The other day I read about a group of banks that were subject to a numerical “stress test” by the US government. A few very notable banks failed the stress test and were “surprised” by the results and challenged the validity of the models. The banks took a swift and painful hit on Wall Street, vaporizing millions of dollars in market value.
Unfortunate for the banks, but in fact this should never have happened. In my work building simulation models I’ve discovered that the more progressive firms stress test themselves – sometimes many times a month, just to make sure that no important performance metric has been overlooked. They “stress” an abstract version of themselves, looking for the perfect storm of conditions that would give rise to bad result for the company. A few even model competitors, to see if similar conditions might weaken them enough to be exploited in a fair market.
If you know where the perfect storms are, there are plenty of things you can do about them. The very definition of perfect storm implies a chain of several conditions coming together at once to create a unique result. Want to prevent perfect storms? Simply take steps to ensure that one or more of the links in the chain never happen in parallel. You won’t completely eliminate risk, but you’ve just pushed the tail far enough to the right to raise your odds of success considerably.
So make it a practice to stress test all of the important systems in your company using a simulation model. Allow external voices to offer thoughts on what might happen in the environment, and test those “might happen” conditions in a systematic way. You’ll be glad you did when the government or some other actor tries to test you themselves.
I’m going outside now, and I’ll take my umbrella. Looks like a storm brewing.