OptiNod Academy

Trend Following vs. Mean Reversion: Two Ways to Make Money in Opposite Markets

Trend following and mean reversion make money in opposite market conditions with opposite P&L profiles. Mixing them inside a single trade combines the weaknesses of both.

Trading strategies broadly fall into two camps. Trend following means buying something that is rising and selling it after it rises further. Mean reversion means betting that a price stretched far from its average will return toward that average. One assumes price will keep moving in the same direction. The other assumes an extended move will reverse.

Their P&L profiles are opposites. Trend following is wrong often, but wins big when it is right. Mean reversion is right often, but can lose big when it is wrong. These are the two shapes discussed in the earlier article on expected value: low win rate with large payoff ratio, and high win rate with small payoff ratio.

Many traders mix the two without knowing which camp they are in. They enter on a breakout as a trend-following trade, then when price pulls back, they hold on and hope for mean reversion. In effect, they enter with trend following and exit with mean reversion.

That mix is a common source of large losses. If you attach a mean-reversion expectation to a trend-following entry, large winners get cut short and large losers are allowed to grow. The weaknesses of both methods end up inside one trade.

The Two Methods Have Opposite P&L Profiles

Trend following has a low win rate. You cannot know in advance where a trend will begin, so you take a series of small losses while waiting for one real trend. That one winner has to be large enough to more than cover the earlier losses. A typical profile might be a 35% win rate with a 4:1 average payoff ratio.

Mean reversion is the opposite. Price often moves away from an average and then returns, so the win rate is high. But when a real trend breaks out, a position held on the belief that price will return to the average can turn into a severe loss. A typical profile might be a 70% win rate with a 1:1 average payoff ratio, with one large loss giving back many small gains.

Both methods can have positive expected value. They just reach it in opposite ways. That is why you need to know which profile you are trading. Only then can you treat frequent small losses in trend following, or occasional large losses in mean reversion, as normal and keep following the system.

The numbers make the difference clear. If a trend-following system has a 35% win rate, an average win of 4R, and an average loss of 1R, its expected value is 0.75R. If a mean-reversion system has a 70% win rate, an average win of 1R, and an average loss of 1R, its expected value is 0.4R. Both are positive, but trend following gets there while losing more than six times out of ten. Mean reversion gets there while winning seven times out of ten.

The real danger in mean reversion is hidden inside that average loss. A loss that was small inside a range can become three or four times larger the moment the range turns into a trend. If you underestimate this tail loss because the win rate is high, one shift into a trend can wipe out a year of profits.

The two methods also drive traders out in different ways. Trend following produces a long run of small losses before the real trend arrives, pushing traders to abandon the method. Mean reversion works well for a long time, then one large loss takes back the accumulated gains. Unless you accept each method's losses in advance, you will abandon the system at exactly the moment it starts losing.

Position Size and Holding Period Are Different

The two methods also size and hold positions in opposite ways. Trend following trades less often and holds longer. Because small losses can continue for a long time before a real trend appears, position size must be small enough to survive a losing streak. Once a position catches a trend, it is held with a trailing stop so the winner is not cut short.

Mean reversion trades more often and holds for shorter periods. Positions are closed quickly when the target is reached, so turnover is high and fees and slippage accumulate. Because of tail losses, position size cannot be increased casually. If you size too large just because the win rate is high, one occasional large loss can damage the account badly.

Even with the same 1% risk, the sizing logic is different. Trend following sizes based on how many consecutive losses the account can withstand. Mean reversion sizes based on how large a single tail loss can become.

They Work in Different Markets

Trend following makes money only in markets that extend in one direction. Mean reversion makes money in range-bound markets where price moves back and forth within a defined band. Even in the same asset, most of the time may look range-bound, while the biggest moves are concentrated in short trending periods.

You can judge whether the market is trending or ranging in a few ways. If highs and lows are rising or falling in sequence, the market is trending. If price is moving between similar highs and lows, it is range-bound. An ADX above 25 points toward a trend; below 20 points toward a range.

In the summer of 2023, BTC was a mean-reversion market. From April through September, BTC moved between $25,000 and $31,000. Mean reversion worked in this period by selling near $31,000 and buying near $25,000-$26,000. Trend-following breakout attempts were repeatedly caught by false breakouts.

Starting in October 2023, the market changed. After BTC broke above the top of the $31,000 range, it moved in one direction to $73,777 in March 2024. In that trending market, trend following worked very well, while mean-reversion trades looking for pullbacks at each new high repeatedly failed.

Now look at the two markets as concrete trades. In the summer range, a mean-reversion trade would work like this: buy BTC when it pulls back to $25,000-$26,000, sell near $30,000, then buy again when it returns to the lower end of the range. Each trade brings a small profit, and the same range produces those small wins many times over.

In the same period, trend following gets punished. A breakout buy above $30,000 gets stopped out when the breakout fails. A breakout short below $25,000 gets stopped out when price snaps back. In a market without a trend, chasing breakouts only stacks losses on both sides.

In the autumn trend, the roles reverse. A trend follower who bought the $31,000 breakout rides the move through $35,000, $40,000, and beyond. A trader who shorts $35,000 because it looks expensive loses badly while waiting for price to return to the average.

The Same Price Level Requires Opposite Actions Depending on Regime

The same $31,000 level required opposite actions in summer and autumn. In the summer range, $31,000 was a place to sell through mean reversion. In the autumn trend, a breakout above $31,000 was a place to buy through trend following.

So the question "What should I do at $31,000?" has no answer by itself. You first have to decide whether the market is range-bound or trending. Only then can you decide whether that level is a sell or a buy. Read the market regime before you read the level.

The hardest area is the transition from range to trend. When price first broke above $31,000, it was difficult to know in the moment whether it was another false breakout or the start of a real trend. In that ambiguous zone, neither method gives a high-conviction signal. It is usually better to wait until the breakout holds on a closing basis and support is confirmed on a pullback. If you rush at a regime change, you can carry the old regime's rules into the new one and take a large loss.

Mean Reversion Is Riskier in Crypto

Mean reversion bets that price will return to its average after moving far away. But crypto trades 24 hours a day without pause, and it often extends sharply in one direction. Trends occur more often and travel farther than in many range-bound stock or FX markets.

That is why fading extremes with mean reversion in crypto can be hit hard by a single trend. Shorting new highs or buying against a sharp selloff falls into this category. If you use mean reversion in crypto, use it only in assets and periods where a range is clearly confirmed, and keep stops tighter.

Mixing Them Combines Only the Weaknesses

The largest losses usually come from mixing the two methods inside one trade. There are two common forms.

• Trend-following entry with mean-reversion holding logic: You buy a breakout, but when price drops, you delay the stop because you believe it will return to the average. The moment you abandon the small-stop rule of trend following, a small loss grows into a large one.
• Mean-reversion entry with trend-following holding logic: You sell at the top of the range, but when price breaks above the range, you hold the short and say a trend has started. The moment you abandon the mean-reversion stop rule of exiting on a range break, you are run over by the trend.

In both cases, the trader attaches one method's exit rule to the other method's entry rule. Whichever method you use to enter, the exit must follow the same method's rules.

The real reason traders mix them is usually that they do not want to accept a loss. They use mean reversion as an excuse to avoid taking a trend-following stop, and they use trend as an excuse to avoid taking a mean-reversion stop. Either excuse turns a small loss into a large one.

Which One to Use, and When

First decide whether the current market is trending or range-bound. If ADX is high and highs and lows are rising in sequence, it is a trend. If ADX is low and price is moving within a defined range, it is range-bound. Use trend following in trends, and mean reversion in ranges.

Apply only one method to a single trade. If you enter with trend following, hold until the trend breaks and exit with trend-based rules. If you enter with mean reversion, exit when the target is reached or when the range breaks. Keeping entry and exit logic within the same method is how you avoid mixing the two.

Beginners are usually better off choosing one method and learning it well. Trying to trade both at the same time makes judgment unstable near regime changes, and eventually leads to mixing the two. A safer sequence is to build a consistent record with one method in one type of market, then add the other method later.

If you want to use both, keep them as separate systems and run each in its own trades. Keep trend-following setups and mean-reversion setups separate, and run each only by its own rules. Large funds often run trend-following and mean-reversion strategies as separate books because one can support the other when its own regime is weak. The key is that they are not mixed inside the same trade.

Separate Setups for the Two Methods

Trend-Following Setup (Trending Market)

• [ ] Entry condition: The weekly chart is trending: highs and lows are rising in sequence, and ADX is 20 or higher. The daily chart closes above the previous high, or rebounds from a pullback that holds the previous low.
• [ ] Exit: Hold until the trend breaks. Exit if price closes below the low of the previous pullback.
• [ ] Stop: Place it below the low that justified the entry.

Mean-Reversion Setup (Range-Bound Market)

• [ ] Entry condition: The market is range-bound: ADX is low and price is moving sideways within a defined band. Price touches the bottom of the range and prints a rebound signal such as a long lower wick.
• [ ] Exit: Exit when price reaches the middle or top of the range.
• [ ] Stop: Exit if price closes below the bottom of the range.

There is no single answer to which method is better. The whole job is to choose one method that fits your temperament and schedule, use it only in the market it is built for, and avoid mixing rules inside a single trade. If you choose trend following, accept frequent small losses as part of the system. If you choose mean reversion, accept occasional large losses as part of the system. Trying to do both and ending up poor at both is the worst outcome.