Breaking with Tradition – Why I Don’t Use Technical Analysis: Part 2

Breaking with Tradition – Why I Don’t Use Technical Analysis: Part 2

Traditional technical analysis is purely inductive reasoning that purports to observe reality, but because it lacks the most important input for any statistical model—that of the testable hypothesis—it lacks any ability to say anything concrete.

The scientific method is a form of positivist methodology, which centers on the falsification of a hypothesis or theory. If, through observation or experimentation, you can produce a predictable result that is in conflict with your hypothesis, then the hypothesis has been proven false. If you cannot produce such a result, then the hypothesis has the potential to be true and remains true until proven false.

The most famous example of this form of thinking is the black swan thought experiment. Our hypothesis is that, because we have never seen a black swan, all swans are white. The minute that we observe a black swan, our hypothesis has been falsified—that is, proven false—and we must necessarily throw it out, no matter how many times we’ve observed white swans. Just because we saw 100,000,000 white swans before observing a black swan does not mean that the black swan’s existence can be ignored simply because it does not fit our hypothesis (or model). The single instance of a black swan cannot be outweighed by our previous 100,000,000 observations.

Technical analysis cannot produce with any certainty falsifiable hypotheses, as it is based completely on inductive reasoning. As philosopher Karl Popper concluded, no hypothesis, theory, or proposition can speak about an observation unless it is falsifiable. Ultimately, the rules of technical analysis are more akin to observational studies that note correlations but cannot prove causality. When two things correlate with each other, it does not mean that one necessarily caused the other to happen.

In addition, the hypothesis itself must come from a deductive reasoning process, which should be based on our descriptive statistical data. Sherlock Holmes is the epitome of deductive reasoning, and it is what makes him such a seminal figure in literature. The deductive, methodical detective is something that we all strive to be in our own areas of keen interest. For me it is trading and investing. For others their passions may be inflamed by base- ball, the cello, or pancreatic cancer.

Inductive reasoning without the grounding of a deductively created hypothesis lacks any anchor to the human condition, because it is just data. Data can tell you nothing in and of itself. Data without context is like a graph without units on the axes. Therefore, it is only capable of providing observations about data that are self-referential and fundamentally faulty.

Generating an investment or trading thesis is central to my view of markets; therefore, these need to be tested.

Technical analysis relies on our ability to recognize patterns and it is rife with the problems outlined above. Since the markets are nothing more than the behavior of everyone who is trading those commodities, a chart of price and activity (volume traded) through time paints a picture for you of what everyone thinks of the subject of that chart. The technical analyst then strives to be an expert in pattern recognition.