Rortybomb’s run-your-own-stress-test spreadsheet

By Felix Salmon
June 11, 2009

Mike at Rortybomb has the blog entry of the year I think; the background is here. Essentially he’s found a Rosetta-stone like table on page 6 of the stress test results, and has used it to create a fabulous spreadsheet which allows you to plug various different unemployment rates into a cell at the top and see what kind of capital needs result.

Of course there are lots of caveats — linear extrapolations are pretty down-and-dirty things. But the upshot is startling: if unemployment does rise more than Treasury feared it might, there are still quite a lot of banks which will be able to withstand the economic downturn and require no new capital. But let’s say that a realistic adverse scenario today has unemployment at 12.2% rather than 10.3%. What happens to banks’ capital requirements?

AmEx, BoNY, Goldman, and MetLife all remain at zero. A few banks require relatively modest cash infusions: FifthThird, for instance, sees its hole grow from $1.1 billion to $4.9 billion. JP Morgan requires $39 billion, which is probably doable; Morgan Stanley needs $15.6 billion, which might be a stretch. But look at Bank of America: rather than needing $33.9 billion, it suddenly needs to raise over $100 billion. In fact, BofA alone accounts for more than 25% of all the excess capital needs under this exercise.

Says Mike:

It really seems if you go out a bit all the losses are with a few specific banks, and maybe we should look into breaking them up before they become even more of a rotting albatross on our economy’s neck…

I was actually surprised – I assume turning up the numbers a bit would cause everyting to start leaking red ink. Instead it seems that if there is an additional slight downturn in the economy, we know the firms that will have all the problems. They are the ones that are too big to fail.

It seems that the result of the government’s ad hoc financial engineering over the past year or so has been to shove hundreds of billions of dollars of tail risk into a handful of enormous banking institutions. Which isn’t reassuring at all.

7 comments so far

The Stress Tests, like all government actions, create their own reality. By that, I mean to say that govt statistics, plans, etc., are intended to be real, but also useful. That’s why some less precise measures are often used by the govt. It’s easier for the govt to make use of them.

I’m glad that Mike has run his own stress test, but Mike’s not the govt.

“It seems that the result of the government’s ad hoc financial engineering over the past year or so has been to shove hundreds of billions of dollars of tail risk into a handful of enormous banking institutions. Which isn’t reassuring at all.”

I’m sure that it would have been better for the govt to say that. Pure FDR. Let’s all scream together. Look, we know that things could bad. These are projections. But when the govt says it, oddly, they carry more weight. There comes a point when worrying about tail risk becomes millenarian.

So thats why Paulnanke forced the tie up between Merril and BoA, better to deal with one big Albatross instead of 2.

Posted by Greg | Report as abusive

yes, pretty nice.

this shows the necessity of having a plan B which involves a rapid restructuring option in the available legal framework (good bank / bad bank type recapitalization).

oh — and who would you rather own the tail risk? someone’s got to own it.

Posted by q | Report as abusive

Such extrapolation may give you a (very) rough idea of the relative position of banks, but I’d put exactly zero credibility on the actual numbers.

1) Each bank’s portfolio looks different. Just look at the huge variability in the difference between the indicative loss expectations and the actual stressed losses for each bank.

2) A linear extrapolation ties all the economic variables together, so they move in lockstep. That clearly will not happen in practice. You can easily imagine house prices bottoming out at some point while unemployment and GDP continue to rise, for instance.

2) If we’ve learned anything over the last couple of years, it’s that many bank assets do not perform in a linear manner. A moderate deterioration in the economy may cause no additional losses, while a small further deterioration could lead to a wave of defaults.

Posted by Ginger Yellow | Report as abusive

Kudos to Rortybomb, this gives a good “spit in the wind” test. As for accuracy, this looks like someone estimated some crude partial sensitivities to each of the risk factors, and I can’t put a lot of confidence in that. Makes me think even less of the original stress tests.

We do significantly more complex calculations every year in the life insurance industry. I once served on the risk-management committee of a small bank, and the difference was amazing — the risk management of the bank was very crude.

Results of calculations whose methodology is hidden must be rejected as nonsense. Hidden calculation methods ALWAYS hide an agenda. But have fun!

Posted by Jimster | Report as abusive

We’re a group of volunteers and starting a new scheme in our community. Your website offered us with valuable information to work on. You’ve done an impressive job and our whole community will be grateful to you.

Post Your Comment

We welcome comments that advance the story through relevant opinion, anecdotes, links and data. If you see a comment that you believe is irrelevant or inappropriate, you can flag it to our editors by using the report abuse links. Views expressed in the comments do not represent those of Reuters. For more information on our comment policy, see