Mean reversion trading

What is mean reversion?

Mean reversion is the idea that a price which has moved unusually far from its recent average tends to move back toward it. Take a $100 stock that falls to $92 in two days, well below where it traded all month. Mean reversion says the next few days are more likely to be up than down. It says nothing about where the stock will be in a year.

The time frame matters. Over a week to a month, stocks that fell the most tend to win back part of the loss (Lehmann 1990, Jegadeesh 1990). Over three to twelve months the opposite holds: recent winners keep winning, which is called momentum (Jegadeesh and Titman 1993). Mean reversion trading in stocks works the short window: trades that last days, not months.

Why sharp drops tend to bounce

Part of a sharp drop in a healthy stock often has little to do with the company. It comes from selling pressure: a fund that has to sell quickly, traders cutting positions, a crowd reacting to a headline. That pressure pushes the price below what patient buyers would pay. When it eases, the price tends to recover part of the drop.

Economists describe the buyer in these moments as a liquidity provider, someone who takes the other side of a hurried trade and is paid for it on average (Nagel 2012). That also explains when the bounces are biggest: in nervous markets, when few buyers are willing to step in. In our test, dips bought while the VIX was above 30 made about five times the average trade.

A mean reversion strategy for stocks

The best-known version for stocks is Larry Connors' RSI(2) pullback rule (Connors and Alvarez 2008). In plain terms: buy a strong stock after a sharp two- or three-day drop, and sell as soon as it bounces.

1. The stock is in an uptrend: it closes above its 200-day moving average. This keeps you out of stocks that are falling for a reason.
2. It just dropped hard: its 2-day RSI closes below 10. RSI(2) measures how one-sided the last two days were; below 10 means almost all selling.
3. It trades enough: at least 1 million shares a day on average, and a price above $10.
4. Buy at the close of the day the signal appears.
5. Sell at the first close above the 5-day moving average, or after five trading days if that never happens.

The exit is what makes this a mean reversion trade: the 5-day average is the "mean" the price is expected to return to. The RSI(2) strategy guide explains the indicator and the settings in more detail.

Example: a trade from start to finish

Valero (VLO) fell from $413 to $377 in two sessions in September 2026 while its long-term trend was still up. Following the rules:

Bought at $377.14, sold at $387.18 three days later: +2.7% before costs. Note what the rule did not do: it did not sell on the first up day, because the price was still below its 5-day average, and it did not wait for the old high of $413. It took the bounce back to the average and got out.

When it goes wrong

Some dips are the start of a bigger fall, and nothing in the rule tells the two apart. J.M. Smucker (SJM) is an example from the same month:

It never closed above its 5-day average, so the rule sold it at the five-day limit: −5.7%. About 20% of trades in our test ended this way, losing about 4% on average. The winners pay for them.

Three things are worth knowing before you trade this:

• Losses come in bunches. Dips appear in many stocks at once during a sell-off, so several trades can go wrong together.
• There are losing years. In our test the rule lost money in 2011, 2018 and 2022, years with repeated waves of selling.
• Stops are a trade-off. A tight stop-loss cuts off many trades that would have recovered. Connors reported that stops hurt this rule on stocks (StockCharts); the five-day limit is the exit instead.

What to expect

We tested these rules on every S&P 500 stock from 2008 to 2026. In round numbers:

• About 67% of trades were winners.
• The average trade made +0.32% after trading costs. Small, because most bounces are small.
• Trades lasted 3.2 days on average, so the same money can be used again and again.

The strategy works through many small trades, not a few big ones. That makes costs and position size matter a great deal: a broker fee of a few dollars per trade can eat the whole edge on a small account.

Other kinds of mean reversion

The same idea shows up in other forms. Pairs trading watches two related stocks, such as two oil majors, and bets on the gap between them closing; on U.S. stocks from 1962 to 2002 a simple version earned excess returns of up to 11% a year (Gatev, Goetzmann and Rouwenhorst 2006). Whole markets also tend to partly reverse, but over several years, not days (Poterba and Summers 1988). And in statistics, a series is called mean-reverting when it is pulled back to a fixed long-run level, as interest rates and volatility tend to be. Individual stock prices mostly are not, which is why the effect in stocks is short-lived.

How Closing Note uses it

Closing Note runs this scan on the S&P 500 after every U.S. close. The full list of qualifying stocks is free on the swing trade candidates page.

The daily email goes a step further: it ranks that list and marks the top-ranked names, then tells you when each one hits its exit. Every top-ranked trade and its result, losers included, is in the track record (806 trades since January 2, 2026; a model record, not account statements).

How much to put into each trade, and how many to hold at once, is covered in position sizing.

Questions

What is mean reversion?

The tendency of a price that has moved far from its recent average to move back toward it. In individual stocks it shows up over days to a few weeks. Over three to twelve months the opposite effect, momentum, is stronger.

Is mean reversion trading profitable?

It can be, but the edge per trade is small. In our test on S&P 500 stocks from 2008 to 2026 the average trade made +0.32% after costs, about two in three trades were winners, and 2011, 2018 and 2022 were losing years.

What is the difference between mean reversion and momentum?

Mean reversion bets that a recent move partly reverses; momentum bets that it continues. They work on different horizons: short-term losers tend to bounce within days or weeks, while winners over three to twelve months tend to keep winning.

Which indicators are used for mean reversion?

Short-lookback RSI (two or three days), distance from a moving average, Bollinger Bands and z-scores. All of them measure how far and how fast the price has moved away from its recent average.

How long does a mean reversion trade last?

A few days. In our test the average trade lasted 3.2 trading days, and no trade was held longer than five.

References

1. Connors, L. and Alvarez, C. (2008). Short Term Trading Strategies That Work. TradingMarkets Publishing. ISBN 978-0-9819239-0-1. openlibrary.org
   Origin of the RSI(2) rules; tests on 1995–2007 data.
2. Gatev, E., Goetzmann, W. N. and Rouwenhorst, K. G. (2006). Pairs Trading: Performance of a Relative-Value Arbitrage Rule. Review of Financial Studies 19(3), 797–827. doi:10.1093/rfs/hhj020
   Pairs of U.S. stocks, 1962–2002: excess returns of up to 11% a year.
3. Jegadeesh, N. (1990). Evidence of Predictable Behavior of Security Returns. Journal of Finance 45(3), 881–898. doi:10.1111/j.1540-6261.1990.tb05110.x
   One-month reversal in individual stock returns.
4. Jegadeesh, N. and Titman, S. (1993). Returns to Buying Winners and Selling Losers: Implications for Stock Market Efficiency. Journal of Finance 48(1), 65–91. doi:10.1111/j.1540-6261.1993.tb04702.x
   Three-to-twelve-month momentum.
5. Lehmann, B. N. (1990). Fads, Martingales, and Market Efficiency. Quarterly Journal of Economics 105(1), 1–28. doi:10.2307/2937816
   Weekly winners and losers reverse the following week.
6. Nagel, S. (2012). Evaporating Liquidity. Review of Financial Studies 25(7), 2005–2039. doi:10.1093/rfs/hhs066
   Short-term reversal returns as pay for liquidity provision; rise with the VIX.
7. Poterba, J. M. and Summers, L. H. (1988). Mean Reversion in Stock Prices: Evidence and Implications. Journal of Financial Economics 22(1), 27–59. doi:10.1016/0304-405X(88)90021-9
   Index returns: positive correlation over short periods, negative over several years.
8. StockCharts ChartSchool. RSI(2). Accessed September 30, 2026. chartschool.stockcharts.com
   Summary of Connors' rules, the lower-threshold finding and his view on stops.