Read the dated measurement, settings and limits below. A historical backtest does not establish future or live execution performance.
35.9%
break-even 28.3% at its realized average win/loss
re-measured (top 10 coins · 1H · 2026-11-20): 39.45%
Percentage of trades that were profitable
1.42
re-measured (top 10 coins · 1H · 2026-11-20): 1.1
Gross profit / gross loss ratio
42.1%
Largest peak-to-trough decline
50 coins · updated 2026-05-04
Marks next to a metric: ~ = borderline (PF between break-even and 1.5; win rate ≤5 pts above break-even, or 50–55% if none shown) · ! = below the bar, or the strategy's research status (dated on its page) is below the bar — that dated record outranks this number. No mark = clears it and its research status is not below the bar.
No live tracking — measured, not guaranteed
Card re-measured: 2026-11-20 · 730 trades analyzed
Overview
ATR Breakout SHORT fires when the previous close breaks below an ATR-scaled band — the close from two bars back minus 2.0× ATR(14) — signalling a downside volatility expansion, with an EMA trend filter confirming downward bias. Unlike BB Squeeze which waits for compression first, this strategy catches the early momentum of a directional volatility burst.
Last test: 2026-06-28 out-of-sample re-test. The recorded cost/tail-risk check covered 36/40 coins. Finer-than-annual walk-forward exposed −321% / 62% max drawdown in 2024-H1, a loss hidden by annual aggregation. These measurements do not support unsupervised real-money use; the bull-market stand-down filter and position size remain unresolved constraints.
How It Works
- Band breakout — the previous close breaks below the lower ATR band (close from two bars back − 2.0× ATR(14))
- Trend filter — EMA20 below EMA50 (previous bar) confirms downward bias
- Entry — SHORT on the next bar’s open after the breakout signal
- Exit — TP 10% / SL 3% / 48-bar (2-day) timeout
- Risk — asymmetric: 1:3.3 R:R (3% risk, 10% reward)
Why It Works (Thesis)
ATR spikes in crypto typically mean one of two things: a genuine directional move, or a wick-spike that reverts. On a 1H timeframe with the 3% SL, you get stopped out quickly when it’s a wick. When it’s a real downward expansion — an altcoin breaks support, a leverage cascade starts — the 10% TP captures the bulk of the move.
The 2026-06-28 measurements must be read alongside the finer 2024-H1 loss window above. They do not establish the same behavior across market regimes or in future windows.
Results (2-year backtest, IS/OOS split, measured 2026-05-04)
| Metric | IS (May24–May25) | OOS (May25–May26) | Combined |
|---|---|---|---|
| Total trades | 1,513 | 1,326 | 2,839 |
| Win rate | ~38% | ~34% | 35.9% |
| Profit factor | 1.57 | 1.28 | 1.42 |
| Coins profitable | 48/50 | 36/50 | — |
Low win rate is by design — the 1:3.3 R:R means each win covers 3+ losses. The combined PF (1.42) is reproducible, but read the caveats: a finer walk-forward shows it is regime-dependent, not all-weather.
Caveats
- 2026-06-28 measurement limits. Finer-than-annual walk-forward exposed −321% / MDD 62% in 2024-H1. That observed loss does not support unsupervised real-money use; a bull-market stand-down filter has not resolved the recorded risk.
- OOS PF (1.28) is lower than IS (1.57) — some IS outperformance did not carry forward.
- Tight 3% SL generates frequent stop-outs. Requires accepting a ~64% loss rate.
- Not live-tracked. Backtest only.
- All LONG configurations remain unprofitable on OOS data.