The Stress Test: Time for Transparency

Like many people, I was disappointed by the Financial Stability Plan announced on Tuesday. But I think there is one glimmer of hope: the “stress test.”

A key component of the Capital Assistance Program is a forward looking comprehensive “stress test” that requires an assessment of whether major financial institutions have the capital necessary to continue lending and to absorb the potential losses that could result from a more severe decline in the economy than projected. (Emphasis added.)

The stress test is supposed to indicate which banks are healthy and which aren’t (so they can be fixed or closed). We need this for the reason most of you already know: nobody thinks the banks (meaning, mainly, the big ones) are healthy. The New York Times has a good summary of the situation. Nouriel Roubini thinks U.S. banks are facing another $1 trillion in write-downs. The IMF thinks it’s more like $500 billion. The only people who think the banks are healthy are the bankers themselves:

“Our analysis shows that the banks have varying degrees of solvency and does not reveal that any institution is insolvent,” said Scott Talbott, senior vice president of government affairs at the Financial Services Roundtable, a trade group whose members include the largest banks.

Edward L. Yingling, president of the American Bankers Association, called claims of technical insolvency “speculation by people who have no specific knowledge of bank assets.”

The first reason to ignore Talbott and Yingling is that whenever someone says something, but he would say the exact same thing even if it weren’t true, you should ignore what he says. (When Stan O’Neal said Merrill Lynch was great at risk management, what did you expect him to say? That they sucked?) The second reason is that even if they are right, it doesn’t matter, because no one believes them. The banks don’t face an immediate liquidity crunch because so many of their liabilities are guaranteed by the government: expanded FDIC insurance, guarantees on newly issued debt, the Fed accepting their junk assets as collateral. But even so, as long as people think the banking sector is sick, the economy will continue to stagnate, and the deepening recession will make banking assets worth less and less, until finally the banking sector really is sick. If Talbott and Yingling have some secret knowledge that their banks really are healthy, it won’t help them.

The stress test theoretically will prove once and for all who is right, and whether the banks can withstand a worst-case scenario (remember, every month the forecasts for the recession are always worse than the previous month). This is a core regulatory function, for two reasons. First, as long as the government is backing up bank liabilities, it has the right to monitor them in order to protect itself; this is what the FDIC does, and why it closes banks when they become insolvent. Second, insofar as the health of the economy as a whole depends on the banking system, all of us, in the form of our government, have the right to know if the banking system is working properly.

When the stress test is applied to Some Big Bank (SBB), it will have one of two outcomes. If the regulators determine that SBB is insolvent and has to be taken over, that would be a good outcome, first because the outcome will be credible to everyone (except maybe Talbott and Yingling), and second because then the government can clean SBB up (that is, transfer the bad assets off its balance sheet) without having to negotiate with anyone. If, however, they determine that SBB is healthy (or can be made healthy with a little extra capital), that could be good or bad. If they just assert it (”believe us, we’ve looked”), then we are right back where we started, because that outcome will have no credibility. Fairly or unfairly, there are already too many people who think that Tim Geithner is in Wall Street’s pocket. The only way to gain credibility in the event that SBB passes would be to publish all the details of the test: details about all of SBB’s assets (including all of the assets that underlie those assets) and all of the assumptions and models that were used in the test. That way independent researchers could come to their own conclusions. Of course, I’m not sure how feasible a stress test on this magnitude itself is, let alone such a massive exercise in transparency, and SBB’s traders would all scream. But I don’t see any other way out.

Calculated Risk seems slightly optimistic about this exercise.

Yves Smith, not so much.


Originally published at the Baseline Scenario and reproduced here with the author’s permission.

2 Responses to "The Stress Test: Time for Transparency"

  1. Guest   February 14, 2009 at 11:20 am

    Blah blah blah. Did it ever occur to you that these are simply Mussoliniesque moves in an oligopolistic shift to corporatism? Read your fascism, dingbat.John RyskampP.S. And be prepared for the opposite outcome–banks fail the stress test and then…take them over? NO WAY! Banks fail the stress test and then…place 401k funds, pension funds and money market funds under Federal control. THAT is what comes next. Stop living in your well-padded dreamworld, idiot.

  2. Anonymous   February 20, 2009 at 9:36 am

    I can imagine tremendous political pressure to either (1) rig the test standards, or (2) debate the results in order to protect special interests and jobs.Most of us expect at least some of the biggest banks will fail, but we also know that the financial service industry has tremendous lobbying power. I suspect the political ‘wind’ will determine how the stress test standards are set. If Geithner and company have already decided to break up those banks they can easily rig the test and use it for political cover to justify their action. If it’s a fair test, how do you decide how hard to make the ‘stress’? If you say real estate goes down another 25% most banks are in a world of trouble. If you say things will stabilize here, most will make it OK. So the critical issue will be how the stress test is designed and what assumptions are used. I hope they get it right.