Mean Reversion is a statistical phenomenon in which extreme events are followed by those closest to the mean.
In this post, I’ll use regression, mean reversion, and reversion to mean the same term. Don’t get confused.
To understand regression to the mean, you must first understand correlation, which is the strength of the relationship between two variables. Variables that are positively correlated move together in the same direction.
Variables that are negatively correlated move in the opposite direction.
Regression to the mean is a statistical phenomenon in which extreme events are followed by those closest to the mean. It occurs whenever the correlation between two factors is less than one. You can think of regression to mean and correlation as two sides of the same coin. The weaker the correlation, the greater the regression.
Mean Reversion example
For example, height is based on genetics and other factors such as nutrition and childhood illnesses. Due to the complex combination of factors that determine height, the correlation between parental height and child height is less than one. (This study in the British Medical Journal estimates it to be 0.5). Since the correlation is less than one, a regression to the mean occurs. This suggests that although tall parents usually have tall children, their children will likely be shorter than they are and closer to average height. But the reverse is also true: Tall children are likely to have parents who are shorter than they are.
Don’t fall into the trap of thinking that the prices of some financial assets can go up (or down) indefinitely or instead that long-term historical averages are ineluctably written in stone and cannot change in any way over time. As always, we need moderation and analyzing each specific case well.
The height of the world’s population is growing on average thanks to optimal nutrition, better healthcare and less strenuous work from an early age. If tomorrow there were a global cataclysm this trend would change.
Here the Macroeconomic conditions can change long-term trends and even reverse them. If tomorrow the USA were no longer a technological superpower where would big tech move?
Would they still drive the market?
I don’t know, but until it happens we shouldn’t worry, as the trend points in that direction, our investments will regress towards Big tech as world leaders in market capitalization.
Fact of the matter
My speech wanted to dwell on the futility of pursuing future trends that have not yet been realized, nor of seeing past performances to understand future performances. (See tall parents with children shorter than them).
As the dotcom bubble has receded, other bubbles will also return to their natural level. Just have to wait.