes only. Accordingly, for purposes of determining compliance with any legal requirement, the statements and views expressed in this publication do not constitute an interpretation of any laws, rule or regulation by the Board or by the officials or employees of the Federal Reserve System.
Supervision, Regulation & Credit