Prediction Markets Need Clearer Definition Before States Tax Them

States seeking to tax prediction markets are attempting to regulate an activity they haven’t yet clearly defined, which risks taxing the wrong base entirely. Until that question is resolved, state taxation efforts should be restrained to avoid building policy on a categorial error.

Prediction markets—online platforms that let people bet on event outcomes are a booming business. Monthly trading volume rose from roughly $1 billion to more than $20 billion in just a few years. This surge has garnered the attention of state and federal lawmakers, eager to regulate and tax the activity.

But what exactly are prediction markets? The answer will determine not just how they should be taxed, but who should do the taxing.

If they are gambling platforms, existing state-level sports betting frameworks offer a template. If they are financial instruments more akin to derivatives, they fall more naturally within the Commodity Futures Trading Commission’s domain. If they are neither, policymakers must ask whether it’s more confusing than clarifying to link them to something that already exists.

The state policymaker position is defensible. Because prediction markets allow users to “bet” on the outcome of real-world events—economic indicators, sports, even elections—it’s not surprising that many states view them as a natural extension of legalized sports wagering.

The CFTC, however, is asserting jurisdiction on the grounds that these platforms operate much more like derivative exchanges. These are marketplaces for trading futures and options—contracts that derive their value from other underlying assets such as stocks or currencies.

Both derivatives and prediction markets draw value from an outcome—the former when the underlying asset goes up, the latter when an event happens.

One factor against treating prediction markets as sportsbook analogs is that platforms such as Kalshi don’t function as “the house” because they’re not taking the other side of users’ bets. Their business model doesn’t require setting odds, so they don’t absorb risk in the way a traditional sportsbook does. Instead, they operate more like a go-between or broker, matching buyers and sellers who take opposing positions of the same event. Derivatives exchanges work on a similar financial model.

If one user believes inflation will exceed expectations, or a major tech CEO will wear a hoodie while testifying before Congress, the platform simply facilitates a trade between those users and others who take the opposite side. The platform’s revenue comes from transaction fees levied on the exchange made when inflation decreases or the chief executive testifies in a velvet track suit—not from the outcome itself. The platform takes no position on either wager and doesn’t make money based on its ability to win bets against users.

Even if states manage to impose prediction market tax systems, those systems may rest on uncertain legal ground and provide just a short-term cash grab (perhaps with later required refunds).

Photographer: Graeme Sloan/Bloomberg via Getty Images

If states approach prediction markets by comparing them to existing sports betting frameworks, they would likely tax operators based on gross gaming revenue—the amount retained by the house after paying out winning bets.

In such a model, sportsbooks assume the underlying risk, as their revenue is tied to the outcome of their own wagers. A tax on gaming revenue assumes that sportsbooks’ revenue depends on society losing their bets. But that logic breaks down when applied to prediction markets, which only earn fees on transactions. There’s no meaningful gaming revenue in any traditional sense.

Thus, trading volume in this regard is misleading. The headline figures cited often represent the total value of contracts traded by users, rather than the revenue earned by platforms. Conflating the two risks inflating the perceived tax base and making the opportunity seem larger for states than it really is.

The CFTC meanwhile has asserted that prediction markets fall within its exclusive federal jurisdiction and has begun suing over state efforts to regulate the space. As challenges move through the federal district and appellate courts, the potential for conflicting rulings increases. At that point, resolution may shift from regulators to judges—and possibly to the US Supreme Court.

Even if states manage to impose prediction market tax systems, those systems may rest on uncertain legal ground and provide just a short-term cash grab (perhaps with later required refunds).

Regulating and taxing prediction markets faster than they are understood would present a structural problem. When the underlying activity resists easy classification under existing categories, forcing it into state-level frameworks risks misidentifying the tax base in a way that can’t withstand legal scrutiny.

There are also compelling reasons to think prediction markets are ill-suited for fragmented state-by-state regulation. These are inherently borderless digital platforms, and if the CFTC is correct, they have more in common with massive financial exchanges than with local gambling operations. That combination points to a potential need for regulatory and tax coherence at the federal level.

A federal regulatory system with a tax keyed to platform fees would be a more defensible and durable choice. But that should come with an explanation of why prediction markets shouldn’t be left to existing income tax policies and how these platforms are distinct enough from ordinary brokerages to warrant bespoke treatment. Once you move outside of the gambling analogy, that burden is no longer carried by the traditional vice tax frame.

None of this forecloses a role for states. Traditional police power areas of enforcement such as consumer protection, fraud protection, and regulations around certain specific areas of event contracts may remain squarely the domain of the states. But the core questions of what prediction markets are, how they are generating income, and what a proper tax base would be are national in scope and may need to be treated as such.

Andrew Leahey is an assistant professor of law at Drexel Kline School of Law, where he teaches classes on tax, technology, and regulation. Follow him on Mastodon at @andrew@esq.social.

Read More Technically Speaking