Trend Following in Five Minutes

If you only have five minutes, this is enough.

In one sentence

Trend following does not predict where price is going. It waits for price to move, then follows.

That sounds like giving up the most important skill. In fact it concedes one thing: short-horizon price forecasting almost never beats chance reliably, whereas following a move that has already happened requires no forecasting — only rules and execution.

What it looks like

Suppose a stock trades like this:

price
 ↑                         ●───────●        ← trend, hold
 |                       ╱           ╲
 |        ╱╲    ╱╲     ╱               ╲
 |   ╱╲ ╱   ╲╱    ╲  ╱                   ●  ← stop hit, exit
 |  ╱                ╲╱
 +──────────────────────────────────────────→ time
     ↑ range: repeated small losses  ↑ entry

The left half is a range: price chops, the strategy enters on each upside breakout and is stopped out shortly after, producing a string of small losses. The right half is a trend: one entry, held the whole way, exited only when price falls from its high to the stop.

All of this method's return comes from the few episodes that look like the right half. The small losses on the left are its cost, not its failure.

Three questions you must answer

Every trading rule set, however elaborate, answers the same three questions.

1. What to buy

Trend following is not picky about the instrument, but it needs adequate liquidity and volatility. Something that does not move cannot trend, and something illiquid will cost you dearly in slippage when the stop fires.

Beginners should start with broad ETFs (QQQ, SPY): moderate volatility, low gap risk. Get the rules working before worrying about stock selection.

2. When to enter

You need an objective trigger that requires no judgment. A common pairing is "direction filter plus breakout trigger":

• Direction filter — close above the 100-day moving average. This confines you to a medium-term uptrend and stops you from repeatedly catching a falling knife.
• Breakout trigger — close makes a new 22-day high. This confirms price is actually moving up, not merely sitting above the average.

Both conditions must hold to enter. Note that both use only completed bars and contain no subjective input.

3. When to exit

This is the most important question, and the one most people never answer properly.

The trend-following exit is a trailing stop: a line below price that rises when price rises and stays put when price falls. Touch it and you are out, unconditionally.

The critical part is "stays put" — the stop never moves down. That fixes each trade's maximum loss at the moment of entry while placing no ceiling on gains: as long as the trend continues, the stop simply keeps trailing beneath it and never calls a halt on its own.

The chandelier exit used by default on this platform is one specific algorithm for that line, using ATR (Average True Range) to decide how far below price it sits. The next tutorial walks through the arithmetic.

Three things to accept in advance

1. The win rate will be low. 35%–45% is normal. Five or six losses in a row does not mean the strategy is broken.
2. Most of the time you are waiting. In a given year perhaps two or three trades produce the return; the rest of the time you are taking small losses or holding nothing.
3. In a strong one-way advance it will underperform simply holding. Its value is a shallower drawdown, not a higher return.

None of these is a defect. They are intrinsic to the method. If you cannot accept any one of them, do not use it — no backtest can help with that.

Next steps

• Open the backtest tool, run the defaults on QQQ, and check what share of total return came from the two best trades in the trade list.
• Read the next tutorial, The Chandelier Exit, Step by Step, and learn how the stop is computed.
• Then read How to Read a Backtest to judge whether a pile of numbers supports a conclusion.