View Full Version : Stock Market Errors

01-15-2002, 12:22 PM
Okay, suppose we go for the EMH, which states you can't hope to pick stocks, or time entries and exits. This basically frees you to make only about four errors, including

1) improper diversification, where you miss out on 2+2=5 opportunities, and instead put all your eggs in one basket and get 2+2=3,

2) trading in and out, and between stocks too much, to where commissions eat up your positive expectation,

3) failure to take advantage of opportunities that are fairly priced for everyone else, but offer a special value to you because you find yourself in a unique tax situation, and

4) taking on too much risk.

Just like there are errors in a zero-sum poker game which we cannot take advantage of directly as counter-parties - playing at limits that are too high for your bankroll, playing in a game where the drop is too big - the only ways to take advantage of these errors in the positive-sum stock-market game are to 1) market a mutual fund, 2) start a brokerage, 3) become an investment advisor, or 4) - and this is where you may begin to see an opportunity - play for washouts when downdrafts approach people's margins, and for parabolic runups when people betting more than 100% of their bankrolls causes immediate returns to have an exaggerated effect on future demand.

More generally, you can try to pick stocks - either by outsmarting everyone or getting some unique information or perspective - you can cook a new slice into the positive-sum pie by transforming one thing in which there is supply into another thing for which there is demand - by buying corporate bonds and then selling short to people who are only allowed to own goverments (assuming your utility for their expected divergence does not change), or by buying calls from covered-call writers, selling stock, and then writing puts for natural hedgers - or you can trend trade.

Of these different strategies, trend trading - or betting on whether other traders will extrapolate too much or too little as a given low-visibility demand situation develops - is the only one where you are really taking advantage of someone else's errors in the poker sense. And believe me, there is a game there, if only because people think there is.

Sure, there are some people who say trend-trading won't work simply because of commissions, and they will be wrong only to the extent commissions have plummeted, and because trend trading can be practiced over a very long timeframe. But there is another group of people who think trend trading is doomed to fail simply because it is the nature of markes to chop up and down without going anywhere. And these people - as well as trend traders themselves who get a little too bloated, a little too habitual - will pay your rent.

There is also the mathematical fact that the laws of perfect diversification dictate you should sell winners and hold losers - meaning trend traders, if nothing else, facilitate diversifiers in a positive-sum way!


01-15-2002, 12:51 PM
I rarely take positions that last more than a few weeks(aside from the SFranc/Dmark spreads I mentioned in an earlier post)But I will state the following.There are traders who I know who are "trend followers".The successful ones have made TONS of money.They go long or short and keep averaging up or down as the case may be.One woman who I have known is "incredible" at picking trends.I have seen her do it in everything from the cattle market to the S&P.She always tells me to follow her trades but it is not my style.I have rarely seen her be wrong over the last 20 plus years and she has made tens of millions of dollars.There is definitely something to be said

for trend following although I believe many of the funds get stopped out frequently when they try to take advantage of trend following.