comments
What are Bank Stress Tests?
Email
Print
|
During the financial crisis, government officials often talked about performing “stress tests” on the financial institutions to see how they would fare if the financial crisis worsened. I didn’t really understand what they meant by “stress tests” because I didn’t see how you could do traditional stress testing, as you would on like a chair, on a bank. If you’ve ever been to an IKEA, you’ve probably seen them demonstrate durability with stress tests (the classic robotic pushing down on the POÄNG Chair and the labeled floorboards that talk about the number of people who have walked over them), but how do you stress test a bank?
In reality, a bank stress test is simply scenario analysis. What would happen if the stock market, say the S&P500 index or the Dow Jones Industrial Index, fell by 20%? 50%? What happens if the yield on the 30 year Treasury increases or if the Fed increases the federal funds rate? Analysts run these various scenarios and look to see how the bank will perform (i.e. survive) under these extreme conditions.
In learning that’s what stress tests were, I was surprised these weren’t part of the usual risk analysis process. I’ve always understood risk analysis to be a look at the probability and severity of a event happening, it appears that banks, perhaps in their zeal, ignored anything with a low probability even if it might sink the firm. As an investor, I’d like to know what the various risks to a company should things go north or south.
Then again, when your bonus is tied to annual performance, big wins in 2010 followed by catastrophic losses in 2011 still means a huge 2010 bonus.






Very interesting look at this. You’d think they would be doing this kind of risk analysis stuff anyway?
Just imagine if they stress tested the home borrowers during the subprime debacle.
It would be interesting to see the correlation between the banks level of vulnerability in contrast to the profit margins received (or executive bonuses issued). Because even though the stress tests are designed to insure stability the overall incentives for executives still condone risky behavior…IMO
I think that what needs to be appreciated is that this was the worst economic downturn since the 1930s. The default rate on mortgages and loans in general – the assets held by banks – has never come close to approaching the rates in 2007 – 2009. The Federal Reserve pushed short rates to 1% from 6% and then back up to 5%. This in an environment where homeowners took out all kinds of crazy mortgages and the percentage in adjustable rate mortgages skyrocketed.
The equivalent would be to tsk tsk over a leveled house and question why the homeowner had never tested the structure for withstanding a tornado.
I recently heard that the idea of stress tests for several states has been proposed to see how their finances will fair in the future. Something you would think that was done and understood anyways.
Part of the problem was that the assumed likelihoods of various negative events were lower than reality.
For Jim’s line “big wins in 2010 followed by catastrophic losses in 2011 still means a huge 2010 bonus”, maybe there should be a corresponding salary deduction in 2011; but then people would just quit after 2010; but if the bonus was in stock options that don’t vest for several years, maybe that’d work better.