Kalshi CEO Tarek Mansour on The Case for Prediction Markets | Ep. 48
The origin of prediction markets, the long fight to build a regulated exchange in the US, and how Kalshi aims to improve forecasting and hedging by letting people trade directly on real-world events.
I sat down with the co-founder and CEO of Kalshi, a prediction market reportedly worth over $20 billion.
After years of struggle, Kalshi won a regulatory battle, culminating with their successful suit to overturn a CFTC decision, giving them the ability to allow consumers to bet on politics, sports, weather, and more.
Tarek shared the history of how he got here and what kind of company he had to build to weather all the uncertainty.
I also got a chance to push him on Kalshi’s place in the world and its potential for positive or negative impacts. I was persuaded by his arguments that it can be a big force for good.
We closed out by talking about his ideas around company building, hiring, and the future of Kalshi. Hope you enjoy.
Timestamps:
(0:00) Intro
(0:23) Kalshi’s genesis
(5:05) Regulation-focused from inception
(11:06) Suing the government
(18:02) Gambling vs. financial markets
(20:58) Defining insider trading
(25:38) Incentive structure of the system
(32:40) Investing vs. trading
(35:31) Hedging use cases
(41:38) Scaling a lean team
(44:02) Defining Kalshi’s culture
Links:
https://kalshi.com/
https://x.com/mansourtarek_
Watch on YouTube; Listen on Apple Podcasts or Spotify
Transcript
Disclaimer: Transcript generated with AI assistance and lightly edited for clarity and accuracy.
From Lebanon to Wall Street
Jack Altman
I’m really excited to be here with Tarek, CEO of Kalshi. Thanks for doing this. I’ve been looking forward to it.
Tarek Mansour
Thanks for having me.
Jack Altman
I want to start with the history of Kalshi. Yeah. Can you take me through the genesis, how the idea came together, how the company got started?
Tarek Mansour
A little bit of background before that. I was born in California. I grew up in Lebanon. Lebanon was a rough terrain to grow up in. It’s super volatile, a lot of uncertainty. I found refuge in math. It was a mix of different things. I grew up with a single mom, my mom was like, “I want you to be successful. Maybe math is a thing to get you back to America.” So I got really into math and then a lot of my decisions at that point were basically like, “What do the best, smart math people do? Oh, they’re going to MIT so that’s what I need to get into.”
I got into MIT and then the next stage was the same question and the answer was finance. I started spending time in finance. I worked at Goldman, I worked at Citadel. I worked at some small prop shops. Luana did Bridgewater and Citadel as well.
There was a pattern that was emerging in a lot of these places. This is the example I always love to give. In 2016 at Goldman, I was working on this desk and there were two questions that were bothering everybody in Wall Street. These are two questions that people are figuring out how to trade on: will Brexit happen and will Trump win the 2016 election? People wanted to have the Trump hedge or the Trump trade. The thing that really stuck with me was that…
So Brexit happened and it was a shock. People were really completely shocked. The polls were saying this was not going to happen. People had all these smart trades about how to hedge against Brexit, but then a bunch of desks on Wall Street blew up and they lost money and all the bad stuff.
When it came to Trump, a very similar thing happened. The trade that we sold at Goldman, the very common Trump trade, was to short the S&P. Because if he’s going to win, the S&P is going to go down. Everyone bought that trade. That was the trade. It was a horrible trade because Trump won and the S&P actually… I think it was sort of like the single biggest rally in the S&P’s history ever essentially. It’s like the worst trade.
Jack Altman
The exact wrong way to trade the idea because what you were trying to trade was this underlying thing that you got right, and then you expressed it backwards.
Tarek Mansour
They were right about the prediction and they lost money.
Jack Altman
This kind of is like when a company has quarterly earnings and people are like, “Oh, it’s going to beat earnings so I’m going to try to buy the stock. And then it beats earnings and the stock goes down.”
Tarek Mansour
And they’re all smart in retrospect. If you actually trace the plot of whether the stock went up or down after an earnings beat, I think it’s pretty much 50-50. It’s mostly priced in.
I realized two things. The smartest traders and institutions, a lot of their trading ideas, the things they’re trying to do, originate from a simple human view about the future. “I think Trump is going to win.” “I think there’s going to be some change in diplomatic relations between these two countries.” “I think COVID is going to come back.” The way they thought they were trading on it with traditional markets is the event, but what they’re really trading on is the reaction function, how the market was going to react to an event.
They weren’t trading on whether Trump was going to win. They were actually trading on how the S&P was going to react to Trump. In retrospect, we can’t really predict that. It’s so impossible to predict. It was a very exciting idea because what if we just build this marketplace where what you’re trading is—
Jack Altman
The specific thing that you’re thinking about.
Tarek Mansour
Yeah. It’s just things that people care about, whether politics or economics or weather, or really any topics that people naturally walk around the street and think about. People don’t think about, “What is Cisco going to print? What are their financials next quarter?” They don’t think about that. They just think about, “The Fed might raise interest rates,” or things that are simpler. So that is exciting, because the TAM could be much larger, a large number of people would care.
The second thing that was really interesting… If you believe in markets, what markets really do is aggregate information. They are a very good weighing function. They can figure out how to take information from a bunch of people, aggregate it, and get a single price. So what if we apply that to all these specific events or questions about the future? In theory we should get a smarter, more accurate, market-based answer about all these questions.
That got me really excited. At the time we were thinking, if we could get a little bit smarter about the future, that’s a very worthwhile product to build. You don’t need to be 100% smarter. Even if you get 10% smarter about the future, that’s more than enough. That’s how I got into prediction markets. We can talk about the whole history of prediction markets and all of that. I got really obsessed.
Starting Kalshi
Jack Altman
So when you got started with the company, what were the first couple years of building like? What did you do when you got started?
Tarek Mansour
I was actually going to go work at Citadel, because I had spent a summer there and I actually loved it. It was one of those situations where Luana and I started talking about it, and the idea was just bothering me. I could not get it off my brain. I’m a little bit OCD and I get obsessed with things. I couldn’t get it off my brain. One thing I always say is that we were not entrepreneurs trying to figure out what product to build to build a company. That was not how Kalshi started.
Jack Altman
You just had this one idea,
Tarek Mansour
The idea forced itself on us. I was like, “No, forget about that. Let me just go to Citadel and they’re going to pay me all this money.” But then I remember we had a friend who was going to a YC hackathon. I don’t know if they still do them, but they used to do these hackathons and bring a bunch of builders. He was like, “Oh, I’m going to this thing, you should come.” I was like, “I think the deadline has passed.” He’s like, “No, you should just email the guy.” I forgot who the organizer was at the time but we emailed something like, “Hey, we were trying to figure out flights or something.” He’s like, “Yeah, fine, you can just come.” So we were like, “Okay, we should just go.”
We did the hackathon, we put together a front end for what the v1 of it was: a bunch of questions in a list format with yes/no, and then some probability. It was like an order book. Literally, we just copy-pasted what the New York Stock Exchange order book would look like. It’s funny, because we had judges who were going to judge the different teams and pick the finalists. Our judges were Michael Seibel and Christina from Vanta. I don’t know if she had started Vanta at the time, or she was in the early innings of it.
Jack Altman
When was it?
Tarek Mansour
It was October 2018.
Jack Altman
Yeah, I think she had started it
Tarek Mansour
She had started it?
Jack Altman
I think so.
Tarek Mansour
We started pitching the idea and it was great. I remember Michael was like, “Oh, everything is great about this idea, except for the fact that it’s totally not allowed in the US and this has absolutely no way of existing in any way, shape, or form.” We were like, “Look, we tried, move on.” I remember I drank a bunch of beers at that hackathon. “We’re done.”
And then this guy ends up picking us to be finalists. He just dunks on the whole thing and then we end up winning that hackathon. We were like, “Maybe we’re onto something.” That got us into YC. We were like, “We have to give YC a shot.” At the time I was like, “Wow, I never expected to get into YC.”
The first year was crazy. You know how there’s this thing about YC where the cool companies are building products and getting all these investors excited? We were the total opposite of that. We had no product, no customers. Week to week, we’d go to office hours, and everyone was like, “Here are my KPIs and here’s traction.” We were like…
Jack Altman
“We’ve got nothing.”
Tarek Mansour
Nothing. We were just like, “We talked to this lawyer who said no,” and then the next week was—
Jack Altman
It’s tough. I did YC, and one of the most notable parts of the experience—which I think is a very positive part—is every week you come back with your group and everybody else has grown 7% week over week, and how much have you grown? You’re just like, none. Doesn’t feel good.
Tarek Mansour
None. I don’t even know what the product is yet. But the vision was always clear. For us, the key question was, how do we get it regulated? From the beginning, we made a decision. We’re not going to launch a product outside of the realm of the law or regulation. We wanted to figure out how to get this regulated in the US, onshore, no matter how long it took, no matter what.
The early days were really tough because we had no progress on regulation. Regulation is not linear. It’s not like you get encouraging status updates from regulators.
Jack Altman
There’s big bang moments almost.
Tarek Mansour
Exactly. It’s zero all the way up until approval, nothing in between.
The Regulatory Desert
Jack Altman
Can you talk about what those were for you?
Tarek Mansour
So that was 2018. We spent the first two years essentially figuring out which lawyer would take this on. I became somewhat of an expert in the law around commodities, which is where I thought this would get regulated. Essentially, there was no proof in the law of why this shouldn’t exist, but there were a lot of things to figure out.
Jack Altman
What was the law at the time? What was not allowed?
Tarek Mansour
It says that a commodity could be an occurrence or a contingency.
Jack Altman
So was the issue with the definition of the commodity?
Tarek Mansour
Yeah. The issue is, we’re used to futures like grain, things that are tangible.
Jack Altman
Right. An election outcome is hard to put a thing around.
Tarek Mansour
Imagine two 21-year-old, 22-year-old MIT kids just walking into the regulator. We got this lawyer Jeff, who’s ex-CFTC. The regulator is the CFTC. He got us the first meeting with the regulators. Two kids are like, “Here’s our 40-page deck plan of how we could regulate this thing.” They’re looking at us like, “What are you talking about? Here are all the issues. How are you going to police for manipulation? How are you going to list this sheer number of markets?”
Usually in our markets, like the CME or New York Stock Exchange, those exchanges list one or two new things a year, or every few years. You’re talking about listing hundreds. Tens of markets sounded crazy.
So it was a lot of these hurdles, none of which felt impossible. But if you add them all up, it started to look like Mount Everest. The problem is that you start climbing Mount Everest and then somehow…
Jack Altman
You see a higher peak.
Tarek Mansour
Yeah, it keeps going. We had no certainty it was going to end. That was the toughest part. It wasn’t actually the work. The work was regulatory all day.
Jack Altman
Usually you could fix this thing, but it might be like, we haven’t made a dent.
Tarek Mansour
It’s like a desert. You don’t know if it ends, and so psychologically it’s very taxing. Because you’re walking in that desert and you have no idea if this thing is ever going to end. You may actually just die and you just walked for years in this desert.
Jack Altman
So what happened?
Tarek Mansour
We were just so stubborn. In those situations, you’ve got to stop thinking. You have to have no introspection.
Jack Altman
Marc Andreessen.
Tarek Mansour
Yeah, he got a little flame for that one. But it’s more like you have to have a bit of tunnel vision. We are going to keep going until we’re proven wrong, i.e. we die. Or we find the other side.
Jack Altman
Can you talk about finding the other side?
Tarek Mansour
Towards the end of 2020, we started seeing… They would send an issue to us, one after the other, and then it started fizzling out. These guys started to be like, “Wait, maybe these people are actually serious.” At some point you run out of issues to find. We walked through all of them and worked through all of them. Think of thousands and thousands of legal documents and pages, et cetera. We started angling towards an approval, and we got approved in November 2020 to get the first regulated exchange for prediction markets.
On the same day as our approval, the election happened. The new administration came around. Our approval was bipartisan, it was Democrat and Republican, but the new administration was like, “Wait, pause, pause. We’re going to have to think through this.” In some ways we were like, “Oh, we are finally through the desert.” But all of a sudden, “Oh shit, we got dropped back into the middle of the desert.” It was like, “We’re going to have to see if we can let you do all these things. Maybe we’ll let you do four economic markets, but not this broader vision.”
That was disheartening. It was really hard because we had spent two years, we were finally there, we were going to launch. So we started the first battle to launch the exchange, and we were like, “Fine, you know what? We’ll launch with the four economic markets.” We never thought that would get product-market fit, which it didn’t. But we were like, “We’ve got to get out the door now and launch and see what happens.” By the end of 2021, we launched with a few economic markets, no traction whatsoever. At the end of 2021 is when we were like, “Okay, we have to open up the space to get all the markets we want.” This is when we started talking about the election market.
For prediction markets—we can talk about the dynamics— we always thought that you need the diversity of markets, but you also need a catalyst. You need something that is enough of a driving force to get people noticing, so that you can break through the supply-and-demand problem. You need something strong enough to get the chemical reaction going. Because you have the supply-and-demand problem, the chicken and the egg. The chicken would come and there’s no egg, and vice versa. You need something strong enough to get everyone at the same time to start trading, and then after that the thing can get going. But also, I think it was one of the best ways to explain to people why prediction markets are powerful.
You need an event that everyone cares about and where we can provide a better product, a better forecast. At the end of 2021, we start talking to the CFTC, and they say, “Maybe, et cetera. Maybe we’ll do it by the end of 2022, or for the midterms.” A whole year of just regulation again. So now we’re three to four years in, just doing regulation. By the end of 2022, the regulator sort of nudges the approval past the deadline. Essentially they just didn’t make a decision.
That was a very hard time at the company, because we thought the approval was going to come. We’d done all the work possible, very tunnel-vision. Again, we didn’t get it. In those circumstances, what happens is people blame the execution of the strategy. It’s never that things are outside of your control. It’s like, “We made the wrong decisions, and it was bad.” We lost a bunch of the team at the time, we had to do some layoffs.
It was a really, really hard time. I think back at that time, it was one of the most painful times I’ve ever experienced in my life. By the way, I went through war in Lebanon. I’ve had missiles drop next to my house, things like that. It doesn’t compare. I think there’s some form… because you feel shame as an entrepreneur. You get it. That feeling of—
Jack Altman
Yeah, no, I’ve had to do layoffs. It’s awful.
Suing the CFTC
Tarek Mansour
You have responsibility, and people are trusting you with all this. We come out of this in January, February, we’re reconvening, what do we do as a company, et cetera. I remember Luana has this sort of dogmatic belief in the vision. Should we pivot? Clearly we’re not going to be able to do this. Luana was like, “We’re going to try again.”
So the strategy for 2023 was, we’re going to do the exact same thing as what we did in 2022, and we’re going to try again. It was unpopular, but we did it, a whole other year of the same thing, talking to policymakers, regulators, the same regulator, et cetera. Then they block the election market at the end of 2023. The same thing happens again. This was the point…
Sequoia likes to call these crucible moments, but I think it was the key decision that got the company to where it is today. We sat down and we were like, “What do we do from here?” Again, Luana was driving a lot of this, but it was like, “Look, we strongly believe we’re right on the law. We also strongly believe this thing should exist. We have come so far. We’re now five years in. We’ve just got to sue the government, and we’ve got to sue our own regulator.
We took this to the board, talked to Alfred. The interesting thing that came out of that conversation is that it’s definitely an anti-pattern for a company to sue its own regulator, or the government in general. It’s even more so for a company of 20 people that has no real product. We’re a nobody company at the time.
Jack Altman
Is it an anti-pattern? I think a lot of great consumer companies have gotten into… I don’t know if it’s a full dispute with their own regulator or a full lawsuit, but at least some legal battles.
Tarek Mansour
Maybe it’s the sequencing.
Jack Altman
Yeah, maybe they got big.
Tarek Mansour
Airbnb and Uber were really big. We’re talking about a platform that has—
Jack Altman
Coinbase, maybe.
Tarek Mansour
Coinbase got really big, right? We’re talking about a platform that—
Jack Altman
Yeah. It was tiny.
Tarek Mansour
We had hundreds, maybe thousands of users a week. It was the evangelists, the really early adopters. In terms of, this is the unlock that will get us going…
I remember Alfred was saying, “Even if we win, even in the offshoot crazy shot that we win, we may still lose because the regulator could kill you in the meantime.” It’s like death by a thousand paper cuts, and it was real. This notion of lawfare, people just coming after you for all these unrelated things, but you know it’s because you sued. You sued your own government and a lawsuit is a big battle, of course. It’s a big deal.
But then I remember Alfred saying, “But sometimes some of the best companies I’ve ever seen start with an anti-pattern. There’s something weird that happens in that company that is unusual. And maybe this is yours.” So we decide to sue. All the bad things that were predicted happen, all of them. All the little things, like, “Oh, we’re not going to let you do this. We’re going to delay this. We’re going to kill you on this.” The audit that was supposed to be two weeks is now like 18 months.
Jack Altman
Oh my God.
Tarek Mansour
It’s nonstop, just knife after knife. But the most important thing is we won, in October 2024.
Jack Altman
How long did that suit take?
Tarek Mansour
A year.
Jack Altman
A year. During that year, you’re just stressed out of your mind?
Tarek Mansour
Yes, but not more than the other ones, because at that point it’s—
Jack Altman
“It’s my last shot. What else am I going to do?”
Tarek Mansour
Walking the desert had become our life, and it was our last shot. It was a bit of desperation. Your back is against the wall.
Jack Altman
“If I don’t do this, I’m not going to make it anyway, so who cares?”
Tarek Mansour
Basically. I don’t think we thought there was any shot at making… We had to get the election market. At that point it was more than just business, it was a mission thing. We wanted to deliver this to the world. We were so dogmatic about it. We wanted the world to see this market in action and see how powerful this thing can be.
Jack Altman
Okay, so you win the lawsuit. Now we’re ready.
Tarek Mansour
Now it’s interesting because we won the lawsuit.
Defining a Financial Market
Jack Altman
Specifically what the lawsuit enabled is what?
Tarek Mansour
The lawsuit was, in some ways, redefining what constitutes gaming or gambling versus a financial market. It’s interesting because there was a lawsuit prior to that—
Jack Altman
So basically it’s now saying things like betting on the president or the outcome of a sports game…
Tarek Mansour
Can be a financial market. There are two elements to that. Very simply put, one is, what is the structure? Are you an open and free market where people are just trading against each other? Versus a house where you’re accepting bets from someone and your business model is gambling, your revenue’s equal to customer losses. The second thing is, is this a real thing that’s happening in the world, where some people may benefit from hedging?
Jack Altman
Basically, by being the marketplace where other people are betting against each other, that’s critical in not being gambling.
Tarek Mansour
That’s a very critical thing societally, but also legally. But also, very simply, there’s a difference between two people transacting on, “Hey, what is this dice going to land on?”—because that’s an artificial risk that you are creating just for the purpose of trading or betting on it—versus if you’re trading on a stock, which exists, a company exists, or oil, or an election.
Jack Altman
Got it. So it’s also the event existing, whether or not—
Tarek Mansour
It’s a natural thing. It’s a natural event, not an artificial event. That’s very important. It’s interesting because the decision that came out of that lawsuit is very similar to one that came out close to 120 years earlier, in 1905, in the Supreme Court, the one that legalized grain futures, the most boring financial market, the OG hedging market.
At the time there was a state versus federal fight, where the states were claiming, “This is gambling, because some people are speculating.” Farmers were going and betting on the price of grain, so that’s gambling. It must be gambling. The Supreme Court said, “Look, a lot of people will speculate. But there are some people that are using it for hedging or getting smarter about the price of grain over time.” That’s why it’s a financial market, and actually, in many ways, the speculation is necessary for that market to exist.
If you want a market—if you want the stock market to exist, if you want commodity markets to exist, if you want prediction markets to exist—you need speculation. You cannot just have people insuring themselves against stuff, because the person on the other side needs to be a speculator. In some ways it’s kind of like history repeating itself, but it redefined the aperture of what’s allowed.
Jack Altman
It would be legal for somebody to speculate on whether or not we were going to say a certain word in this podcast? Would that fit the definition, if people wanted to trade on that?
Tarek Mansour
Yeah and it’s happening.
Jack Altman
It’s happening anyway.
Tarek Mansour
It’s happening anyway. They can bet with each other, and they’re trading against each other. It would not be legal for us to and we can talk about that. If we had a position, we cannot go and trade on that, and then you just go say something. That would be market manipulation.
Insider Trading and Fairness
Jack Altman
Is it illegal to bet on something where you see a bet playing out in the world, but you know the answer for sure?
Tarek Mansour
It depends. That’s the whole conversation we’ve been having around insider trading. There’s a whole long history here. We are a regulated financial market. We’re a regulated exchange and clearinghouse. A lot of the rules we have are mimicked after the rules of the stock market.
In the stock market, the line is drawn: you cannot trade on material non-public information. The way that’s defined is, you have a piece of information that you acquired under certain rules. One of those rules is that you cannot disclose it. So material non-public information is information you’re not allowed to disclose.
If you, as an executive of Tesla, went and said it to the press, you would get in trouble. You’re not allowed to do it. Trading is a form of disclosure. That’s the whole point of prediction markets. Prediction markets are a way to disclose information. When you trade, maybe you’re not saying it on Twitter, but you’re actually trading that information, and you’re moving the price in a way that discloses the information.
Jack Altman
Let’s say this morning I saw a trade that was, “Will Jack and Tarek see each other today?” I’m like, “I know the answer to that, so I’m going to make a big bet.”
Tarek Mansour
Are you Jack, or are you someone else?
Jack Altman
I’m me.
Tarek Mansour
You have influence on that. You have direct control over that. So that wouldn’t be insider trading, that would be manipulation.
The majority of participants in grain futures are grain farmers. The way that the price of grain futures is done… This is going to be a little surprising. Can you guess how they determine the price of grain?
Jack Altman
I don’t know. How?
Tarek Mansour
You literally survey a bunch of the farmers that are trading in the market. It’s a little weird, right? They definitely have inside information then, because they’re the guys and they know what the price is. It’s interesting, because the line there has been, “Okay, insider trading is very hard to define in grain futures. But where we’re going to draw the line is, you cannot put a position and then artificially manipulate the underlying price. You cannot move the event or the underlying in a way that will help your profit.”
So it’s the same thing here. If you have direct control over an event—if you’re a politician that was looking to pass a bill, you take a position, then you tank the bill or you try to pass it even harder—that’s illegal. That’s banned on Kalshi.
Jack Altman
I could see that it could get a little gray, though. Back to us talking on the podcast today, let’s say you’re a friend of mine, and you knew it was happening. You don’t really have control over it. But you could be like—
Tarek Mansour
It’s the same as the stock market. Is the cousin responsible? If someone on the street heard an executive talk about some MNPI, are they allowed to trade it or not? These lines have always been—
Jack Altman
Does this stuff come up a lot or not?
Tarek Mansour
Yeah, definitely. Look, this is a conversation that Kalshi alone will not have all the responses to. It’s a conversation we’re having with regulators and, over time, policymakers. But my principles are generally simple when it comes to this. I always go back to, why is insider trading banned on the stock market? There are some people that would argue, maybe we should let insider trading happen in the stock market because it would make the prices more accurate.
Jack Altman
Yeah. My reaction to it is, it feels unfair.
Tarek Mansour
Exactly.
Jack Altman
People got money for no reason other than they were told something, and that shouldn’t be a source of people generating wealth.
Tarek Mansour
Exactly. It’s all about unfairness.
Jack Altman
It feels like there was no skill and no effort that went into it.
Tarek Mansour
Yes, and it’s actually even worse than that. There’s a very practical implication, not just a moral implication, to this. The practical implication is that if people believe the stock market is rigged, they stop trading. The liquidity dries up. There’s this nice property of markets. Markets that have a lot of insider trading, in some ways, are not going to exist.
Jack Altman
At some point, at the limit, you’d be like, “If I don’t have insider information on a stock, why am I trading? I’m definitely the sucker.”
Tarek Mansour
There’s a mix. I think you have to have information on something. Seeking information is very good, doing research. For example, I don’t know if you’ve ever heard this, apparently Two Sigma used to—I don’t know if it’s Two Sigma specifically—use satellite images of the parking lot at Walmart to figure out how many cars were coming in.
Jack Altman
Probably even before that, people would just hire somebody to sit outside of Walmart and just see what the foot traffic is.
Tarek Mansour
That’s not insider information. Exactly. That’s information.
Jack Altman
That’s just work.
Tarek Mansour
That’s work. Exactly. But I think there’s a balance. To me, insider information, the way to define it is information that you could not get access to with work, if you’re not an insider.
Jack Altman
That’s right.
Tarek Mansour
The reason you should ban it—and the rules that you have to build in the marketplace—is, how do we keep it fair? It’s very practical. If it’s not fair, then people will stop participating. The message I always say is that there’s something nice about insider trading, which is that if the marketplace is really not fair, you can trust that people will just stop doing it. They won’t want to do it. That’s why we take this very seriously.
The Steelman Against Prediction Markets
Jack Altman
I have a topic I wanted to get your take on, because I can tell that you’re extremely thoughtful about fairness and the way it should work. What’s the better future? One of the discomforts people have with prediction markets is that they look like gambling. Some people are not comfortable with them.
It’s this new idea, and it looks like people just pulling their phone out and starting to gamble. Obviously that is not at all what your conception of it is. There’s all sorts of functions outside of that. But I’d be curious to hear your steelman version of, what’s the way this goes wrong? What’s the version of it that is—
Tarek Mansour
The bad version?
Jack Altman
Yeah. What’s the bad version of gambling that you’re not comfortable with? That you think crosses some line that you don’t think is good. Obviously, I think you and I share a baseline libertarian value that adults should be able to do what they want to do, obviously with some balance of we shouldn’t expose people to things that are short-term addicting and long-term bad for them. There’s some balance here. Where are you not happy?
Tarek Mansour
Look, I’ll tell you. I’m a risk taker. I’m a trader. I don’t consider myself having really gambled ever. I trade, I speculate. There’s a lot of similarities between these. There are a lot of arguments. The argument I hear most is… People talk about ducks a lot, like it “quacks like a duck.” That’s usually the argument. It looks like gambling, so maybe it is gambling.
It’s interesting, because the thing I always say is that that argument has been made about every single new type of financial market that has ever come to the US, or really anywhere. The argument has been made about grain futures, like we just discussed. When we started with life insurance back in the day, the headlines at the time were like, “Oh, this is morally horrible. You’re gambling on people’s lives. This is terrible. We should not have this at all.” Now I think a lot of us would agree we should have grain futures. We should have the stock market. We should have life insurance. We should have all these different things.
There is a basic… Yes, speculation has a flavor. It sometimes looks like gambling, but it doesn’t make it as such. I like this frame of, let’s actually play this out. How does it go wrong? In my opinion, the thing that ends up contributing to this bad perception in gambling is the incentive structure in the system. How is the system built, and what are the incentives built into the system?
When you think about a gambling business model, it’s a business model where the primary KPI—the thing that will not just predict your net income but will be pretty much equal to your net income—is your customer’s losses. If that’s your business model, that’s what your incentive is, what are you going to do? You’re going to promote—
Jack Altman
Losses.
Tarek Mansour
Losses. What else are you going to do? If you do a great job at stopping losses, you’re going to lose money.
Jack Altman
More throughput, bigger rake.
Tarek Mansour
Yes. It’s just inevitable. What a lot of these businesses do… If you’re sitting at a casino and you’re making money, what do they do? The bodyguard comes and takes you aside and says, stop. If you’re doing something informed, if you’re doing something smart, if you’re seeking information—the very point of financial markets—you get blocked. You get banned, or at least limited.
Jack Altman
Part of that happens because they’re trading against you. So in a casino—
Tarek Mansour
That’s the business model. Yes.
Jack Altman
Blackjack’s against the house. So you winning is exactly them losing.
Tarek Mansour
Exactly. Online casino, all these models, the counterparty is the house.
Jack Altman
Versus for you in a marketplace, you don’t care who—
Tarek Mansour
I don’t care. That’s fundamentally different. The way that it goes bad is when the house is trading against its own customers. Inevitably the algorithms—
Jack Altman
One thing that does persist though… Theoretically you don’t care if on net the two of them leave the day with more or less money between the two of them. You don’t care that two people are betting against each other.
Tarek Mansour
I’m neutral.
Jack Altman
You don’t care. Exactly. If they both start with $100, you’re okay. If at the end of the day one has $110 and one has $80, that’s still okay.
Tarek Mansour
I actually in some ways prefer if they’re both at $100. To me, it’s the incentive in the system. Going back to how it goes bad, if you’re the house, what are you going to do? You’re going to build algorithms. Whether it’s a physical algorithm… The casino has all these smart ways of figuring out who the big losers in the room are. You’re going to figure out how to get them hooked, get them to come back even when they’re losing. They know that they’re losing. You give them a suite, give them all these different things.
Jack Altman
Make the lights more flashy.
Tarek Mansour
All of that. The waitress comes in with a cocktail. That is where the unhealthy behaviors emerge. You have an algorithm that’s just promoting unhealthy behaviors. We’ve seen it in a bunch of other contexts, even tech platforms, et cetera. What I like about a free and open marketplace—and that applies to the stock market, it applies to crypto, it applies to options—is that that dynamic doesn’t exist.
What is my incentive as a company? My company takes a small fee, a transaction fee. So what I want is volume. I want people to just trade more. I’m actually incentivized. If I want people to trade more, I want the platform to be fair, and perceived to be fair. I want the platform to be neutral, as neutral as possible. I want it to be transparent, which is another key thing. All the trades are public, everyone can see what they do. I like this model significantly more, because when we think about what our responsibility is as a platform, I have a much better shot structurally at creating healthy feedback loops into the product.
We do this a lot. We have limits on how much people do, how much they trade, et cetera. But it’s not just something that we just say. Our business model is tied to it. If someone is engaging in excessive behavior and losing too fast, we as a business will not be hurt as much as the other types of business models if we tell this person, “Hey, maybe you should pause.” Now, I don’t know if it’s our job to block this person, that’s a different story. But all the incentives would say, “Hey, maybe you should stop. Maybe you should self-exclude. Maybe you should put limits on yourself.” Because again, they’re not losing it to us. They’re losing it to someone else, and that’s not a positive thing for me if that’s happening.
That’s what’s exciting me. I hope over time, not just Kalshi… Kalshi spends a lot of time thinking about customer protection in the context of how we limit unhealthy or excessive behaviors. I hope this gets applied to all the other financial markets where retail participation is going higher, whether it’s crypto, options, all these different things.
Jack Altman
Let’s say that what people were trading was just like stocks, instead of the outcome of an election or something like that. In that world, even with a fee, over 10 years of trading the stock tends to go up. Versus with an election, it’s just a trade back and forth with a little…
Tarek Mansour
Yes.
Jack Altman
So the outcome that you’re producing is not incrementally more net worth over time. The outcome you’re producing is better information and people being able to express their views.
Tarek Mansour
Yes.
Jack Altman
It does seem to me like it would be cool if you also had the other type of product where you were helping people—
Tarek Mansour
Like the investment-type product? There’s a difference between investing and trading, and that has always existed. Holding a stock for five years, that’s investing. If you trade a stock in and out over the next few days, that’s trading.
Jack Altman
You’re going to get crushed by all the fees.
Tarek Mansour
Not just fees, also your directional view.
Jack Altman
It’s not enough time.
Tarek Mansour
Yeah, you’re trading. That’s a zero-sum game. Options are a zero-sum game. Crypto, in my opinion… We’ll see over time, but most of crypto trading… If you hold Bitcoin for five years, that’s investing. But if you’re trading Bitcoin in and out, you’re zero-sum.
My mental model for this is, yes that’s true. But it’s interesting, because we ask a lot of our customers, “Hey, do you trade?” Not invest. They’re different. “Do you trade, for example, S&P, or do you trade traditional options?” Consistently, for nine out of ten of our customers, their response is no. The reason is, “I don’t gamble.”
In people’s minds, elections trading or betting sounds more like gambling than trading options, because that sounds financial. But actually, if you ask people, it’s “I don’t have a way to win an option. I don’t have an edge. There is no way for me to do it.”
Jack Altman
Those markets are so efficient.
Tarek Mansour
They’re efficient. The hedge funds have way more information than I do. There isn’t a way for me to be truly informed. If I put in a lot more research. So the whole point is, if you put in a lot of research, can you get better and can you win? The reality is, in a lot of traditional markets, the answer is no. Wall Street will always beat Main Street. Wall Street will always beat above the average person. The beauty of what we’re building is that it’s just not the case. The average person is winning more than Wall Street. Our best inflation forecaster is not a Wall Street person.
Jack Altman
I guess by definition the average person is neutral with you, because there’s a buyer and a seller of everything.
Tarek Mansour
But it’s more the point that there isn’t that structural advantage that Wall Street has in our markets.
Jack Altman
What I would argue is they might be neutral with you, and they’re definitely going to be negative if they try to trade against Citadel or something like that.
Tarek Mansour
Generally, yes. Our average user is neutral, but I’m talking more about—
Jack Altman
The people who want to put in real work.
Tarek Mansour
Yes. If you put in real work in figuring out how people vote on bills, or… Back in the 2024 election, there’s the guy who put a lot of money on Trump because he did the neighbor poll. It’s amazing. That’s the markets working exactly how they should. They’re rewarding someone going out there doing the research and doing the truth-seeking on behalf of society. You get rewarded for it. You’re doing a reward mechanism for someone to do research, which does not exist in a lot of the traditional markets.
That’s why when you talk to these people that are on Kalshi… I don’t know if you saw the New York Times article about the rise of the prediction market trader, that class of people that are doing this as a full-time job. They’re excited about this because it’s a way for them to get rewarded for all the things they are learning about the world.
Hedging and Institutional Adoption
Jack Altman
By the way, one of the things that I think is very interesting is whenever there’s a new financial product, there are all these emergent behaviors and properties. An example with yours is insurance and hedging and things like that. When I first learned about that, I was like, “Oh, that’s surprising, but it makes sense with a hurricane or something like that.” It’s getting used for those types of things too, right?
Tarek Mansour
Yes. The trajectory over time is that this is becoming an increasingly bigger part of the platform. Obviously we started with retail—people, individuals—but now as we’re getting into the institutional side, that’s becoming a bigger and bigger piece.
Let me talk about retail, then let’s talk about institutions. There are two functions of the market. One is what we call price discovery, which is predicting all these events. That’s one of the benefits of prediction markets. You are giving people an incentive to do the price discovery, which is to predict all these events. That’s working. I think a lot of people now at least understand it increasingly more. I don’t know if you saw the Fed paper that came out, you saw that?
Jack Altman
Yeah. It’s cool.
Tarek Mansour
“The Rise of Macro Markets”. The Fed itself is saying this is the best gauge we have on the economy.
Jack Altman
It’s crazy.
Tarek Mansour
It’s amazing. By the way, it’s not Wall Street. Again, it’s Main Street. We’ve figured out how to build this community of people dispersed across America that are making us smarter about the economy.
Jack Altman
It’s like, it turns out that if you ask a big enough crowd of people how much a particular cow weighs, they get really close.
Tarek Mansour
That’s the OG original prediction market.
Jack Altman
Yeah. It’s pretty cool.
Tarek Mansour
That’s how it started. Literally, you bring in the wisdom of the crowd.
Jack Altman
This happened with the elections too, with Trump and stuff like that. Everybody’s like, “No, Trump won’t possibly win” and maybe…
Tarek Mansour
If you have an incentive to actually do the research, I think you may actually do it.
The second prong is hedging. Hedging is a little different from insurance. Insurance is usually regulated at the state level because there’s also a house. So you go to an insurance company and they give you a price. Hedging is on the open market.
Jack Altman
So hedging is just, “I’m on a coast and I’m going to bet that a hurricane is going to come and knock my house over.”
Tarek Mansour
Yes. But the key thing is it’s an open and competitive market. You say, I want to buy X amount of something that protects me, and then people can fill you at whatever price and they compete for that price. We see this a lot. For example, in Florida in the Keys, insurance companies have pulled out because they don’t know how to price hurricane risk anymore. It’s really expensive. So we get calls—
Jack Altman
As of when?
Tarek Mansour
It’s been like two, three years. We get a lot of calls around hurricane season where people are like, “Hey, I want to buy X amount of hurricane hitting this town. I don’t want to deal with the insurance process. Oftentimes they don’t pay me back. There’s deductibles, there’s all this. I just want it so that if the hurricane hits the town, I get paid.” That’s a hedge. That’s one really clear use case that we’ve seen. The other one is the time with Biden and the forgiveness market. A lot of students were hedging, smoothing out their student loans. They were super worried about having to pay back.
But the interesting thing about the vision long term is we’re now seeing a massive acceleration in institutional adoption of prediction markets and Kalshi. Think about it this way. You own S&P as an institution, but you’re really worried about an upcoming election. You’re really worried about the midterms one way or the other. If the Republicans win, or Democrats win, you think it’s going to impact your portfolio in a certain way. Today, you don’t have any options. You may just have to sell your position before the event happens.
If you want to protect yourself with prediction markets, you can actually put on the hedge. So if you’re worried about, for example, Republicans winning or Democrats winning, you can buy Republican or Democrats. If you think that’s going to impact your portfolio, your hedge is going to basically compensate for that. You don’t have to sell your position anymore, you can put on the hedge. That’s where the next generation is: AI, COVID, elections, bills passing, regulatory changes. All these different things can become insurable risks. What Kalshi has provided for that is the layer one of being able to think about those, research them, and price them.
Any of these risks now we can put on the platform. We basically have all the top super forecasters on the planet that will give you the price. You send this thing to the system and it’ll come back and spit out, “Hey, will the Citrini scenario happen?” I don’t know if you read that. Have you seen the research report from Citrini?
Jack Altman
Oh, Citrini, yes.
Tarek Mansour
Yeah. We put it on Kalshi and it started at 11%, now it’s at 33%. This thing has gone up. It’s amazing because you have a market that you can probe now instead of listening to different pundits and people battling on Twitter.
Jack Altman
And then based on that, if you believe that, then you can go do other trades against that, I suppose.
Infinite Markets
Tarek Mansour
You can either use that to hedge if you’re worried about that impacting your portfolio. Or the interesting thing is, pricing this thing can enable us as a society to make the prices of all of our traditional assets better. Now, there is this theory about infinite markets. Have you ever heard about that?
Jack Altman
No.
Tarek Mansour
The idea is that as society gets increasingly more complex, the number of dimensions that matter for asset prices increases. A hundred years ago, you maybe needed to understand supply and demand, industrial labor, and maybe the agricultural economy in the US. Now you have to understand those things, and you also have to understand what’s happening in Iran, what’s happening with COVID, AI and what it’s going to do, how technology is rapidly progressing, cyber, all these different things. Society is just getting increasingly more complicated and more interconnected. So everything impacts everything.
The interesting thing is, as the number of dimensions increases, our pricing of traditional assets like the S&P or home prices, the entropy there goes up. We have less and less of the relevant information. So you have to, over time, price all these different dimensions so that you can then price the S&P more accurately.
One of those dimensions, for example, is this Citrini report, what will happen with AI, what will happen with COVID. That idea of infinite markets is very tied to prediction markets, because prediction markets fill that gap. Prediction markets can actually price all these different things, and if you get smarter about all the sub-components, then you can be smart about the actual component.
If you want to price Tesla stock, you have to price whether Elon is going to leave, whether they’re going to over- or underdeliver on deliveries, how fast autonomous vehicles are going to come around, all these different questions. Prediction markets can price all these different factors that then feed into the stock price. It’s very important for us to keep being smart about resource allocation over time. Otherwise, our model of the world is going to just get worse, get less smart.
Staying Lean
Jack Altman
As for the last topic I’m interested in, I didn’t realize until chatting with you that your company is small relative to the scale you’re at. You’re not much over a hundred people.
Tarek Mansour
Yeah, we just hit 127 now.
Jack Altman
So how does that work? That is very small relative to what you’ve accomplished.
Tarek Mansour
What’s interesting about it is that we didn’t sit down proactively and write a doc about how we were going to build a small, lean company. It just happened.
Jack Altman
Did you do any particular other things that this was a byproduct of?
Tarek Mansour
Yeah, I think a few things, and I’m not 100% sure. I’m still figuring out why all these other companies are so much bigger. I’m still trying to figure out, am I missing something?
One is, Luana and I work very hard. Very hard. Until today, we really had a chip on our shoulder. We’d like to think we’re the underdogs. What I learned over time is, if you look at the kind of… We’re generally first in, last out of the office, and work on weekends. I think the output per person in the company is heightened because the leader is really on the frontline doing a lot.
Number two, we have a lot of direct reports.
Jack Altman
Like how many?
Tarek Mansour
There’s not really a managerial layer in the company yet.
Jack Altman
It’s like a hundred?
Tarek Mansour
If you ask Luana what maybe 80-85 of the people in the company today are doing, she knows because she probably checked with them on Slack in the last 48 hours. The rest is maybe me. Well, there’s a chunk of people where you don’t need to know what they’re doing. You know that they’re just doing. You’ve got to let them cook. So that’s number two.
Number three is, we don’t think about org charts much. I don’t know how that will scale. We’re still thinking about that. But we keep dynamically listing, “Here are the top X problems of the company today, and who do we have on those problems?”
Jack Altman
And people move between problems? Do people self-organize to the problems?
Tarek Mansour
Yeah. It’s like cells in an organism. You know how, if you get cut, your cells will just come around the cut and do their thing? It’ll be a bit like that.
Jack Altman
I mentioned to you, I really want you to read the Valve Employee Handbook.
Tarek Mansour
Yeah, I’m excited about reading it.
Jack Altman
It’s very good.
Tarek Mansour
But it just sort of happens, and it’s making sure there’s as few constraints or bottlenecks to that sort of self-organization happening as possible.
Jack Altman
Is this what people expected when you hired them? Did you hire types of people that you thought could only function in a place like this? How did you end up with that culture, when it’s what I would describe as extremely uncommon?
Tarek Mansour
We don’t have all the answers, obviously. One is, we do bias on slope versus intercept, because people who have an intercept, I think generally can land in this culture and be like, what on earth is going on? This is crazy. That’s just happened. We’ve had this happen.
Jack Altman
Someone who’s used to a big, more structured organization comes in—
Tarek Mansour
Or not even big, an organization that’s just structured in a different way. They show up and they’re like, this is crazy. This is complete chaos. Because it looks like an organism where these organisms are moving around. Slope is, they don’t know, they’re just super smart, very high agency, they don’t even think about what’s happening. They just think it’s normal. That’s one.
Number two is, I always say Brian Chesky put it better. I didn’t verbalize it until he put it into words, but we don’t manage people, we manage work. People that just generally have high agency, we never have to check on whether they’re doing something. Sometimes we have to reorient a bit, “Hey, this is not actually that useful, we should do something else.” Or be very much in the details of what they’re doing.
Jack Altman
It’s the work, not the people.
Tarek Mansour
They’re just doing stuff. They have this sort of high agency, I want to always be doing something. And I bias honestly toward execution over strategy.
Jack Altman
Me too.
Tarek Mansour
I really do. Look, strategy is hard, but what I’ve found over time is the natural next step for a company is generally natural. You know what I mean? For a public company, for example, the CEO lays out a strategy, it’s—
Jack Altman
It’s not like, what should we do? It’s, can we do it?
Tarek Mansour
Most of the time.
Jack Altman
How quick can we move? And all of that.
Tarek Mansour
Most of the time.
Jack Altman
Periodically there’s probably a non-obvious strategic decision that the founder needs to make. But that’s probably a couple times a year kind of thing.
Tarek Mansour
Yes, max, and it’s usually a little bit longer horizon than a year. So I think of our role, Luana and I, as, at a very high level, are we directionally in the right direction? What are the big risks in the next three to four years? Make sure that we are thinking about those and executing against those. I really mean three to four years. I’m pretty paranoid, again from my time from Lebanon. I always think, “What is the thing that’s going to go wrong in three, four years? Let me work through that.”
And then very much in the details. I literally am often in specific copy in the product. A lot of it Luana and I wrote, still to this day. Or even ads, we get into the copy, is this good? Very specific things. Everything in between, we try to not spend any time on, which pushes other people to not spend any time on it. Sometimes there’s a subset of people that doesn’t work well for.
Jack Altman
And then people can just opt out.
Tarek Mansour
Basically. There’s not a clean solution, building a company is hard. But a lot of people like it. They just don’t want to be managed.
Jack Altman
Are you going to be able to keep the company small even though you’re scaling fast and everything? Small as possible as at least.
Tarek Mansour
I really hope so. That’s one of the worries I have. I think about this a lot. That’s one of the worries we have. I would say maybe the trade-off… People sometimes walk away with, “Oh, this is a perfectly run company.” The answer is, no, not at all. The trade-off is usually, I think we take on more organizational chaos. Does that make sense?
To me, there’s a bit of a decision you make as a company. Either you’re more chaotic, or you have more process, which means bureaucracy and some degree of slowness. Kalshi is very comfortable with putting something out there and getting bashed for it, but three, four weeks later it gets much better. Then you’re much better than if you had waited two months. We’re very comfortable with that.
Jack Altman
That’s great. Tarek, this is super fun. Thanks for your time.
Tarek Mansour
Thanks for having me. This is awesome.