| start | 2018-01-01 |
| end | 2025-12-31 |
| total_return | +889.1% |
| cagr | +33.2% |
| sharpe | 0.87 |
| sortino | 1.36 |
| max_drawdown | -72.4% |
| volatility | +44.0% |
| Check | Result | What it means |
|---|---|---|
| Out-of-sample degradation | fail | Sharpe 1.59 before 2023-08-07 vs -0.42 after (-27% retained). Edges that survive usually keep at least half. |
| Year-by-year consistency | pass | 5 of 8 calendar years were profitable. |
| Deflated Sharpe ratio | pass | Probability the true Sharpe is above zero: 99% before, 99% after deflating for 49 configuration(s) tried, skew 2.40 and kurtosis 58.9. Below 95% means the result is not distinguishable from the best of many lucky tries. |
| Monte Carlo | warn | 2,000 resampled histories (bars). P(loss) 4%. Backtest max drawdown -72%, bad-case (p5) -80%. Plan for the bad case, not the backtest. |
| Parameter sensitivity | pass | 49 parameter combos. 80% sit within 25% of the best sharpe (the 'plateau'); 100% are positive. A lone bright cell is a red flag; a broad plateau is reassuring. |
Sharpe 1.59 before 2023-08-07 vs -0.42 after (-27% retained). Edges that survive usually keep at least half.
5 of 8 calendar years were profitable.
Probability the true Sharpe is above zero: 99% before, 99% after deflating for 49 configuration(s) tried, skew 2.40 and kurtosis 58.9. Below 95% means the result is not distinguishable from the best of many lucky tries.
| sr_period | 0.0453 |
| skew | 2.4016 |
| kurtosis | 58.9009 |
| n_obs | 2921 |
| n_trials | 49 |
| sr_benchmark_period | 0.0005 |
| psr | 0.9946 |
| dsr | 0.9941 |
2,000 resampled histories (bars). P(loss) 4%. Backtest max drawdown -72%, bad-case (p5) -80%. Plan for the bad case, not the backtest.
| total_return | cagr | max_drawdown | |
|---|---|---|---|
| p5 | 9.7% | 1.2% | -80.4% |
| p25 | 289.0% | 18.5% | -66.9% |
| p50 | 872.2% | 32.9% | -58.6% |
| p75 | 2310.6% | 48.9% | -50.1% |
| p95 | 9395.8% | 76.7% | -41.1% |
49 parameter combos. 80% sit within 25% of the best sharpe (the 'plateau'); 100% are positive. A lone bright cell is a red flag; a broad plateau is reassuring.
OOS split: last 30% unless a split date is given. Monte Carlo: bootstrap of closed-trade returns (or stationary block bootstrap of per-bar returns for equity input). Deflated Sharpe: Bailey & Lopez de Prado (2014), benchmark = expected max Sharpe of N null trials. Parameter sensitivity requires the optimization grid CSV. Engine: Crypto Backtest Toolkit Pro.
Backtests are hypothetical, use historical or synthetic data, and do not guarantee future results. Not financial advice.