Strategy Performance
The real backtested track record of our signal engine — no cherry-picking, no lookahead.
Win Rate
44.3%
share of trades that hit target
Profit Factor
1.20
gross profit ÷ gross loss
Net Result
+169.9R
total profit in risk units
Max Drawdown
26.2R
worst peak-to-trough dip
Equity Curve
How an account following every signal would have grown, in risk units (R).
Cumulative profit in R (risk units) across all backtested trades — +169.9R total. 1R = the amount risked per trade.
By Timeframe
The edge is strongest on 4H (profit factor 1.36). Higher timeframes filter out noise.
| Timeframe | Trades | Win % | Avg R | Profit Factor | Net R |
|---|---|---|---|---|---|
| 15m | 447 | 46.1% | +0.133 | 1.25 | +59.6R |
| 1H | 420 | 43.8% | +0.086 | 1.15 | +36.1R |
| 4H | 348 | 47.7% | +0.184 | 1.36 | +64.0R |
| 1D | 332 | 38.9% | +0.031 | 1.05 | +10.3R |
Are our confidence scores honest?
We calibrate the confidence % to the real win rate. When we say a signal is ~45%, it historically wins ~45% — verified below.
A ~45% win rate is profitable here because winners are larger than losers — the average trade still nets a positive return (profit factor 1.20). Confidence measures how often a call is right; risk/reward does the rest.
Methodology
- No lookahead. Each signal is computed only from candles up to the decision bar; exits are checked on later bars.
- Conservative fills. When a bar touches both stop and target, the stop is assumed to hit first.
- Real data. 1,547 trades over 1,000 candles per market on 5 assets (bitcoin, ethereum, solana, ripple, dogecoin), via Binance.
- Same engine. The backtest runs the exact production signal engine — what you see live is what was tested.
Past performance does not guarantee future results. Backtests can over-fit and exclude slippage, fees and funding. This is research tooling, not financial advice.