Prediction Markets Traded $75.5B In September As Kalshi's Lead Narrowed

American sports betting used to mean one thing, an app where you place a wager and the house sets the odds. A second version has grown up next to it, where people trade yes or no contracts on an exchange. September's numbers show how far that second version has come. Prediction market competition has its own scorecard.

Two volume numbers that look nothing alike

Sportsbook handle across 15 states came to about $11.11B in September, down an estimated 0.3% from a year earlier. Prediction-market notional volume across four exchanges reached $75.5B over the same month, with Kalshi at $60.28B in notional volume. Beneath the near-flat total, seven states posted year-over-year gains, seven reported declines, and Iowa was virtually unchanged.

Those two figures aren't apples to apples. Handle measures what bettors actually wagered, while notional volume counts every contract at its $1 settlement value, so a contract bought for five cents still adds a full dollar.

Parlay-style combination contracts widen the gap further, since they trade for pennies and settle at a dollar.

Kalshi's lead is narrowing

Kalshi's share of regulated prediction-market volume fell to 79% in September from 86% in August as Polymarket and other exchanges picked up ground.

Stifel put single-event sports contracts at $22.8B in September, up 23% from August. Combination contracts more than doubled to $41.6B, and together the two made up 83% of Kalshi's volume.

The same research found sportsbook handle rose an estimated 4% year over year on Stifel's narrower 13-state count, with gross gaming revenue down 9% as results favored bettors.

The cost of competing is rising

Flutter EntertainmentFLUT, which owns FanDuel, generated group revenue of $3.87B in the third quarter, up 2%, and cut its full-year outlook. Adjusted operating profit fell 25%.

Dan Taylor, who took over as Flutter chief executive on Oct. 1, made the case in his first shareholder letter.

Defending that position costs money. Category marketing spend rose 60% or more in September and 50% or more in October, driven largely by the prediction operators, Taylor said on Yahoo Finance. Flutter, by contrast, intends to put at least as much money behind its own prediction market in 2027.

DraftKingsDKNG reported third-quarter results Thursday evening. Its 2026 guidance of $700M to $900M in adjusted earnings before interest, taxes, depreciation, and amortization already absorbs $200M to $300M of prediction-market spending, and analysts expect $660M of that EBITDA this year.

FanDuel is handing the permanent chief executive job to Christian Genetski, effective Jan. 1, after he ran the business on an interim basis since May.

Regulators are still drawing lines

FanDuel stopped letting Utah users of FanDuel Predicts trade sports, politics, and novelty contracts after the state filed felony criminal charges against Kalshi. That exit shows how quickly state action can reshape the category.

That uncertainty is why operators have largely confined their own prediction products to states without legal sportsbooks, which limits how much their earnings reveal about the competitive damage.

Who collects as the two markets merge

Two groups sit on the right side of the convergence. The first is operators that run both a sportsbook and a prediction product, since a customer trading contracts is still a customer. The NFL betting shake-up has its own playbook.

The second is the plumbing. Stifel also flagged the tie-ups with Kalshi and Polymarket as a source of upside for SportradarSRAD and Genius SportsGENI, pointing to Genius's acquisition of Legend and room for more advertising revenue.

Data and odds suppliers get paid whichever venue the wager lands on. Stifel kept buy ratings across DraftKings, Flutter, Rush Street InteractiveRSI, and Sportradar, and warned that further earnings misses could push out any recovery.