Forex Trend Following: A 20-Day Breakout Strategy Explained

Reviewed: 12 October 2026. Examples are educational; they are not live trading signals.

Trend following attempts to participate in a sustained price move after evidence of that move appears. It does not require buying at the exact bottom or selling at the exact top. The trade-off is that entries can arrive late, and a market that repeatedly changes direction can produce several failed signals.

This article preserves the original framework: a 20-day breakout, a 25-day versus 300-day moving-average filter, and a 10-day exit. Those numbers define a strategy to investigate; they are not a verified optimal combination or a claim of profitability.

Synthetic daily prices crossing the previous 20-day high and later the previous 10-day low — Synthetic channel example, not a backtest or a full trade simulation. Each threshold uses only completed earlier bars; the article explains the other strategy rules.

What a 20-day breakout means

A Donchian channel uses the highest high and lowest low within a selected lookback. TradingView documents its calculation and use in analysing breakouts. For a trade decision, compare today's price with the previous 20 completed daily candles, excluding the candle being evaluated.

That exclusion matters. If today's high is included in the upper boundary, the boundary moves whenever a new high appears. A finished chart can then make the entry rule ambiguous. Record the broker's daily candle timezone too, because different daily sessions can produce different highs and lows.

The original strategy framework

Component	Long direction	Short direction
Trend filter	25-day moving average above the 300-day average	25-day moving average below the 300-day average
Entry reference	Break above the highest high of the previous 20 days	Break below the lowest low of the previous 20 days
Exit reference	Break below the lowest low of the previous 10 days	Break above the highest high of the previous 10 days

The filter is the relative position of two averages, not simply whether the shorter average slopes up or down. It can disagree with the most recent price move. When the averages are equal, the framework supplies no directional permission.

Specify what the older outline left open

The original outline did not fully define moving-average type, fill timing, initial protective stop or position size. You must specify these before interpreting a test. Using simple moving averages of closing prices is one possible implementation, but it should be labelled as that choice rather than presented as an original rule.

For example, a close-confirmed version would evaluate all conditions after the daily candle closes and enter at the next available price. An intraday stop-order version could enter as the level is crossed. The resulting trades will differ; do not mix them within one performance record.

An initial stop also requires a separate decision. A hypothetical implementation might use two times the daily ATR(20) measured at entry, with position size adjusted to its distance. This addition is an untested example. MetaQuotes defines ATR as an average of true range; it measures price movement rather than guaranteeing a suitable stop.

Why a trend-following trade can feel uncomfortable

A trailing 10-day exit gives the market room to fluctuate, so an open profit may shrink before the exit triggers. In a range, a 20-day breakout may quickly return inside the old boundaries. These are consequences to measure, not reasons to change the exit opportunistically after each trade.

Consider a hypothetical market whose prior 20-day high is 1.1000. If the chosen close-confirmed rule triggers at 1.1020 and the next available fill is 1.1030, use that fill in the record. Crediting an entry at 1.1000 after waiting for the higher close would overstate what the strategy could have achieved.

Evaluate the complete strategy

Start with enough history for the 300-day filter, then keep a separate evaluation period. Include spread, commissions and any overnight financing applicable to the instrument and direction. Test how results change with different execution assumptions without selecting only the most flattering outcome.

Also review the portfolio. Several dollar pairs can repeat one currency exposure, and simultaneous signals can create more combined risk than a single-trade limit suggests. A win rate alone will not capture that problem; compare average gains, losses, costs and drawdowns with the expectancy guide.

If you prefer entries around pullbacks rather than new extremes, study swing levels and price action as a separate approach. Whatever you choose, keep the rule version stable long enough to evaluate it and use the strategy-review process when results or circumstances change.

How to test a forex trend-following strategy without hindsight

For a useful trend-following comparison, decide in advance whether the entry is triggered by a completed candle, an intrabar break or a stop order. Record the timeframe, symbol, broker server time, spread assumption, stop method and what happens when the breakout reverses. Do not choose entry conditions after looking at the winning chart.

Write the 20-day breakout, 25/300-day moving-average filter and 10-day exit as mechanical rules, then use the MT4 backtesting workflow for a historical sample.
Inspect losses, drawdown and modelling limitations with the backtest report checklist; include realistic trading costs.
Check how those frozen rules behave in a demo forward test before drawing any conclusion about live suitability.

A tweezer-top setup is a different chart hypothesis. Test it separately rather than mixing reversal signals into the trend-following sample after a losing streak. All examples are educational; no win rate or return is implied.