Trend Following: Main Points
- Trend following takes long positions in rising markets and short positions in falling markets, keeping them open until a predetermined rule identifies the end of the trend.
- Five components underpin a trend following strategy: a trend filter, entry and exit rules, position size and risk limits.
- Trailing stops are generally preferred by trend followers, since a substantial winning trade can cover numerous small losses, while a fixed profit target would close it prematurely.
- Fewer than half of trades are often profitable, although a 35% win rate still produces positive expectancy when gains are three times the size of losses.
- Range-bound trading creates whipsaw: successive false signals that constitute the chief weakness of trend following.
Trend Following Responds to Established Price Moves
Rising and falling prices provide the signals for trend following: the strategy buys markets already moving upwards and sells those already moving downwards. A position stays open until a rule indicates that the trend is over. Rather than anticipating a price reversal, the rules respond after prices have moved.
Neither market direction excludes trend following. Buying, or going long, follows an upward trend; selling, or going short, follows a downward one. Its underlying principle is captured by the longstanding maxim “cut losses short and let profits run”.
By comparing current market prices with their own history, the most basic version of trend following measures what researchers call time series momentum. How quickly the rules respond determines whether positions remain open for days, weeks or months. Commodity trading advisors (CTAs), large funds operating in numerous markets simultaneously, use trend following. The term trend follower applies to anyone trading in this manner.
Because profitability requires trends to persist, it is useful to understand what sustains them.
Slow Price Adjustment and Collective Behaviour Sustain Trends
Two characteristics help markets develop trends: information takes time to be reflected in prices, and traders often act collectively. The following three forces prolong these moves.
- Delayed responses to news mean prices can drift for days or weeks after a report as traders adjust their views incrementally.
- Herd behaviour among traders extends the movement when buying by others draws more participants into the market.
- Large funds' risk limits compel them to buy as markets rise and sell as markets fall, reinforcing the trend.
Rules determining participation in a trend and the subsequent exit together constitute a trend following strategy.
Fixed Rules Govern Five Trading Decisions
With a trend following strategy, five decisions are codified in advance rather than left to guesswork.
- A directional trend filter establishes whether to buy or sell; for instance, purchases might be permitted only above the 200-day moving average (MA).
- A specified entry rule determines the opening trigger, such as a closing price above the highest high of the preceding 20 days.
- A specified exit rule determines the closing trigger, such as a stop that trails the price throughout the trend.
- The size of the position determines the quantity traded; each trade could, for example, be sized to lose 1% of the account if its stop loss is reached.
- Overall risk limits constrain aggregate exposure, perhaps by setting a maximum proportion of the account that all open trades may put at risk.
Delayed entry and exit are unavoidable features of trend following. Confirmation is required before entry, leaving the opening stage of the trend untraded. At the exit, confirmation of a reversal means surrendering some of the final gain. This lag is intentional: the middle of a trend, rather than either turning point, is what rules can confirm.

Establishing the trend filter, the first rule, therefore begins by identifying a trend.
Trend Following Indicators Apply Objective Trend Tests
An objective rule replaces visual chart judgement when a trend follower identifies a trend. The following three tests are common.
- Trading above an MA that slopes upwards identifies an uptrend; trading below one that slopes downwards identifies a downtrend.
- A sequence of higher highs and higher lows identifies an upward trend, whereas a sequence of lower highs and lower lows identifies a downward trend.
- Closing above the preceding N days' highest high identifies a new upward trend; reversing this test identifies a downward trend. N denotes the lookback period, which might be 20 days.
If none of these tests is satisfied, the market is range-bound, where signals from trend following often prove unsuccessful.

Different tasks are assigned to the five indicators most widely used in trend following: the MA, Moving Average Convergence Divergence (MACD), Average Directional Index (ADX), Donchian channel and Average True Range (ATR).
| Technical Indicator | Measurement | Application in Trend Following | Principal Limitation |
|---|---|---|---|
| Moving average | Price averaged over a defined number of days | An upward-sloping average beneath price identifies an uptrend; crosses provide entry signals | Delayed signals, with false readings in range-bound conditions |
| MACD | Separation between two exponential moving averages | Provides confirmation of directional trend and momentum | Delayed readings and repeated crosses in sideways conditions |
| ADX | A 0 to 100 measure of trend strength, excluding direction | Values above 25 suggest a trend is strong enough to trade | Cannot distinguish an upward trend from a downward one |
| Donchian channel | The highest high and lowest low within a defined daily lookback | An upper-band closing breakout triggers entry | Breakouts frequently prove unsuccessful |
| ATR | The market's daily trading range, averaged | Determines stop placement distance and trade size | Provides no directional signal |
None of these indicators leads prices. All use historical price data, confirming trends only once they are under way rather than forecasting the next move. ADX therefore needs an MA or Donchian channel alongside it, because it cannot establish direction.
Using a trendline to connect the ascending lows offers a further view of an upward trend. Before treating the line as reliable, most traders require at least two touches; a third adds strength.
After defining the trend filter, attention turns to entry. Three principal trend following strategies provide distinct entry triggers.
Three Trend Following Strategies Differ in Entry Timing
Entry timing separates moving average crossovers, Donchian channel breakouts and trend pullbacks, the three principal trend following strategies. Otherwise, each follows the same structure: a trend filter, entry, stop loss and trailing exit.
| Trading Method | Entry Condition | Exit Condition | Advantage | Drawback |
|---|---|---|---|---|
| Moving average crossover | An upward cross of the slower average by the faster average | A reverse cross or stop that trails price | Rules are clear and straightforward | Delayed participation may miss much of the move |
| Donchian channel breakout | A closing price above the 20-day high | A closing price below a shorter-period low, for instance 10 days | Early participation in emerging trends | A high incidence of failed breakouts |
| Trend pullback | A retracement towards the moving average during a trend | A stop that follows price | A tighter stop improves reward relative to risk | Powerful trends may provide no retracement |
Moving Average Crossovers Trigger Entries and Exits
Under a moving average crossover strategy, an upward cross of a slower MA by a faster MA triggers a purchase. A downward cross back through the slower MA closes the position or initiates a short. Traders most commonly use the 50-day and 200-day MA combination, whose bullish crossover is widely described as a golden cross.
Lag is the price paid for this approach. The slow response of a 200-day MA means much of a trend may have passed before a crossover develops. Using shorter lookback periods makes the pair more responsive, but also increases false signals during range-bound trading.
Donchian Channel Breakouts Capture Emerging Trends
Buying on a close above the upper band is the basis of the Donchian channel breakout strategy. The channel marks the highest high and lowest low over a defined number of days. For a 20-day entry channel, a long position opens on a close above the preceding 20 days' highest high. A shorter channel governs the exit, for example a close below the lowest low of the preceding 10 days.
Among breakout traders, the best-known example is the turtle traders, whom Richard Dennis trained in the 1980s. Their position sizing made a 1 ATR movement equivalent to 1% of the account. Placing the stop loss 2 ATR away consequently put about 2% at risk on each trade.
Early access to new trends comes with a cost: numerous breakouts reverse and return inside the Donchian channel.

Trend Pullbacks Offer Entries Within Established Moves
Rather than entering at a fresh high, a trend pullback strategy first requires an established trend and then buys a retreat towards the MA or a trendline. Price reaching the MA and beginning to recover provides the entry.
Placed just beneath the pullback low, this strategy's stop loss is nearer the entry than a breakout stop would be. That shorter distance reduces the loss on an unsuccessful trade and improves reward relative to risk if the trend continues. However, some strong trends offer no pullback at all.
An exit is required for every entry. In trend following, the exit has the greater influence on the eventual result.
Volatility-Based Stops and Trailing Exits Govern Trade Closure
Volatility determines the distance to the initial stop loss; thereafter, trend followers move it behind the price. The trade thus ends in response to price, not a predetermined profit target. ATR, which measures the average daily market range, is the standard volatility measure. Setting the stop loss 2 to 3 times ATR away from entry places it beyond ordinary daily fluctuations.
Most trend following systems rely on four exit methods.
| Exit Method | Closing Rule | Main Application |
|---|---|---|
| ATR stop | Place the stop 2 to 3 times ATR away from entry | Keeping the initial stop beyond routine fluctuations |
| Trailing stop | Raise the stop as price rises, without lowering it | Securing gains during an ongoing trend |
| Channel exit | A closing price beneath the preceding 10 or 20 days' lowest low | Systems based on breakouts |
| Opposite signal | Close on a downward cross of the slower average by the faster one | Systems based on crossovers |
Fixed profit targets are avoided by most trend followers. Taking profit at, say, twice the initial risk shuts down substantial winners too soon, excluding the very trades on which the system depends. By contrast, a trailing stop allows a profitable position to remain open until the trend reverses.
With a trailing stop, adjustments are permitted only in the trade's direction: upwards for long positions and downwards for short positions.
Position sizing also depends on the gap between entry and stop loss, which determines the quantity to trade.
Stop Distance Determines Trend Following Position Size
To limit a stop-loss exit to a small, fixed proportion of the account, trend followers adjust the size of each trade. Calculating position size involves multiplying account size by risk per trade, then dividing the result by the distance to the stop.
Position size = (Account size √ó Risk per trade) √∑ Stop distance
At 1% risk per trade, a $10,000 account permits a $100 loss on a single position. If entry is 4.00 away from the stop loss, dividing $100 by 4.00 produces a position of 25 units.
This calculation automatically reduces trade size in volatile markets. Higher ATR places the stop loss further away; dividing the same $100 risk by that wider stop distance results in fewer units.
Risk per trade is often restricted to around 1% or less, preserving sufficient capital to continue after consecutive losses. Such restraint is important because losing trades are frequent in trend following.
Large Winners Offset Trend Following's Frequent Losses
In sideways conditions, most breakouts and crossovers fail, leaving trend followers with more losing trades than winners. What sustains the strategy is the size of its average gain, which is several times its average loss. Consequently, the proportion of winning trades is often below half.
The average outcome of a trade is measured by expectancy: multiply the win rate by the average gain, then subtract the loss rate multiplied by the average loss. R denotes the sum risked on a single trade. Both tables use rounded figures solely for illustration.
| Measure | Illustrative Value |
|---|---|
| Proportion of winning trades | 35% |
| Mean winning trade | 3R |
| Mean losing trade | 1R |
| Expected result per trade | (0.35 x 3R) minus (0.65 x 1R) = plus 0.40R |
An average return of 0.40R per trade therefore results when 35% of trades win and each winner is three times the size of a loss.
Extended sequences of losses also accompany a low win rate. Below, eight consecutive losses are illustrated at three settings for risk per trade. Each successive loss is calculated against the reduced account balance.
| Risk on Each Trade | Account Remaining After 8 Consecutive Losses | Cumulative Loss |
|---|---|---|
| 0.5% | approximately 96.1% | approximately 3.9% |
| 1% | approximately 92.3% | approximately 7.7% |
| 2% | approximately 85.1% | approximately 14.9% |
Winning infrequently is a normal feature of trend following. Keeping individual trade risk small allows traders to maintain the rules and remain in the market until a substantial trend emerges.
The following worked trade brings together the trend filter, entry, stop loss, position sizing and exit.
A Worked Trade Demonstrates Trend Following Rules
Six stages take a trend following trade from its initial filter to closure. Rounded numbers are used purely for illustration; they refer to no actual market, date or price.
- Establish the permitted direction by checking that price exceeds its 200-day MA, allowing long trades only.
- Open a breakout position following a close above the 20-day high at 100.00.
- Place the initial stop loss at 2 ATR beneath entry. An ATR of 2.00 puts the stop loss at 96.00, a distance of 4.00.
- Calculate the trade quantity using 1% risk per trade. For a $10,000 account, that means risking $100; dividing $100 by 4.00 yields 25 units.
- Adjust the trailing stop to remain 2 ATR beneath the highest closing price as the market rises.
- Close the position following a close beneath the trailing stop.
Two outcomes are possible in this example. An early failure takes price back to 96.00, where the stop loss closes the position at a loss of 1R, or $100. If the trend persists instead, the stop trails upwards behind price, retaining some of the gain when the eventual reversal occurs.

Applying identical rules to different markets can nevertheless produce markedly different outcomes.
Liquid Markets Offer Scope for Sustained Trends
Sustained, extended moves make liquid markets suitable for trend following. Examples include major forex pairs, gold, oil and stock indices.
Research into trend following draws its results from portfolios trading numerous markets concurrently. Applying identical rules to just one or two markets produces more uneven outcomes: an individual market may remain range-bound for months, whereas a broad portfolio often contains a trend somewhere.
Crypto markets can exhibit strong trends, but their high volatility requires more distant stops and smaller positions to maintain the same risk per trade.
Even these markets experience lengthy periods without a trend, creating difficult conditions for trend following.
Sideways Trading Exposes Trend Following's Weaknesses
Repeated signals followed by reversals, known as whipsaw, make range-bound markets the most challenging setting for trend following. Four weaknesses account for most periods of losses.
- Range-bound whipsaw produces accumulating small losses from unsuccessful breakouts and crossovers.
- Abrupt trend reversals surrender some accrued profit while the exit rule waits for confirmation that the trend is over.
- The cost of trading absorbs more profit in faster systems: shorter lookback periods generate more trades, each incurring the spread and fees.
- The responsiveness of signals involves a compromise. Faster rules respond sooner but produce more false signals; slower rules reduce false signals at the expense of later entries.

These limitations naturally prompt the question of whether trend following remains effective.
Research Records Positive Average Returns From Trend Following
A long record of research continues to support positive average returns from trend following. That does not, however, mean the strategy is profitable in every period.
Across 58 liquid futures markets, Moskowitz, Ooi and Pedersen (2012), published in the Journal of Financial Economics, identified time series momentum. A market’s historical returns tended to persist, providing the pattern on which trend following operates.
Using data stretching back to 1880, Hurst, Ooi and Pedersen (2017), published in the Journal of Portfolio Management, examined a trend following strategy across 67 commodity, equity index, bond and currency markets. Over that extended history, the study recorded positive average returns for the strategy.
Diversified, multi-market portfolios were tested in both studies, rather than individual markets. Trading costs and adherence to the rules during losing periods also affect results, and past results do not guarantee future results. During prolonged market sell-offs, trend following has frequently performed well because its rules allow a switch to short positions as prices decline. It has not, however, performed well in every sell-off.
How traders implement the rules accounts for most trend following failures, making the recurring errors important to recognise.
Early Profit-Taking and Abandoned Rules Undermine Trend Following
Closing profitable trades prematurely and dropping the rules after successive losses are the leading trend following errors. Six mistakes recur particularly often.
- Premature profit-taking eliminates the substantial winners needed to cover the numerous small losses.
- Widening the stop loss away from price allows a deliberately small potential loss to become a large one.
- Selecting the best-performing backtested lookback period amounts to curve fitting. A period optimised for historical data often performs poorly on new data.
- Repeatedly adding filters to tidy up a backtest excludes so many trades that the system no longer captures the trends it was designed to trade.
- Abandoning the strategy during a drawdown means discarding the rules just before another trend might recover the losses.
- Opening positions ahead of trend confirmation replaces trend following with speculation about direction; these premature entries fail more frequently.
Before putting rules into practice, running a backtest against historical data reveals the length of past losing sequences and the depth of drawdowns.
Comparison with related approaches helps establish where trend following sits among the available trading strategies.
Trend Following Differs From Mean Reversion and Momentum Trading
Buying strength and retaining the position characterises trend following. Mean reversion instead trades against extended moves in anticipation of a return to the average, while momentum trading follows recent returns or ranks them.
| Comparison Point | Trend following | Mean reversion | Momentum trading |
|---|---|---|---|
| Underlying premise | An established move persists | Prices move back towards their average | Recent leaders continue to outperform |
| Most suitable conditions | Markets with sustained trends | Range-bound conditions | Markets featuring distinct leaders |
| Usual winning proportion | Frequently fewer than half | Frequently higher | Dependent on strategy construction |
| Pattern of gains and losses | Numerous small losses offset by a few substantial gains | Numerous small gains with an occasional substantial loss | Determined by strategy construction |
| Frequently used indicators | Moving averages, Donchian channel, ADX, ATR | RSI, Bollinger Bands | Relative strength ranking, rate of change |
| Principal exposure | Repeated reversals in range-bound trading | A persistent strong trend with no reversion | Sudden reversals among leading markets |
Trend Following Trades Continuation Rather Than Reversion
The expectation of continued movement distinguishes trend following from mean reversion, which anticipates a return towards the average price. Trending conditions favour trend following, whose few substantial winners accompany a low win rate. Sideways conditions favour mean reversion and its higher win rate, although a strong emerging trend that never returns can cause a large loss.
Because favourable conditions for one approach tend to challenge the other, some traders use both in parallel.
Trend Following Overlaps With Some Momentum Strategies
Although the terms overlap, momentum trading and trend following do not invariably describe identical approaches. Comparing a market with its own historical prices is time series momentum, an approach closely related to trend following. Cross-sectional momentum is different: it ranks multiple markets, buying the strongest and selling the weakest. Longer holding periods and mechanical exits, such as trailing stops, are also usual features of trend following.
Frequently Asked Questions
Trend Following Suits Traders Prepared for Frequent Losses
For traders willing to tolerate low win rates and extended losing sequences, trend following offers the possibility of a few substantial winners. There is no promise of profit: costs, individual trade risk and adherence to the rules during drawdowns determine results. Those who need most trades to succeed often struggle to maintain this approach.
Different Trend Following Indicators Perform Different Tasks
Different functions mean that no single indicator is best for trend following. Moving averages are the most widely used because a single MA can provide both the directional filter and the entry trigger. ADX is often added to assess trend strength, alongside ATR to determine stop distance.
Trend Followers React to Price Rather Than Predict Turning Points
Whether a trader or a fund, a trend follower buys markets moving upwards, sells those moving downwards and retains positions until a rule identifies the trend's end. Price drives the response, with no attempt to forecast peaks or troughs. Among funds, commodity trading advisors (CTAs) are the most widely recognised trend followers.
Trend Following Usually Implies a More Rule-Based Approach
Trading with the prevailing trend is often described interchangeably as trend following or trend trading. Usually, however, trend following denotes a fixed system with mechanical entry and exit rules. Trend trading has a wider meaning, encompassing discretionary chart-based decisions as well.
Shorter Timeframes Increase False Signals and Trading Costs
Hourly charts and other short timeframes can accommodate trend following rules, but faster trading brings more false signals and greater costs. Since every trade incurs a spread, the much higher turnover of a short-timeframe system absorbs more of each gain. Shorter-timeframe trends also tend to be smaller, reducing the scope for the substantial winners on which trend following relies.










