Mean reversion: what it is and how traders use it

Written by the Tactfolio research teamPublished 6 min read

Mean reversion is the idea that a price that has moved far from its usual level tends to drift back toward its average over time. A mean reversion trader buys after a sharp drop and sells after a sharp rise, betting on the snap back. It works often enough to be studied, but some prices fall for good reasons and never come back.

Key points#

  • The mean is the average. Mean reversion means “moving back toward the average.”
  • It is the opposite bet to momentum, which assumes winners keep winning for a while.
  • Research finds both patterns, at different time spans: momentum over months, reversal over very short and very long spans.
  • Tools like RSI and Bollinger Bands measure how stretched a price is.
  • The big risk is a price that keeps falling because something is truly wrong.

What is mean reversion in trading?#

Mean reversion in trading is a strategy that bets a stretched price will move back toward its recent average. It is one of the oldest ideas in algorithmic trading. Say a stock usually trades near $50 and suddenly drops to $40 with no lasting bad news. A mean reversion trader buys and hopes for a return toward $50.

Three questions define any mean reversion rule:

  1. Average of what? Usually a moving average of closing prices over a set number of days, such as 20.
  2. How far is too far? For example, 10% below the average, or two “standard deviations” below it. A standard deviation is a measure of how much prices usually wander from their average.
  3. When do you get out? When the price is back at the average, after a set number of days, or at a stop if it keeps falling.

Mean reversion vs momentum#

Mean reversion and momentum are opposite bets. Momentum says recent winners keep winning. Mean reversion says recent extremes fade.

Mean reversionMomentum
The betBig moves partly undo themselvesBig moves keep going for a while
BuysRecent losersRecent winners
Typical holding timeDays to weeks, or years for long-term value betsMonths
Hurts most whenA drop turns into a long slideA trend suddenly flips

Both can be true at once, over different time spans. In 1993, Jegadeesh and Titman looked at US stocks that rose most over 3 to 12 months. Those winners tended to keep beating the losers over the next 3 to 12 months (Journal of Finance). Over longer spans, the research points the other way.

Trend following is the momentum idea applied to whole markets. Some investors hold both kinds of rule, hoping one does well when the other struggles.

What does the research say about mean reversion?#

Two classic studies found mean reversion in stock prices over long periods.

De Bondt and Thaler (1985) asked, “Does the stock market overreact?” They grouped US stocks by how they had done over the past few years. The past losers went on to beat the past winners in the years that followed. They argued that investors overreact to dramatic news and push prices too far (Journal of Finance).

Poterba and Summers (1988) tested stock markets in the US from 1871 and in 17 other countries. They found returns tended to continue over short spans and partly reverse over long spans of several years (NBER). They suggested slow-fading “fads” might push prices away from fair value for years at a time.

Neither study says reversal is fast or reliable for any one stock. They describe tendencies across many stocks over long periods.

How do traders measure “too far”?#

Traders measure how stretched a price is with indicators, which are formulas run on past prices. Two common ones:

RSI. The relative strength index runs from 0 to 100. It compares recent up days with recent down days. A reading under 30 is often read as “oversold,” and over 70 as “overbought.” Our RSI guide covers how it is calculated.

Bollinger Bands. John Bollinger’s bands draw a line two standard deviations above and below a 20-period moving average (Bollinger Bands). A price at the lower band is far below its recent norm. A price at the upper band is far above it.

Neither indicator predicts a bounce. Both only tell you a price is unusual compared with its recent past.

A worked example#

Say a broad stock fund has averaged $100 over the past 20 days. Your rule:

  • Buy when the 10-day RSI drops below 30.
  • Sell when it rises above 70.

The fund slides for a week to $92 and its RSI falls to 25. You buy. Two weeks later it is back at $99 and RSI reads 72. You sell, up about 7.6% before costs.

Now the other case. The fund slides to $92, you buy, and the slide keeps going. A month later it is at $80 and RSI has stayed low the whole time. You are down 13%, and the rule still says “hold until RSI rises above 70.” This is the trade that hurts mean reversion traders. Decide in advance how you would handle it.

When does mean reversion fail?#

Mean reversion fails when the price was not “stretched” at all. Sometimes a price falls because the thing itself changed.

  • Broken companies. A company losing its business can keep falling until its shares are worth nothing. There is no average to return to.
  • Long bear markets. In a slow crash, “oversold” can stay oversold for months. Buying every dip means buying all the way down.
  • New normals. When interest rates or a whole industry change, the old average stops mattering.

Funds that hold many stocks are less likely to go to zero than single stocks, as our guide to ETFs vs stocks explains. That is one reason many people test mean reversion rules on broad index funds rather than on individual companies.

Common mistakes#

  • No exit for losers. A rule that only sells on a bounce can hold a loser forever.
  • Too many small trades. Short-term rules trade often, and costs add up.
  • Testing only on survivors. Stocks that went bankrupt are often missing from free price data. That makes buying dips look safer than it was. Our backtesting guide covers this bias.
  • Ignoring the trend. Buying dips in a falling market is fighting the bigger move. Some traders only buy dips while the longer trend is up.

Try it on Tactfolio#

You can test simple mean reversion rules in the strategy builder. It has an RSI reading and an Enter / Exit block. That block is made for rules that buy when RSI drops below 30 and hold until it rises above 70. There is no built-in Bollinger Band reading today. You can build something close with a moving average, a price standard deviation reading, or a 21-session change z-score. The builder guide lists every reading. Stock and ETF strategies run on daily prices, so they cannot trade within the day. They are long only, so they cannot bet on falls by short selling. Every backtest is hypothetical.

Common questions#

What is mean reversion in simple terms?#

It is the idea that prices that move far from their average tend to move back toward it. A mean reversion trader buys after big drops and sells after big rises. It does not always happen, and it can take a long time.

What is a mean reversion strategy?#

It is a set of rules that buys when a price is unusually low compared with its recent average and sells when it returns. Common signals include RSI below 30 or a price below the lower Bollinger Band. Good strategies also set an exit for trades that keep losing.

Is mean reversion the opposite of momentum?#

Yes, in the bet it makes. Momentum expects recent winners to keep winning, while mean reversion expects extremes to fade. Research finds momentum over months and reversal over very short and very long spans.

Does mean reversion work in the stock market?#

Studies have found long-run reversal across many stocks, and short-term bounces after sharp moves. It is far less reliable for any single stock. A stock can fall and stay down if the company is failing.

What is reversion to the mean?#

Reversion to the mean is the statistical idea behind mean reversion. Extreme results tend to be followed by more ordinary ones. In markets, it means unusually large moves are often partly undone later.

Sources

Put it into practice