local _ADJ dividend column · 500 names · raise 4849 / cut 695 events · entry next-session close, excess vs EW panel · IS<2016/OOS>=2016
| horizon | IS excess | OOS excess | OOS net10 | OOS hit | boot p | placebo p | n OOS |
|---|
| 5d | +0.39% | +0.11% | -0.09% | 48% | 0.185 | 0.037 | 351 |
| 21d | +1.17% | +0.41% | +0.21% | 48% | 0.069 | 0.001 | 350 |
| 63d | +2.55% | +1.36% | +1.16% | 48% | 0.005 | 0.000 | 346 |
| horizon | IS excess | OOS excess | OOS net10 | OOS hit | boot p | placebo p | n OOS |
|---|
| 5d | -0.08% | +0.04% | -0.16% | 50% | 0.328 | 0.704 | 2366 |
| 21d | -0.21% | +0.01% | -0.19% | 50% | 0.130 | 0.863 | 2357 |
| 63d | -0.47% | +0.40% | +0.20% | 48% | 0.395 | 0.687 | 2316 |
| side | bar | result |
|---|
| cut (h21) | mean<0, boot/placebo p<0.05, IS same sign, n>=50 | REJECT (OOS +0.41%, IS +1.17%, boot 0.069, placebo 0.001, n_oos 350) |
| raise (h21) | mean>0, boot/placebo p<0.05, IS same sign, n>=50 | REJECT (OOS +0.01%, IS -0.21%, boot 0.130, placebo 0.863, n_oos 2357) |
| horizon | beta | raw excess | beta-adj excess | raw OOS | adj OOS | up-window | down-window |
|---|
| 21d | 1.55 | +0.79% | -0.33% | +0.41% | -0.54% | +1.95% | -1.58% |
| 63d | 1.65 | +1.95% | -1.89% | +1.36% | -2.22% | +3.08% | -2.69% |
Cutters carry beta ~1.6; the raw positive excess is a high-beta / recovery-regime artifact — beta-adjusted excess is negative in both windows. The classic cut->negative drift is absent too. Null after the audit.
- Baseline = median of up to 4 prior payments (special-vs-regular resets still leak as noise; stated, not fixed here).
- No DSR yet; a full study would add K over horizons/thresholds and reconcile specials.