Her cite isn't exactly a distortion, but it does greatly simplify what went wrong at LTCM.
Without getting into endless detail, the models they used at LTCM (which were substantial extensions of the B-S option pricing model) made certain key assumptions. When the assumptions failed, the models failed. Their mistake was believing that they has accounted for all possibilities, and the resultant refusal to get out of their positions and sit it out when things were obviously not working for whatever reason.
Not the first time, nor the last. In 1987 we had this thing call portfolio insurance. Another good, and still very useful model, so long as you understand that it presumes the existence of a liquid market and will fail catastrophically if you try to apply it in an illiquid one. When everybody ran for the doors at once on that Tuesday morning, the model failed. When politics intervened in international finance in 97-98 and caused some unexpected and irrational behavior, the LTCM models failed.
Same is true of virtually all financial models, including the B-S option pricing model. (Which is why it can't even be applied to a startup company with illiquid stock.)
Years from now some new "perfect" money making scheme will be invented by people who forgot the lessons of 98, 87, the early 70s (nifty fifty explosion) and all the others. And their models will ultimately fail catastrophically too... |
