An indicator that doesn't work can cause more harm to your portfolio that no indicator at all. Lesson: The simpliest explanation for most indicators being occasionally correct is coincidence.
A Market Indicator That Predicts Nothing
By Barry Ritholtz at Bloomberg View
Excerpt:
Every now and again, a way of looking at markets suddenly gains traction. Data gets assembled, analyzed, reviewed. Eventually, it becomes the basis of traders' decision-making process. It even can become part of Wall Street lore.
The problem that arises all too often is that this approach is statistically bogus. The data gets cherry picked; backward-looking analysis gets form-fitted to what just happened and has no meaning for what is most likely to happen in the future. Confirmation bias and selective perception can lead an investor to lose objectivity, choosing an approach that justifies an existing portfolio mix, as opposed to objectively evaluating the data.
Consider as an example the ominous-sounding Hindenburg Omen, a technical analysis that purports to signal the likelihood of a market crash. That's exactly what it's been doing — unsuccessfully — since 2010. This sort of recession porn allows people to confirm their existing prejudices. After five years of money-losing forecasts, the Hindenburg Omen’s following among traders is fading.
