A class of strategy that assumes "price in motion tends to stay in motion" — entries are placed in the direction of an established recent move.
Trend Following
A class of strategy that assumes "price in motion tends to stay in motion" — entries are placed in the direction of an established recent move.
In plain English
A trend-following strategy bets that once price starts moving in a direction, it keeps going in that direction for a while. The opposite philosophy is mean reversion, which bets that price will return to an average.
Trend-followers typically:
- Win a minority of their trades (often 25–45%).
- Have large average winners and small average losers — winners run, losers get cut.
- Make money in trending markets (sustained directional moves).
- Lose money in choppy markets (sideways noise) — they enter every false breakout.
Why it matters for this fleet
EMA Cross (ema cross) is a textbook trend-following signal. The whole 210-strategy EMA-cross dossier inherits its characteristics: low-to-moderate win rates that can still produce positive returns when the winners are bigger than the losers. The fleet median win rate (the share of trades closing in profit) is only ~30% — below half — which is normal and expected for this class. The catch is that only 5 of 210 rows actually clear the edge-significance bar; a sub-50% win rate is the signature of trend-following, not proof that it works.
Examples from the live fleet
- id523 (EMA 21/50 · SOL · 1h · 2× · long) — the textbook trend-follower. Win rate just 31.9% (it loses about two trades in three), yet profit factor 1.46 (gross wins ÷ gross losses) because of positive skew (a few big winners carry the PnL): winners average ~3.1× the size of losers. This is the one signal family in the dossier with a real in-sample edge — and it still wins a minority of its trades. Low win rate, positive expectancy.
- id628 (EMA 9/21 · BTC · 1m · 2× · short) — the same class, failing. Win rate 10.6%, but here there is no fat-tail winner to carry it: profit factor ~0.11 and a −98% drawdown. A 1-minute crossover is reading noise, not trend, so every entry is a false breakout. Trend-following blown up in the chop — the exact failure mode the class is prone to when the noise-to-trend ratio is too high (see noise to trend ratio).
When trend-following fails
- Choppy / sideways markets — every entry is a whipsaw.
- High noise-to-trend ratio symbols (ETH, SOL in this fleet) — see noise to trend ratio.
- Slow signal on fast timeframe — 1d EMA cross on a market that moves in 1h cycles is too laggy.
Related
Sources
wiki/qa-sessions/2026-05-17-session.md#q1 (first asked here)/api/analytics