Sportsbook handle stays flat as prediction markets set September records

In September, prediction markets posted their best trading month on record. Across fifteen major states, sportsbook handle was nearly unchanged from the same month a year earlier. Those two facts sit in the same sentence, and the conclusion most operators are drawing from them is wrong.

The consensus read is reassuring: traditional sportsbooks held their ground while a new category expanded around them, proof that the two products serve different needs and different customers. Flutter and Entain both saw their share prices rise in October as Q3 results came in, and the market has apparently decided that flat is fine.

Flat is not fine. Flat in a September with the NFL back, in a period when the broader wagering population was demonstrably active enough to move a hundred million dollars through prediction market contracts, is a number that deserves more scrutiny than it is getting.

Here is the mechanism the optimistic read misses. Prediction markets do not just expand the wagering universe — they absorb attention and capital from within it. A sports bettor who opens a Kalshi account to trade a Senate race does not stop being a sports bettor. But the next time he has $200 and a Sunday afternoon, he has more options than he did in September 2025. Sportsbooks have always competed for share of wallet. They are now competing for share of a wallet that has a new and legally distinct pocket in it.

The football calendar is the stress test that matters most. September is when handles should climb, when re-engagement after the summer is structurally guaranteed, when even a mediocre marketing quarter produces growth. The industry did not grow. The industry that did not grow is the one whose product has no federal preemption argument, no institutional sports league partnership framework, and no line of Supreme Court cases developing in its favor.

I have watched an incumbent product category treat a flat period as stability before. The error is always the same: mistaking the absence of decline for the presence of health. The question is not whether sportsbooks lost customers in September. The question is whether the growth that should have arrived did not, and where it went instead.

The SBC data on operator share prices is real, and the Q3 relief is real. But a rising share price in October reflects Q3 results, not September handle. Investors rewarded what already happened. The handle number is about what is happening now, and it is telling a different story than the equity market has priced.