Economaton

Technical analysis and fundamental analysis have been at war for a hundred years. Each has its zealots and adherents who disparage the other entirely and argue endlessly about which is the better way to trade the markets. Many people sit somewhere in between and use both in their trading decisions, which seems like the best bet to me.

Technical analysis consists essentially in looking for patterns in the path of volume and price over time (and in the paths of mathematical derivations thereof) to make trading decisions. Technicians have created a vast array of purported patterns and indicators for different purposes. To fundamental analysts, trained to look at factors like profitability and cash flow and book value to come up with a theoretical “real” value for a share of stock, this seems little different from tea leaf reading.

I take a middle view: I believe both technical and fundamental analysis can be useful tools. The fundamental value of a stock will dictate price movement (in the long run, and as Keynes said, “the market can stay irrational longer than you can stay solvent”), but this is not mutually exclusive with the existence of chart patterns that can uncover telling information on current buying and selling that can have predictive and tradeable value.

When we approach the financial markets, we are approaching a fundamentally human system. The markets are vast and complex, but still, in the end, run by people, who all have similar brains. Even with the advent of algorithmic trading, the programs are all still written by people to trade strategies designed by people — all that automated trading systems can do is trade these strategies faster.

Humans are not random. People like patterns. We like looking for them, we like creating them, and we like using them. This is what our big brains are designed for. In this sense, technical analysis is a closed loop feedback system: if people trade based on technical propositions, they will create a self-fulfilling prophecy by the very act of trading. In this sense, the question of whether these patterns are “real” in some deeper sense becomes irrelevant and meaningless; if people believe that technical signals are real and act as though they’re real by buying and selling based on them, then they’ll move price and become “real” in some sense.

That’s not the entire picture, of course. Not everyone is a technical analyst. The degree to which technical patterns influence price action will obviously vary greatly from market to market, depending on how thin the market is and what percentage of participants are using technical analysis. This brings us to a more interesting question: aside from thin markets where technical analysts move price by trading with each other, do technical indicators bear any relationship to the activities of non-technical market participants?

If the markets were completely random, there would be no way to invest intelligently. Results would be random. Throw darts at the stock page of the Wall Street Journal and buy whatever hits; it’s as good as any other strategy in a completely random market. Both technical and fundamental analysts therefore reject the efficient market hypothesis, and accept that it is possible to achieve better than random results, because there is some way to describe price movements besides random Brownian motion. The markets may still be considered mostly random, they are just not considered entirely random, and by learning the non-random parts, one can gain an edge and consistently make money.

Technical analysis provides a set of structures within which to conceptualize the movement of value. That’s it. There’s nothing magical about it — you can’t just learn a few chart patterns and then expect to become a millionaire in a month. Not all proposed patterns work, and the ones that do don’t work all the time; they just work a little bit more often than they don’t. It’s a tool, not a religion. Like any tool, it can help you, or it can hurt you if it’s misused. And like any tool, sometimes you can take every precaution and still hurt yourself accidentally.

There is plenty of anecdotal evidence that technical analysis works. Many traders at least claim to have made lots of money with it; examples in Jack Schwager’s Market Wizards series, among many other sources, suggest that even big investment banks use it internally. Opponents claim that these proposed success stories are just cases of survivorship bias.

I think the reason fundamentalists disparage technical analysis so much is that they are expecting too much from it (no doubt encouraged by the many TA snake oil salesmen who advertise miracle systems and seminars on the Internet and in trade magazines.) Flipping a coin (and assuming for the moment that the market moves randomly), the trader gets a 50% win rate over time. If price movement consists of mostly, but not entirely, random behavior (indeed, a basic proposition of fundamental analysis, too), then the patterns and indicators of technical analysis are ways to try and pick out the non-random components. The win rate doing so might increase from 50% to 55%. Or it might win 35% of the time with a 3:1 payoff. These win rates are profitable in terms of expected value over time, but only slightly. The skilled technical trader still loses quite a lot in absolute terms; he just wins more than he loses.

The argument against centers around the idea that technical analysis is inherently subjective, since more than one interpretation can be arrived at from any given chart setup, depending on what set of patterns and indicators is used, and therefore technical analysis is a kind of unscientific magical thinking.

Part of the problem is that there is a vast array of disparate tools all lumped together under the blanket term “technical analysis,” whereas fundamental analysis, by its nature, deals with a much more limited and well defined set of concepts. It’s true that, in this sense, “technical analysis” per se is not a testable and falsifiable theory — it’s an amorphous mass of sometimes unrelated theories. This does not imply that no particular elements of what is usually termed “technical analysis” are falsifiable and scientific, though; and indeed, there has been some recent progress in mapping out this area, such as David R. Aronson’s Evidence Based Technical Analysis.

Technical analysis is best regarded as a protoscientific discipline. At our current level of scientific knowledge, phenomena that affect prices and their description such as randomness, market forces, statistics, and human psychology are all minimally understood at best. The world is still trying to figure out how to piece all these things together coherently, and there are bound to be a number of false starts along the way — along with successes. I liken it to alchemy in the middle ages. Much of what was developed then became the core of modern chemistry once it was better understood. That’s happening even now with quantitative analysis. It’s just technical analysis under another name; a more precise mathematical formulation of similar pattern-seeking concepts.

There are no sure bets in the markets. Based on the empirical evidence of people who have used it, it seems likely to me that employing carefully chosen tools from the broad and diverse field of technical analysis can allow the clever trader to get a slight edge, but they’re not going to magically funnel money into your account any more than a hammer will build a house for you. It’s all about the trader.