In mid-March, the Federal Reserve Board announced the results of stress tests it conducted in early 2012 on 19 large U.S.-based banks. Although most of the banks under review passed the tests, the exercise raised a number of concerns within the banking sector.
Recall that in early January, the 19 banks were required to submit plans regarding how they would manage their capital under a severely adverse macroeconomic scenario over a nine-quarter horizon. The scenario was designed in November of 2011 and took into consideration the problems in the euro zone and in the mortgage markets. The goal was to ensure that the banks had adequate capital planning processes that would enable them to weather periods of economic hardship.
The Federal Reserve then evaluated the plans submitted by the banks, and made its own projections of each bank’s probable losses, revenues, expenses, and capital ratios under the hypothetical stressed environment. Banks whose calculations at least matched the Fed’s were allowed to carry out their capital plans. After passing the stress test, U.S. Bancorp (USB) increased its common stock dividend by 56%. Regions Financial (RF) sold new common stock and repaid $3.5 billion of preferred stock sold to the government in 2008 under the bank bailout program.
However, in a number of cases, the government’s estimates did not match the banks’ more positive figures. The banks have sought more information regarding the models used by the government in the stress tests. Although the Federal Reserve says it is open to suggestions regarding how it might improve the process, it has refused to reveal how it arrived at its calculations. In a recent article in the American Banker, reporter Donna Borak indicated that the Fed may fear that “banks could reverse-engineer an outcome that is more favorable than it should be’’.
But not knowing exactly which factors the Fed took into consideration in computing its figures makes it difficult for the banks that failed the test to correct any shortfalls in their capital positions. And in the case of Citigroup (C), the government calculated the company’s basic capital ratio (the Tier I ratio) at 4.9%, just short of the 5% bar required for a bank to be allowed to carry out its plans to buy back stock or increase dividends paid to shareholders. Many question whether falling short of the 5% mark by only ten basis points is meaningful.
Many also criticize the assumptions behind the severe economic scenario used in the stress test. These included unemployment in the United States peaking at just over 13% in mid-2013 (as compared with slightly over 8% as of this writing) , U.S. equity values falling by 50% from their 2011 September-quarter values through late 2012, and U.S. housing prices (which may be close to stabilizing) declining another 20% by the end of 2013. The government has admitted that such a scenario is unlikely, and most consider the assumptions highly unrealistic.
Too, the banks, which had only six weeks to submit their capital plans to the Federal Reserve, say they need more time to prepare for the complex and important stress tests.
Meanwhile, banks that are not large enough to be included in the top 19 are worried that they will be eventually required to undergo similar stress tests even though most don’t engage in activities that are as risky as those common at the largest financial institutions. Moreover, the personnel and other resources needed to prepare the stress tests aren’t as extensive at the mid-sized and small banks, so the tests might impose a heavier expense burden on the smaller players in the industry. At present, the tests are mandatory for institutions with at least $10 billion of assets and recommended for those with at least $500 million of assets.
Certainly, in a world fraught with economic and financial risks, and in light of JPMorgan Chase’s (JPM - Free JPMorgan Stock Report) recent trading blunder, determining a bank’s capital strength is a valuable exercise that may serve to head off potential problems. Nonetheless, there also appears to be some room to improve the process.
At the time of this article’s writing, the author did not have positions in any of the companies mentioned.