Is Kalshi a Casino? Its CEO Makes the Case It’s a Market

I'm excited.

Yeah. So, you know, I was born in California and my parents, my dad was studying for med school. Then we moved to Lebanon when I was sort of young and my dad sort of left. I grew up with a single mom. You know, and Lebanon, the way I describe it, is a very volatile country. It's a very unpredictable country. And the thing that's interesting about unpredictability is it makes it hard for you to plan for the future. You're just, when you don't know what you're going to wake up to the next day, you're not really thinking three, four years ahead. And that's kind of a hard place to be because, you know, when I always think about whether when people want to succeed and they kind of build great outcomes for themselves, you have to plan, you have to think ahead. You shouldn't think basically day-to-day. And that volatility sort of made me an anxious kid, you know, a single mom added to it, and my refuge was math. And what I liked about math is it's elegant, it's clean, it's deterministic, there's not a lot of debate. Either the thing works or it doesn't, right? And you know, you're not going to debate a proof because it's undeniable, right? And that led me to go to MIT, which is where all the math guys, you know, the smart math geeks were going to. And you know, inevitably when you get to MIT, you see where the math geeks were going to afterwards was finance. And you know, when I went to Goldman in 2016, I got really struck by a few things that really struck me. One, just sheerly how complicated the products that we were trading were. Like it's credit default swaps and swap options and more, like all these very fancy names, and the math was so complicated. It's stuff that the average person, really like the layman, has nothing to do with. It's completely far removed from what an average person thinks about. But the second thing that really struck me is the opinions, the views that led to trades were actually very simple. It wasn't like, oh, this swap is mispriced in one way or the other. It was Trump is going to win the election or respect that it's going to happen or not happen. And you know, the Trump trade was a very interesting one because, you know, a lot of people came in and said, 'Look, I think Trump is going to win the election.' And we're like, 'Okay, I want a position on it. How do I basically put a trade on it? How do I bet on it?' And we said like, 'Okay, he's a populist candidate, so he's going to be bad for the market.' So the way to do it is you short the S&P. If he wins, S&P is going to go down. That's the way to do it. Everybody, basically a lot of our customers did that. And, you know, come election day, he won. People were right about the prediction but then surprised they lost money because the S&P went up. It was actually, I don't know if you remember this, but there was a record rally that was... and so clearly it was a very bad product. Like we were using proxies to get the specific thing that we were getting. And so that got me into this: what if you build the New York Stock Exchange or whatever, the Chicago Board of Trade, the traditional financial markets where you're matching buyers and sellers with different opinions on something, but instead of doing it, we broaden it. Instead of the same structure that like, okay, you're predicting the price of a stock in the stock market, predicting the price of oil in the derivatives markets, the commodities... what if in this case you were predicting whether an event is going to happen or not, whether Brexit is going to happen, Trump is going to win, whether elections, politics, all of these different sort of topics. And the more I read about it, I realized actually the idea wasn't new. And you might, I don't know if you've read Friedrich Hayek from like World War II, but I think he's the inventor of prediction markets. He didn't call them prediction markets at the time. He called them information markets. And you may have come across the knowledge problem. It really is the idea of, okay, we make decisions. Generally it's like a small group of elites that make decisions that impact society, but the decisions that they're making need a certain like information. That information doesn't sit with them. It actually sits with people, you know, an average person in Hawaii and California and Texas. I mean, everybody, everybody. And you need a way, like you need a magnet to get this information to the decision makers. And Hayek is saying market prices are the best way to get truth. Why? Because people buy a little bit, people sell a little bit. You get all millions of opinions that clash against each other in an open market. And out of that clash, you get the price that's sort of constantly fluctuating. You get like an averaging of everybody's opinion. And it works because there's skin in the game. Markets don't lie. When people have money on the line, they kind of cut the... they tend to be more truthful. And he's like, 'Oh, we need an information market.' And it wasn't up until way later that we started sort of this idea. And like when we, when I, you know, we were kind of started working on prediction markets, I got obsessed with it because I was like, if we could build a market where there's enough people predicting the price of all these different things, that's incredibly useful because we're going to get more signal in a world where, you know, I started two years in 2018 where polarization was starting to go crazy. I think our source of information, like we have more information than ever, but the filters are increasingly worse basically because social media, mainstream media, like their filters is clickbait. They don't get paid based on truth, right? Like you don't get paid if you put out an article or tweet that is super smart, thoughtful, well analyzed. Actually, in fact, you probably don't get very many likes if you put that sort of tweet out. But if you put a tweet that's, you know, hate-driven or crazy or extreme, you're probably going to get way more likes. And the likes leads to more impressions, which leads to more ad revenue and so on and so forth. And I found that markets will be the counterbalancing act to that where you're actually rewarding people for being right and you're punishing them for being wrong. And I had to start it.

It's really the stock market. There's no question about it. I mean, maybe let's just sort of break it down in a few sort of modules. I think that like there's kind of this gambling question, is it the same as DraftKings or a sportsbook, and let's get to that in a second, but then I want to talk a little bit about the polls and you know, you've seen maybe some of the recent studies around like how much more accurate the prediction markets are than the polls, and maybe the... and then there's the elites. So the best way to think about prediction markets versus polls or versus elites is I think a source of information, it's kind of in many ways one of the only sources of information that flows bottom up from people rather than top down. There's no gatekeepers, no pundits, no filter of, you know, a small group of five people in a news organization deciding this is truth and this is not truth and this is worth your attention and this is not worth your attention. It's millions of people with millions of different opinions trading against each other in an open, transparent marketplace and out of that they basically make the prediction. But that to me is a bit of a kind of reshuffling of power from like the small group of elites that told us and are telling us this is your truth to actually like the truth coming now from people, millions of people are essentially battling it out in the open market competing. And I think that when it comes to polls, it's a bit of a like, it's a different structure because polls are asking people what you want to happen, what are you going to do. Markets are not asking the same question. They're actually asking people what you think will happen. And when you're asking that question now, it's a very different thing because now people are not thinking emotionally about what they want or don't want. They're thinking more rationally. Now you're giving people an incentive of like, look, if you have a good piece of information and you go out and you do research, you may trade based on that knowledge and make money if you're right. And if it's a good piece of knowledge, and if you have a bad piece of knowledge, you're going to lose money and get punished. And that incentive is why markets work so well. So maybe I'll get back to the stock market, but when it comes to politics, it's been interesting to me because our politics have been, we've been dealing with it in a very sort of, with very, very much of a lack of rigor. You know, it's like people on both sides are saying a bunch of different things. They're usually very kind of, they're saying the opposite things basically from each other. Social media has added fuel to fire. You know, my feed may be different from somebody else's feed in a way that like our realities of the world are completely distorted and different. And that has been always like interesting to me because, you know, okay, if we're thinking about what the value of Apple is today, we have a ton of rigor around figuring that out. You have the quants, the analysts on Wall Street, you have traders that are, you know, and hedge funds and we have data infrastructure around it. We've studied every small piece of how the Apple stock moves and why it's going to move and so on and so forth. We have an insane amount of rigor. Where does that come from? From the existence of a market. Because these people that are doing all that research, they're incentivized. They get paid for that research by trading it and making money off of it. And we're bringing that same thing to politics and weather and sports and economic. We're bringing that same level of rigor because now on Kalshi, all these traders are doing all this amount, copious amounts of research and they're incentivized to do it. Some of them are doing it as a full-time job and have set up funds to do it, because they get rewarded if they end up being right.

That's right.

Yeah. So there's a few kind of pieces here. I mean, I think one is how to evaluate prediction markets. They're not a poll. They're a probabilistic model, right? They're telling people like, okay, we think there's like 80% chance of something happening. Now that doesn't mean we're saying it's not going to happen. It's going to happen 20% of the time, right? So if you play that race five times, four times, like in four of them one of the candidates should win and in one of them the other candidate should win, right? And the way to evaluate them is like what we call a calibration plot, which is you have to take a large number of markets over time and say, okay, what is the probability that the markets are saying something is going to happen, and you compare it to how often those things happen, and the 45-degree line is basically the perfect thing. So if we say something has a 20% chance of happening, it happens two out of 10 times basically. And when you look at prediction markets from that lens, it's actually close to perfectly calibrated. It works very, very well. Now, people always mistake that because, you know, they look at like, okay, we're saying something is 80% chance of winning. They think it's a poll, but it's not a poll. 80% could mean a 51-49 poll, right? Because, and so it's different. And what prediction markets do is they're not creating a new insight, they're aggregating all the information that's out there. Whether it's the polls, the pundits, the social media, they do their own independent poll, like traders do their own independent research. It's a way to aggregate all the information that's available and constantly pricing it in real time. That's what it does. Like a candidate goes out on stage, says something really bad, you'll see the market tank. The same way that like if Apple releases bad earnings, the stock is going to tank, right? That same, it prices in real time all the available information. Now, here's what's interesting. Like any financial market, prediction markets are exposed to insider trading and market manipulation, right? And market manipulation, the way to think about it is like, okay, if Ben has, you know, 10% chance to win the 2028 election, you know, maybe you commission a few friends or a big hedge fund guy to come in and buy in bulk to try to move 10% to 25%. And actually, it's interesting. It actually happened in the mayor race in LA with Spencer Pratt where it was actually at 10%, it moved to like something like, I don't know, 35%. Because someone put $2 million to move the market up and it was amazing because within 9 seconds it moved back to 10%. Why? Because there's now a 25% incentive for people to correct. It's a little bit like, you know, me, you know, like if I, you know, if a stock is worth $100 and somebody moved it unnaturally to $60, what are you going to do? You're going to keep buying it...

...all the way back to $100. You have an incentive to basically correct the mispricing. And in the meantime, you know, that's, we're a regulated exchange. We spent the first four years of the company getting regulated up front and adding the right guardrails. So we have to police, we ban that sort of activity as financial crime. And so, you know, in the meantime, we had 100 whistleblower complaints from our customers saying, hey, something is weird with that trade, you should go investigate it. And I don't know if you've seen it, we caught George Santos, we caught, you know, White House teleprompter, like we caught a bunch of people that were trying to cheat in the markets. It's, I always think about it, it's like, you know, that $2 million would have been much better used if it was given to the campaign, or you could have, I don't know, commissioned a bunch of AI, you know, I don't know, like videos, or you could have commissioned your own polls to say whatever you want to say about that race, and that would have been better used in terms of influencing public opinion, which by the way, I think people do anyways.

So that's a great question. It kind of comes to the gambling question, which I think we will address. But look, I think it depends where you fall on the spectrum, right? Like I'm generally a free markets person. Like I believe that if people want to transact and speculate on difference of opinions, they should be able to do it. But I think there should be sensible guardrails. Like, you know, that's why I chose upfront to get regulated. I like having guardrails. And, you know, one of those guardrails, what are the types of markets that shouldn't exist? So like, you know, war, terrorism, assassination, violence, things that can create moral hazards. So we don't do wildfires, for example, because yes, predicting whether there's going to be a wildfire or hedging against wildfires is very useful, but it's very easy to create a wildfire. So the moral hazard is too great. And then there's like stuff that's a bit more on the line where, you know, we have a standard as a regulated exchange of is something too manipulable, it's too easy to manipulate, and like you have to kind of set the line, and that's a constant, you know, push and pull between us and the regulator. And where do you fall? You know, it's interesting, teleprompter markets are one of the most criticized markets on Kalshi because they seem at face value a bit gimmicky. Like it's like, well, you're just, you know, betting or trading on like whether someone is going to say something. But, you know, it's so interesting, we put out some research on this, like whether Trump is going to say the word Iran or Bitcoin or Tesla, it actually moves that market. It's actually oftentimes when he mentions a specific topic. And, you know, the reason we came up with it, I used to work in finance, like we discussed, at Citadel, we would read Powell's transcripts to a T and Janet Yellen, same thing. Like, you know how when he goes out and does the press conference, if he says the word austerity or if he says, you know, soft versus hard landing, like these specific words, I would say, move the Treasury futures market by billions and tens of billions of dollars, right? Because, you know, the words of these kind of elected officials really matter. And then the questions actually exist for all markets or not, and that's where the regulator draws the line. And so, but, you know, what's interesting, and maybe this kind of brings to the question like gambling or betting and what's different. So people ask me like, oh, you know, but people are betting on Kalshi, and I say certainly, like, you know, certainly people are betting on Kalshi the same way they're betting on the stock market. Like, do you believe that people gamble colloquially or bet on Robinhood? Certainly. I mean, you know, zero-day expiry options, a lot of customers are betting on these things. The same way that initially, how did these markets start, the derivatives commodities markets, was the farmers, the grain farmers. What were they doing? They were betting on the price of grain. And so is that gambling or is it not gambling? And what's interesting about this is like, you know, there was this recent court opinion. Let me actually, I don't have my phone right now, but I remember the recent, there was a court opinion that was essentially, it goes like this: we believe that this type of activity is contrary to public interest and it's void by the common law because of that. And it's not only contrary to public interest, it's actually a crime against society. Crimes against humanity is one of the preeminent national sins. The species of gambling is one of the worst things that's happened to society. And what's interesting is that that opinion, it was from the Illinois Supreme Court and it's from 1888 about grain futures. And it was interesting to me because, you know, they were seen as gambling at the time. Why? Because people were betting on the price of grain and some farmers would lose their shirts on that thing. So how is the difference? How do we, how is grain futures, why are grain futures not regulated now state-by-state as a gambling business model versus a market? And the difference is the structure. Like one of them is happening on an open marketplace where the incentive of the market is to match Ben with Kelly, and whoever is smarter or whoever has better opinion will win on an open and free competition. And the market is like the stock exchanges, like Kalshi, we take a small percentage fee out of the transaction. Or are you betting against a house whose sole purpose is to make you lose, right? And that's the whole distinction. And why do we treat it differently? Why does the law treat it differently? Is because on a market it has positive externalities because we get a forecast out of it, right? We bring in and we incentivize people to do smart things and to do research. And the more they do that, the more the markets are efficient, the more accurate the prices are, and so we can predict different things, whether it's weather or sports or other things, which makes us make better decisions and get smarter about the future. But then the second thing is some people are not doing it to speculate to make money. They're doing it to hedge and insure themselves against a bad risk, and that's why these markets are regulated as a federally regulated marketplace. The other one actually roots out the smart people. If you're a sharp or a smarter on DraftKings, what happens? You get limited. Like the card counters get kicked out of the casino, and then they figure out who the losers are and they keep, you know, giving them, they figure out how to make them addicted. So they keep losing. That model is regulated differently because DraftKings sets the odds and you need to make sure the state-by-state, like, hey, don't screw your customers too much. Don't give them too many tools for addiction and whatever, like to promote addiction. Whereas an open market, it has to be the rule. It's more about do you have fair rules and are they uniform across all 50 states because the market is matching a bunch of buyers and sellers and you want a very large diversity of participants to get the accurate forecast and so on and so forth. Which I always say the best analogy is like it's not a question of whether we want to have customer protections or not. That's certainly the case. You should have the same or similar type of customer protection. It's more, imagine New York Stock Exchange was siloed to 50 New York stock exchanges state-by-state where you can only buy Tesla stock in Florida from another person in Florida. Wouldn't make any sense. Like it would, the liquidity wouldn't work. It just doesn't make any sense. And that's how we kind of treat those two things differently.

Right. Exactly.

And the core is about predictability and risk management. And so that is true for the stock market, but the derivatives market and the commodities markets is not about capital allocation. So when we have oil futures, why do they exist? Oil futures don't help people allocate capital from people that have it.

Yeah. And that's not, that's not the purpose of that market. Oil futures or grain futures or traditional commodity futures, they came around and at the time people were saying the same thing, like, well, the stock market, you're allocating capital. So I understand why some of the speculating and the betting that's happening or the gambling that's happening in the stock market is okay because it's justified by the capital allocation. So why do we need grain futures? And so, and here's kind of where it gets really interesting. When farmers basically were like, you know, figuring out what to do in March, they had to basically try to guess what they're going to sell their plants for in September. And what would happen oftentimes, like, you know, let's say they plant a batch and then the prices fall in September, they would end up not being able to sell it and like they would lose, they would not be able to turn a profit in that year. So what would happen is that come the next year, they plant less, and that was bad because, you know, you have less food supply and so on and so forth. So come the derivatives markets and the futures markets, what's the point? And at the time people were saying, what's the point of letting farmers bet on the price of grain? Isn't that gambling? Sounds bad. What it did is actually gave them predictability because now that you have a market for predicting, like grain futures are just a prediction market, right? They're just about the price of future, right? It was giving them predictability over what the price that they're going to be able to sell at in the fall is going to be. So once you have predictability, what can you do as a business? You know, hedge, yeah. You can hedge, but you also, you can make better decisions. You can figure out exactly how much to build, and then now you can get lines of credit. You can, the predictable and forecastable businesses are much better than unpredictable and unforecastable businesses. And that's why risk management is actually one of the core tenets of the economy long term is because if you can predict and forecast things better, you can manage your risk better and allocate resources and capital and make decisions that are more kind of in line with what the future holds, right? So the farmers now have an indicator of like, okay, here's how much I can sell at. So here's how much I'm going to produce this year. And I also can buy a position that offsets some of my losses if the thing moves against me. And the same exact, and this is the same way that we got then commodities futures, and then we got interest rates where people were hedging against interest rate. And on Kalshi, we started getting, you know, inflation markets where people were hedging against inflation, then weather markets where you have, you know, we have a bunch of small ice cream shops, that's one of the biggest use cases that hedge every day against rain because when they have rain, their revenue tanks, and so they smooth out their P&L by putting on the hedge. And then we brought in overtime to politics and elections and sports. And people always ask, well, is there economic impact to sports? And I look at them, I'm like, think about that question again. Certainly. Like the World Cup being played in America, like has actually, if you go to Kansas and you go to a bunch of these cities, they'll tell you about the impact of the games that happened there. And I'll tell you one of the biggest sources of hedging that we see on Kalshi is actually sports markets because, you know, when US, for example, when the World Cup was happening, people were hedging against the US getting kicked out early. Why? Because the advertisers were buying a bunch of inventory, but they know that if the US loses, they lose 50% of their inventory value immediately. So they buy the offsetting position so that they smooth out their P&L, which helps them invest more decisively, more deterministically. You know, last week there was a big, you know, I don't know why that one went viral. We have a lot of hedging on sports, but that one specific one was a restaurant group that had 25 bars where they hedged against the Dodgers not making it past the playoffs because they made a math. They're like, 'Okay, we'll make $120,000 of revenue if the Dodgers make it through and zero if not.' And so they bought an offsetting position where they make $60,000 no matter what. And why is that important? Because it gives them predictability, and businesses that are, the more predictability you have, the more you can invest in businesses. And so long-term, what I always see, what are the prediction markets long term, it's not just about predicting the future for the sake of it. It's about managing risk more effectively. And the things that we need to manage risk on are not just grain futures anymore. Sports has become a big industry, the economy, geopolitics, what's going to happen to straight up horses, all these different questions are increasingly more important to price.

...or 51-47.

Exactly. You're taking that as 100% probability of a thing happening.

Yeah, I mean, you have to look at it. So a few things. I mean, one, you have to look at it a long enough time period where you aggregate all the races and you see how it calibrates. And actually the prediction markets do calibrate very well. And there's a bunch of studies that, you know, a bunch of third-party studies were run about biases. Are there biases? Because people used to think like, oh, it's young men that are biasing prediction markets. And like the studies actually surface that actually any type of bias is corrected by the smart money. There's people that make it, they live like in the stock market where like when people are too excited about GameStop, there are people that are like shorting it on the other side to bring it back to the right price. And like the same thing happened here. When people are too excited about one candidate or the other, there are people that are like, you know, studying that and thinking like, oh, the market is biased in a certain way, so I'm going to like short that bias or buy against that bias, and that's my profit strategy. And that's why these markets work. It's interesting. So a few things. One is I think it's on us to educate people, and this is one of the things that really, these are probabilities, not a poll, and that's something that we need to keep pushing on the message so that people interpret them the right way. And that's why one of the things we do at Kalshi is like we're putting all the information next to each other. So I put the prediction markets and then next to the polls, next to the fundraising history, next to their policy position, next to what people are saying on Twitter, and next to the news so that all of it is looked at holistically. It's like, you know, in the same way that, you know, a lot of people on Wall Street right now thinks that Apple stock is overpriced or underpriced. And that's the beauty of a market. And people come to me and say, I disagree with your probability. And I tell them, guess what? That's amazing. You can buy Paxton now for 30 cents on the dollar. You can 3x your money. And that's what contrarians do. They challenge the status quo and they get rewarded for it, right? And that's the beauty of the market. It's always dynamic. It's always calibrating. If we think it's going too far in one direction, there's people that are going to come and bring it back to the other direction. And the most critical piece of all of this is, you know, I love what you said. It's like a bit like, could you develop momentum towards a candidate now because of prediction markets? And my answer is generally yes. Now, there's a few things that are a little bit, they're very, very like, there's kind of like some structural things that are very important to understand. So let's take out prediction markets and let's think about how does momentum historically build up towards a candidate. It's a small group of elites, four publications, whether it's CNN or Fox or New York Times or Wall Street Journal or Politico, that say these are the four candidates that are worthy of watching ahead of the election. And those five, you know, groups of four elites each, like 20 people have decided who are the people that are going to be worthy of getting the data and worthy of coverage and worthy of getting campaign donations, and they kind of start building up that campaign. And to me, that's why like oftentimes the elites don't like prediction markets because what do prediction markets do? Who's actually deciding this time around? It's millions of people. Like to me, the choice is very simple. It's like, are you just trusting the elite media or the polls? By the way, the polls, do polls influence public perception? Like, you know, the Texas race? Well, yeah. It's so easy to commission a poll. In fact, you know, a week ago, you saw the Wedgewood poll that was... I mean, two dudes basically came up with fake numbers and like it was reported also to mainstream media and they were influencing public perception. But now you have actually something that does two things. One, it's bottom up. It's people-driven information. It's...

Something that flows from people rather than a few gatekeepers, which I think is a good thing. Like, I would rather millions of Americans tell us who is worthy of campaign donation, who's worthy of building momentum, rather than, you know, 20 different people that, like, you know, are sitting in DC. And the second thing is markets, at least, you know, are more resistant. They have a more clean incentive. Like a poll, when somebody runs it, you don't know if they're running it to truth-seek or they're running it to bias our perception one way or the other. And by the way, a lot of polls are run to bias our perception. Like in the next seven days, you're going to see polls from the Republican and polls from the Democrat that are put out. They are saying exactly the opposite things. Why? Because these polls are, like, not commissioned to surface the truth. They're commissioned to, you know, bias public opinion one way or the other. But at least the markets, you know, that people are trading on there, they're just trying to make money, right? They're just trying to truth-seek. And so to me, that's what's so interesting. In many ways, it's actually going to be a solution to a lot of the issues that we do face in our elections right now, which we're flooded with misinformation on social media. We're flooded with AI is going to empower that misinformation even harder. Polls are increasingly less trusted because why would we trust them? People are using them left and right to push a narrative. And so I like, like to me it's so crit- like it's, it's a little bit like I tell people, you know what, you can ignore the prediction markets, but you're ignoring millions of people and where their pulse is at. You should just do it at your own peril.