Thursday, February 26, 2009

“I never thought it would get this bad”

From Free Exchange:

"Can we test stress?
Posted by:
Economist.com | NEW YORK
Categories:
Financial markets

COMMENTER OneAegis asks an excellent question regarding the stress tests federal banking regulators plan to run on banks.

When this simulation is run and the answer is 75% of banks cannot survive this test, does the DOW fall to "merely" 6500, or to the 5000s? And what about the simple hit to confidence over the numbers the test chooses to use?

If there is one common refrain I’ve heard from every credible financial or economic expert about the crisis, from Paul Volcker to George Soros, it’s “I never thought it would get this bad”.

I am with them on that. Yes, we had a big real estate bubble which set the stage for big losses at some financial firms, and of course, America and Britain were over-leveraged. But that, it seems, should have set us up for a good old 1983 style recession. A worse downturn than we’ve experienced in the recent past, but fundamentally, financial markets did not have to collapse the way they did. Markets will come crashing down when investors lose faith, but it takes an extreme series of events for that to happen.

We can, at least partially, blame ineffective, timid, and inconsistent government policy for inciting the outright panic which brought financial markets to a halt. Most economists anticipated coordinated and clear policy reactions and markets that would price in an eventual turnaround. At this point, I am beginning to find the people who claim they predicted how the crisis would unfold incredible. It's tantamount to claiming foresight that the government would save Bear, let Lehman fail, then nationalise AIG, and then markets would freak out.

So as OneAegis points out, can the stress testing accurately test stress? Do the tests imply an omnipotent government policy? Do they build in a major sell off if unemployment reaches 10%? It may not feel like it, but things could get much worse than that. Once one moves to the tails, correlations often go to one and humans (including policymakers) become unpredictable. Perhaps the stress tests account for that. I’ll be mighty impressed if they do.

Unfortunately, the models and data we have are all we have. They provide some roadmap of what might lie ahead for embattled financial institutions and are a better alternative than throwing our hands up and saying “Citi is doomed”! (Though, perhaps, it is). But in the end, they can not alert us to all the potential hazards down the road."

Me:

"Q8: What will be the source of capital if supervisors determine that a banking organization
requires an additional capital buffer?

A: An institution that requires additional capital will enter into a commitment to issue a CAP
convertible preferred security to the U.S. Treasury in an amount sufficient to meet the capital
requirement determined through the supervisory assessment. Each institution will be
permitted up to six months to raise private capital in public markets to meet this requirement
and would be able to cancel the capital commitment without penalty. The CAP convertible
preferred securities will be converted into common equity shares on an as‐needed basis.
Financial institutions that issued preferred capital under Treasury’s existing Capital Purchase
Program (TARP 1) will have the option of redeeming those securities and replacing them with
the new CAP convertible preferred securities."

http://graphics8.nytimes.com/images/2009/02/25/business/stresstest.pdf

I might be missing something, but I don't see this as a test that they can fail. Isn't it simply an assessment to figure out how much money they need, and then simply a matter of how they're going to get it and under what conditions? It should be called " A Test To Determine How Much Money They're Going To Get". Maybe I'm wrong.
2/26/2009 2:58 PM GST

No comments: