The Lab
Every trading day at 9:25 ET the Lab forecasts where SPY will be at 10:00, noon, 2:00 and the close, then learns from what actually happened. The bar it has to clear is the simplest honest forecast: the latest pre-market price holds.
Skill at the close +1.4% ± 1.1% · promising, not proven
Edge hit rate 63% direction beyond the latest price, 59 leaned days
Range coverage 71% target 68%
Days learned 127 first 15 are warmup and not counted
Call
Wed, Jul 15, 2026 Graded All inputs inClose target $754.20, range $750.54 to $757.89, 71% chance of closing above the prior close ($751.83). Closed at $754.81.
| Checkpoint | P(up) | Target | Range | Actual | In range | vs naive |
|---|---|---|---|---|---|---|
| 10:00 | 78% | $753.61 | $751.49 – $755.73 | $752.14 | Yes | Beat (0.299 vs 0.349) |
| Noon | 74% | $753.81 | $751.33 – $756.30 | $753.20 | Yes | Beat (0.195 vs 0.242) |
| 2:00 | 73% | $753.97 | $750.79 – $757.15 | $753.18 | Yes | Beat (0.189 vs 0.240) |
| Close | 71% | $754.20 | $750.54 – $757.89 | $754.81 | Yes | Beat (0.193 vs 0.248) |
pushes downpushes up
- Baseline lean
- IWM minus QQQ yesterday
- VIX futures (VIXY) yesterday
- Net dealer gamma
- Put/call ratio (log)
- Pre-market gap
- High yield (HYG) yesterday
- Implied vs realized vol
Track record
-10.0%-5.0%+0.0%+5.0%+10.0%
naive+1.4%
May 22Jul 16Sep 3Oct 26
40%60%80%100%
68% target70%
May 22Jul 16Sep 3Oct 26
- Kalman: everything, slow memory 13%
- Kalman: options positioning 12%
- Gaps keep going 11%
- Analog days (10 nearest) 10%
- Yesterday's move reverses 10%
- Everyone else 44%
0%50%100%
Apr 27Jun 26Aug 26Oct 26
0%50%100%
22
16
15
19
40
0%25%50%75%100%
Leaderboard
Average CRPS at the close in sigma units (lower is better), and each expert's edge over "latest price holds". Weight is how much the ensemble listens to it today.
| Expert | CRPS | vs naive | Weight | Verdict |
|---|---|---|---|---|
| Kalman: everything, slow memory | 0.4752 | +0.0061 ± 0.0098 | 13.1% | ~1064 more days to confirm |
| Gaps keep going | 0.4763 | +0.0049 ± 0.0076 | 11.4% | ~972 more days to confirm |
| Kalman: options positioning | 0.4772 | +0.0040 ± 0.0079 | 11.5% | ~1578 more days to confirm |
| Yesterday's move reverses | 0.4798 | +0.0015 ± 0.0044 | 10.1% | ~3923 more days to confirm |
| Latest price holds (random walk) | 0.4812 | — | 9.4% | the benchmark |
| Analog days (10 nearest) | 0.4818 | -0.0006 ± 0.0081 | 10.4% | no edge so far |
| Kalman: price action | 0.4828 | -0.0016 ± 0.0096 | 7.3% | no edge so far |
| Kalman: Jev text reads | 0.4849 | -0.0036 ± 0.0035 | 7.4% | no edge so far |
| Dealer hedging drift | 0.4870 | -0.0058 ± 0.0047 | 6.6% | no edge so far |
| Yesterday's direction continues | 0.4897 | -0.0085 ± 0.0046 | 6.1% | no edge so far |
| Kalman: everything, fast memory | 0.4947 | -0.0135 ± 0.0150 | 3.0% | no edge so far |
| Gaps fade | 0.5022 | -0.0209 ± 0.0080 | 3.9% | no edge so far |
By checkpoint
| Checkpoint | Skill | From range | From edge | Coverage | Edge hit rate |
|---|---|---|---|---|---|
| 10:00 | +10.1% ± 1.9% | +2.2% | +8.0% | 74% | 76% (68) |
| 12:00 | +1.5% ± 1.0% | +0.6% | +0.9% | 69% | 65% (37) |
| 14:00 | +1.9% ± 0.9% | +1.7% | +0.2% | 71% | 52% (31) |
| Close | +1.4% ± 1.1% | +0.7% | +0.7% | 71% | 63% (59) |
How it learns
- Senses. Options positioning from the 9:25 snapshot, yesterday's moves in stocks, bonds, credit, the dollar and VIX, and Jev's reads of the morning headlines and WSB.
- Experts. Simple fixed hypotheses ("gaps fade"), Kalman filters whose weights drift slowly with the market, and an analog expert that asks what happened on the most similar past days.
- Ensemble. Experts earn weight by forecasting better (Hedge), and every expert keeps a small share so one that fell out of favor can come back when the market changes.
- Range. The range widens after a close lands outside it and narrows after one lands inside, so over time about 68% of closes fall inside it.