Prediction Markets: Real Financial Assets or just Sports Betting?
A new asset class or just gambling run amok
Are event contracts financial assets, or are they just sports bets with a different name?
Prediction market companies claim the former. Kalshi CEO Tarek Mansour has said “If we are gambling, then I think you’re basically calling the entire financial market gambling,” adding that his company’s event contracts have “economic utility behind the speculative activity, and that’s what makes it a financial instrument and not a gambling instrument.” His long-term vision, in his own words, is to “financialize everything and create a tradeable asset out of any difference in opinion.”
Sportsbook operators, state regulators, and certain members of Congress see it differently. Circa Sports CEO Derek Stevens has called prediction market sports contracts “illegal gambling” that circumvent rules and regulations applicable to licensed operators, adding that these platforms “figured out a way to steal all the data” while paying nothing in state or federal taxes, league fees, or problem gambling funding. Caesars CEO Thomas Reeg put it more bluntly: “To me, this is clearly gambling.” Rep. Dina Titus (D-NV), co-chair of the Congressional Gaming Caucus, has introduced legislation to ban prediction markets from offering sports contracts, arguing that these platforms “are evading these protections, leaving consumers, players, and the integrity of sports at risk” and that they “make no contribution whatsoever.” And Sen. Chris Murphy (D-CT), sponsor of the BETS OFF Act, has calls prediction markets “fundamentally corrupt” and “rife with insider trading,” warning that “when events that involve good and evil, life and death become just another financial product, morality no longer matters.”
Here I’ll walk through the case for both sides.
The Case for Financial Assets
Two primary applications of prediction markets that distinguish them from casual wagers are their use for hedging and monetizing information through a tradeable asset on an exchange.
Hedging. During the 2022 World Series, Jim McIngvale (AKA “Mattress Mack”) placed approximately $10 million worth of bets on the Houston Astros to win at 7.5-to-1. His bet was designed to hedge a $50 million exposure based on a promotion at his furniture store, Gallery Furniture, which promised to refund all purchases made between a certain period if the Astros were to win. The Astros won. McIngvale paid the refunds and collected $75 million from his bet. Both Gallery Furniture and its customers benefited from the promotion, which would have been imprudent — if not impossible — without the ability to hedge.
This is not a hypothetical example. It is a real hedging use case of a sports bet to run a creative marketing campaign, albeit through a traditional sportsbook — prediction markets were not widely available at that time.
Information monetization. When a bet is placed, the market price adjusts in the direction the bet was placed. Therefore, when someone with an edge places a bet, the price moves toward the more accurate side. Prediction markets, therefore, incentivize the sharing of valuable information through the opportunity to profit for being right. In theory, a well-functioning prediction market on an exchange produces a transparent, real-time probability estimate, backed by financial conviction, on any event it lists. The resulting price signal is useful to anyone who cares about the underlying event — businesses, policymakers, journalists, the public — which can and does have downstream economic and financial impact by users of the signal.
But Sportsbooks Can Do Some of This Too
Sportsbooks can also be used for hedging and monetizating information. Mattress Mack hedged through a sportsbook, not an exchange. And many professional betting syndicates do bet profitably through sportsbooks. Moreover, sportsbook odds do tend to track each other, which means they produce a useful informational signal about the events they cover.
The problem with sportsbooks is that they do not willingly serve this purpose. They are willing to work with Mattress Mack because of the PR sideshow he generates. It’s good publicity to take a large bet from a losing bettor. But they want nothing to do with winning bettors. They do not aim to produce useful price signals. In many ways, they go out of their way to avoid providing a useful signal — by only allowing losers to bet with them, they are limiting the quality of information that is reflected in their prices. The fact that sportsbook prices tend to be accurate is incidental to their primary purpose — which is to take money from losers.
Another difference between event contracts on prediction markets and wagers at a sportsbook is their tradability. On a prediction market, the event contract is a tradable asset. On a prediction market, a contract can be bought and sold as prices change. This ability to trade in and out of positions adds a meaningful skill element and real economic value to the proposition. It is, in a very real sense, the difference between a bet and a financial instrument. At a sportsbook, the only way to trade out of a bet is through “early cashout” — which is just another way for the book to extract more vig from its customers. The “asset” value argument for a sportsbook wager is, therefore, thin.
This argument in favor of prediction markets, however, is diminished by how fees are currently charged at most prediction markets. A big improvement that prediction markets could make would be to implement market-level fees, instead of trade-level fees. Currently, prediction markets impose fees on every trade. For simplicity assume a 1% fee on each trade. So if you buy $100 worth of Yes shares, it costs $1 in fees. Offsetting that position with $100 worth of No shares also costs $1 in fees. So trading out of a position requires a double payment of fees for a position that cannot win or lose. This is essentially the same as sportsbooks applying vig both on the initial wager and to the cashout value of a ticket.
The Banning Problem
A major line of demarcation between prediction markets and sportsbooks is how winners are treated. Whereas prediction markets are open to everyone, sportsbooks ban and severely limit participation from winners. (Worth asking: if lawmakers and regulators care about the people in their state, why are they actively trying to ban prediction markets while promoting sportsbooks in their state?)
Banning winners is the operating model of most sportsbooks. The vast majority of sportsbook operators employ what the American Bettors Voice calls a “ban or bankrupt” model: competent customers are banned from participating while losing customers are exploited and enticed to bet to the point of financial ruin. Despite claims by the National Problem Gambling Council and the American Gaming Association about the importance of consumer protection and player health, the “ban or bankrupt” approach is anything but healthy.
Transparency
The transparency of sportsbook pricing is also lacking. Sportsbook odds are set by a bookmaker based on proprietary algorithms and internal information. Because most sportsbooks ban competent bettors from their platforms, they are often late to receiving information. Instead, many operators apply a “copy trading” model by copying odds movements at other competing sportsbooks — and, ironically, using prediction markets as inputs to their pricing models. As a result, a small amount of money bet at the right time at an offshore book in Asia can trigger a wave of market movements worth tens of millions of dollars across the U.S. market.
None of this justifies the claim that sportsbooks are viable outlets for monetizing information. A sharp bettor who consistently moves a sportsbook line is not rewarded for the signal they provide. They are punished for it.
Exchanges, by design, do not have this problem. In a well operated exchange, the operator has no stake in the outcome of any given event. Its only incentive is to facilitate trading activity through open participation and substantial liquidity. There is no reason — and should be no legal basis — for an exchange to restrict trading from eligible participants who follow the rules.
The Liquidity Problem
With all that said, there are some weaknesses in the prediction markets’ claim as “financial assets”.
Prediction market volumes can be very thin. As I write this, the day before the NCAA Tournament Sweet 16, Polymarket shows very low volumes on the next round of games. The game between St. John’s and Duke has had only $17K of volume with small open interest. Most major sportsbooks have significantly more liquidity available on a game of this magnitude. For serious hedging or information monetization, this is insufficient.
The liquidity problem compounds by disincentivizing sharp information. Why would a sharp trader bother to move a market and provide a valuable public signal if the available liquidity means they can only deploy a relatively small amount of capital? Sharp money flows to where it can extract maximal value. Right now, for most sporting events, that is still the sportsbook — despite all the sportsbook’s deficiencies.
The liquidity problem compounds by disincentivizing sharp information. Why would a sharp trader bother to move a market and provide a valuable public signal if the available liquidity means they can only deploy a relatively small amount of capital? Sharp money flows to where it can extract maximal value. Right now, for most sporting events, that is still the sportsbook — despite all the sportsbook’s deficiencies.
Beyond Sports: The Flexibility Advantage and the “Gambling on Everything” Problem
But we’re just talking about sports here, which make up a large amount of prediction market volume. The more compelling financial and informational arguments for prediction markets are on everything else: politics, earnings, crypto, weather, world events.
Beyond sports, the value of prediction markets is that they can run a very large number of different markets on niche topics. Sportsbooks are limited in this capacity by regulatory constraints. Even without regulatory constraints, bookmakers would struggle to price and stay on top of the many different events that prediction markets cover, especially with low liquidity per market. It is simply not a viable product in a sportsbook format.
This flexibility is a real advantage for prediction markets. Markets on elections, geopolitical events, public health, and economic indicators can serve real hedging and informational purposes for people and institutions who have exposure to those outcomes. These applications have real value.
But this flexibility also provides critics a justifiable claim that prediction markets are just “gambling on everything.” And some of the markets on offer make that critique very easy to justify.
What are 5-minute crypto markets besides a creative way of generating a coin toss every 5 minutes? Bloomberg described them as “addictive,” and with as much as $60 million changing hands daily in these ultra-short-duration contracts, the comparison to a digital slot machine is not unfair. One Polymarket team member has teased 1-minute markets as a coming feature. The direction of travel here is not toward financial utility.
What perceivable value can be gleaned from the State of the Union “mention markets” — contracts on whether the President will say the word “discombobulator” during his address, with $62,000 in volume on that single proposition? Or from Rotten Tomatoes score markets, where Polymarket hosts 104 active markets letting users wager on critic scores for upcoming movies, many with volumes in the low thousands? Or — my personal favorite — a market on whether “Clavicular” will be the next Supreme Leader of Iran, which has somehow attracted more than $1 million in trading volume.
It seems like for every genuine use case in politics or geopolitical events — remember all the positive news after the 2024 election? — prediction markets shoot themselves in the foot with markets about Clavicular and the second coming of Christ. So while one can construct an argument for financial or informational value in some of these markets, the arguments are thin as of right now. Not only do these markets feel frivolous, they are also very low liquidity and not closely connected to events that institutions or people outside of market participants care about. The Clavicular market is an absurdity. The 5-minute crypto markets are structured gambling. The mention markets are party tricks. These are not the kinds of markets that support a serious case for prediction markets as financial infrastructure.
State Regulators: “Pseudo-Regulated” — Says the Industry That Lost the Public’s Trust
State regulators and sportsbook operators have taken to calling prediction markets “pseudo-regulated” or effectively unregulated. The American Gaming Association has gone further, framing event contracts as a threat to “the carefully crafted statutory and regulatory safeguards developed in each of our states.” Eleven states have issued cease and desist orders. Arizona has filed criminal charges against Kalshi. The AGA estimates states have lost over $600 million in tax revenue to prediction market platforms.
But state regulators and their licensed sportsbook operators cannot have it both ways.
The same operators who created the sports betting craze are now criticizing prediction markets for cashing in on it. These are the same operators whose sell a dream of winning to customers, only to ban and limit those who actually win. These are the same operators whose use of AI-driven targeting to promote impulsive and excessive customer behaviors has been widely documented.
State regulators and their licensed sportsbooks have betrayed the public trust, and are acting in conflict with the public interest. They have had ample opportunity to demonstrate their ability to serve the public, and they have failed decisively. The idea that these same actors are the proper guardians of consumer welfare is difficult to take seriously.
The irony runs deeper. Many of the same politicians sponsoring legislation to ban prediction market sports contracts have been enthusiastic supporters of state-licensed sports betting. The sports betting industry generates $53 billion annually in state and local tax revenue. That tax revenue comes overwhelmingly from the losses of ordinary customers — losses that the regulatory framework is designed to facilitate, not prevent. When the AGA says that “every dollar prediction markets siphon from the legal, regulated sports betting market drains money from a framework that supports local communities,” the quiet part is: those dollars come from people losing money at sportsbooks that are incentivized to keep them losing.
The Dream Is Alive
What is easy to understand, however, is that the states are threatened. The prediction markets are eating their lunch. Not because they are unregulated or pseudo-regulated — the CFTC is much better equipped to regulate than most states — but because the product is perceived to be better by many customers. In the eyes of many, prediction markets already offer a better product, and they have more potential to improve and innovate from this baseline.