Key Concepts: Reversion to the Mean
Reversion to the mean is a powerful force that investors cannot ignore.
Many market commentators refer in passing to “reversion to the mean” as something they count on for future returns, or as a possible risk. They rarely explain what the term means or signifies, which can be confusing.
This article aims to clear up that confusion.
What is Reversion to the Mean?
The concept of reversion to the mean originally comes from mathematics and statistics. It deals with averages and divergences from averages.
For example, seasonal temperature oscillates around an average. It is very unlikely to see Iraq with a freezing temperature and Sweden with warm weather in December. It’s not impossible to see those phenomena, but when temperature moves away from the seasonal mean we expect it to revert to that seasonal mean.
Measuring the mean and how far we are from it requires some quite complex mathematics.
Instead of looking at the math behind it, investors just need to understand the general idea. It means that any data point – like price, valuation multiples, or volatility – tends to gravitate toward its historical average.
For example, if let’s say semiconductor companies are priced on average at a P/E (Price to Earnings ratio) of 15 over an extended period. If we see semiconductor companies priced at a P/E of 40, we expect a reversion to the mean of 15. That would suggest that it’s not a good time to buy stock in semiconductor companies.
You can see why this is a powerful tool. It can help detect overvaluation or bubbles and find undervalued businesses. It’s no surprise this is a concept that value investors love.
How to Use Reversion to the Mean
First, you need to understand one limit of this tool. As this is essentially a statistical tool, it needs a lot of data.
A company IPOed 2 years ago would not have enough time to have enough data for its average price to be useful. Actually, using it for any individual company can be tricky, as any event specific to that company can modify the average.
The largest the dataset, the better. So for example, looking at the whole market average is likely to produce much better results than individual stocks.
The same is true with duration. If you use data over a multi-decade period, any short-term event gets smoothed out of the calculation. In 1999, looking at just the last 10 years would have shown a misleading constant growth of tech stock values. Looking at the last 40 years would have made you more careful and notice the risks of the tech bubble.
Some of the useful financial metrics for which you can use reversion to mean are:
Stock prices
Multiple valuations
P/E
Price to Free cash flow
Price to Sales
Margins
operating margins
gross profits margins
net profit margins
All of these metrics allow you to establish a mean and note a divergence from that mean. You still have to be aware of the limitations of this technique.
The Risks of Using Reversion to the Mean
So, if reversion to the mean is a known phenomenon, investors should it extensively, right?
Yes, but with caution.
The reason is that this is a pretty “dumb” tool that looks only at numbers, without qualitative data. So first, you need to understand if the conditions are still similar enough for older data to still be relevant.
For example, using reversion to the mean, you could have thought that a company like Kodak would get back on its feet and turn a profit again. That didn’t happen: digital photography was killing film for good and the company was going bankrupt. So being aware of disruptions and changes in the fundamentals of the business is vital.
The other risk is to ignore the question of timing. Yes, the metric you follow might go back toward the mean and that would make you a handsome profit. But this says nothing of when it will happen. An overpriced stock can stay that way for years. A depressed sector can take a decade to make a comeback.
This also implies an extra warning. Any strategy requiring timing should not rely on reversion to the mean. This applies to strategies like shorting a stock or buying put options.
Remember that there have been people aggressively shorting Tesla for the last 5 or 6 years. Even if they are proven right one day, they are more likely to have gone bust a long time before being proven right.
Observing that Tesla is overvalued relative to the mean is a poor basis for a short position. It’s not enough to know that Tesla is overvalued and will probably fall. To short profitably you need to know when it will fall, and the theory of reversion to the mean won’t tell you that.
The last warning might sound trivial but really isn’t. You need to assure that the data you use is of good quality. Do you have access to all the relevant data? If you are looking at foreign companies or markets, this might quickly be an issue. Small-cap stocks or OTC markets can be tricky as well.
If the market is valuing a company at a different level than its industry peers, you may be seeing an opportunity. There may also be a reason for that valuation, and you may not see that reason.
Conclusion
Reversion to the mean is a powerful force that investors cannot ignore. When valuation metrics drift far above their long-term average, caution is advised. And if they are far below average, maybe there is an opportunity to chase.
But this is also a limited tool that must be used with caution. Reversion to the mean might happen but in an unknown timeframe. Or it might never happen, because the industry or economic conditions have permanently changed.
Ultimately, reversion to the mean is a good example of why investing is both an art and a science. It is a very precise mathematical tool, but using properly requires not just knowledge, but also wisdom and experience.
Could You Share This Post?
Not only will you help your friends become better investors but recommendations are the lifeblood of Stock Spotlight’s growth. A few people sharing it on Twitter can really be a huge deal in terms of driving audience growth.
Thanks!
This article originally appeared on FinMasters.com: What is Reversion to the Mean and Why Does It Matter?