A. Pre-earnings drift (OOS 2016+)
| Variant | Sharpe | Ret | MaxDD | p |
|---|---|---|---|---|
| pre-earn [d-10,d-3] (no announcement touch) | 1.05 | +432.16% | -24.30% | 0.0006 |
| pre-earn [d-5,d-1] | 0.99 | +339.43% | -25.97% | 0.0008 |
| market EW (baseline) | 1.01 | +500.92% | -38.33% | 0.0004 |
B. Vol-scaled EqMom (OOS 2016+)
| Variant | Sharpe | Ret | MaxDD |
|---|---|---|---|
| EqMom EW (baseline) | 1.06 | +1004.91% | -38.55% |
| EqMom 1/vol-scaled | 0.92 | +556.04% | -39.02% |
C. Book (EW monthly, 2020-01-02.., n=1629)
| Book | Sharpe | CAGR | MaxDD |
|---|---|---|---|
| 4-sleeve ex-GLD | 1.10 | +15.99% | -25.86% |
| 5-sleeve +GLD | 1.27 | +16.42% | -21.82% |
Verdicts
- Pre-earnings drift: REJECT (1.05/0.99 vs market 1.01, lower total return -- pure beta participation, no edge).
- Vol-scaled EqMom: REJECT (0.92 vs 1.06, maxDD unchanged -- scaling does not help).
- 5-sleeve +GLD book: MARGINAL (Sharpe 1.27 vs 1.10, maxDD -21.8% vs -25.9%, GLD corr 0.01-0.28 -- but paired p=0.50; diversifier properties only, no book change now).