Technically, Fallacies

There's a division in stock market philosophies that gets played up repeatedly: technical versus fundamental analysis. Fundamental analysis, though having its own problems, tries to ground itself in some basic qualities of a business -- how much money it has in its coffers (although the example of Enron and its ilk show that even accounting sometimes can't be trusted). Technical analysis, with its candlesticks, Bollinger Bands, flags and head-and-shoulders, spends a great deal of time looking at graphs, backtesting theories against old data and generally doing a good job at imitating empiricism. Ironically, it appeals to people who believe themselves rational... or maybe scientific?

There's a casino game called baccarat, one that isn't too often played but has a relatively small house edge (blackjack being one of the few that has a lower house edge). There are two sets of cards that come out for every round of the game, and although initially seeming complex, the play boils down to betting on which of those sets is higher or if they are equal. That's fine and dandy, but one of the things that casinos do is produce a baccarat scoresheet, which is essentially just a long rectangular grid of squares. Gamblers take that sheet and through their own systems, attempt to track the results of previous rounds and predict the results of future rounds. Since baccarat boils down to a sort of coin-toss, and anyone with a basic knowledge of odds knows that a series of coin tosses are completely independent (each toss has a 50% chance of being heads or tails), it's obvious that the only thing those meticulously recorded histories are doing is giving people the illusion of control. In fact, it's been proven mathematically that betting solely on the banker for every hand is the optimum baccarat strategy -- but people hate that, because... it's boring. So we discover the crux of gambling -- entertainment. This same urge is what drives investors and traders to develop theories that they feel might assist them. Essentially, they're all gamblers, but with longer timelines and more sophisticated systems.

One might argue that the stock market, being influenced by the actions of other players, means that paying attention to historical trends has a strong basis. After all, in poker games with players that you see consistently and learn to distinguish, knowing their previous play matters tremendously. Why doesn't the same apply to markets? Simple. Markets are composed of a group of people making decisions that are 1. too numerous to easily account for; 2. follow any infinite variety of systems; 3. interact with themselves (when people look at what the market is doing to determine what to do in the market -- they are the market). Computationally, even the idea of trying to model that behaviour is beyond our means. And still, in both cases, poker and the markets, we can only think about future actions as probability -- the passive, fish-like poker player of the past could suddenly turn aggressive, with no prior warning. Even if we had the computational horsepower, it would be pointless.

Ultimately, the investor or trader who forgets that they are making bets on future changes in the stock market and not on history is more naive, in my opinion, than one who admits that very little can be predicted. On the flipside, super-investors like Warren Buffett claim that paying attention to "intrinsic value" of a company and its stock, and ignoring the action of the market until it wakes up and realizing it was paying the wrong price, is the most effective means to make money. I have problems with that argument, too, but I'll address those another day.