312 FOMC dates. Overnight prev close -> open.
| instrument | n | mean | t | IS | OOS | placebo p |
|---|
| SPY | 201 | +0.104% | +2.06 | +0.150% | +0.045% | 0.0790 |
| VUG | 201 | +0.109% | +2.11 | +0.169% | +0.033% | 0.2138 |
| MTUM | 111 | +0.062% | +0.85 | +0.114% | +0.048% | 0.5682 |
| regime | n | mean | t | IS | OOS |
|---|
| VIX Q1 | 50 | +0.050% | +1.58 | +0.029% | +0.074% |
| VIX Q2 | 49 | +0.121% | +2.86 | +0.162% | +0.078% |
| VIX Q3 | 49 | +0.107% | +1.36 | +0.129% | +0.077% |
| VIX Q4 | 49 | +0.579% | +3.53 | +0.574% | +0.588% |
| uptrend (21d>0) | 128 | +0.166% | +4.60 | +0.186% | +0.147% |
| downtrend (21d<=0) | 68 | +0.302% | +2.41 | +0.301% | +0.304% |
| cost bps RT | mean net | t | IS | OOS |
|---|
| 2 | +0.354% | +3.84 | +0.386% | +0.322% |
| 5 | +0.324% | +3.51 | +0.356% | +0.292% |
- A: if the minutes overnight is positive and OOS-stable, it is a
second Fed-information event to add to the sleeve.
- B: a leg that is strong only in one regime is a conditional filter,
not a standalone edge (re-gridding risk).
- C: the beta basket must beat a single ETF AFTER turnover cost.
Time 12s.