Dilemma of Financial Modeling and Current Crisis So depressed by the recent credit crisis, I picked When Geniuses Failed again. Amazing how people could forget so quickly. The book spoke of the exact same crisis about Long Term. And here we are again, repeating the exact same mistake as we did in Long Term, only bigger. It seems that after every crisis, we grow stronger and take bigger risks, so we can fall more bitterly next time. How ironic! This is the exact thing, history repeating itself, that the book was talking all the time what had gone wrong with Long Term and those arbitrageurs. I remember when I first read it when it was published and got fascinated by it. I had friends in the firm. The book mentioned how wrong the mathematical models had been. And what a disaster these models had caused. The book talked about how rare events could happen more likely than what mathematical models could ever model. The book talked about how history could not have been repeated so mathematical models based upon historical data could be so wrong. The book talked about the independence assumption among various financial markets was wrong because in the real world, events should happen with high correlation. Yet at the end, the book left us with no conclusion of what exactly wrong with mathematical modeling. The main problem with mathematical modeling is NOT about mathematics but about economics. In basic economics, we assume that the market is PERFECTLY COMPETITIVE. A layman term for it is that all traders must take securities prices as given. No one, absolutely no one, can single-handedly affect the prices. The sufficient condition for this assumption is that everyone is a small player in the market. Because everyone is so small, his/er action is so tiny that should arouse no attention by others. As a result, if one finds an arbitrage opportunity (like Long Term), he or she can lever up 100 times to take advantage of it and makes tremendous amount of money WITHOUT AFFECTING THE MARKET. But look at what Long Term did. They placed large bids in the market. Many times, they were the biggest player of the market. So the perfectly competitive market assumption surely failed, and failed terribly. But ironically, this was how Long Term could make so much money – by bidding on large positions with ridiculously high leverage. Hence, on one hand, Long Term expects themselves to be a small player so no one would notice their transactions (that was why they always wanted to remain secretive and shy from the press) but on the other hand they wanted large bids that could eat up the entire world’s profits. It is not the mis calculation by the mathematical models that brought down Long Term and recent Lehman and Bear. It was the conflict and dilemma they got themselves in. As Long Term made more and more money and grew bigger and bigger, such dilemma grew deeper and deeper. Many articles cited greed, fraud, and mis practices as reasons of the current crisis. Many also blame financial modeling for worsened such crisis. This reminded me of 1987 Black Monday that the press blamed program trading for it. These accusations could be all true. But the point that I am trying to make is the dilemma of how financial modeling does and should work.