Rejected thesis
VIX regime overlay — sharper macro filter
Code: research/vix_regime_overlay.py
Report: research/vix_regime_overlay_report.json
VIX is a more sensitive macro-stress proxy than SPY drawdown. The finding is even cleaner than the SPY-dd one:
| Drawdown bounce dd ≤ -30% × OOS 2020+ × VIX regime | n | 60d edge | 250d edge |
|---|---|---|---|
| VIX ≤ 20 (calm) | 1,848 | -3.1 % | -2.9 % |
| VIX 20-30 (elevated) | 2,590 | +4.2 % | +14.7 % |
| VIX 30-45 (panic) | 962 | +9.9 % | +42.3 % |
| VIX > 45 (extreme panic) | 185 | +25.3 % | +99.6 % |
The divergence cohort behaves the same way but is more robust in low VIX (still positive +5-6 % at all horizons when VIX < 20). Both amplify dramatically when VIX > 45 (rsi_div +29.7 % at 40d, n=488).
Live implication: drawdown_bounce is essentially a VIX-conditioned trade. Don't fire it in calm VIX (< 20). Divergence cohorts can fire in any regime but are also amplified by elevated VIX.
| Sample | 488 |
|---|---|
| Validator | research/vix_regime_overlay.py |
| Verdict | Rejected |
Every result here is reproducible from the script named above. Reports are in the repository.
- Volatility squeeze
- Pre-FOMC drift
- Uptrend pullback
- Momentum continuation
- Industry-relative momentum (at 4-week lookback)
- Earnings-reaction reversal (anti-PEAD)
- Beat-persistence pre-earnings drift
- Sector ETF mean-reversion
- 52-week-high breakout
- Volume accumulation (above-200DMA +...
- Cross-sector momentum
- Analyst-upgrade cluster