The term "regression toward the mean" refers to the statistical phenomenon where a variable that has deviated significantly from its average is likely to return closer to that average in the future (if you’re interested, this is an excellent article that expounds on the subject and that should take about 10 minutes to read). It’s a concept that has applicability in a wide range of endeavors, including in psychology and economics. For example:
In economics, it is often used to explain changes in economic growth or other macroeconomic variables. For example, countries that experience a period of high economic growth are likely to experience slower growth in the future, and countries that experience a period of low growth are likely to experience faster growth. Economists use regression analysis a lot. This article published by the International Monetary Fund is fun (if you are math geek, that is).
In the investing context, "regression toward the mean" refers to the phenomenon where an asset class's performance tends to return to its long-term average mean after experiencing a period of extreme returns. In other words, if an asset class has had a very high return in a given period, it is likely that its future short-term returns will be less than its average over time. And vice versa.
It’s at the heart of value investing (the idea that one might want to buy underperformers and sell over performers). It also is one of the bases of rebalancing. If you do not know what “rebalancing” means in this context, do yourself a favor and read Investing Basics for Beginners Installment #3: Never Put All Your Eggs in One Basket.
If this is done using a basket approach (i.e., doing it with many stocks or among asset classes), it can make sense. But doing it with individual stocks ignores all the fundamental factors that go into determining a stock’s value.
Or stating it less politely, it’s stupid to try to apply regression analysis to an individual stock. Doing so would suggest that you should sell your stock in a company solely because that company’s stock has been steadily increasing since regression to the mean would suggest that the stock price is likely to return to average levels in the future.
But that’s not reality.
But with respect to well-established (read: not crypto) stock market sectors populated by companies that actually make money? Hell, yes.
|