3 strategies tested. All 3 lost money.
Fixed rules, public Bybit data, fees and slippage on every fill.
Every result
| Sample | Trades | Win rate | Average R | t-statistic | Max drawdown | Verdict |
|---|---|---|---|---|---|---|
| Fair value gap BTCUSDT 15m | ||||||
| In-sample 2025-11-18 to 2026-07-30 | 801 | 31.46% | −0.57 | −4.93 | 66.97% | Clear loss |
| Out-of-sample 2026-07-31 to 2026-11-18 | 304 | 27.96% | −0.83 | −5.23 | 43.19% | Clear loss |
| Liquidity sweep BTCUSDT 15m | ||||||
| In-sample 2025-11-18 to 2026-07-30 | 1,639 | 29.16% | −0.67 | −10.91 | 94.35% | Clear loss |
| Out-of-sample 2026-07-31 to 2026-11-18 | 688 | 28.20% | −0.79 | −8.48 | 69.31% | Clear loss |
| MA cross BTCUSDT 15m | ||||||
| In-sample 2025-11-18 to 2026-07-30 | 572 | 20.80% | −0.31 | −2.95 | 63.33% | Clear loss |
| Out-of-sample 2026-07-31 to 2026-11-18 | 263 | 19.01% | −0.05 | −0.68 | 31.18% | Cannot be told apart from zero |
Fees and slippage included. Funding payments are not.
What R measures
A trade’s profit after costs, divided by what it would have lost at its stop before costs.
The verdict rule
| In this order | Verdict |
|---|---|
| Under 100 trades | Too few trades |
| t not computed | Not computed |
| t ≤ −2 | Clear loss |
| t ≥ 2 | Clear gain |
| −2 < t < 2 | Cannot be told apart from zero |
What t measures
The average trade after costs, against zero. It assumes independent trades, which flatters it.