October 29th, 2002, 6:18 pm
Yeah - mainly I was talking about the near term that business is bad for RM consulting.No, in case of a storm you hedge with weather derivatives, just like you buy life insurance for yourself and fire insurance for your home. Let the reinsurers cover the tail risk and blow themselves up.The reality is that the implementation costs for RM are gigantic, especially if you want to make the most of your diversification, and nobody is going to upgrade things on the capital economics alone right now. Also, the tighter the model, the more sensitive to the input data, and this in itself can be a major problem. The riskiest stuff is either never marked to market - so the timeseries is flat and shows no risk until you got blown up - or they mark it all over the place and you are driving your VaR with crap.My experience in the recent past is that people aren't using quantitative RM to make decisions... In particular, the focus is still on covering the risk of potential big loser positions rather than on optimising capital costs through diversification. That's definitely not to say there is no future in it long-term. On my first job interview ever, one of the senior guys at Susq alerted me to the fact that 25 years passed between Markowitz and the mutual fund explosion - that's an important thing to remember.