Most coverage of this question is about law — whether event contracts are derivatives or wagers, and which regulator wins. That fight matters, but it obscures a more decisive point.
Even if prediction markets win every court case, they do not have the economics to replace sports betting. The two businesses make money in fundamentally different ways, and the difference is large.
The economics, side by side
Sportsbook | Prediction market exchange | |
|---|---|---|
Revenue model | Hold — the share of total wagers retained after payouts. The operator is the counterparty. | Fee on trade volume. Users trade against each other; the venue takes a cut either way. |
Standard market margin | A –110/–110 market implies roughly 4.76% theoretical vig built into the price. | Kalshi charges an explicit per-contract fee, capped at $0.07 per side per contract — roughly 1% on a trade. |
Parlay margin | Around 20%, and by some measures 20–30% — roughly four times a straight bet. | Approximately the same 1%, whether the trade is a parlay or a straight position. |
Actual realised margin | US national hold reached 10.1% for 2025 and 11.1% in April 2026. | Fee-based, outcome-independent. |
Who bears outcome risk | The operator, on every bet. | Nobody at the venue. Traders take both sides. |
Kalshi earns about 1% on a parlay. A sportsbook holds about 20% on the same bet. Volume can move; margin cannot follow it.
The parlay problem
This is the crux, and it is worth spelling out because it explains the entire competitive dynamic.
Parlays transformed sportsbook profitability. Average US hold moved from roughly 6.7% in 2018 and 7.5% in 2021 to 9.2% in 2024 and 10.1% in 2025, reaching 11.1% nationally in April 2026 against 9.3% a year earlier. In Massachusetts that same month, FanDuel reported holding 13.40% and DraftKings 13.12%, with the majority of handle coming from parlays.
Both DraftKings and FanDuel have pointed to a higher parlay mix as the key driver of margin expansion, and it coincided with the industry finally reaching profitability.
Now apply that to an exchange. Kalshi’s CEO has described the model as a fair game — participants trading positions rather than betting against an operator. That is an accurate and appealing description, and it is also the commercial constraint. A fair game does not have a house edge to expand.
So parlay volume growth on a prediction market does not produce the margin expansion it produces for a sportsbook. The revenue is a fee on turnover, and it is the same fee regardless of how exotic the product is.
Which is why the convergence is dangerous for exchanges
Prediction market operators have been adding parlay-style and multi-leg contracts, chasing the products that drive sportsbook engagement.
The trouble is that this imports the legal risk without importing the economics. Every step towards parlays strengthens the argument of state regulators claiming these are wagers in a different wrapper — while generating roughly the same 1% the venue would have earned anyway.
That is an unattractive trade: maximum regulatory exposure for minimal margin gain.
The scale comparison
TWO INDUSTRIES, MEASURED DIFFERENTLY US sports betting, 2025 — handle of roughly $165.4 billion, gross gaming revenue of about $16.6 billion, national hold 10.1%, and around $3.73 billion paid in state taxes. Prediction markets, Q2 2026 — sector volume of roughly $111 billion, with fees measured in the low single-digit percentage points at most. Note the units differ. Sportsbook handle counts money staked; exchange notional volume counts contracts traded on both sides. They are not directly comparable, which is precisely the point. Exchange-style betting platforms have historically accounted for under 1% of US handle even in the states where they operate. |
The last line deserves attention. Peer-to-peer betting exchanges are not new — Sporttrade and Prophet Exchange have operated in a handful of US states, and Betfair has run the model internationally for two decades. Where they compete head-on with sportsbooks under gaming licences, they remain marginal.
Prediction markets are growing far faster than those exchanges ever did, but the structural reason exchanges stay small is unchanged: the product is better priced for the user and worse monetised for the operator.
The finding that reframes everything
Research indicates roughly 1% customer overlap between sportsbook users and prediction market users.
If prediction markets were replacing sports betting, that number would be high. Replacement requires the same customer switching venues. One percent overlap describes two largely separate populations.
That explains the behaviour of the incumbents better than any legal theory does. DraftKings did not litigate the category out of existence — it reportedly invested $200–300 million, saw volumes quintuple between April and July, and reached around 600,000 users. Fanatics launched across 24 states. You acquire a category that expands your market; you fight one that takes it.
What each format is genuinely better at
Prediction markets win on | Sportsbooks win on |
|---|---|
Price. A liquid exchange market typically pays out better than a –110/–110 sportsbook line, a meaningful edge before commission. | Parlays and same-game parlays, with built-in engines and the product depth casual users want. |
Ability to take either side of a market natively, rather than only the offered position. | Promotions — sign-up offers, odds boosts, parlay insurance. Exchanges run a minimal bonus economy. |
Treatment of winning users. Limits scale with liquidity rather than being cut for skill. | State coverage. Around 35 states versus a handful for exchange-style platforms. |
Markets a sportsbook will not list — economics, politics, awards, long-dated outcomes. | Brand, distribution and customer acquisition machinery built over eight years. |
Exit before resolution at a market price. | Scale. DraftKings and FanDuel together held roughly 70% of US handle in Q1 2026. |
For a price-sensitive user, the exchange is better. For most users, the sportsbook is a better product experience. Those two facts have coexisted in international markets for twenty years without either format eliminating the other.
The argument states will make
There is a fiscal dimension that will shape this more than most commentary allows.
US sports betting generated around $3.73 billion in state taxes during 2025. Prediction markets regulated federally as derivatives do not pay state gaming taxes on the same activity.
That gives state governments a direct financial interest in the classification fight that has nothing to do with consumer protection arguments. Whatever the legal merits, a category that moves wagering volume out of a taxed channel and into an untaxed one will meet sustained political resistance — and states have already brought enforcement actions in Nevada, New Jersey, Ohio, Massachusetts, Tennessee, Maryland, Illinois and Connecticut.
The most plausible long-run outcome is not that one format wins. It is that prediction markets eventually get taxed or licensed in a way that narrows the arbitrage.
So can they replace sports betting?
On volume, potentially. Growth rates are extraordinary and prediction markets already handle sums comparable in headline terms to established betting handle.
On revenue, no. A 1% fee cannot replicate a 10–11% hold, and the parlay engine that drives sportsbook margin does not transfer to an exchange model.
On customers, apparently not. One percent overlap suggests expansion of the total market rather than substitution within it.
On product, partially. Prediction markets are superior on price and on markets sportsbooks will not offer. They are behind on the products most users actually buy.
The bottom line
Prediction markets are not replacing sports betting. They are growing alongside it, serving a largely different audience, with a business model that is better for users and considerably less profitable per unit of activity.
That is precisely why the incumbents bought in rather than fought. DraftKings put money into the category; Fanatics launched in 24 states; Robinhood and Crypto.com entered. Nobody with a functioning sportsbook believes an exchange charging 1% is going to displace a book holding 11%.
The real competitive question is narrower and more interesting: whether prediction markets can build a product casual users prefer without adopting the parlay-style features that would make them legally indistinguishable from the thing they claim not to be. So far, they are trying — and the attempt carries all the regulatory risk and almost none of the margin.
Important
This article is industry analysis and general information. It is NOT investment advice, not a recommendation to trade event contracts, and not encouragement to gamble. Handle, hold, volume and fee figures are drawn from third-party industry sources and regulatory filings, reflect the periods stated, and vary by state, operator and methodology; handle and exchange notional volume are not directly comparable measures. The legal status of event contracts is contested and varies by jurisdiction. Both sports betting and event contract trading carry risk of financial loss. If gambling is causing you or someone you know harm, support is available; in the US, the National Council on Problem Gambling helpline is 1-800-522-4700.
Sources
Handle, revenue and hold data from Sportsbook Review, the American Gaming Association revenue tracker and state regulator reports; fee and volume data from Kalshi disclosures and Dune Analytics dashboards; plus analysis from Legal Sports Report, Sports Illustrated, Track360, Tech-Insider and Sportsboom.
Bitnxt tracks prediction markets, exchanges and licensed operators across the US, UK, EU and UAE. Explore the directory at bitnxt.io.
