Technical Analysis: Snake Oil

Let me ask you a question. What is the science of technical analysis called? 

[… crickets …]

Oh, yeah, that’s right. There isn’t a science of technical analysis! You can’t use the tools of science—repeatability, falsifiability, and an objective & systematic approach—to study technical analysis.

Instead, a technical analyst (TA) will look at a chart and begin connecting dots, highs and lows, drawing arcs, shapes, etc., really, by letting their eye find “patterns” in the chart. Once they’re done, they’ll “show” how the lines and shapes on the chart indicate what the future direction of the market will “likely” be. (Often, if they don’t see any good “patterns” in a chart, they’ll pick a different charting timeframe; say instead of a daily chart, they’ll switch to using a 4-hour or weekly chart instead).

And different TA’s will come up with different lines & shapes, and different predictions, unless, of course, they work together, in which case you’ll find a great deal of agreement among them. In short, there’s no objectivity nor a defined system for making these predictions.

And they won’t let you prove them wrong, because, they’ll say, the prediction isn’t certain. It is just more likely (and they can’t tell you how much more likely, just more). But has anyone ever proven that? That it’s more likely? No. No one ever has. Perhaps because the chart and the lines and shapes arbitrarily drawn, aren’t repeatable, and the TA’s story about them prevents you from being able to ever falsify their claims.

The TA will say, “it’s broken through resistance!” But in reality there was no “resistance.” The price just went above a given number—the same way it’s gone above a myriad of other numbers prior to this one.

You can look up at a cloud in the sky and it can look like a face, or a boat, or like some animal. It’s what human brains do: they detect patterns, ceaselessly matching the patterns in front of them with all of the things they’ve ever previously seen and attempting to associate the two. Trouble with TA is that the markets are random, not ordered, so whatever patterns one sees in a chart are, just like the doggie in the clouds, merely an artifact of a pattern-matching brain trying to cope with random data. You can’t predict the future of a market, and there is no dog.