The prediction market bubble is already here

Prediction markets are everywhere right now. The Golden Globes! The Oscars! The airwaves! For the Super Bowl, people can use prediction markets not only to wager on the outcome of the big game but also on which commercials will run and what song Bad Bunny will play first. Sportsbooks, media companies, and even traditional investment entities seem to want to get in on prediction market action. The enthusiasm makes sense given how industry boosters talk about these markets. To them, betting on anything and everything is the next frontier of finance.

But the rhetoric and the reality still don't line up. There's a lot of chatter about big institutions using prediction markets as hedges for everything from inflation to elections to the weather, but the amount of money flowing through platforms such as Kalshi and Polymarket is tiny compared to the stock market or even crypto. While that makes sense given their relatively new entry into the US, even in countries where betting exchanges have existed for years, they're largely a niche product. Thus far, the story of prediction markets in America has mainly been about betting on sports, particularly in states where sports gambling is illegal. In California and Texas, you can't legally bet on an NFL or NBA game on DraftKings or FanDuel — but thanks to a contested regulatory quirk, you can on Kalshi, for now.

The result: a fragile "boom." Prediction markets are having a moment, not because they've solved forecasting or finance, but thanks to a legal workaround and a deluge of marketing money. That combination has fueled rapid growth and breathless valuations, but it's built on a narrow, unstable foundation that could crumble once regulations catch up or if bettors decide the value proposition is less compelling than advertised.

"It's one of those things that could go either way," says Steve Ruddock, a gambling-industry analyst and consultant and the author of Straight to the Point, a newsletter about gambling.

Maybe it's OK if this doesn't work out: It's fair to ask whether we really want to live in a world where everything is a trade.

The regulatory landscape around sports gambling and prediction markets is messy. The Supreme Court struck down a law prohibiting sports betting in the US in 2018, opening the way for a state-by-state patchwork in which over three dozen states have legalized it. Prediction markets have found a way to circumvent this setup: they say they ought to be federally regulated by the Commodity Futures Trading Commission because their markets are futures contracts tied to event outcomes. Since this is "trading" and not "gambling" under the prediction markets' argument, these contracts should therefore be allowed everywhere. The prediction market-friendly Trump administration agrees.

Legal battles over this are shaping up across the country, as state-level regulators take a dimmer view of this activity than the feds, and the issue could eventually land at the Supreme Court. In the meantime, however, if you're in a state where sports betting is illegal, you can place a Super Bowl wager on a prediction market, and a lot of people are. Sports account for an estimated 85% of volumes on Kalshi, according to a 2025 report from Keyrock and Dune Analytics, companies in the digital asset space. On Polymarket, which was banned in the US in 2022 and is coming back online in the country, sports account for 39% of volumes, though Filippo Armani, data research lead at Dune Analytics and one of the authors of the report, tells me once Polymarket is back in the US, he expects increased volume on sports.

"Most of this is just a way to kind of skirt the state sports betting rules," says Chad Beynon, an equity analyst at Macquarie who covers gaming. "If this was a federal sports betting country, I don't think prediction markets would have popped up at all."

After some initial resistance, the major sportsbooks have dipped their toes into prediction markets — business-wise, it's a safety valve in case these really do take off. And if the regulatory landscape changes, they can just go back to business as usual. "In markets where sports betting is legal, we're not seeing an impact from prediction markets, mainly because the product isn't as good," Beynon says.

In the UK, the betting exchange Betfair has been around for 25 years. (A betting exchange is basically British for prediction market.) The company, which is owned by Flutter, FanDuel's parent, accounts for just 5% of the country's online sports betting market.

Sebastien Sabic, the cofounder of Moon Intelligence, a London-based data consultancy that's moving into market-making for sports betting, hopes Kalshi and others will grow beyond what he views as Betfair's "capped' position in the UK, but he also realizes they're not there yet. "In terms of where they are now, especially where they were a few months ago, they got really high valuations when their volume was quite low, and that product was quite poor in some sense," he says. "So they have done really, really well in terms of marketing themselves as the new cool thing."

The broader vision behind prediction markets stretches far beyond sports. They allow people to bet on election outcomes, awards shows, what certain public figures will say during media appearances, and, in a recent jarring example, whether the US would invade Venezuela. Prediction market proponents say they're a useful tool for forecasting outcomes, can beat polls in terms of accuracy, and have the potential for significant real-world value. Ahead of the 2024 presidential election, prediction markets had Donald Trump winning handily, while polls had the race close to tied. The upstarts also aspire to attract institutional money, where big financial players put big money into prediction markets to make forecasts and hedge risks. If they're worried that the value of their real estate loans in a certain city may go down, for instance, they could bet that prices in that market will decline, offsetting some of the potential losses. Goldman Sachs CEO David Solomon recently said they're looking into how they might get involved, though he cautioned, "the pace of change might not be as quick and as immediate as some of the pundits are talking about."

Devin Ryan, head of financial technology research at Citizens Bank, believes there's a meaningful long-term economic opportunity for prediction markets, assuming they can successfully move beyond sports. But it's not something that can evolve overnight.

"You need, I think, a couple of things," he says. One is market integrity rules to root out undesirable insider trading, such as solid know-your-customer guardrails and consumer protections. Until then, the regulatory uncertainty will make decision-makers wary. Second, there needs to be more trading on the non-sports markets. At the moment, there isn't enough volume for the big guys to get involved. Because many individual prediction markets still have relatively shallow order books, a $5 million wager on a specific outcome could swing probabilities significantly, and it might not match with existing orders on the other side. On Kalshi, for example, the market predicting what January's CPI inflation reading will be is under $1 million, and for core inflation, it's under $30,000. That's not enough for an institution to get in on.

"Do we have a clear view on the regulatory aspects of transacting in it? And do we trust the current regime in terms of the market makers and the legality of the outcomes? If we're putting capital at risk, are we going to get paid out?" Ryan says.

These two issues work in tandem. For every "winner" on a prediction market, there's a "loser" — users don't bet against the house, as they do with sports betting or casinos, they bet against each other. A market isn't very valuable if it's a tiny group of randos guessing what the temperature in New York City will be based on their iPhone apps. What you need first is a lot of randos betting on the weather, so that the payouts are substantial enough to make the bets worthwhile. Then you need at least some people who really understand the weather to get involved, which they'll be incentivized to do because of the randos. Some level of insider trading on prediction markets is, in a purist sense, a good thing — if you prioritize accuracy, you want the people closest to the truth playing along. There is a balance, of course. If you calibrate the right mix of "smart" money and "dumb" money so that the market is both liquid and relatively well-informed, then you can get the "big" money from institutions. Maintaining this balance, however, is easier said than done.

"Those other people are going to, on average, make losses if they know less about the subject matter than the experts. So you need them to be willing to trade despite that," Zitzewitz says. "You need them to be overconfident about how much they know, or you need them to be participating for some other reason."

Over time, people may get tired of losing. Once the second-best weather nerd figures out they're second-best, they'll decide they're out.

Ruddock, the gambling industry analyst, explains that peer-to-peer offerings in gambling tend to have a lot of churn. Love them or hate them, sports betting apps and casinos can make shifts to keep people engaged — they can change odds or offer boosts and freebies.

"When you are just facilitating the bets, the people on the other side of those bets are trying to make as much money as possible," Ruddock says. "They don't care about your ecosystem as a whole."

Other types of traditional gambling not only adjust the odds to keep you involved, they also have plenty of other ways to make sure you walk away willing to come back. The casino gets you to lose more slowly at the blackjack table, and once you do, you've had a couple of drinks for a couple of hours and a generally good time.

The outcome of a sporting event is fun to bet on when the payoff is quite immediate, and having some money at stake adds some oomph to an already interesting event. Wagering on when Costco will raise the price of its hot dog combo is just not that thrilling. Plus, if you've locked money up in that bet, those are funds you're not using somewhere else, whether it be buying said hot dogs or investing in the stock market. Some prediction markets pay users interest to try to get them to stick around.

"For them to reach their full potential, you need participants from everywhere," says Amir Hajian, a digital assets researcher at Keyrock.

When reached for comment via email, a Kalshi spokesperson pointed to a a link outlining how the company protects traders and said the company stands by its data on volumes, but declined to comment further. Polymarket did not respond to a request for comment.

The purpose of prediction markets is to find some sense of certainty in an uncertain world, so it's ironic there's so much uncertainty around the future of these markets themselves.

The legal and regulatory landscape around them is shaky, and it's likely to be months or even years before that's settled. Courts could decide they are actually state-regulated, or they could be forced to comply with regulations in a way that's costlier than the current setup. The economics surrounding these products are tough, and many prediction markets don't turn a profit. Some analysts even question whether these platforms are moving as much money as they claim. David Katz, an equity analyst at Bernstein, estimates that volumes on prediction markets are significantly lower than reported because of how these companies calculate the money flowing through them.

Maybe this all does work out for prediction markets — the institutions get in, markets get to keep their desired regulatory blessing, the dumb money draws in the smart money and doesn't figure out it's dumb, and they become mass-market products. That doesn't answer the second big question: Is all this good? Gambling has taken off across America in recent years, thanks to sports betting and the casino-like turn of investing. Does it also need to spread to what Jimmy Kimmel says on TV tonight, and whether anyone will go to jail over the Epstein files?

Timothy Fong, a clinical professor of psychiatry at UCLA and the codirector of its gambling-studies program, tells me he hasn't yet had anyone come in telling him they have developed a problem with a specific prediction markets app, but there's "tremendous growing concern" in the gambling studies field around them. Clarity is lacking around user protections, for example, and he takes issue with prediction market ads he's noticed that call it a side hustle, which may obfuscate how risky it is.

"Do we want to live in a world where you can make money off current events?" Fong says. "Is that what we want? A society based on money contracts to determine whether things happen or not? It's very bizarre to me, it feels dystopian."

Prediction markets may be the story in some corners of finance and media, but it's not clear whether they're the story on the ground. Perhaps they'll ultimately have mass-market value. Perhaps they'll always just have niche appeal. For now, the two questions looming over them are how legal they really are — and how much interest there is outside sports.

Emily Stewart is a senior correspondent at Business Insider, writing about business and the economy.