Risk Management

A big part of my job is risk management: identification of risk, establishing risk plans to mitigate those risks, and reporting all of this to my management and customer. On Sunday, January 4 2009, the New York Times Sunday magazine published an article on risk management as practiced (or malpracticed) in the financial industry, and how one analyst believed it was a key driver for the current economic situation. See:

http://www.nytimes.com/2009/01/04/magazine/04risk-t.html?_r=1&scp=1&sq=risk%20management&st=cse

for the article. I finally finished reading the article this week, and found it interesting, but unnecessarily alarmist. By that I mean that the critics of the practices offered neither realistic alternatives nor identification of any key middle ground between qualitative and quantitative risk analysis, and seemed to condemn the risk tools along with their misapplication.

First, there are folks who will say that unless you have a 100% solution it is no good. By those criteria, we would not even be riding horses, no less putting someone into space. The answer to that is obvious, and implicit in its definition: it's called RISK, which means you don't know something, and CAN'T know. See a similar argument in today's health care debate: unless you can insure 100% of the people, you don't have a viable solution. I think we all know that isn't true.

There's also a huge contingent that says they don't believe in numbers, ever…except that their pay checks are too small and their taxes are too high. Any attempt to use numbers for something is by definition flawed to them. Of course, there's no alternate path to gathering or analyzing data, but that doesn't matter until something goes wrong and we need someone to blame…or things go very well, and they want their share. We call people who fear data 'English majors'. There are too many of them.

Last, VaR (the risk tool criticized in the article) may not be a perfect tool, but it sounds like a good first approach. Yes, it only watches for 2 sigma risk (95%) on assets, with the criticism that it doesn't watch for 3 sigma (99%). Even assuming it did watch for 3 sigma, once that is achieved the criticism will be that we don't have 4 sigma. Pardon the pun, but 'go figure'.

Risk, at least as practiced by the finance industry, is a quantitative measure of qualitative events, and qualitative measure of quantified events. What they ignored was that the rules were changing. Make it easier for people to get loans and more people will borrow, and more will default. Taking away the safeguards puts us at risk. Strip away redundancy (for instance, fewer but larger banks) to achieve more efficient markets, and we are more at risk. Marketing risky investments as safe so that all the gains go to the few, and with impunity, is the final ingredient in the recipe for economic disaster.

There's something to be said for tradition. 20% down, no more than 35% of your income towards housing and 39% total debt, were initially established qualitatively…but they worked.