Have a little faith in markets

Introduction

Prediction markets have exploded in popularity over the past 12 months. The function of these markets is relatively simple: users are able to purchase binary contracts which will resolve at some future time to pay either a fixed amount of money or nothing. Generally, these contracts are denominated in either one dollar or one-hundred-dollar increments. Mechanically, the function of a market is to establish a “market price” between the limits of zero and the contract price. Academics theorize that these contracts’ market price represents a prediction as to the likelihood of an event occurring through price discovery.

As this segment has increased in popularity, two flavors have emerged. The first, which most notably includes Polymarket, operates on cryptocurrency rails without regulatory licensure, and generally prohibits individuals in the United States from participating through a variety of strategies. The second, spearheaded by KalshiEX LLC (Kalshi) and North American Derivatives Exchange, Inc. (Crypto.com), offers these contracts as derivatives through Commodities Futures Trading Commission (CFTC)-registered designated contract markets (DCMs).

It is this second group that has generated significant legal intrigue over the last year, and which we will discuss here. While event contracts are not new — Kalshi has offered essentially the same core product since 2020 — a series of legal events have conspired to broaden their scope dramatically.

Registered derivatives are regulated according to the Commodities Exchange Act, and work basically like this. The DCM is permitted to self-certify a market, and after doing so, may list the market publicly the next business day. Since Dodd-Frank in 2010, the CFTC may then review the market according to Section 5c(c)(5)(C) of the CEA and CFTC regulation 17 CFR § 40.11.

The language of Section 5c(c)(5)(C) of the CEA says that the CFTC can then prohibit a contract from trading if it determines that the contract is “contrary to the public interest” because it involves:

1. activity that is unlawful under any Federal or State law;
2. terrorism;
3. assassination;
4. war;
5. gaming; or
6. other similar activity determined by the Commission, by rule or regulation, to be contrary to the public interest.

For a long time, the CFTC interpreted this definition in a way that gave it plenary authority to ban more or less any market that it wanted. In general, this meant that it refused to permit markets relating to elections, sports, and award shows – so, the juicy stuff – from trading.

That’s not what the law says though, and after the Chevron doctrine was overturned in Loper Bright Enterprises v. Raimondo, 603 U.S. 369 (2024), federal agencies have generally been required to follow the law. So when, in late 2023, the CFTC moved to prevent Kalshi from listing contracts related to elections, Kalshi sued. And in August 2024, in KalshiEX LLC v. Commodity Futures Trading Commission, No. 1:23-cv-03257 (JMC), 2024 WL 4164694 (D.D.C. Sept. 12, 2024), Kalshi won.

The proximate effect of this ruling was ostensibly narrow. Kalshi, and other DCMs, were allowed to offer election contracts in the lead up to the 2024 election. This lucrative market promoted their growth, but they were dwarfed in scale by their off-shore competitor, Polymarket, who ultimately gained the greatest attention share that season.

After Donald Trump was elected, as the implications of the Kalshi ruling began to sink in, practitioners began to think about the margins of the new regime. Crypto.com was the first mover, self-certifying contracts based on the outcomes of sporting events in December 2024, and Kalshi soon followed. While the lame-duck Biden CFTC did move to review the Crypto.com contracts in January, they did not issue a final determination prohibiting the contracts prior to Mr. Trump’s inauguration, and the incoming CFTC has appeared more receptive to the segment.

Indeed, in May the CFTC dropped its appeal of the Kalshi case, in effect accepting the limits to its authority to prohibit election contracts. This came a midst speculation that incoming Commission Chair Brian Quintenz will view DCM prediction markets favorably, given arguments he made as far back as 2021 in favor of sports-based prediction markets. Things are looking good for prediction markets on the regulatory front.

As sports markets have grown in popularity, states have taken note. Post-PAPSA state-licensed sports betting regimes are major money-makers for states that permit them, and federally licensed DCMs sit outside of this framework. Naturally, traditional sports-betting friendly states like Nevada and New Jersey see this development as a threat to their fiefdoms. In the Spring of 2025, this dynamic led to a flurry of state cease-and-desist letters targeting the new markets.

At the time of writing, this litigation is ongoing. However, Kalshi has made significant strides and won preliminary injunctions against Nevada and New Jersey. See e.g. KalshiEX LLC v. Flaherty, No. 1:25-cv-02152, slip op. at 16 (D.N.J. Apr. 28, 2025). Section 2(a)(1)(A) of the CEA provides that “[t]he Commission shall have exclusive jurisdiction [over covered derivatives]” and that is about the strongest pre-emption case you are likely to see litigated in your career. There are caveats and counterarguments, but brass tacks, if you were betting on the outcome, you would put your money on Kalshi and Crypto.com winning these cases.

That doesn’t mean the field is settled, however. In the coming years there will be numerous avenues for policy change, as the CFTC likely retains rulemaking authority to prohibit contracts. As event contracts become an increasingly attractive substitute for state-licensed sports betting, it is also plausible that the CFTC could attempt to take action against gambling companies like DraftKings and FanDuel for offering unregistered derivatives.

While the time for legal arguments will no-doubt come again, for now, as a few rule makers in Washington hold unusual sway over the future of an industry, normative arguments come to the fore. What should the future of DCM-based prediction markets be in the United States? 

What is gambling?

The most common aspersion cast against prediction markets is that they are simply gambling by another name. This forced Kalshi CEO Tarek Mansour to take the other side, arguing in an interview with Axios that prediction markets cannot be gambling, because of everything else that would encompass.

“I just don’t really know what this has to do with gambling,” he said at an office in Washington. “If we are gambling, then I think you’re basically calling the entire financial market gambling.”

But Mr. Mansour is being obtuse. He, you, and I all know what gambling is and know that when a user purchases an event contract on Kalshi, that is exactly what they are doing.

Industry commentator Dustin Gouker argued as much in a piece earlier this year.

This is all a bit of a straw man. I think we all understand that every decision we make in life comes with some amount of risk, either large or minuscule. That does not inherently make everything we do “gambling.” Gambling generally implies betting on something that is out of our control. Most of the decisions we make that have risk, we also have some measure of control over them.

As an attorney representing prediction markets, I have been asked this question for years and now it’s time to put this conversation to rest. Yes, all of this is gambling, but that definition misses the point.

Since the demise of PAPSA in Murphy v. National Collegiate Athletic Association, 584 U.S. 453 (2018), the definition of gambling has been left to the states, and generally, it is quite similar. New Jersey, for example, defines it thus:

Staking or risking something of value upon the outcome of a contest of chance or a future contingent event not under the actor’s control or influence, upon an agreement or understanding that he will receive something of value in the event of a certain outcome.

You’ll notice first that this definition is quite broad, and second that it unarguably covers every single event contract, sports or no. Indeed, it covers much more than just event contracts. Virtually any derivative or insurance contracts are equally “gambling” by this definition.

This is not a flaw of the definition, but it is not legally meaningful because we recognize that these other types of contracts – grain futures, insurance, and, yes, event contracts – are socially valuable and so they are regulated under different regimes.

Et tu, Brute?

Indeed, the same logic that suggests that prediction markets are “gambling” applies in the other direction as well. Why aren’t state gambling contracts just unregistered, federally regulated derivatives? Consider the CFTC’s definition of “binary option”

A binary option is a type of options contract in which the payout will depend entirely on the outcome of a yes/no proposition… When the binary option expires, the option holder will receive either a pre-determined amount of cash or nothing at all. Just as clearly as state-law “gambling” would apply to Kalshi’s event contracts, so too could the Commodity Exchange Act (CEA) apply to DraftKings or FanDuel.

The United States has the capacity to regulate this type of contract on a state or a federal level, and by successfully registering as a DCM, Kalshi and Crypto.com find themselves within the federal regime. Because of the language of Sec. 2(a)(1)(A) and the supremacy clause of the United States Constitution, they are outside of the state regimes.

The qualitative distinction between prediction markets and state-regulated betting, then, is not that one or the other is gambling. They both are, by the state definition at least. To draw conclusions about prediction markets, policy-makers and stakeholders should think at one less level of abstraction.

A normative framework for evaluating the role of prediction markets in American society should consider from first principles whether the new prediction markets modality or the old state-regulated gambling model are better for the United States. A simple and principled way to evaluate this choice is to determine which option maximizes benefits while minimizing harm.

The whole naked truth

The benefits of prediction markets are novel but well documented. In the first instance, they create signal. Markets are incredibly efficient tools for price discovery: by incentivizing participants to aggregate their knowledge, a price represents the synthesis of all of that knowledge. This, Justin Wolfers and Eric Zitzewitz remarked in their 2004 paper “Prediction Markets”, might make them a powerful tool for predicting events. This suggestion was borne out in 2024, when Polymarket correctly predicted Donald Trump’s election in the face of traditional polling aggregates showing a coin flip.

Famously, during this period a French trader named Théo conducted his own polling before betting for Trump. This is the paradigmatic case of prediction markets developing new information. The market created an incentive, and traders responded by learning true things about the world. This information was then passed on through the market price. This process is fundamentally generative.

Beyond just price discovery, predictions markets are natural hedging vehicles. Have a business with lots of exposure to the price of chicken? Well, you can hire Ray Dalio to build you synthetic corn-soy futures, as McDonalds did in the 1980s, but it is probably easier to purchase a large contract betting that the price of chicken will go up. These types of contracts are the whole basis of the insurance industry, and as long as concave utility functions exist, they will continue to add massive value to society. Prediction markets will make them more available and less expensive where they thrive.

To be fair, these arguments mostly concern commercial or election markets, which are not the principal locus of dispute today. At the limit, prediction markets concerning relatively trivial events probably are far less valuable along these vectors. This should not be overstated — the Super Bowl and the Academy Awards do have meaningful economic consequences that certain parties may wish to hedge — but it is safe in this domain to flatten the calculus and narrow the benefit we consider to the enjoyment that users derive from using the product.

Here’s the truth. People have a fundamental desire to wager on the outcome of events. This may be hubris, entertainment, or opportunism for the intelligent or knowledgeable. Whatever the reason, these markets are inherent to human society and will arise independently anywhere that currency and contingent events meet. It is an entertainment product. This is the benefit of many markets, and if prediction markets don’t do it better than state-licensed betting, they do it at least as well.

On the benefit side of the ledger there is no reason to prefer state gambling over prediction markets. In many cases, quite the opposite.

One bite and I was gone with the wind

But what of the harm? This is a second hard truth: these markets lead many to ruin. There is a squishy dark spot deep in the recesses of the human mind that triggers reinforcing dopamine in response to uncertainty. That is science.14 You might have experienced it if you’ve ever felt yourself glued to your phone waiting to hear from a school, job offer, or lover — or played golf. Some people exposed to uncertainty react compulsively and destroy themselves.

Even more broadly, there is a gradient of sophistication and knowledge in human populations, and, where we are allowed to gather and wager against each other freely, the strong end up extracting money from the weak. This isn’t fair and it isn’t good, but it is the fundamental dynamic of sports betting, day trading, and it will be of prediction markets too.

These qualities, combined with a tendency to attract cheats and gambling’s natural association with dissipation and waste, have historically made the practice unseemly, so interlopers have often tried to marginalize and ban it. But this doesn’t work. Markets are not constructed, they are inherent. Bans on betting only create unregulated black and gray markets.

This is probably part of the reason that gambling laws have liberalized in the United States in recent years — people recognize that prohibition can cause more harm than good. After all, the problem gambler may lose his house in either a regulated or unregulated betting market. But at least in the regulated market he will keep his teeth.

A clear-eyed perspective takes both the downsides and benefits of these markets together. If this is “gambling,” then that is no invective.

Weighing these benefits, we can construct a rough harm reduction hierarchy. Prohibition is the worst option, maximizing harm. But state regulated markets likely still cause more harm than do federally regulated prediction markets. That is because of incentives.

State regimes have mostly flowed from pre-existing casino regulation, typically favoring the house. That is why sports books are permitted to limit bet sizes of any “sharp” bettor while giving their losers exorbitant perks to keep betting.15 This practice maximizes the damaging aspects of these markets, while simultaneously removing the most sophisticated participants, thereby kneecapping price discovery. Extractive behavior flows naturally from sports books’ posture as their customers’ counterparty. They can only make money as long as they are taking it from users.

Prediction markets do not have this problem. They are simple intermediaries where bids and asks can meet, and from that process, a signal may emerge. This is the point Mr. Mansour tried to make in his Axios interview, but if you can’t first admit that these products are close substitutes, then the principled arguments melt into noise. Prediction markets are like gambling, just better. It is simply an artifact of history that state-regulated sports books came first, so have an incumbency advantage.

Conclusion

The CFTC has expressed reservation in the past over the idea of directly regulating new markets. Its capacity will have to increase to do so effectively. This is reasonable, but in accepting the status quo and abdicating the regulation of risk assets to the states, it would be materially harming American consumers.

NCAA v. Murphy made progress, but we should not stop there. Prediction markets’ ascendance is an inflection point in the regulation of risk assets. The CFTC will have an opportunity to choose, or at least influence, whether popular betting markets remain regulated by the states or come under the federal umbrella. In making that choice, the natural question is which structure will maximize the benefits, hedging and price discovery, while minimizing the harms, predation and compulsive behavior.

This doesn’t necessarily mean that the CFTC should move to  ban state-law gambling. While such a move might be legally plausible, it would provoke a significant legal battle and have some legitimate Major Question Doctrine concerns.

But it is time to stop pearl-clutching when it comes to prediction markets. This new modality demonstrably produces high quality information, something state-law gambling has never claimed to do.

At the same time, it short circuits a perverse incentive structure that suffuses traditional book-making. People respond to incentives, and this relative platform neutrality should prove a boon to consumer welfare.

There will still be harm. There is no doubt about it. But the pertinent question is not whether that harm is present, but whether it is more or less than the alternatives — and it is difficult to see how it could be more than the present regime.

During the Obama years, Cass Sunstein was the Administration of the Office of Information and Regulatory Affairs (OIRA). As part of OIRA’s efforts, Sunstein implemented cost benefit analysis for all federal policy, dictating that agency action may only proceed where “the chosen approach maximizes net benefits.”16 Years have passed, administrations come and gone, but the analytic frame still pertains.

The age of state-licensed betting is over. The time of the prediction market has come.