One of the points that the administration's bank stress-test summary makes again and again and again is that the results are all rather hypothetical. "The estimates are not forecasts or expected outcomes," says page two of the results (emphasis in the original). They are "the products of a two‐year‐ahead 'what if' exercise conducted under two alternative macro scenarios." The first what-if scenario is the expected path of the economy. The second what-if scenario is "more adverse" -- a deeper and more protracted downturn.

The stress-test report says it is "virtually certain that the economy will not evolve in lockstep with either." Really? As far as I can tell the economy is evolving in lockstep with the adverse scenario. Or maybe it's a little worse.

Here's what the adverse scenario looks like, from last month's white paper:

So for 2009 it's 8.9% unemployment, a 22% decline in housing prices, and a 3.3% decline in GDP.

Boy, bad timing on the first one! The April unemployment figures just came out today, and they show an unemployment rate of exactly 8.9%. And at this point annualized GDP growth is actually worse than the adverse scenario anticipates. (Though no one really expects it to stay that bad.) And unless I'm badly misreading the Case-Shiller index (which I'll stick below), it shows a fall in housing prices of about 20%. So why is it "virtually certain" that this adverse scenario won't occur?



Update: I see Dean Baker has more along these lines.