Many investors believe there may be a way to predict when to buy and sell securities.
One theory behind this belief is referred to as “Mean Reversion” – which can be defined as the idea that higher-than-average returns are followed by a period of lower-than-average returns, and vice versa. The extension of this idea is that if you can identify when this unusually high or low return period will end, then you could sell or buy securities at advantageous times and profit from this.
Applying the scientific process to the study of financial markets has provided us a much deeper understanding of them over the last 50 years
We believe an idea like this can be tested in this way to determine if it can be used to reliably enhance investor returns. This is one of the things we rely on our partnership with Dimensional Fund Advisors to provide. The embedded video from Jim Davis at DFA presents their study of Mean Reversion. Jim’s experience as a former academic comes shining through as he does an excellent job of introducing the idea and then describing their test of the theory. It’s worth a few minutes of your time. Summarizing, tests were run in 15 stock markets around the world and in multiple asset classes, using several different trading ‘rules’ in each to discover whether any of these rules returned promising results.
The vast majority of the tests produced negative results (implementing the trading rule was harmful to returns)
There are just so many challenges to profiting from a trading strategy like this; and Jim talks about some of these. While it’s interesting to note that some of these rules returned positive results, the mathematical reality is that so many tests were run that a few would be expected to produce positive results, simply by random chance. This is much like giving 1,000 people a quarter and asking them to flip it ten times. Approximately 1 of these 1,000 will flip ‘heads’ all ten times. This obviously doesn’t reveal the one person with skill in coin-flipping, it demonstrates probability in action.
While the existence of mean-reversion cannot be ruled out, there just doesn’t appear to be enough evidence of a way to implement this strategy in a profitable manner.
