Evidence by cohort
Signal performance laboratory
Fixed-horizon outcomes grouped without removing failures. Medians reduce outlier distortion; MAE and drawdown expose risk that a headline win rate can hide.
Signal set
Sample
32
Win rate
28.13%
Median return
-3.59%
Median excess
-1.67%
Median MFE
+1.89%
Median MAE
-6.21%
Cohort drawdown
-11.82%
Evidence
30+ sample
8 Sept 2026 to 11 Sept 2026Regime benchmark: Nifty 500
| Sector | Sample | Win rate | Median return | Median excess | MFE | MAE | Drawdown | Evidence |
|---|---|---|---|---|---|---|---|---|
| Financial Services | 6 | 33.33% | -3.95% | -2.68% | +1.80% | -6.19% | -9.64% | Under 10 |
| Chemicals | 4 | 0.00% | -2.61% | -0.61% | +4.22% | -6.39% | -7.07% | Under 10 |
| Industrials | 4 | 50.00% | -1.44% | +0.14% | +2.40% | -7.19% | -8.82% | Under 10 |
| Pharma | 4 | 25.00% | -2.45% | -0.89% | +0.68% | -7.07% | -13.15% | Under 10 |
| Energy | 3 | 33.33% | -8.79% | -7.49% | +4.03% | -10.43% | -11.91% | Under 10 |
| Auto | 2 | 0.00% | -5.41% | -4.12% | +1.21% | -6.72% | -5.41% | Under 10 |
| Consumer | 2 | 50.00% | -0.03% | +1.98% | +1.90% | -4.00% | -1.23% | Under 10 |
| Services | 2 | 50.00% | -1.56% | -0.28% | +3.42% | -5.76% | -5.12% | Under 10 |
| Materials | 1 | 0.00% | -8.28% | -7.03% | 0.00% | -8.28% | -8.28% | Under 10 |
| Media | 1 | 100.00% | +1.89% | +4.61% | +2.66% | -5.80% | 0.00% | Under 10 |
| Metals | 1 | 0.00% | -7.13% | -5.27% | +1.68% | -9.94% | -7.13% | Under 10 |
| Power | 1 | 0.00% | -3.10% | -1.24% | +0.53% | -5.28% | -3.10% | Under 10 |
| Real Estate | 1 | 0.00% | -8.20% | -6.95% | +0.43% | -8.65% | -8.20% | Under 10 |
Returns: Median is used for return, excess return, MFE, and MAE to reduce outlier distortion. Return greater than 0% at the selected fixed horizon.
Drawdown: Signals from the same publication date are equal-weighted as one cohort, then cohort returns are compounded chronologically.
Market regime: Bullish, bearish, or sideways uses the benchmark close versus its trailing 50-session average and trailing 20-session return, using only data available by publication day.
Sample discipline: Groups below 10 outcomes are insufficient; 10-29 are early evidence; 30 or more are established samples, not proof of future performance.