On Wednesday, the Treasury Department will release some details of the much-ballyhooed “stress test” it plans to apply to all banks with over $100 billion in assets. While it’s possible, I suppose, that Secretary Geithner will disappoint the markets yet again, I doubt that will happen. What’s more, I expect that investor concern surrounding the banks will ease palpably once the details of the Geithner stress test (assuming they’ll be what I think they’ll be) become known.
I should say up front, though, that I continue to believe that the whole idea of a one-size-fits-all stress test is a really dumb idea. I also believe that the likely introduction of a new capital ratio as a key bogey is a dumb idea, as well.
Why? First, the big banks already are examined and stress-tested by their on-site examiners-in-charge. The EIC is in a much, much better position to take into account a bank’s unique characteristics in sizing up its inherent risk than a 40,000-foot stress test could possibly be. And if it turns out the EICs don’t do a better job than Geithner’s new test, one of the first things he should do is fire them all.
Second, I’m not crazy about the likely adoption of a new ratio, tangible common equity to risk-weighted assets, as a key new regulatory standard. The only reason Treasury has decided to look at TCE/RWA now is that spooked investors, determined to view banks in the worst possible light, have become semi-obsessed by the number. As general rule, in my view, investor anxiety should not drive regulatory policy.
Regulators typically don’t just pull new measures out of the air, after all. In the 1980s, the Fed, the OCC, and the FDIC all had their own capital ratios and minimums; it took years of analysis and debate among them to settle on which measures were truly the most important. (They were Tier 1, total capital, and leverage.) It doesn’t seem to me to be good regulatory practice to adopt a new one in the current crisis atmosphere.
I’m not a fan of the test, or the new standard Treasury plans to use. Nor do I have any inside knowledge about what the test will look like. Here’s my best guess, though, of what to expect:
- The economic assumptions. I expect the “stress” in the Geithner stress test will translate into an economic forecast that includes a 10% peak in unemployment rate, a 40% decline (peak-to-trough) in U.S home prices, and a recession that lasts until late this year. It’s no coincidence, I believe, that these assumptions are what popped up in the New York Times yesterday, and are the ones being used by JPMorgan Chase (JPM) in its in-house stress scenario.
- Maximum cumulative loss assumptions by loan category over some period (I believe through 2010). I only hope the government will adjust these cumulative loss assumptions by institution, to take into account factors such as loan geography, experience, underwriting practices, pricing, and so forth. If it doesn’t, the output of the test is apt to be highly arbitrary.
- Maximum cumulative loss forecasts will then be measured against an institution’s existing loss reserves, pre-tax, pre-provision earnings, as well as various measures of capital.
- Two capital ratios will be the focus. The resulting pro forma maximum income or loss over the test period will then be used to calculate estimated stressed Tier 1 capital and TCE/RWA ratios.
- The moment of judgment. For a bank to pass, it will have to show pro forma, stressed ratios above certain minimums. I believe those minimums will be 6% for Tier 1 capital and 3% for tangible common equity to risk-weighted assets.
- The fate of those that fail. If an institution fails to meet either of those two minimums, it will have to raise new capital by some deadline, perhaps April 15. It might raise new equity in a secondary offering, for instance. Under another option, the bank might convert its existing TARP preferred shares into mandatory convertible preferreds at an exchange ratio based on the common’s price as of a certain date. That date will be important, since it could be the date that Geithner first discussed the plan (when bank stock prices were higher) or it could be the then-current date.
If the capital hole still can’t be filled, the Treasury might require an investment of new capital via a newly issued mandatory convertible preferred with even more severe operating restrictions attached. I assume, for instance, that any institution that needs to issue new preferred over and above the preferred investment already will have to replace its CEO.
This table shows the effect that conversion of existing TARP preferred into mandatory convertible preferred would have on 2008 year-end tangible book value per share and tangible common equity to risk-weighted asset ratios of 102 large-bank TARP recipients. Obviously, a full conversion at today’s prices would be very dilutive to book value. But for most companies, conversion would provide a big cushion to absorb whatever losses result from the stress test.
Again, assuming full conversion of the TARP preferred for all 102 companies (which will not happen) this group of banks would end up with an average 10.3% tangible equity to risk-weighted assets ratio. The group is correctly trading at just 1.1 times pro forma book values.
I repeat, I’m no fan of the stress test concept. It’s apt to rely way too much on cumulative loss forecasts by loan category, and will likely be overgeneralized and too severe. That said, I expect that tomorrow’s disclosure of the test’s details should be good for bank stocks, since investors are likely to gain confidence in the capital strength of our largest banks to weather even the most severe credit storm.
So much for my speculation; we’ll all be tuned in for details today. Please, Secretary Geithner, no more disappointments!