Trend following buys strength and accepts that most trades lose a little. Mean reversion does the opposite: it buys weakness, wins most of the time, and occasionally meets a fall that does not stop. The two are not competing opinions about markets — they are bets on different behaviour, and they work in different conditions.
Stretch a rubber band and it snaps back. Pull a rope and it simply gets longer. Both are elastic in casual language, and betting on one behaviour while holding the other is how people get hurt.
A range-bound stock is a rubber band. A stock in a genuine downtrend on deteriorating fundamentals is a rope. The chart at the moment of purchase looks similar, which is the entire problem.
The shape of the returns
| Property | Trend following | Mean reversion |
|---|---|---|
| Win rate | 30–40% | 65–80% |
| Average win vs loss | Winners much larger | Winners smaller than losers |
| Where the money comes from | A few very large moves | Many small, repeated moves |
| The failure mode | Long grinding drawdowns in ranges | One position that never comes back |
| Feels | Frustrating most of the time | Excellent, right up until it is not |
| Best regime | Expanding volatility, clear direction | Settled volatility, no trend |
What makes a genuine setup
- 1A trend filter above the setup
Only buy weakness inside strength — price above the 200-day average, or the sector holding up. Buying oversold readings in a downtrend is how mean reversion earns its reputation for catastrophe.
- 2A measure of how stretched, not just a direction
Two standard deviations below a 20-day average, or RSI below 30, or price at the lower Donchian band. The point is a quantified extreme rather than "it has fallen a lot".
- 3A reason the fall is not news
A stock down 20% on a guidance cut is not overextended, it is repriced. Mean reversion works on movement without information; it does not work on information.
- 4A time stop as well as a price stop
If the snap-back has not happened within the expected window — often five to ten sessions — the premise was wrong even if the loss is small. Exiting on time is what keeps the rare large loss rare.
Where it works in Indian markets
- Large caps and the index far more than small caps. Liquidity is what produces the snap-back; a thin small cap that fell 15% may simply have found its new level with no buyer waiting.
- Intraday and over a few days, rather than weeks. The effect decays quickly. Most of the measurable edge in Indian equities sits inside a week.
- In settled volatility. When India VIX is climbing, the rubber band stretches further than the model expects, and stops that were adequate last month are not.
- Not around results. An earnings-driven gap is information. Fading it is a bet against the fundamentals, made with a chart.
- Pairs, more safely than outright. Buying the laggard and selling the leader within a pair removes the market direction and leaves the divergence — which is what you actually wanted to bet on.
The oversold reading
A large-cap bank has fallen five sessions in a row and RSI reads 24. It sits 6% below its 20-day average but still 4% above its 200-day. The sector index is flat and there has been no company-specific news.
A mean-reversion system wins 78% of trades. Average win ₹4,000, average loss ₹19,000. Is it profitable?
Rubber band kheencho toh wapas aata hai. Rassi kheencho toh bas lambi hoti hai. Chart pe dono kheenche hue lagte hain — wahi poori dikkat hai. Range wala stock rubber band hai; girti hui company rassi. Isiliye kamzori tabhi kharido jab bada trend upar ho.
- Mean reversion wins often and loses large; trend following does the reverse.
- A high win rate is the danger, not the evidence.
- Require a trend filter above the setup — buy weakness inside strength.
- Movement without information reverts; movement caused by news does not.
- The stop must come from outside the trade's own logic, and averaging down removes it.
Mark it done to track your progress through the curriculum.
Common questions
Short, direct answers to what people ask about this topic.
- mean reversion trading meaning
- Mean reversion is a style that buys after an unusually sharp fall and sells into an unusually sharp rise, on the premise that price snaps back toward a recent average. It is the mirror image of trend following: it wins on most trades and loses rarely but large, because the one fall that never stops is open-ended in a way the many small gains are not.
- a strategy that buys weakness and expects price to snap back to its average is called
- Mean reversion. It is normally built on a quantified measure of how stretched price has become — two standard deviations below a 20-day average, RSI under 30, or the lower Donchian band — rather than a subjective sense that something has fallen a lot.
- is a 78% win rate enough to make a mean reversion system profitable
- No — a hit rate says nothing on its own without the payoff ratio beside it. A system winning 78% of trades at ₹4,000 a win while losing ₹19,000 on the other 22% has an expectancy of 0.78 × 4,000 − 0.22 × 19,000, a loss of about ₹1,060 per trade. That is the characteristic trap of the style: it feels successful four days out of five while losing money.
- why does averaging down ruin a mean reversion trade
- Because the setup already says cheaper is more attractive, so adding on every further fall makes the position largest exactly when the evidence against it is strongest. Mean reversion has no stop built into its own logic — the stop has to be imposed from outside it — and averaging down removes the only protection the trade had.
- does mean reversion work better on large caps or small caps in india
- The measurable snap-back in Indian equities is far stronger in large caps and the index than in small caps, because liquidity is what produces the bounce at all. A thin small cap that fell 15% may simply have found a new level with nobody waiting to buy. The effect also decays fast — most of it sits inside a week rather than over months.