---
name: Kalshi market volume vs strategy ROI - open question
description: Hypothesis that higher-volume Kalshi markets have less edge; data is inconclusive due to EPL being a single-league outlier. Re-check at n~40+ per league.
type: project
---

Open research question: **Do higher-volume Kalshi soccer markets erode the CLV edge?**

**Why:** EPL paper P&L over 2026-04-22 to 2026-05-25 is -65% ROI on n=19 — worst of 22 leagues by a wide margin. EPL is also currently the only whitelisted league for the auto-trade rollout (`AUTO_TRADE_LEAGUES=['E0']`). Need to know if this is structural before flipping `AUTO_TRADE_ENABLED=True`.

**How to apply:** When user revisits auto-trade league whitelist, EPL performance, or asks about market-efficiency effects, refer back to this analysis. Re-run the study when EPL paper sample reaches n>=40 (probably ~3 months out from 2026-05-26).

## Method
- Pulled median `volume_fp` per Kalshi soccer series via `/markets?status=settled` (Kalshi has both `volume_fp` total and `volume_24h_fp` daily; used total).
- Cross-referenced against paper-strategy ROI from `bet_history.json` (304 Kalshi bets, 22 leagues, +29% ROI overall).

## Volume tier on Kalshi (median contracts per market)
EPL is its own tier:
- EPL: 446k (alone at the top)
- La Liga: 292k
- Serie A: 184k
- Bundesliga: 159k
- Ligue 1: 122k
- MLS: 120k
- ... down to EFL L1 at 5k

## Correlation results (median Kalshi volume vs paper ROI)
| Sample | Pearson r | Pearson r(log) | Spearman rho |
|---|---|---|---|
| All 22 leagues | -0.157 | -0.018 | -0.021 |
| Excl EPL | +0.084 | +0.131 | +0.126 |
| Excl EPL + Arg Premier (n=1 outlier) | +0.260 | +0.170 | - |

## Conclusion (as of 2026-05-26)
**Hypothesis is unfalsifiable in current sample.** The whole negative correlation collapses (and flips sign) when EPL is removed. EPL is alone in its volume tier, so we cannot tell:
- (A) "EPL is structurally sharper because of Kalshi volume" vs
- (B) "EPL is a P=2.5% bad-luck draw at n=19" vs
- (C) "Something else about EPL — line-move timing, MM activity, fee % at low asks"

Strongest counter-example to volume-erodes-edge story: **EFL L1 has lowest volume (5k) and 2nd-highest ROI (+72.5%)**. If volume were the driver, low-vol leagues should systematically outperform — they don't (USL 9k vol -> -5%, SL Greece 18k -> -33%).

## Practical takeaways embedded in plan
1. EPL whitelist for auto-trade soak test is risky — both worst paper performer AND highest Kalshi volume. Consider switching whitelist to MLS (n=37, +71% ROI, 3rd-highest volume, still a real test of plumbing) before enabling.
2. Re-test when EPL sample reaches n>=40. Spearman should stabilize one way or the other.

## Snapshot of full league table (2026-05-26)
| League | MedVol | BetsN | P&L | ROI% |
|---|---|---|---|---|
| EPL | 445,818 | 19 | -48.04 | -65.4 |
| La Liga | 292,346 | 18 | +47.39 | +53.4 |
| Serie A | 184,435 | 21 | -5.26 | -6.3 |
| Bundesliga | 158,666 | 11 | +56.05 | +100.1 |
| Ligue 1 | 122,390 | 11 | +18.14 | +59.8 |
| MLS | 120,494 | 37 | +141.54 | +71.4 |
| Liga Portugal | 45,976 | 22 | -5.30 | -4.0 |
| Brasileirao | 45,342 | 6 | -15.53 | -51.5 |
| Arg Premier | 44,563 | 1 | +3.55 | +355.0 |
| EFL Champ | 35,709 | 13 | +30.24 | +36.7 |
| La Liga 2 | 34,933 | 23 | +39.85 | +32.5 |
| Super Lig | 33,099 | 8 | -8.02 | -19.9 |
| Eredivisie | 30,117 | 16 | +18.59 | +19.7 |
| Bundesliga 2 | 28,726 | 15 | +62.30 | +61.1 |
| SL Greece | 18,093 | 17 | -30.33 | -32.6 |
| Scot Prem | 16,999 | 7 | +34.38 | +83.4 |
| Serie B | 16,637 | 8 | +8.84 | +18.4 |
| Belgian PL | 13,029 | 9 | -3.34 | -8.4 |
| Eliteserien | 12,422 | 7 | +10.73 | +49.4 |
| Allsvenskan(SE1) | 12,228 | 6 | +10.41 | +32.3 |
| USL | 9,114 | 10 | -2.71 | -5.4 |
| EFL L1 | 5,483 | 19 | +101.72 | +72.5 |