The people betting on catastrophic world events

0:00This is The Guardian.

0:08Today, betting on the end of the world.

0:31In the early hours of January the 3rd, Donald Trump ordered a surprise attack on the Venezuelan capital, Caracas, to kidnap the country's leader, Nicolas Maduro. Neither the White House nor the Pentagon has issued any comment. Millions of Venezuelans' lives were thrown into uncertainty. Politicians at home and abroad scrambled to respond. And it's obviously a fast-moving situation. And we need to establish all the facts. It seemed like this was something no one had seen coming. The public needs answers. Across America, people are just saying, what the hell is going on? Or almost no one. Just hours before news broke of the military operation, an anonymous user on the prediction market placed a significant bet that Maduro would be out of power by the end of the month.

1:23Because it turns out that one person had literally predicted that this would happen. The facts are not known publicly yet. However, this particular trade really has all the hallmarks of an insider trade. In the hours before the attack, someone, and we have no way of knowing who, put a series of bets that Trump would oust Maduro on a prediction market platform, netting them nearly half a million dollars when it happened. These platforms allow their users not just to bet on whoever's going to win the Super Bowl, but on catastrophic world events. I just read about this last week. Kalshi is the first company that lets you legally put money on who you think is going to become president of the United States.

2:06Heavily regulated under the Biden administration, these apps have enjoyed a huge boom in popularity since Donald Trump came to power. That's amazing. Bro, there's$6.5 million of betting volume on this. Man, why are so many people betting on this? And almost anything is fair game. Prediction markets are the future. I think they're the future not just for traders, but also for news and information. You can bet on the political, if or when the US will strike Iran. To the trivial, how much will a Pokemon card sell for? To the far-fetched. Yep, people are even betting on whether Jesus Christ will be resurrected before 2027.

2:51But who are the real winners and losers in this game? The long-term vision is to financialize everything and create a tradable asset out of any difference in opinion. And I think if you do build a general purpose exchange... From The Guardian, I'm Annie Kelly. Today in Focus, the dizzying and dystopian rise of prediction markets.

3:17Sahil Desai, you are a senior editor at The Atlantic. Welcome to Today on Focus. I wonder if you could tell us, when do you remember the first time you heard about prediction markets? I remember, I think, during the lead up to the 2024 presidential election, seeing lots of billboards for these prediction markets. There's a billboard in Times Square showing live bets that people are making on the election. Would you bet your hard-earned money on that? I would bet someone else is hard hurting. I remember walking to work here in New York, where I live, and seeing, you know, a billboard just plastered with odds about whether at that point Biden or Trump would win.

3:58And it's Kamala. You think she would win? Yes. Yeah, I hope so. The subway for a little while here in New York was also sort of plastered with kind of ads for these sites. Our billboards. These are live trades, live bets, and people are actually betting on the exchange as we speak. So you can see the numbers, the amounts. But since then, they've only become exponentially more mainstream and more popular than that. Okay, and these platforms, and the best known ones are Polymarket and Kalshi, they're called prediction markets. What does that mean? Prediction markets are basically platforms where you can bet on anything and everything.

4:32You know, whether Trump will say a certain word in a press conference, whether Democrats or Republicans will win control of Congress later this year, who will win the Super Bowl next week. And so they're called prediction markets because these sites are sort of thought of as different from betting, mostly, honestly, for legal and financial reasons. They're thought of as more akin to stock markets. The idea being that, you know, you put money based on what you think will happen. And in that sense, prediction markets let you forecast the future. That's at least how these platforms like to talk about themselves.

5:12But in effect, it's just a fancy way of betting. So unlike in the UK, where the bookies are setting the odds, on prediction markets, there are no intermediaries, are there? So the websites themselves don't get involved. It's just the users who are betting against each other on a potential outcome. And it's interesting that they're also using the language and the aesthetics of financial markets, isn't it, to bet in favor of something, you quote, buy, yes. And then you, to confirm this, you click on a button that says trade. But isn't this just gambling? Yeah. I mean, they say it's not gambling.

5:50In my opinion, I think it is. It's functionally gambling. And what sort of things are people betting on? Honestly, I think almost the better question is what they're not betting on, because you can really put your money down on anything and everything. I think the funniest bets, in my opinion, is on Polymarket, where you can bet on the chances that Jesus will return in 2026. There's a 3 % chance of that, apparently. You can bet on anything and everything related to sports. You can bet on whether Donald Trump will throw up a middle finger at any point this year. Another really common bet, which almost has$20 million in trades right now, is how many times Elon Musk will tweet this week.

6:29And what are people predicting? There's a 27 % chance that he will tweet between 360 and 379 times, and a 69 % chance that he will tweet 380 to 399 times. Almost$20 million worth of bets have been placed just on that alone in the last week. Where would you put your money on that one? So thankfully, I suppose, for my insanity, I'm not up to date on Elon Musk's prolific tweeting habits. So I'll stay out of this one. But really, like, if you can think of it, there is probably a bet for it out there. And when it comes to politics, are there any kind of big geopolitical bets going on? Totally. There's bets on how long the government shutdown will last, whether the U.S.

7:25will strike Iran and by which date. There's bets on whether the Iranian regime will fall by the end of the month. According to Polymarka, there's a 1 % chance of that. Another example of this was earlier this month. We knew the media would be covering January 6th quite a bit because they think it's something that the American people are still believing their lies on. When White House Press Secretary Caroline Lovett pretty randomly just ended her press conference right at the cutoff of 65 minutes. In November of last year. And so I would encourage everyone to take a look at the website. It's quite well done.

7:58And thank you all very much. It's great to be back with you. And on the prediction market Kalshi, the traders had an almost 98 % chance that she would talk longer than 65 minutes. And so they lost a lot of money and they ended up sort of conjuring up a conspiracy theory that Caroline Levitt had put some money down on this and was inherently skewing things to make money. There's no evidence of this. But all of this is to say, it's really staggering to see every conceivable news event just reduced to numbers and to bets. God, it's so dark, isn't it? And when you listen to the founders of these companies in interviews, they'll often, like, herald their products as new ways to predict the outcome of future events.

8:47And that they're even more accurate than traditional polling. It's the most accurate thing we have as mankind right now until someone else creates some sort of super crystal ball. What do you make of that? I will say their logic, the idea that these markets can predict something is not totally nonsense. They do have some logic undergirding them. You might have an opinion, but if you are forced to put your money down on, say, whether it will be sunny tomorrow or whether it will be snowy, then you will probably bet more on what you really think will happen versus what you want to happen. And so if you can marshal the wisdom of the crowds that way, then hypothetically, you can use prediction markets to forecast real world events.

9:36Whether that functionally happens in practice in a real helpful way is a lot murkier. So it seems like they're kind of also, I guess, like scooping up lots of people that wouldn't like to think that they were gambling, but do like the idea that they're involved in this kind of more complicated trading. But the numbers you're talking about are kind of mad in terms of just how much money is being traded on these sites. How big is this in the US? It's big. Based on one estimate I actually just saw today, there's like a good chance that there's$1 billion that ends up being betted on Calgene Polymarket just on the Super Bowl.

10:16Just on one sporting event. So yes, there's a ton of money coursing through this.

10:27And so you're talking about the massive growth in these companies, the massive amounts of money that are being gambled, traded, however you want to call it, on these sites. But who are these companies? Do we know that much about them and who's running them? We do, yeah. So Kaoshi and Polymarket are the two biggest prediction markets. Kaoshi is different than Polymarket because it does have a certain level of regulatory approval that Polymarket doesn't. So if someone that's and there's there ends up being questions of insider trading, for example, Kalshi does have information on who this person is.

11:08They have to abide by what's called know your customer laws. But Polymarket, interestingly, doesn't. So it is it's a crypto site. Right. And it actually technically does not have approval yet to launch in the U.S. So many people still use in the US, but they use, you know, a VPN or virtual private network to do so. But what all of that means is that we don't know who these people are, right? So when someone placed some suspiciously well-timed bets on Polymarket, on whether Nicolas Maduro would be ousted as the leader of Venezuela about like an hour before Trump ordered the raid, even the company doesn't know who this person was.

11:51It's all sort of obscured through a layer of blockchain anonymity. And what about some of the big names that have been linked to these sites? Yeah, so Polymarket and Cauchy are both pretty integrated in Silicon Valley. Some of the biggest investment firms are investing in these prediction markets. It's quite notable to me as well that Donald Trump Jr. is both an advisor to Polymarket and Cauchy. And I think that's telling for one particular reason, which is that these companies were very intensely regulated under the Biden presidency. And once Trump went into came into office, he's really let these companies run wild.

12:34And so since Trump came into the office, these these companies have really, really, really boomed. And in fact, the Trump media group is set to launch its own prediction market sometime soon, which will be definitely sort of an inflection point for this industry as a whole. The other thing that we've seen is that media companies are starting to strike partnerships with these prediction markets. So we've seen CNN partner with Calci. You know, you can watch CNN talking heads sort of integrate the Calci odds into their coverage. The Europeans taking it seriously. What about the prediction markets?

13:15Yeah, are people in the public taking it seriously? The people who are putting their money where their mouth is. They're absolutely taking it seriously. I mean, take a look here. The chance that Trump buys any of Greenland by the end of his term on Friday, it was just 12 percent. You've seen The Wall Street Journal announce its own deal with Polymarket. And both of these just have been in the last month and a half or so. And so it's doing the same thing, integrating Polymarket odds into its coverage. I think perhaps most interestingly, the Golden Globes had a big Polymarket integration when that happened a few weeks ago.

13:50So, you know, you could watch the Golden Globes and hear the announcers tell you about the betting odds about who will win, you know, best podcast, which is quite a surreal thing to watch. And obviously these are commercial partnerships, but it also seems like media organizations are using this almost like the opinion of the man on the street. Yeah, I would say like many orders of magnitude of more people are hearing about the odds on the news or, you know, watching the Golden Globes and are actually putting money down themselves. Like, you know, it's become fairly inescapable over the last few months.

14:27So actually, basically, the way that they are almost like being called to be barometers of the news or current affairs is actually kind of skewed in terms of the actual number of people who are going on and making these predictions in the first place. Totally. Yeah. And I think that's one particularly concerning aspect of prediction markets is if you're just watching CNN and hearing about the odds of whether Trump will invade Greenland on Calci, which is one bet that has been mentioned in recent weeks by CNN directly, you are perhaps likely to think that this is a pretty helpful and accurate forecasting method when in fact it's not that clear at all whether these sites are telling us anything.

15:13especially as these media outlets strike partnerships with these sites. Polling, for example, is in a way much better than prediction markets. You know, you are you're having a pollster reach out to lots of different likely voters versus, you know, random people placing bets. You're making sure that you have a representative sample versus like maybe it's just like 3 ,026-year-olds who are hypotensitiously putting money down on some bet on polymarket. And so there are some clear reasons why polling is generally more accurate. Does that concern you, you know, as a journalist, that you've got that coming together of these prediction markets and journalism and news reporting?

15:59Yeah, it does. I don't definitely think that there can be some role for prediction markets. For example, to give maybe a helpful way that they could work, if the Wall Street Journal is writing a story on OpenAI's next financial earnings reports, and they could look at Polymarket or Calci and use the betting odds at that to get a sense of what people think might happen. I see some limited value in that, but I think what's particularly alarming to me is the political aspect of this. So let me make that a little bit more tangible. If you are a donor to a major Senate candidate, let's say, you could put millions of dollars into the polymarket prediction market for whether your preferred candidate would win later this year and swing the odds that way in a way that you can't with a traditional poll, right?

16:55And so you can really shape media coverage in a way that you can't with traditional polling. And all of that is exacerbated as media outlets start to incorporate this into their coverage.

17:13Coming up, the Trump family enter the prediction market.

17:31You mentioned before that Donald Trump is launching his own prediction market, and you mentioned that Trump's son is involved with both of these companies as well. Is it not really concerning then that there's no boundaries between political figures potentially making a bunch of money off a prediction market that they are very directly linked to being able to influence. Yeah, I think the Donald Trump senior, his coming prediction market is interesting and concerning for a completely different reason, I think, which is that once there is a Trump affiliated prediction market out there, it's really not that hard to see what happens next, which is that lots of Trump fans are going to put money down on the 2026 election and almost inherently they're going to say that Republicans will have a huge chance of winning the House, let's say.

18:27That seems fairly inevitable to me. But then what happens if Democrats end up winning the House? Then it seems like these prediction markets and the Trump-affiliated prediction market in particular gives yet another way for the president to potentially say that something went awry. Maybe the election was stolen. Of course, this is a completely hypothetical, but it's not hard to envision the sort of steps along the way to that happening. Wow. Yeah. And is there a real risk here that someone might influence real world outcomes, like make things happen so they can make a lot of money? It's a real possibility.

19:10Do we know that it is indefinitely happening all the time? We don't. But I think the concern that I see is that as these prediction markets go mainstream and as they become integrated into every inch of news coverage, it just continues to heighten the chances that something goes seriously wrong. That could mean, you know, a government official learning about some intel a few minutes before the public and putting money down and coming away with a lot of money and alarming for obvious reasons. You know, it just raises the potential graft at play here. I think the other concerning outcome is just how this could warp understanding of politics and end up becoming a self-fulfilling prophecy, I suppose.

20:03I mentioned earlier that CNN is talking about the odds that Trump will invade Greenland, for example. Say a massive Trump donor puts in tons of money into that market, and suddenly the odds that Trump will invade Greenland go from 33 % to 66 % or something like that. Maybe Trump sees that on CNN and uses that to make his own policy decisions. And these accusations of insider trading that these companies have already faced, what are they saying in response? Yeah, I think it's quite interesting. So Calci and Polymarket are different on this front. On Calci, insider trading is illegal. But I will add that it's sort of difficult to decipher insider trading all the time.

20:52You know, to use another hypothetical, let's say how many times Elon Musk will tweet this week. What if Elon Musk himself doesn't place money on CalShare Polymarket about how many times he himself will tweet, but he tells someone to do so or sort of hints at one of his pals that he's only going to post 380 times or something like that. On polymarket, insider trading is murkier. As I said earlier, we don't know who is betting on the platform, which makes it nearly impossible to root it out. But what is sort of undergirding the two different platforms is a larger sort of ideology to prediction markets, which is that insider trading is good, actually.

21:41Yeah, I think that people going and having an edge to the market is a good thing. Obviously, you need to curate them and you need to be really clear and stringent on where the line is drawn and like sort of ethics. And we spend a lot of time on that. But it's sort of an inevitability that this will happen and there's a lot of benefits from it. And people will adapt. And so what you've described to me is kind of this like Wild West situation. Like they don't seem to be really regulated by much at all. Like they're not regulated under gambling laws that necessarily want or are being regulated by insider trading laws at all.

22:15Is there anything at all that could rein these sites in? Yeah, so they're being regulated in a small way. So Kalshi in particular, right, because it has U.S. approval, they do have to abide by some insider trading laws. But in effect, like, they are mostly running wild, I will say. So just the other day, the Trump administration official who is in charge of regulating these markets has basically, like, pretty directly said that he's encouraging prediction markets to go mainstream. There is something that's slightly more sinister going on here as well, isn't it? I mean, here in the UK, it would be very difficult to find a betting company that would allow you to put a bet on whether the US is going to bomb Iran and potentially kill lots of people.

23:07It's not just betting on a basketball game, is it? It's kind of quite astonishing that people are making money and sometimes quite a lot of money out of this stuff. What do you think it says about this moment in the U.S. that these markets have become so dominant? In a social sense, it's alarming to see any news event be reduced to a derivative. It just adds a layer of abstraction from what's really happening in the world. You know, Calci got in a little bit of hot water late last year for letting people bet on whether there would be some formal announcement of a famine in Gaza by a governing body.

23:44And just think about that, to step back from that a little bit. Betting on people going hungry, right? It just reduces everything that's happening around us all the time into just dollars and cents and percentage points. And there's something inherently concerning and a bit dystopian about that. And so there's a lot of consequences to the boom in prediction markets that we're only just now beginning to understand. And maybe we don't even yet really know how this can go wrong. It's sort of just like we're, you know, we're opening Pandora's box and we don't know what will happen next. Sahil, thank you so much.

24:31Thanks for having me. And that's it for today. That was Sahil Desai, a senior editor at The Atlantic. My thanks to him. If you'd like to read more about the world of prediction markets, my colleague Niamh Rowe has just written a great piece called Aunt and Polly Market, Privileged Users Made Millions Betting on War Strikes and Diplomatic Strategy. What did they know beforehand? You can find that on theguardian.com. This episode was produced by Isaac Tompkins, Alex Atak, and Eli Block. Sound design was by Ross Burns, and the executive producer was Homa Khalili. We'll be back this afternoon with the latest.

25:16This is The Guardian.