Prediction-Market Volume Fell for the First Time in a Year. Everyone Arguing About It Should Read the Split.

Editor's note: TrueEdge builds odds tools and earns affiliate commissions from licensed sportsbooks. This piece argues a position on prediction-market regulation, and our commercial interest runs toward the sportsbooks. Weigh it accordingly.

Combined trading volume at Kalshi and Polymarket was $45.33 billion in August, down 14.5% from July. The Block reported it on September 2 off its own data dashboard. It is the first month-over-month decline in more than a year, and it is the first number in this entire fight that neither side gets to spin without breaking something they said last week.

The interesting part is not the total. It is how the drop divides.

Kalshi did $37.17 billion, down 7.3% from $40.1 billion in July. Polymarket, including its US app, did $8.16 billion, down 36.7% from $12.89 billion. One of these venues gave back a rounding error. The other gave back more than a third of its business in thirty-one days.

Two companies, one category, different products

The reporting treats Kalshi and Polymarket as a single sector because they are regulated the same way and sued by the same attorneys general. On the demand side they are not remotely the same company.

Kalshi's volume is overwhelmingly sports, and sports have a calendar. Baseball plays every day in August. Football's futures markets stay live all summer. The floor never really drops out, and a 7.3% dip in the deadest month of the American sports year is what a business with recurring demand looks like when the schedule thins.

Polymarket's volume needs an event. The men's World Cup ran in North America from June 11 to July 19, and it inflated July at a venue whose international user base cares about soccer far more than the American exchange's does. When the tournament ended, a third of the volume left with it. That is not a company losing share to a competitor. That is a company whose product is speculation on things that are happening, in a month when comparatively little was happening.

Hold both numbers next to May's combined $25.66 billion and the shape becomes clear: the category is still enormously larger than it was in the spring, and the August figure is a fall from a World Cup peak, not a fall through a floor.

Why this is a bad month for the trade group's argument

The American Gaming Association projected on September 4 that $29.5 billion will be wagered on this NFL season with legal sportsbooks, against $29.4 billion last year — flat, for the first time since 2018. Bill Miller's explanation, in the second paragraph of the AGA's own release: "Since the widespread launch of backdoor sports betting on so-called 'prediction markets,' the growth of legal handle has stalled." We took that release apart yesterday, mostly on the grounds that contract volume and handle are not the same unit and cannot be stacked in one chart.

August gives the argument a second problem, and it is a cleaner one. If exchange growth were mechanically draining sportsbook handle, exchange volume would be a fairly smooth line, because the thing it is allegedly stealing — habitual sports bettors — is a fairly smooth population. Instead the line moves 14.5% in a month on the presence or absence of a soccer tournament. Volume that swings that hard on event supply is not a stable substitute for a sportsbook account. It is closer to a venue people arrive at when there is something specific to trade.

You can hold that view and still think the substitution is real at the margin. What you cannot do is point at $36.8 billion of projected NFL contract volume, call it evidence of a structural transfer, and then decline to explain why the same series fell off a cliff the moment the World Cup ended.

And a bad month for the exchanges' argument

The other story is the one Kalshi and Polymarket have been telling investors: that prediction markets are a new asset class with a permanent, compounding demand curve, and the only thing standing between them and the whole sports-wagering market is a patchwork of state gambling statutes.

A single month does not falsify that. But this is the first month anyone got to check, and the check produced a decline in the middle of the most aggressive legal offensive the states have run yet. Washington's attorney general, Nick Brown, announced a final King County Superior Court order on August 13 requiring Kalshi to stop offering wagers on sports, elections, politics, entertainment, culture, tech, science and "mentions" in the state, and to stand up an IP-and-residency geofence by August 19 and a multi-source geofencing solution by September 2. On August 28, a Ninth Circuit panel ruled for Nevada in No. 25-7516, holding that Kalshi had not shown the Commodity Exchange Act likely preempts state gaming law as applied to its sports contracts. Connecticut's attorney general, William Tong, sued on August 26, arguing that sports event contracts "are no different than sports betting and are not magically shielded by federal law from Connecticut's commonsense consumer protection laws."

Every geofence removes users. Nobody has published state-level volume, so we do not know how much of August's 7.3% at Kalshi is Washington and Nevada residents being switched off versus August simply being August. That is a genuine gap, and it is the single most useful number nobody is disclosing. An exchange that wanted to win this argument in public would break out volume by state and show that enforcement is costing it nothing. Neither has.

The strongest version of the other side

Here is the case that August means nothing, stated as well as we can state it.

August is the trough of the American sports calendar. There is no NFL, no college football until the last weekend, no NBA, no NHL. The relevant comparison for a growth business is year over year, not month over month, and on that basis prediction-market sports volume in August was several times what it was a year earlier. September will have the NFL, college football's full slate, an MLB pennant race and the US Open, and the combined figure will almost certainly set a record. Reading a July-to-August decline as a trend is the kind of thing people do when they want a decline.

The best single piece of evidence for that view sits inside the same month. Kalshi cleared $10 billion in a single week for the first time in the week ending August 30 — the week football came back. A business posting a record week in the last week of a declining month is not a business in retreat; it is a business whose customers went away for a tournament-shaped reason and came back when the sport they actually care about started.

That is right, and it is the reason we are not calling this a top. It is also why the AGA should not have leaned on a monthly-volume comparison in the first place: the metric is too seasonal to carry the weight either side keeps putting on it.

What would change our mind

Two checkable things. If September's combined volume does not clear $45.33 billion by a wide margin — with the NFL, college football and the US Open all live, and the World Cup comparison gone — then the growth story is in real trouble and we will say so. And if either exchange publishes volume by state showing that Washington's September 2 geofencing deadline and Nevada's post-Ninth-Circuit posture cost it materially, the enforcement campaign becomes the explanation for August and the seasonality argument shrinks.

Until then the honest reading is narrow and unsatisfying to both lobbies: prediction-market volume is large, still growing year over year, extremely sensitive to what is on the calendar, and concentrated in a product — Kalshi's sports book — that behaves much more like a sportsbook than like an exchange. The AGA is describing that resemblance accurately and using the wrong number to prove it. The exchanges are describing a durable new market and just posted a month that looks like an events business.

The genuinely unresolved question underneath all of it is still the one from Fortune's August 31 report: 18-to-21-year-olds traded an estimated $5.4 billion on Kalshi this year, $3.9 billion of it on sports, in an age band that cannot open a sportsbook account in most legal states. Kalshi told CNN that cohort is 3.14% of its volume. Both things can be true. Only one of them is a policy problem, and it does not get smaller when the monthly chart ticks down.

Watch