Mean Reversion Bias
Mean reversion bias is the cognitive tendency of investors to assume that extreme outcomes — whether high returns, low prices, or poor performance — will automatically reverse toward average levels, even when there is no fundamental reason for the trend to reverse. This bias can lead to premature profit-taking, averaging down into declining investments, and systematic underweighting of trend persistence.
Key takeaways
- Mean reversion bias is related to the gambler's fallacy — the erroneous belief that past events influence the probability of independent future events.
- Investors exhibiting mean reversion bias often buy 'falling knives' prematurely, expecting a bounce that may not materialize, or sell winning positions too early.
- The bias reflects an overextension of the valid concept of economic mean reversion into contexts where it does not apply, such as individual stock price momentum.
- Mean reversion bias interacts with the disposition effect — investors who hold losers too long may do so because they expect mean reversion to eventually bail them out.
- Professional investors are not immune: academic studies document evidence of mean reversion bias in mutual fund manager trading patterns and analyst forecast revisions.
Explanation
Mean reversion bias represents the behavioral finance counterpart to the valid statistical concept of mean reversion. While prices, valuations, and macroeconomic variables do exhibit genuine mean-reverting tendencies over appropriate time horizons, the cognitive bias occurs when investors apply this pattern indiscriminately — seeing mean reversion where it does not exist, applying it over inappropriate time horizons, or expecting it to occur faster than the economic forces driving it can act.
The psychological root of mean reversion bias lies in representativeness heuristic — the human tendency to expect small samples to reflect the properties of larger populations. If a stock has declined for three consecutive months, an investor prone to mean reversion bias may believe the probability of an up month has increased, even if each month's return is essentially independent. This is structurally identical to the gambler's fallacy, where a roulette player believes that a long run of red increases the probability of black on the next spin.
In practice, mean reversion bias manifests in several characteristic investor behaviors. Value investors who screen for low P/E, low P/B, or high dividend yield stocks are engaging in a disciplined, historically supported form of mean reversion investing — but many retail investors who 'buy on dips' without fundamental analysis are simply exhibiting the bias, hoping for price recovery without analyzing whether the price decline reflects genuine fundamental deterioration or a temporary market overreaction.
The interaction between mean reversion bias and actual market dynamics creates both trading opportunities and pitfalls. During trending markets — particularly extended bull markets or sustained sector downtrends — mean reversion bias causes investors to fight the trend, repeatedly entering counter-trend positions too early and losing capital before the trend eventually does reverse. Research by Jegadeesh and Titman (1993) documenting the stock price momentum effect can be partly interpreted as the market's exploitation of mean-reversion-biased investors who sell winners and hold losers, allowing momentum traders to profit by doing the opposite.
Example
During 2021–2022, many retail investors purchased shares of Peloton Interactive (PTON) repeatedly as its stock fell from $145 to $120, then to $90, then to $60, to $30, and ultimately below $10. Many of these purchases were driven by mean reversion bias — the belief that such dramatic declines must be temporary and that the stock would 'bounce back' to its peak. In reality, Peloton's fundamental business model deteriorated as pandemic-era at-home fitness demand reversed, making the price declines partially or fully justified by deteriorating fundamentals rather than temporary mispricing.
Related terms
Behavioral Finance Dividend Dividend Yield Familiarity Bias Herding Behavior Home Bias Mean Reversion Overconfidence Bias Representativeness Heuristic Stock Yield