Combos
The combo/parlay product, on its own: what it earns, why the number depends entirely on which denominator you pick, and how fast it's growing. Volume is measured from each venue's own tape; Kalshi's fees are modelled, because no venue publishes them.
Per-market breakdown, plus the fee chart: /volumes.
Combo share of allKalshi notional — 1.8% in Oct 2025 to 41.0% MTD in Aug 2026 — and, inside combos themselves, the share of that notional trading under 2¢. One price band explains almost the entire gap between combos' huge notional and their tiny share of real cash: Sep 2025 – 13 Aug 2026, combos are 25.8% of notional but only 6.3% of handle.
| Notional | Handle | Fees (modelled) | Trades | Avg price | |
|---|---|---|---|---|---|
| Combos (KXMVE) | $46.6B | $4.0B | $177.8M | 92,279,527 | 8.5¢ |
| Everything else | $133.7B | $59.0B | $1.4B | 991,205,907 | 44.2¢ |
Combo share of Kalshi: 25.8% of notional · 6.3% of handle · 11.2% of fees.
Retail pays roughly 3x the fee rate of a 50¢+ trade — not because Kalshi charges small accounts more, but because retail buys longshots, and 0.07 × P × (1−P) is highest at the extremes of the price range.
| Notional | Handle | Fees | Fee on stake | Tier | |
|---|---|---|---|---|---|
| Kalshi | $32.7B | $2.6B | $115.0M | 4.48% | MODELLED |
| Polymarket | $257.0M | $46.4M | $1.0M | 2.2% | MEASURED |
Polymarket combos launched 22 May 2026 (17 transactions in beta); June is the first full month. Kalshi's combo book is roughly 127x larger on notional over the same window.
Every executed trade, tradeType = COMBO. Notional is the sum of qty on TAKER rows, handle the sum of cost. Quoted by liquidity providers through an RFQ rather than matched on the book. Source: Novig public trade tape (data.novig.com).
Markets whose report ticker is a combo series (KXMVE). Volume is a contract count; handle is each market's volume priced at its own price. Kalshi also runs an RFQ; both venues quote multi-leg contracts as well as listing them. Source: Kalshi public daily tape (kalshi-public-docs S3).
No combo/parlay series built yet. Slot held open. Nothing is charted until the series exists — an empty panel, not a zero line. Source: —.
Handle is cash actually staked and is the column to compare across venues. Notional counts $1-payout contracts; combo contracts clear at a few cents, so notional flatters parlays everywhere. Each venue is charted on its own scale — the series are not measured identically and are never summed. Novig detail: Novig.
Fees row-level: 0.07 x contracts x P x (1-P), P in [0,1], never on an average — Jensen's inequality overstates by ~40% on a price average. Kalshi fees are MODELLED (no fee field published); Polymarket fees are MEASURED on-chain. Combo = report_ticker LIKE 'KXMVE%'.
| Metric | Ours | Independent check |
|---|---|---|
| Oct 2025 combo share of notional | 1.8% | Part 1 (Predicted): “under 2%” |
| Apr 2026 combo share of notional | 22.6% | Sportico: 22% |
| Jun 2026 combo notional | 36.1% of $33.0B | Prior draft: 36% of ~$33bn |
| Q2 2026 combo notional | 30.6% of $65.7B | Predicted Q2 report: 30%, $65.7B |
| Fee rate on handle | 4.48% | Sportico: 4.5% |
Research as of 2026-08-13. Combos Article research (Dune, kalshi.trade_report), third piece in the series.