# Kalshi says its fee model avoids sportsbook loss incentives

Markdown copy of https://realalpha.app/s/kalshi-says-its-fee-model-avoids-sportsbook-loss-incentives/ on Real Alpha, the best minutes from long interviews with the people building AI. Cite and link the page, not this copy.

- Speaker: [Tarek Mansour](https://realalpha.app/people/tarek-mansour/) · Tarek Mansour is cofounder and CEO of Kalshi. · Built Kalshi, the federally regulated exchange for event contracts, with co-founder Luana Lopes Lara.
- Host: [Scott Galloway](https://realalpha.app/people/scott-galloway/)
- Host: [Ed Elson](https://realalpha.app/people/ed-elson/)
- Show: Prof G Markets
- Episode: [Prediction Markets vs. Gambling: Where’s the Line? | Prof G Markets](https://realalpha.app/v/J31lDCBRZgg/), Feb 27, 2026
- Moment: 12:29 to 18:18 of the episode (6 min), playing on https://realalpha.app/s/kalshi-says-its-fee-model-avoids-sportsbook-loss-incentives/; the episode from there: https://realalpha.app/v/J31lDCBRZgg/?t=749

## Quote

> “Like when you go to a casino or you go to, you know, a traditional sports book, the revenue of that company is **equal to the customer's losses**.”
> — [Tarek Mansour](https://realalpha.app/people/tarek-mansour/) at [12:48](https://realalpha.app/v/J31lDCBRZgg/?t=768) · Prof G Markets

## Takeaways

1. Sportsbooks earn from customer losses; Kalshi charges transaction fees.
2. Kalshi says user-to-user trading avoids a built-in house incentive.
3. Its safeguards include limits, pauses, self-exclusion and surveillance.

## About this segment

Tarek Mansour contrasts Kalshi’s transaction fees with sportsbooks that earn from customer losses, then describes limits and monitoring. The distinction shows how Kalshi says it can curb harm without blocking trading.

From “Prediction Markets vs. Gambling: Where’s the Line? | Prof G Markets”, Feb 27, 2026. [Watch the full episode](https://realalpha.app/v/J31lDCBRZgg/)

## Transcript

Transcribed by machine (Whisper) from the episode’s audio, so a word may be off here and there.

[12:39](https://realalpha.app/v/J31lDCBRZgg/?t=759) is a much healthier mechanism to engage with something than a lot of these other mechanisms. Like, I think a lot of the problems that we've seen with dopamine hits and addiction A lot of the issues that we, you know, people generally pertain to as gambling have come from the gambling industry. And I think, and you ask why, well, you've got to look at the incentives in that industry. Like when you go to a casino or you go to, you know, a traditional sports book, the revenue of that company is equal to the customer's losses.

[13:05](https://realalpha.app/v/J31lDCBRZgg/?t=785) The way that that whole system, the business model is like, well, if ad comes to me, the business model is figuring out how much money can I take from ad. That's how it works. So if that's your business model, what are you going to do? Well, what you're going to figure out is like, okay, I have Ed and Scott as customers. If Scott is the winner, I'm going to block Scott from participating because anything that Scott wins, it's going to take from me as a business. And if Ed is a loser, I'm going to figure out how to create a habit, how to get Ed hooked in the platform and get them to come back pretty consistently.

[13:37](https://realalpha.app/v/J31lDCBRZgg/?t=817) And that to me is a lot of where, like kind of a lot of where, it's a perverse incentive. And, you know, show me the incentive. I can show you the outcome. Well, what the outcome is going to be is basically you're going to get people hooked and dopamine hits are going to basically be, you know, increasingly a bigger part of the platform and so on and so forth. Prediction markets just don't have the inherent incentives.

[13:53](https://realalpha.app/v/J31lDCBRZgg/?t=833) You know, I take a small fee and Ad is not trading against me. Ad is trading against Scott. It's inherently just more social. It's inherently more, like I think the model doesn't have this sort of embedded perverse incentive. And that enables me as a business model to just have much more versatility around solving that problem, around creating really good guardrails around that problem. Like, you know, when Ed comes, he's not basically facing an algorithm that's just sort of optimized to get Ed hooked.

[14:17](https://realalpha.app/v/J31lDCBRZgg/?t=857) Ed is facing Scott, is facing you and is figuring out whether Ed can be smarter than you, whether he can do research, how can he beat you? Intrinsically more social, intrinsically more competitive, intrinsically more interesting. In terms of the guardrail, we focus on this a lot. And I think that like a lot of the guardrails are really what you want to prevent is excessive behaviors. I think anything taken to the extreme is bad.

[14:39](https://realalpha.app/v/J31lDCBRZgg/?t=879) And, you know, I talked about some of the trading, like retail trading, some of these examples, but, you know, you see it in other places. You see it with drinking. You see it with online shopping. Like you probably have a, I mean, I'm the type of person, like I have a bunch of shit in my house, like because I have these Instagram ads that keep feeding me all these like random trinkets that I keep sort of like clicking on and buying.

[14:58](https://realalpha.app/v/J31lDCBRZgg/?t=898) And I think that's a form of excess. Like I somehow like find my, you know, I do the thing and then five minutes later, I'm like, why did I just buy that? Like, why, you know? And I do this on Amazon. I do this on Instagram a lot. And all of these things take into the excess tend to be bad. And so how do you, you know, you have to create measures on like, okay, how do you prevent excessive behaviors?

[15:16](https://realalpha.app/v/J31lDCBRZgg/?t=916) You have to do a lot of customer education. We have to. tools around self-exclusions, limits, pauses. And then we also have surveillance where like when we do find someone, especially on the younger side, you know, we do have age gating. So you cannot participate as a minor, obviously. But someone on the younger side, you know, repeatedly losing, repeatedly kind of taking aggressive actions, doing too much of their portfolio, we start showing warning signs.

[15:35](https://realalpha.app/v/J31lDCBRZgg/?t=935) Like, hey, you should not put more than X percent of your portfolio in one thing. Because, you know, again, like long-term, if you want to build a healthy ecosystem, you don't want these sort of behaviors to go off the rails. But then also our incentive is not misaligned with those guardrails. In some ways, it is aligned. We want people to be long-term participants in this model. We want them to be engaged socially.

[15:56](https://realalpha.app/v/J31lDCBRZgg/?t=956) We don't win a lot if they come and lose a lot of their money very fast. That's actually a bad outcome for us. The pushback to that point, though, you make the point, which I agree with, show me the incentive, I'll show you the outcome. The pushback for a lot of people to that would be actually the incentive might not be that consumers need to lose, or that traders need to lose money, but they need to trade.

[16:18](https://realalpha.app/v/J31lDCBRZgg/?t=978) And if you can get them addicted, if you can get them to keep trading and get hooked, ultimately, that actually is a boon for your business. So the guardrails you're describing, they are inherently counter to the incentives that are built into your business, which is, we need people to trade as much as possible, because that is our business model. We're taking a transaction fee on all the trades that happen.

[16:44](https://realalpha.app/v/J31lDCBRZgg/?t=1004) We don't want you to lose money, but we want you to put some money up. What would you say to that? I think there's a presumption in your question that's just assuming that all the trading is basically dopamine-type behavior, right? Which is not the case, right? Actually, again, a majority of power users, the people that are really controlling the most volume, are the people that are making these, are actually necessary.

[17:06](https://realalpha.app/v/J31lDCBRZgg/?t=1026) Their trading is necessary to this, because they're making the forecast more accurate, right? Like the stock market doesn't get efficient if nobody trades. You need the trading and you need the informed trading. You need the well-researched trading. And actually the trading that grows the most over time is the winners, is the people that like are essentially very well-researched. They predict the weather because they're scraping satellite data.

[17:27](https://realalpha.app/v/J31lDCBRZgg/?t=1047) These are people that are building models. They're in the spreadsheets. They're doing math. They're building systems. These are people that are actually rejected from traditional, for example, like sports betting because they're making too much money. They get blocked and limited there, but they can come to prediction markets. It's a good home for that. So, like, yes, you want the trading, but, you know, the dopamine level trading is not necessarily the thing that makes the forecast the most.

[17:48](https://realalpha.app/v/J31lDCBRZgg/?t=1068) accurate, right? Like it's, it's not, you know, it's, you know, it's part, you need that vibrant ecosystem. You need speculators, you need, you know, uh, forecasters, you need hedgers, which we can, which we can talk about, but, um, and then, but you don't necessarily need the excessive behaviors and the excessive behaviors are the things that you can cap over time to build a well-balanced ecosystem. We'll be right back after the break. And if you're enjoying the show so far,