A Fama–French-style decile-cohort event study: does YUCLAW's composite signal score carry forward information about subsequent realized returns? Cohorts are grouped by score decile or signal label and tracked as equal-weighted research cohorts. Derived statistics only — no raw prices. This is an event study, not portfolio management.
Forward Day 0 = 2026-05-18; signals 2026-05-20 → 2026-06-11. A 16-trading-day out-of-sample window is far too short for any statistical inference — this is a directional illustration shown for transparency as the forward record accrues, not evidence of skill.
| Cohort | Cumulative return | Max drawdown | Volatility (periodic) | Hit-rate vs SPY | n (min/med/max) |
|---|---|---|---|---|---|
| Top-decile cohort (by composite score) ⚠ thin | +6.27% | -3.83% | 2.21% | 40% | 1/8/8 |
| Bottom-decile cohort (by composite score) ⚠ thin | -0.66% | -6.50% | 1.24% | 40% | 1/8/8 |
| SPY benchmark (broad-market reference) | +0.39% | -4.49% | 1.03% | — | — |
⚠ The evidence-extraction model's training cutoff overlaps this window — in-sample results carry an unavoidable parametric look-ahead bias and are systematically optimistic. A replay, not a forecast.
| Cohort | Cumulative return | Max drawdown | Volatility (periodic) | Hit-rate vs SPY | n (min/med/max) |
|---|---|---|---|---|---|
| Top-decile cohort (by composite score) | +19.79% | -3.34% | 2.40% | 62% | 8/8/8 |
| Bottom-decile cohort (by composite score) | +8.50% | -6.09% | 3.51% | 54% | 8/8/8 |
| SPY benchmark (broad-market reference) | +7.50% | -5.47% | 1.87% | — | — |
| Cohort | Cumulative return | Max drawdown | Volatility (periodic) | Hit-rate vs SPY | n (min/med/max) |
|---|---|---|---|---|---|
| Bullish-labeled cohort (STRONG_BULLISH + BULLISH) ⚠ thin | +70.44% | -5.92% | 8.28% | 46% | 1/3/10 |
| Cautious-labeled cohort (WEAKENING / NEGATIVE_EVENT / BEARISH_WATCH / RISK_ALERT) | +17.43% | -5.36% | 3.11% | 77% | 4/12/26 |
| SPY benchmark (broad-market reference) | +7.50% | -5.47% | 1.87% | — | — |
⚠ Label cohorts can be as small as a single name on some dates (see n column); small-n figures are statistically noisy and shown for illustration only. The decile cohorts above are the robust comparison.
Equal-weighted cohorts, rebalanced at each signal date, ranked by composite total_score; top/bottom decile (~10%). Returns are close-to-close from internal price_history (derived statistics only — raw prices never shown). Benchmark: SPY. Two panels (forward OOS, in-sample replay) are never blended. Annualized figures are intentionally omitted — annualizing a ~3-month window is misleading; cumulative return over N trading days is shown instead. Full methodology, including the in-sample look-ahead disclosure and the forward-window data limitation, is in methodology/validation_lab.md.