Trend Following Strategy Explained: How Trend EAs Work and Which Timeframe to Run Them On

The idea in one paragraph

A trend follower assumes that when a market has moved a long way in one direction, it is more likely to keep going than to reverse, at least often enough to pay. It does not know when a trend will start or end. It defines a trend by a rule, such as price breaking above its highest high of the last 20 days, or a fast moving average crossing above a slow one, enters in that direction, and exits with a trailing stop that follows price. The system accepts many false starts as the price of catching the real moves.

Why the profile looks strange: low win rate, big winners

New traders judge systems by win rate, and a trend follower looks terrible by that measure. It might win 30 to 40 percent of its trades. The rest are small losses, because the stop is placed close enough that a false start costs little. The profit comes from the trades that turn into real trends, where the trailing stop lets the winner run to several times the initial risk. The total is positive because the few large wins are bigger than the many small losses added together.

The arithmetic is worth doing once. If a system wins 35 percent of the time, with an average win of 3 units and an average loss of 1, expectancy per trade is 0.35 times 3 minus 0.65 times 1, or +0.40 units. The calculator lets you try other combinations and shows the break-even win rate, which falls as the payoff rises.

How trend EAs enter and exit

Most of a trend system's quality sits in the exit. A system that takes profit too early turns a trend follower into a poor scalper, and one that never trails gives a whole move back. The cleanest way to see the difference is to test the same entry with two exits and compare the average win. If you are reading a vendor's test, look for the average win and average loss, not only the profit factor.

What drawdown a trend system has, and why it is hard to hold

Trend systems have long drawdowns because they are wrong in clusters. In a range, every breakout fails and the system takes a small loss repeatedly. Ten or fifteen losses in a row, each small, is normal for a 30 percent win rate. The drawdown is shallow per trade and long in time: the equity curve can sit below its previous high for many months while the system waits for the next trend. The decision that ends more trend-following careers than any other is switching the system off during that wait, just before the next move.

The table is an approximation from the probability of consecutive losses, and the point is the direction: at a 30 percent win rate, a run of a dozen losses is something you should expect to see, not something that means the system is broken. Plan the account and your patience for it. Our guides on what to do when an EA is in drawdown and when to turn off your EA give a way to decide ahead of time.

Best timeframe for trend following: H1, H4 or D1

Unlike scalping, trend following is not squeezed by costs, because the targets are many times the spread on any of these charts. The choice is a trade-off between trade frequency, noise and how long you must wait to know whether the system works.

The last column is the one buyers forget. A daily-chart trend system on one symbol takes a dozen or so trades a year. To see 150 trades, enough for any statistic to mean something, you need a decade or more of history, and a six-month live record tells you almost nothing. That is why serious daily systems trade many markets at once. A single-symbol H4 system reaches a meaningful sample in a few years, which is why most EAs aimed at retail traders sit there.

• Choose H4 if you want a trend EA on one or a few symbols, a sample you can evaluate in a reasonable time, and moderate noise.
• Choose H1 if you want faster feedback and can tolerate more false starts and a lower payoff per trade. Watch the spread on tight stops.
• Choose D1 if you can run many markets, have years of data, and can sit through a long flat spell.

Which markets trend, and when trend systems fail

Gold, equity indices and some currency crosses have produced long trends in many years, and trend followers have concentrated on them. No market trends all the time. Currencies that spend long periods in ranges, such as many crosses of Australian, New Zealand and Canadian dollars, punish breakout systems with repeated false starts, which is why range-bound pairs attract grids instead. Trend systems also suffer when volatility collapses, since a stop of a few ATR becomes too small to trade, and when news gaps jump a stop. Our guide on the best currency pairs for EAs goes into the characteristics of each.

A practical consequence: when you test a trend EA, make sure the history contains a long range, not just a strong trend. A system tested only on a trending decade will look excellent and tell you very little. The long, flat period is the test.

Buying or building a trend EA: what to check

For example, our Quantum Commander verdict reviews a trend-following US30 EA with a stop on every trade. It is a useful case study in the maths above: the vendor's live signal won 69 percent of its trades and still lost money because its average loss was 2.6 times its average win, the reverse of the classic trend profile. If you want to see the principles applied to products, our EA verdicts set out the evidence for each. Browse gold EAs and forex EAs, where each listing states its design.