688. When Is a Bet Not a Bet?

Freakonomics Radio

This episode explores the legal and economic battleground of prediction markets, specifically focusing on whether platforms like Kalshi are legitimate financial

Key takeaways

• Tarek Mansour argues Kalshi is a peer-to-peer exchange where users trade against each other, unlike casinos where users bet against 'the house'.
• Economist Robin Hanson uses a 'wolves and sheep' metaphor to explain how information-rich traders (wolves) drive price accuracy in markets with many uninformed participants (sheep).
• Gary Gensler warns that legally classifying sports bets as 'swaps' could create massive legal instability for all financial derivatives.
• The distinction between regulated onshore platforms like Kalshi and offshore, crypto-based platforms like Polymarket is a central point of regulatory concern.
• Insider trading remains a primary risk in prediction markets, though Kalshi employs KYC (Know Your Customer) and surveillance to mitigate it.

Main topics

• Prediction market regulation
• Federal vs. State jurisdiction
• Market efficiency and information theory

Notable quotes

The wolves have information and want to trade on it, and the sheep don't.

Conclusion

The long-term viability of prediction markets depends on establishing

Transcript preview

Speaker 4 (0:00) U.S. voters abroad, have you got your ballot yet? All states are now sending ballots to overseas voters. So if you want to vote this year, you should request your ballot right away. Go to votefromabroad.org and click request my ballot. Ask to receive it by email. Questions? Ask our support team directly. Don't miss this chance to shape the future. Don't delay. Votefromabroad.org. Speaker 8 (0:29) When it comes to prediction markets, there are a lot of questions. Last week in part one of this two-part series, we asked a big question right in the headline. Are prediction markets the best forecasting tool ever or just another casino? In today's episode, we have more questions like when is a bet not a bet? Or more precisely, when is a prediction market bet gambling, in which case it would typically fall under state regulation? And when is it a trade on an event contract, which is regulated by the federal government? These are really great questions, and I really enjoy talking about these questions. That is Tarek Mansour, who co-founded and is CEO of Kalshi, the biggest prediction market in the U.S. As for these thorny questions about prediction markets. I like to think of them as risks Speaker 5 (1:17) rather than downsides. Like AI comes with a bunch of risks and we know that there's risk of online addiction. A lot of people kind of having mental health issues because of AI and there's job displacement. New technologies come with a series of risks. I come with the view that thinking critically about these risks and embracing regulation ahead of time up front actually is an enabler of growth. Speaker 8 (1:39) That, at least, is the path that Kalshi has taken, collaborating with federal regulators to create a new kind of financial platform. And Mansour's ambitions go well beyond betting on sports and politics and culture. He sees Kalshi as a new kind of financial exchange. Today on Freakonomics Radio, we ask Mansour and other proponents of prediction markets to talk us through the risks. On some topics, I might not recommend making a market. We also hear from a regulator. To some extent, even the stock market itself is a prediction market. And we consider whether this whole enterprise may at some point be shut down. Speaker 1 (2:18) I would not like to see prediction markets banned in the United States. Speaker 8 (2:22) Part two in our series, The Price of Prediction, starts now. Speaker 3 (2:37) This is Freakonomics Radio, the podcast that explores the hidden side of everything with your host, Stephen Dubner. Speaker 8 (2:54) Tariq Mansour is the 30-year-old CEO of Kalshi, and on paper, he's already a billionaire. When he talks about the power of prediction markets, he can make it seem as if they've got magical properties. For Speaker 5 (3:07) instance... Kalshi is really the most accurate way to predict the future, which is pretty cool. And... Allocation of capital. That's incredibly important. And even this. It will also depolarize the Speaker 8 (3:20) conversation across the board. Mansour also argues that prediction markets can be a more nutritious alternative to social media. People on prediction markets, Speaker 5 (3:29) they're reading about politics and they're reading about the economy and they're getting engaged and they're getting smarter and they have these nuanced, smart takes, taking them away from whatever they're being fed on these social media platforms that is driving them to extremes more and more over time. Speaker 8 (3:46) So how well do these claims hold up? That depends on who you speak with. Let's start with another advocate of prediction markets. The George Mason University economist Robin Hanson has been exploring and promoting this type of market for decades. On accuracy, they just Speaker 6 (4:05) consistently win. It's pretty clear that when you have a market on a topic and you compare it to another mechanism at the same time on the same topic, the markets are just... as accurate or much more accurate. Hanson Speaker 8 (4:17) has by now heard all the arguments against prediction markets, like the claim that they are a slightly classier version of gambling. His point is that if you want to call prediction markets gambling, then you need to think about whether our stock markets are also gambling. Ordinary financial Speaker 6 (4:34) markets really give people pretty much as many gambling opportunities as they might like. I guess you're just going to say, oh, but the sports betting you're allowing here is maybe more fun for them. So it's the added fun that's actually what you're objecting to, unless the losing of the money. But you might say, I don't want people to have so much fun because they lose money when they have fun. And I might go, okay, fine, you're going to be a killjoy. Then I might say, look, there's other purposes of markets, which is aggregating information, but we can distinguish roughly which markets add more information versus less. So a market on some small sporting contest really isn't very likely to add information about very many. practical decisions out there. I might just want to say, if you're going to cut back on people having fun because you think they're wasting too much money having fun, I'd say, well, just allow the markets on questions that matter. What comes to the top of your mind? What kind of topics? Well, there's world events, there's foreign policy, there's national policy, there's politics, there's business, there's tech. Look on Kalshi and Polymarket, you will see the headers of categories and many of those headers. are not as much fun for most people. I'm okay with people having fun, especially since I think we just allow people to spend a lot of money having fun in a lot of other ways in our society. Even risky fun. Like we allow people to take risky careers to be actors or musicians or even journalists. A lot of those wannabes fail Speaker 8 (5:59) and we let them take those risks. We let people go skydiving and drive race cars and all kinds of things. Indeed. And they kind of know they're risky when you... talk to them, but they still seem to do it. It sounds like you're saying there is an ethical or moral taint to some betting activities. Is that the case? There are definitely people Speaker 6 (6:20) who find them as a distasteful kind of fun. I think the idea that there's greed or money involved is part of their distaste. I mean, somebody who wants to be an actor is in some sense greedy for success, but people don't think of Speaker 8 (6:32) that as greed so much for the money, I guess. I find there's a distinction, though, between betting on yourself or taking a risk yourself. I want to go skydiving. I want to become an actor, etc., versus betting on the activities of other people, because I'm not actually doing anything. I think it's Scott Galloway who once said that especially the success trajectory of a young American male is directly proportionate to the amount of time they spend sweating versus the time they spend watching other people sweat. Well, again, people Speaker 6 (7:05) who succeed often then become financial market traders and enter the finance industry. So it is a way to try to see if you would be well-suited for that. Think about a bar bet, basically, a classic bar bet, right? People hanging out at the bar, some controversial topic comes up, somebody makes a drunken claim, and somebody else says, no way, want to bet. Okay, now that's not just a waste of time. That's a matter of who's right about this. And that's an assertion of your own confidence and knowledge and commitment to what you know. That's not just about other things besides yourself. That is about you and the other person and who actually knows Speaker 8 (7:45) better. Speaker 8 (7:49) Robin Hanson, we should remember, is an economist, and most economists believe that these differences in information are what makes any kind of speculative market work by putting a price on that difference. Here's how he explains it. Our standard simple model Speaker 6 (8:04) of financial markets is that they have two kinds of traders. One kind of trader knows something and they are trading to profit from their information. The other kind of trader is they're trading for any other reason and it doesn't really matter what it is. They are a worker who has to deposit so much for retirement, every paycheck, or they're a retiree who needs to withdraw money, every paycheck. Either way, they have to make trades, but they don't have information. These are the two kinds of traders in the standard simple model of financial markets. We call those noise traders. If they are noise traders, what are the first ones? You calling them signal traders? I'll call them wolves, and I'll call the other one sheep to give you a vivid picture here. Okay, perfect. The wolves have information and want to trade on it, and the sheep don't. Now, the wolves want to trade against sheep, not wolves. If two wolves trade against each other, then they've got information that cancels, and they are not going to profit from that. But if a wolf trades against a sheep, then the wolf on average knows something the sheep doesn't, and then they will profit on average from their trade. So wherever the sheep accumulate in the world of markets, the wolves go there. That's why Google stock has more accurate prices than some penny stock, because there's a lot more sheep there, so there's a lot more wolves there. Presidential election markets are more accurate than senatorial election markets, because there's more sheep there, so there's more wolves attracted. There's one last thing. To finish the analysis here is just to say, manipulators are people trying to affect the price without actually having information. They are sheep. For the purpose of this model, they don't actually know something. So a wolf is happy to trade against a manipulator. In fact, the more manipulators are expected in the market, the more wolves show up to counter them and the more accurate the prices Speaker 8 (9:54) get. Let's say I am an ambitious sheep. I want to do. some research or I want to gain some deeper understanding of the way this market works. How do I tell the difference between a wolf and a manipulator? Speaker 6 (10:06) The whole thing is that when you trade in the market, it's really hard to tell. Usually you just can't tell. So you have to go on the averages. Wolves on average expect some of their trades will be against wolves, but they want to go where there's more sheep. So the odds are better. As a sheep, you would rather trade against other sheep than against a wolf. But hey, you can't help the fact that there are sometimes wolves in the markets and you have to accept Speaker 8 (10:27) that. Speaker 8 (10:31) Robin Hanson's wolves versus sheep metaphor will be familiar to every gambler and especially to every odds maker at every casino ever. But Hanson makes the case that every market has wolves and sheep and maybe even manipulators. So how about Kalshi? Tariq Mansour likes to say that prediction markets can help democratize our financial markets by letting outsiders in. So you might think, well, aren't outsiders inevitably sheep? And won't they inevitably lose to the people who are setting the odds? Not necessarily. I asked Mansour to explain the difference between the Kalshi model and the casino model. Speaker 5 (11:10) It's very different. One of them, there's a house that offers lines and you're betting against the house and the house is betting against you. And that's the gambling business model. The house sets odds and then the house makes money from the users. The losses of the customers are equal to the wins of the house and vice versa. Over time, what the house does is it figures out who are the big losers and makes sure that they come back. And then it figures out who are the winners and makes sure they don't come back. It's very different from a financial market, which some people will lose, some people will win. That is a fair and open marketplace where the platform itself is a neutral arbiter of what's happening there. Speaker 8 (11:44) So you're describing not only the gambling industry historically, but you're also describing, as far as I understand, at least the current online sports gambling industry as well, correct? Yeah. You don't want the smart Speaker 5 (11:54) sharps. You don't want the people that are truth-seeking because those people are taking away from your bottom line. Prediction markets, you want those guys. These are the people that will do a lot of volume. They're making the markets accurate. They're the ones that are doing the research. They're the crux of what we do. Speaker 8 (12:11) So how well do the smart sharps do on prediction markets? The Wall Street Journal recently reported that on Polymarket, Kalshi's nearest rival, two-thirds of the profits are made by 0.1 % of the traders. What about Kalshi? Here is Nicole Kagan, who is Kalshi's head of research and market structure. Speaker 1 (12:31) The ratio is like 1 to 2.9. Speaker 8 (12:33) What's that mean? Speaker 1 (12:34) One person will make money off a trade and 2.9 will not. Empirically, that puts us ahead of basically every other kind of financial derivative transaction. It puts us well ahead of the traditional metric of comparison, which is a sportsbook. It would be unreasonable to expect every participant in an effectively zero-sum market to be profitable all the time. That would