Equity · cumulative % from the start of the record
Shaded: out of sample, days never used to build or tune the strategy. Band: the middle 90% of 2,000 paths with the same daily returns in a resampled order.
Statistics
- Total return
- +117.1%
- CAGR
- +46.8%
- Sharpe
- 4.42annualised
- Sortino
- 7.15annualised
- Max drawdown
- −5.93%peak to trough, % of peak equity
- Worst day
- −3.18%16 Apr 2025
- Worst month
- −1.86%Aug 2026
- Days
- 738
Total return, Sharpe, max drawdown and days are the record’s own figures. Sortino, worst day, worst month, CAGR are computed on this page from its daily series, annualised at the series’ own 365 observations a year.
In sample and out of sample
| Stretch | Dates | Return | Sharpe | Max drawdown | Days |
|---|---|---|---|---|---|
| Out of sample | 30 Sep 2024 – 31 Mar 2026 | +95.3% | 4.95 | −5.9% | 548 |
| In sample | 1 Apr 2026 – 29 Sep 2026 | +20.6% | 2.94 | −3.4% | 182 |
| Out of sample | 30 Sep 2026 – 8 Oct 2026 | +1.3% | 6.73 | −0.2% | 9 |
Monte Carlo · 2,000 resampled paths
| Over the next | Chance of a loss | Return, 5th / 50th / 95th percentile | Max drawdown, typical / 1 in 20 |
|---|---|---|---|
| 90 days | 1% | +3.7% / +14.2% / +32.1% | −2.2% / −6.5% |
| 180 days | 0% | +13.4% / +29.8% / +54.8% | −3.2% / −6.7% |
| 365 days | 0% | +35.4% / +61.6% / +96.9% | −4.0% / −6.8% |
Daily returns of the out-of-sample days, resampled in blocks of about ten days. Resampling reorders the days this record had; it cannot add a market it never saw.
Drawdown · peak to trough, % of peak equity
Rolling Sharpe · 63-day window, annualised
Monthly returns
Outlined months are partial: the record starts or ends inside them.
How this was tested
- This is a backtest: the strategy simulated over historical market data. No money was traded.
- It covers 30 Sep 2024 to 8 Oct 2026, 738 days in the record.
- Returns are measured as % of starting capital.
- Costs: Maker fees on every fill. Not charged: slippage or funding. $8,000 starting capital, the size of the live demo book. No compounding. Open trades are marked to market at the end of every day.
- Partly out-of-sample: the shaded stretches, 557 of the 738 days, were never used to build or tune the strategy; the rest overlaps the data it was tuned on, which tends to flatter the result.
Record: Backtest · 738 days · net of costs · in-sample and out-of-sample. Exported 9 Oct 2026 from platform commit b143faa4 (ML-Stat-Arb-Signal-Service research pools, Bybit incl. delisted coins, point-in-time monthly rotations; the live demo bot's sizing and pair limits; out-of-sample = outside the tuning window).