For the past few months, the federal government has been conducting the much publicized “stress tests” on the nineteen largest bank holding companies (BHCs) under The Supervisory Capital Assessment Program (SCAP). Agencies participating in the SCAP program included the Board of Governors of the Federal Reserve System, the Federal Reserve Banks, the Federal Deposit Insurance Corporation (FDIC) and the Office of the Comptroller of the Currency. BHCs were selected based on their relative size and combined significance in the U.S. financial system as illustrated by the fact that they hold two-thirds of the assets and more than 50% of the loans in the U.S. banking system. With the release of the stress test results on May 7, 2009, we have a clearer understanding of the process, findings and initial reactions of the participating BHCs.

While there is much skepticism regarding the process, the SCAP has several relevant principles that can be applied to enterprise business intelligence (BI) initiatives for organizations in financial services and other industries. This article examines some of the key aspects of SCAP for executives who lead other enterprise-wide BI initiatives.

SCAP Overview

The program was instituted in response to concerns that the country’s largest financial institutions have sufficient capital to provide a buffer against losses in the event of a severe economic downturn. The reason these assessments are significant is that regulators can require the BHCs who do not possess sufficient reserves to raise additional capital or convert government loans into stock, thus diluting shareholder equity and potentially triggering a drop in share price. As a result, the findings of the assessment process are being closely watched by banks, investors, consumers and other interested parties. Expectedly, shortly after the findings were published, the ten BHCs who were found to have insufficient capital began announcing steps they were taking to address the government’s findings.

While there is no shortage of guidelines for evaluating the nation’s banks, this process includes several departures from traditional approaches. Typically, examiners from each agency conduct their own evaluation in an independent manner. Findings from each assessment are unpublished and limited to distribution within a BHC’s business unit and the reviewing agency. Results are rarely shared across and between agencies. In contrast, SCAP is a collaborative effort involving multiple federal agencies who reviewed the overall health of each of the 19 BHCs, the macroeconomic influence of other economic drivers and the BHC’s influence on each other. Unlike other efforts, the agencies were very open in publishing the details of the process and the assessment results. The SCAP was carefully positioned as a forward-looking “what-if” exercise, versus a “forecast” of the two-year period 2009-2010 and considered each BHC’s viability under two scenarios. The first scenario (i.e., the baseline scenario) was based on a consensus of opinions derived from the government, the BHCs and professional forecasters regarding the length and depth of the recession. The second scenario (i.e., the adverse scenario) was designed to describe a more severe recession than the consensus-based scenario.

The SCAP Process

SCAP was organized as a collective effort across the participating agencies with the intent of determining each BHC’s ability to withstand a severe economic downturn. The assessment process was conducted with more than 150 senior supervisors, examiners, analysts and economists. Supervisory teams (“Teams”) were formed and organized to conduct evaluations by asset classifications, revenues, reserves and other factors. At the outset of the SCAP, the senior supervisors developed a methodology that included a common evaluation framework, consistent approach and clearly defined assumptions (e.g., the regulatory environment that existed on December 31, 2008; limited time horizon; focus on forward-looking losses rather than lifetime losses; etc.).

The process began with the collection of projections, data and other analyses from the government agencies, analysts, and the BHCs themselves. Collected data included estimated losses on loans, assets held in investment portfolios, trading-related exposures, and other information. The Teams reviewed the analyses and developed benchmarks that were used to evaluate each BHC’s own projections. The Teams then conducted a cross-firm comparative analysis using a variety of qualitative and quantitative methods at both macro- and micro-economic levels. Within each area, the Teams had to address and reconcile fundamental differences in accounting assumptions, data quality, timeliness, contextual relevance and other factors that might influence the assessment. These differences, as well as the findings, were compiled by the various Teams and reviewed in depth by the senior supervisors before being aggregated for final review. Once the analysis was completed, results were validated and aggregated for publication. The senior supervisors ultimately made the determination of the necessary capital buffer for each BHC. 

It is important to note that many of these types of analyses and findings had been done by various agencies in the past. This approach differed in several respects in that the findings were collected and shared across agencies and product lines, and eventually with the general public. While there were many challenges and significant discomfort with the idea of disclosing the results, there are some very good lessons for organizations with enterprise-wide business intelligence and risk management programs.

Lessons Learned

In every industry there are many examples of BI projects that have driven outstanding results while similar projects have resulted in total failure.  In addition, every vendor with any market traction has an extensive history of “wonder stories” and “skeletons in the closet.” The margin between success and failure is often more influenced by differences in process, people, culture, sponsorship and other factors than which vendor’s technology is employed. It is in these areas where key lessons can be learned from the initial rollout of this program.

Conclusion

While the industry can and has followed published guidelines, the complex and changing nature of the global financial ecosystem shows that traditional ways of thinking are likely to fall short of addressing new problems brought about by the dynamics of a globally connected economy. There are valuable lessons to be learned from the initial SCAP efforts and as the results are evaluated against the test of time. Among those lessons will be the determination of what blind spots were inherent in the initial assessment and how to improve the effectiveness of the follow-up process.

There continue to be many questions that have not been answered to the satisfaction of the American public such as: How did the billions in bailout money actually get used? Why aren’t the banks lending more money? How do changes in one area of a company or ecosystem ultimately influence the other areas of that system? Whether we are dealing with the nation’s banks, healthcare or the manufacturing sector, the opportunity for business intelligence and analytics to take on a greater role is evident. If the Fed continues to expand its role in the financial services industry, it is likely that future exercises will be conducted that go beyond the initial 19 BHCs. It would be prudent for executives in other financial services organizations to consider the adequacy of their processes, skill sets and technology strategies in order to provide the best chance for success.

Recent articles by Michael Brooks

• The Future Role of Business Intelligence within the Global Financial Community