Generated 2026-05-04 08:44 UTC (7.6 h ago) · methodology · raw JSON · audit log
Today's CME Signals
The Coherent Markets Engine (CME) V0.2.0 pipeline ran across 5,000 Polymarket markets, evaluated 623 cross-market probability-axiom constraints, and emitted 105 Convexly Signals through the three-gate filter (statistical significance, cost-adjusted expected value, resolution risk). Total modeled net EV at prescribed size across surviving signals: $74,747. Methodology pre-registered at AsPredicted #288046; backtest evaluation runs ~2026-07-29 (resolved-PnL pending; capacity coverage gap closing in V0.1.6).
Funnel
Top 30 signals by tension
75 more in the Researcher tier| Event | Type | n markets | Tension (pp) | p-value | Net EV (%) | Days to res. | Est. PnL |
|---|---|---|---|---|---|---|---|
| Event #1 | mene | 12 | 97.95 | <1e-4 | 91.0% | 57 | $418 |
| Event #2 | mene | 8 | 88.65 | <1e-4 | 78.7% | 240 | $197 |
| Event #3 | mene | 3 | 85.91 | <1e-4 | 78.4% | TBD | $356 |
| Event #4 | mene | 15 | 84.90 | <1e-4 | 75.0% | 240 | $323 |
| Event #5 | mene | 6 | 82.90 | <1e-4 | 67.0% | 605 | $1,422 |
| Event #6 | mene | 19 | 82.40 | <1e-4 | 72.7% | 240 | $3,938 |
| Event #7 | mene | 17 | 76.09 | <1e-4 | 66.4% | 240 | $6,620 |
| Event #8 | mene | 13 | 75.40 | <1e-4 | 68.0% | 105 | $4,930 |
| Event #9 | mene | 36 | 73.90 | <1e-4 | 64.0% | 240 | $629 |
| Event #10 | mene | 24 | 69.40 | <1e-4 | 59.5% | 240 | $261 |
| Event #11 | mene | 34 | 66.25 | <1e-4 | 59.8% | 28 | $9,222 |
| Event #12 | mene | 22 | 65.40 | <1e-4 | 55.5% | 240 | $1,927 |
| Event #13 | mene | 9 | 64.70 | <1e-4 | 54.8% | 240 | $137 |
| Event #14 | mene | 13 | 63.90 | <1e-4 | 54.1% | 241 | $3,028 |
| Event #15 | mene | 12 | 62.23 | <1e-4 | 55.3% | 57 | $1,104 |
| Event #16 | mene | 5 | 60.40 | <1e-4 | 50.4% | 241 | $76 |
| Event #17 | mene | 10 | 59.40 | <1e-4 | 52.0% | 86 | $106 |
| Event #18 | mene | 6 | 59.05 | <1e-4 | 49.9% | 209 | $887 |
| Event #19 | mene | 10 | 57.93 | <1e-4 | 48.0% | 241 | $610 |
| Event #20 | mene | 27 | 56.90 | <1e-4 | 47.0% | 240 | $2,979 |
| Event #21 | mene | 8 | 56.10 | <1e-4 | 49.2% | 56 | $403 |
| Event #22 | mene | 5 | 54.90 | <1e-4 | 48.5% | 23 | $266 |
| Event #23 | mene | 17 | 48.70 | <1e-4 | 39.6% | 240 | $5,165 |
| Event #24 | mene | 21 | 48.08 | <1e-4 | 38.1% | 241 | $60 |
| Event #25 | mene | 28 | 47.99 | <1e-4 | 38.0% | 240 | $73 |
| Event #26 | mene | 14 | 46.90 | <1e-4 | 37.0% | 240 | $338 |
| Event #27 | mene | 11 | 45.95 | <1e-4 | 36.0% | 240 | $167 |
| Event #28 | mene | 3 | 45.71 | <1e-4 | 35.8% | 241 | $91 |
| Event #29 | mene | 8 | 45.15 | <1e-4 | 34.2% | 299 | $125 |
| Event #30 | mene | 12 | 44.08 | <1e-4 | 37.2% | 54 | $29 |
What is on this page
- Daily three-gate signal counts + funnel summary (counts only)
- Per-signal aggregate characterizations (constraint type, tension magnitude, p-value, net EV%, est PnL) — event identities redacted
- Methodology + hash-chain audit log + AsPredicted #288046 pre-registration
What the Researcher tier ($99/mo) adds
- Event identities (which Polymarket markets each signal references)
- The full 105-signal feed (not just top 30)
- Trade construction (which markets to long/short, sizing, capital lockup)
- Day-over-day signal delta (which signals appeared, persisted, resolved)
- Search across the full dated archive (signals from any past day)
- Daily email digest (Resend; opt-in)
- 5-15 min delay relative to cron generation
The CME methodology is documented in full at /research/coherent-markets-engine-v1. The 90-day walk-forward backtest is pre-registered at AsPredicted #288046 (filed 2026-04-29; methodology frozen at commit 8616a63). The CFTC ANPRM comment on probability-axiom- feasible price coherence as a regulatory transparency standard was filed at regulations.gov on 2026-04-29 (docket CFTC-2026-0331-0001, comment tracking mok-lrhf-09it).
Audit-log row 7ac8d5f689bf88a3… cryptographically links the daily input artifacts (gamma metadata, four-price CLOB snapshots) to the published outputs. The full hash chain is at /research/cme/audit_log.jsonl and verifies via python -m services.api.scripts.cme_v0_1.provenance verify.
This work used AI tools (Claude, GPT-4) as research aids during methodology design and pre-publication review. All claims and quantitative results on this page are reproducible from the public data bundle (cron output atapps/web/public/research/cme/) with the documented seed; no claim is taken as true on the basis of an AI tool's output.