Strategies

What is Mean Reversion?

Reviewed by Karthikeyan Anbarasan (Founder, TradeLyser) · Updated 19 Oct 2026 · Editorial policy

DefinitionMean reversion is a trading approach built on the idea that a price stretched far from its recent average tends to move back towards it. A large-cap NSE stock with a 20-day average of ₹1,500 and a standard deviation of ₹30 that drops to ₹1,440 is two standard deviations below the mean, a typical buy signal for a mean-reversion trader targeting ₹1,500 (illustrative).
Formula
z-score = (Price − Moving average) ÷ Standard deviation of price over the same period
Example
(₹1,440 − ₹1,500) ÷ ₹30 = −2.0, the same level as the lower Bollinger Band (20, 2)

How mean reversion is identified

A mean-reversion setup needs three things: a mean, a way to measure how far price is from it, and a reason to believe the market is ranging rather than trending. The mean is usually a moving average or VWAP. Distance is measured in standard deviations, ATR multiples or an oscillator reading. The regime filter is often ADX, with readings below about 20 a common choice for “no strong trend”.

Within ±1 SD
Ordinary movement around the average; no signal.
±2 SD
Stretched. Many mean-reversion rules start looking for an entry here.
Beyond ±3 SD
Rare in a range. Often a sign that the regime has changed and a trend has started.

Common triggers include a close outside a Bollinger Band (20, 2) followed by a close back inside, RSI (14) below 30 or above 70, or price two ATRs from VWAP intraday. Settings vary between traders; whatever you use, write it down so the trades can be compared.

A worked example on an NSE stock

Illustrative. A large-cap stock has a 20-day average of ₹1,500 and a standard deviation of ₹30, so the bands sit at ₹1,440 and ₹1,560. ADX is 16. After a weak session the stock closes at ₹1,440, a z-score of −2.0, and the next day opens back above it.

Buy 200 shares at
₹1,440
Stop at z = −3 (₹1,410)
−₹6,000
Target at the mean (₹1,500)
+₹12,000
Reward to risk
2.0

The target moves while you wait. If the 20-day average has slipped to ₹1,485 by the time price gets there, the exit pays ₹45 a share, ₹9,000 or 1.5R. Mean-reversion traders usually exit at the current mean, not the one at entry, and some use a time stop, closing the trade if it has not reverted within a set number of bars.

How traders use it and log it

  1. Define the mean, the stretch and the regime filter in writing, for example: 20-day SMA, close below the lower band, ADX under 20.
  2. Place the stop beyond the stretch, not at a round number, and size the position from that distance.
  3. Tag each trade with the z-score or band reading at entry and the ADX value.
  4. Once a month, split results into range weeks and trend weeks. A mean-reversion book that loses mostly in trend weeks needs a stricter regime filter.
  5. Track the largest single loss next to the average win; this style tends to have many small wins and a few large losses.
Common mistake
Treating a stretch as a guarantee. A stock falling on bad results can stay below its lower band for weeks while the average follows it down. The stop is what separates a mean-reversion trade from averaging down.

Mean reversion vs momentum

Mean reversion

Bets that a stretch will fade. High win rate, small wins, occasional large losses on trend days.
Best regime: ranges

Momentum

Bets that strength will continue. Lower win rate, a few large winners, many small losses in chop.
Best regime: trends
For traders in India
Short-side mean reversion in the cash segment has to be closed the same day, since retail short sales of delivery shares are not carried overnight; overnight shorts use stock futures or options. Stocks also have daily price bands, and a stock locked at its lower circuit may have no buyers to exit into. Delivery trades pay STT of 0.1% on both buy and sell.

Reference guide

ContextValueReading
Fading a stretch inside a range, with a stop beyond itDefined mean reversionSmall, frequent wins; the stop caps the trend-day loss
Buying every dip in a falling stockCatching a trendPrice can keep moving away from a mean that is itself falling
Adding more as the stretch growsAveraging downOne trend day can erase many weeks of small wins

Questions

What is mean reversion in trading?

Trading on the expectation that a price far from its recent average will move back towards it. Traders buy when price is stretched below the mean and sell or short when it is stretched above, with a stop beyond the stretch.

How do you measure mean reversion?

Usually with a z-score: (price − moving average) ÷ standard deviation. A z-score of −2 is the same as touching the lower Bollinger Band with the standard 20-period, 2-standard-deviation settings.

When does mean reversion fail?

In trends. When news or a change in flows starts a sustained move, price keeps moving away from the average and the average follows it. A regime filter such as ADX and a hard stop limit the damage.

Is mean reversion better for intraday or swing trading?

It is used on both. Intraday traders often use VWAP as the mean; swing traders often use a 20-day average. Costs matter more intraday because the targets are smaller.

What is a good win rate for a mean-reversion strategy?

There is no standard figure. Mean-reversion systems often win more often than they lose, but the win rate only means something next to the size of the average loss. Check expectancy, not win rate alone.

Related terms