A trader takes on portfolio theory.

The pseudo-science hurting markets? Or just economics in general. I've used the tools that he talks about ARCH and BS theory, and they handle strong discontinuities quite well. He mentions the stock market crash of 1987, and yet use of these mathematical tools precisely at the time were able to predict that crash. I know because I did it.

Now. Nassim Nicholas Taleb has some very strong words, some of them deserved: an excessive reliance on poor use of mathematical tools has, indeed, lost a great number of people a great deal of money. But what he puts out as the alternative, in essence an appeal to authority of his career is more or less the same argument that he excoriates when used by the other side. Some people are going to win, and others are going to lose. His career is also in a period of time that has been friendlier to finance and hedging of risk than any other. Financial genius is a rising market place and leverage JKG intoned, and he was right, then and now.

So what's really the problem with modern portfolio management? Is there a problem? Yes there is a problem, but it lies not in the mathematical tools that people are being taught, but with something that Nassim rightly points out as "self-serving" - that is the failure to understand that economics is the study of the games people play, and it breaks down, not because the mathematics are bad, but because the assumptions it is used within are. And the fault lies in economics, as Nassim argues.

ARCH and GARCH are tools that have proven their worth in the form of the VIX - the volatility index, which has quite reliably predicted major turning points in the market, and major peaks in local market runs, two of the hardest things to do. However, their utility for doing this is precisely their non-utility in a large number of other circumstances. Think of them as the National Hurricane Center, it isn't something you need to check every day, but on those occasions when a tropical cyclone is threatening specific interests, the NHC is must read material. GARCH most of the time doesn't tell us anything that "simple rules" tell us. Except when it does.

"Later, Robert Engle received the prize for “Arch”, a complicated method of prediction of volatility ..."

This is not really so. ARCH is not particularly complicated once you get used to it, and can be computed on spreadsheet by a high school student. So are all measures of volatility once you get used to them, and that is mathematics: you get used to things.

This failure to understand what these tools mean is at the heart of the failure of economics, the discipline, not economics the tool set.

The important thing to realize about economics is that it is about the boundaries of what people will put up with in the abstraction of exchange. As long as the operation of the abstraction of exchange produces results that people can manage and are willing to accept, the tools of economics perform admirably well. However, it is precisely at the points where people are no longer willing to put up with economic behavior, that risk management leaves the economic world, and enters the world of politics. When economics predicts outcomes, people want government out of the picture so they can gamble on uncertainty. When it does not, they want a government bailout of some kind. Right now governments are very prone to bail out wealthy financial interests, and so the tools of wealthy financial interests look much better than they ought to: "heads I win, tails you lose" is great work, if you can get it.

Nassim's article is a rant. It offers not one shred of proof for its blast of assertions, and as such, it has limited value. Yes there is a great deal of pseudo-science in economics, which wants to be the "physics of social sciences." However, the very tools he is talking about support, not deny, the thesis that he is arguing for: namely that crisis disturbs equilibrium, and during the break down of equilibrium there is the greatest need for real understanding of market failure. Market failure is not adequeately modelled, simply because market failure can be summarized as "situations where linear mathematics does a crappy job of predicting the market."

For example, GARCH could have been utilized by traders in 2002 to predict a steady long term decline in the dollar without a precipitous drop that would be recognized as a climax or event. Standard macro-economic theory, as present by, for example, Paul Krugman, would have had a rapid equalization of the imbalances in the US foreign accounts deficit and budget deficit. This event has not happened. The reason this event has not happened is because of the very fear of a significant drop itself. Everyone understands that if the dollar meltsdown, they are much worse off. However, strategically, there are ample chances for brinksmanship: sell as much as can be sold without pushing everything over the edge. GARCH has done an admirable job of predicting when volatility was so high that selling the dollar would abate. It has also done an admirable job of predicting peaks in the oil market, and in job creation.

Thus, while I share many of his problems with academic economics, he is shooting at some of his friends.