Bank Stress Test - A Flawed Exercise

The bank stress test of the nation's 19-largest financial institutions is a flawed exercise that threatens to elevate the very economic-system stress it was designed to relieve.

The U.S. Treasury Department isn't scheduled to release the results of the much-ballyhooed bank stress tests until Monday. Little has been revealed so far, but one fact seems certain: Whatever information is disclosed is likely to be either too much - or even more likely - not enough for analysts, investors and the public to determine the soundness of a banking system upon which the nation's economic growth is predicated.

We're already starting to see bits and pieces leak into the public domain. And the response hasn't been positive.

Although the tests reportedly concluded that only one of the 19 banks that received a stress test would require additional capital, the government's own bailout-fund watchdog has questioned whether it was really much of a test at all.

The reason: The "adverse scenario" used for the test was "disturbingly close" to current economic conditions, said Elizabeth Warren, the chairperson of the Congressional Oversight Panel for the Troubled Asset Relief Program, and a frequent critic of the government bailout programs.

Now the government is urging foundering giants Bank of America Corp. (BAC) and Citigroup Inc. (C) - which have already taken in a combined $95 billion in taxpayer-provided bailout money - to raise more capital.

Flawed Assumptions Lead to Flawed Results

As outlined to the public, the stress tests were to pre-suppose a set of declining economic circumstances that would negatively impact bank balance sheets. For example, one scenario assumes that U.S. unemployment rises to 10.3% by the end of 2010. How or why the 10.3% assumption was chosen - as is the case with other scenario parameters - is unknown and is supposedly not to be revealed.

The assumption of testing through the end of 2010 means only that a two-year window was established for definitive calculations. Under this scenario, examiners assumed two-year cumulative losses of 8.5% on mortgage portfolios, 11% on home-equity lines of credit, 8% on commercial and industrial loans, 12% on commercial real estate loans, and 20% on credit card portfolios. The results are then totaled and weighed against assumptions - again unknown - about the capital positions of the banks at that time.

The fact that the assumptions themselves are a constantly moving target in our current crisis doesn't lend comfort to any baseline conclusions that may be reached. On the other side of the equation, the tests don't assume any revenue forecasts - either negative or positive - and may not assume further equity capital destruction or changing capital structures at the banks.

The idea is to take an expansive look at capital adequacy in the face of inherent credit risks, and exposure to off-balance-sheet liabilities, derivatives, or counterparty agreements. Once those risks are quantified, the government examiners will attempt to determine about how much capital is needed to support bank balance sheets and fluid liquidity. The net result of the tests is to identify which banks need to raise additional capital now to meet assumptions that may never happen, or may in fact be more destructive than assumed.

What's missing, unfortunately, is an assumption of how much additional capital would be necessary to facilitate credit expansion - which, in turn, would serve to fuel economic growth. That, after all, should be the ultimate stress-test objective.

In fact, the key problem with the whole stress-test exercise is that it does nothing to improve financial-system transparency - something I've said would be key to a true reformulation of the U.S. banking system. As currently conceived, there will be no clear assessments possible as a result of these tests upon which private investors can rely to provide the necessary capital to make up for any shortfalls. The stress tests may, in fact, have the opposite effect - and could discourage new equity investment in any of the banks.

Where Are the "Toxic Assets" Hiding?

Whatever is revealed through the convoluted prism of the stress tests should be compared to the relatively straightforward data available from other institutions, such as the Federal Deposit Insurance Corp. (FDIC). The FDIC's "Quarterly Banking Profile" offers a cut-and-dried summary of financial results for all FDIC-insured institutions.

According to The Financial Times' review of the data as of Dec. 31, loans outstanding were $7.87 trillion. The Times noted that Goldman Sachs Group Inc. (GS) economists estimate the shortfall in Tier 1 capital, given certain liberal assumptions, to be at least $753 billion.

The International Monetary Fund (IMF) has estimated that the potential losses of U.S.-originated credit assets held by banks and financial institutions around the world is $2.7 trillion, The New York Times said last week. While the percentage of those losses that sit directly on the books of U.S. banks isn't known, it is widely assumed that the value of impaired assets exceeds $1 trillion. To further complicate and obfuscate necessary transparency, new accounting rules replace mark-to-market reality with subjective internally modeled accounting of the value of distressed assets. And until those unidentified and convolutedly accounted for assets are removed from bank balance sheets, they will weigh down the worldwide banking system for years to come.

Perhaps the greatest danger the stress tests will cause may result from seemingly healthier banks pressuring the government to take back taxpayer-funded capital, while at the same time facilitating a dangerous lopsidedness to the entire banking landscape.

The fallacy that some banks are doing well and want to pay back government money is easily pierced. Whether it's for general liquidity, for credit spreads, or to backstop their efforts to raise additional capital, the truth is that there isn't a single bank that isn't currently using government support in some form. The only way to determine if a bank is healthy is to require it to stand entirely on its own and to not incorporate any cheap government capital or any government liquidity enhancing facility.

That will never happen in the current crisis. No bank is going to give up the preferential, cheaper cost of borrowed money provided by government vehicles when their competitors remain at the trough soaking up cheaper, government-backed capital resources. Until the entire system is self-sustaining, privately funded and supportable - and that distinctly assumes that some banks need to die a quick death and be buried - the system should be looked at as just that - a system.

Attempting to measure the health of every patient in the hospital by sticking a thermometer in only the sickest patients isn't going to do anything to stem the epidemic. The Treasury Department made a fundamental mistake in offering to reveal the results of stress tests. What it should have done is conduct system-wide tests on all banks, after which it systematically merged and shut down institutions that were either desperately threatened, or downright insolvent.

Yes, equity capital would be lost. But, as a lesson in moral hazard, it would be the clearest signal possible that shareholders are responsible for controlling boards even more so than boards are responsible for controlling management and risks at corporate institutions.

Until private-equity investors are confident that the nation's sick-and-injured banking patients are truly curable, there will be a decided reluctance to invest necessary capital into an uncertain future.

Nothing has been accomplished by these "stress" tests - except to further stress the banks, as well as the already-badly stressed U.S. economy.