Prediction markets traded $75.5 billion in notional volume in September. Sportsbook handle across 15 states came in at $11.11 billion, down an estimated 0.3% year over year, as football returned.
Prediction markets reached a record $75.5 billion in September notional trading across four major exchanges. In contrast, sportsbook handle in 15 states was nearly flat at $11.11 billion. Kalshi alone accounted for $60.3 billion. Combos made up 58% of its September activity, while straight sports contracts added 24.9%.
- Prediction Markets Set September Records
- Sportsbook Revenue Diverges From Handle
- Why Notional Volume Misreads the Rival
- The Evidence on Customer Cannibalization
Prediction markets reached $75.5 billion in notional trading volume in September. The figure covers four major exchanges and marks a record month. Traditional sportsbooks accepted $11.11 billion in handle across 15 states, down an estimated 0.3% year over year. Seven states rose, seven fell, and Iowa was nearly unchanged. Football returning pushed both markets upward, but the sportsbook side barely moved.
That result follows a rough August for sportsbooks. Online handle fell about 3% year over year nationally, with the American Gaming Association recording a nationwide decline of 0.7% as sportsbooks took in $10.89 billion. Excluding the World Cup months of June and July, monthly handle had declined year over year since December 2025.
Prediction Markets Set September Records
Kalshi posted $60.28 billion in September notional volume, up 56.4% from August, according to TickerTracker. Across Kalshi, Polymarket, Polymarket US and Novig, the total reached $179.6 billion for the quarter, up 84.3% from Q2. Kalshi contributed $60.7 billion in September on Prediction Frontier’s dataset, a slightly higher number than the tracker’s $60.3 billion.
The momentum came from football weekends. Kalshi generated $4.89 billion over September 12-13, its first full NFL weekend, a then-record for two days. TickerTracker put the same weekend at $5.66 billion across seven platforms. On September 26-27, Kalshi recorded consecutive single-day records near $3.03 billion and $3.04 billion, while the next weekend’s eight exchanges generated $7.64 billion, with Kalshi taking in $5.83 billion.
Sportsbook Revenue Diverges From Handle
The divergence is concentrated in New York. The state generated $159.4 million in sports wagering revenue in September, down 17.7% year over year, even as handle rose 5%. New Jersey’s revenue fell 14.3% to $77 million with volume down just 3.6%. Maryland and North Carolina followed similar patterns.
Ohio and Pennsylvania moved the other way. Ohio’s sportsbook revenue rose 12.8% to $87.7 million despite a 2.2% handle decline. Pennsylvania posted a 50.5% revenue increase to $44.7 million as volume fell 6.1%. Massachusetts and Nevada showed similar trends.
The gap points at hold, the share operators retain after paying winning bets. New York’s hold dropped from about 8.47% in September 2025 to 6.64% this year. Maryland’s statewide hold fell from 11.2% to 9.0%. Better outcomes for players can shrink revenue even when wagering stays stable or climbs.
AGBrief covers state-by-state revenue filings and handle reporting.
Why Notional Volume Misreads the Rival
Prediction markets report contracts at their $1 settlement value regardless of the price paid. A contract bought for five cents adds $1 to notional volume while requiring five cents from the buyer. Repeated trading and popular low-priced combos amplify the number further.
Prediction Frontier put traders’ actual payments at $56.1 billion for the $179.6 billion of Q3 contracts, while Kalshi’s $60.7 billion September notional figure represented about $14.7 billion in dollars paid. Macquarie estimated roughly $20.1 billion in taker-side prediction-market volume in September, including about $17 billion tied to sports. That is still not comparable to 15-state sportsbook handle.
The Evidence on Customer Cannibalization
A flat handle does not prove prediction markets have seized customers. It may indicate that football’s demand simply offsets other weakness. The AGA projected $29.5 billion in legal NFL wagering for the season, almost identical to the $29.4 billion estimate for last year, and September broadly matched that expectation.
The next data points will be Q3 earnings from DraftKings and Flutter, FanDuel’s parent. Both have largely kept sports event contracts to states where regulated sports betting is unavailable, so their results offer only a partial view. October’s sportsbook figures should sharpen the picture, but they will not prove whether customers are switching between the two markets.
Frequently Asked Questions
What sportsbook handle was reported for September?
Sportsbooks across 15 states recorded approximately $11.11 billion in handle, down about 0.3% year over year. Seven states grew, seven shrank, and Iowa was virtually unchanged as football returned.
How large was prediction-market activity in September?
Four major exchanges recorded $75.5 billion in notional trading, a record month. Kalshi led with $60.28 billion, up 56.4% from August. Across Q3, the four platforms generated $179.6 billion, up 84.3% from the prior quarter.
Why does prediction-market volume not equal sportsbook handle?
Prediction markets value every contract at its $1 settlement price, so a five-cent purchase adds $1 of notional volume. Dollars actually paid were about $14.7 billion against Kalshi’s $60.7 billion September notional total, and neither figure covers the same states.
Why did New York’s sportsbook revenue fall while handle rose?
New York generated $159.4 million in revenue in September, down 17.7%, with handle up 5%. Hold fell from 8.47% a year earlier to 6.64%. New Jersey followed a similar pattern, with revenue down 14.3% as volume dropped 3.6%.
Are prediction markets winning sportsbook customers?
The evidence is mixed. A flat 15-state handle alongside record prediction-market trading shows demand is not collapsing, but it does not show customer switching either. Q3 earnings from DraftKings and Flutter and October’s results should provide the next clues.
Will October change the picture for sportsbooks?
October should bring a second month of football-season results as prediction-market trading continued to expand. A higher handle would point to stronger demand. It would still not establish whether customers are moving between sportsbooks and prediction markets.
This article has been thoroughly researched and reviewed by the CasinoBait editorial team to ensure accuracy and relevance for Asian casino players.
