Is the stress test stressful enough?

Guidelines revealed Friday renew questions about whether the government's stress tests are stressful enough to account for the losses that might hit banks in a deepening downturn.

Regulators at the Federal Reserve, the Federal Deposit Insurance Corp. and the Office of Comptroller of the Currency have decided whether 19 giant banks - reportedly ranging from giants Citigroup (C, Fortune 500) and Bank of America (BAC, Fortune 500) to regional players such as KeyCorp (KEY, Fortune 500) - need to raise new capital.

Regulators are expected to begin revealing their decisions to banks next week, and to the public the week of May 4. Analysts are bracing for turbulent trading as speculation builds about how individual institutions may fare.

"We expect significant trading volatility in these 19 stocks between now and May 4," FBR Capital analyst Paul Miller wrote in a note to clients Friday.

Policymakers say the stress tests will help to ensure that big institutions are healthy enough to continue lending even if the economy -- which fell by 6.3% in the fourth quarter of 2008 -- weakens further over the next two years.

There were few specifics about individual banks in Friday's documents, though that will change.

"We're not prepared to discuss what will be released," a senior federal official told reporters Friday. But "we recognize it's important that this be a credible, transparent process and are working toward that end."

Yet the credibility of the tests will hinge in part on whether investors believe the tests capture the full brunt of the losses banks might face in a deeper economic downturn. And on that count, there are still reasons to be skeptical.
How bad will it get?

The Fed's paper outlined two stress cases. The so-called baseline scenario is based on the consensus expectation of economists, and the adverse scenario reflects a tougher environment.

Federal officials laid out the assumptions underpinning those cases in February. Since then, unemployment has continued to rise and many economists have downgraded their expectations for economic growth for this year.

But Friday's stress test guidelines show that despite signs of a deepening slump, the government's assumptions haven't changed.
0:00 /5:56Stressing the stress tests

The decision to hold the assumptions steady is sure to raise eyebrows among forecasters who have already questioned whether policymakers aim to produce test results that support a policy of forbearance -- letting the banks earn their way through the recession.

"Both the baseline and more adverse scenarios...are so optimistic that actual data for 2009 are already worse than the adverse scenario," New York University economist Nouriel Roubini wrote earlier this month.

To take one example, the adverse scenario envisions the U.S. jobless rate gradually rising from 6.9% at the end of 2008 to 8.9% at the end of 2009.

But the actual unemployment rate was already at 8.5% in March, according to government data. That's above the rate called for in the baseline scenario and up 3.4 percentage points over the past year.

Policymakers admitted that the odds have increased that joblessness will reach adverse-case levels of 10.3% by 2010.

Still, officials shrugged off concerns that the tests may not be stressful enough. They said that while the unemployment and economic growth trends had indeed darkened, there was no change in what they called a far more important factor in determining banks' credit losses: the pace at which house prices are falling.
Risk of an 'undershoot'?

Not everyone buys into this idea. FBR's Miller said in an interview Friday that his research suggests that if the unemployment rate hits 12%, "all these guys are going to have to raise capital," regardless of what happens to house prices.

The plunge of home prices in the U.S. has been well documented. The most widely watched barometer of U.S. housing price trends, the S&P Case-Shiller national index, dropped 29% between its mid-2006 peak and the end of 2008.

Regulators' adverse case envisions an additional 22% tumble in 2009 and a further 7% decline in 2010.

As it happens, declines of that magnitude would bring the national Case-Shiller index down to 100 -- equal to its level a decade ago, before factoring in inflation.

Buttressing the regulators' case, a decline that steep sounds unlikely.

"To be most useful," they noted in a 21-page document released Friday, "stress tests should reflect conditions that are severe but plausible."

But a plunge in house prices beyond the stress case may not be implausible.

Over the long haul, house prices tend to move in concert with factors such as rental rates and household income -- and some of those numbers suggest a decline that steep may not be unlikely.

The Census Bureau's personal income data show that median household income actually declined, in inflation-adjusted terms, between 1999 and 2007. That was before unemployment began its surge and the economy turned sharply negative. Lower income, if the relationship holds, could mean lower house prices.

Of course, income is only one factor in determining home prices, and efforts to revive the economy may well keep prices from tumbling that far.

But some well-known economic bears have said they expect housing prices -- having risen well above their long-run average during the bubble earlier this decade -- to fall further than such data would predict.

The housing crisis "is likely to get a lot worse than currently anticipated because markets do overshoot," billionaire financier George Soros said last year in an interview in the New York Review of Books. "They overshot on the upside and now they are going to overshoot on the downside."

Of course, officials are aware of the havoc a steeper housing price drop could wreak on banks' balance sheets. The Fed is trying to shore up prices by keeping interest rates low and buying up privately issued mortgage securities, for instance.

Federal Reserve chief Ben Bernanke said last week that policies aim to avoid a "major undershoot" in house prices.

An undershoot could be particularly painful for big banks with substantial mortgage portfolios in big cities where prices haven't fallen as much so far.

In New York, for instance, prices fell 9% in 2008. But they would have to drop an additional 24% to reach the year-end level of the S&P national index -- and more than 30% to reach their inflation-adjusted 2000 level.