Episode 128: "We're Getting Big, Fast In Prediction Markets"

[Music] Hello and welcome to Zero, Lentz, the podcast from Ilas and Crycheck Gaming. I'm your host, Brad Allen and this episode is brought to you by Optimove, the creator of positionest marketing and the number one player engagement solution for eye gaming and sports betting operators. My guest today is Bernard Marantelli, the founder of White Swan Data among many other companies but we're going to talk mainly about prediction markets. As Bernard said to me the other day, we're getting big quick in prediction markets so I felt like he was the man to share some insight from behind the scenes, from behind the trading streets. Bernard, how are you doing? Thanks for joining me. Very good Brad, thanks for having me. Where should we start? Let's start with your background and how you've got how you've got involved with prediction markets? What are you doing there? Just so people know what we're dealing with. Yeah, well I think my background grew up in a gambling firm as I think people have heard me on podcasts before gambling family. My father was a bookmaker, started Greyhound videos and place markets in my teams. Really first got into gambling was declining. I went off and started a biotech career and came across our bet fair in 2003 or 2004 and it's kind of funny because people say, oh prediction markets they're brand new. We've had them for 23 years in Europe, 24 years in Europe. They broadly failed. We can talk about why they failed. Now it's kind of revisiting a lot of those exuberance of growth of 20 years ago. The moral dilemmas of asymmetric information, etc. Through to what's the potential upside and obviously I think they're significantly different this time around because of regulation and now the movements in society. So I think it's a new world but yeah I'm dusting off software and you know mathematical solutions that I was using 20 years ago that have kind of been in hibernation for a while. So fun times. Yeah. So I've seen your calendar. It seems like you've got meetings with everyone in the prediction market space exchanges. So what are you guys trying to do? What's what's one data trying to do? Well I think you've got to look at our DNA and we started really even before we were whites wanting. We were just a two-man shop. We were trading that fair, matchbook, etc. As those declined we moved into much more exotic type betting and a lot of the what you generally find in life when something is new it's mispriced. The market tends to adapt to that mispricing and fix it and you've got to move on. So I think over time in the last 20 years is we've been sort of an informal and now formal hedge fund. We've had to pick apart aspects of the gaming gambling business that are mispriced and then that tended to become more exotic things like same game parlies where firms that are just trying to price for everybody don't get the correlation right and you pick that apart. So when the prediction market space came in and they started to go after the RFQ or the combo or parli market and we've got the DNA of trading and managing risk. We've got the DNA of pricing exotic parlies. It just made sense to become an RFQ market maker or combo market maker. So look that's what I think Kelshi launched that around last August September. We probably entered in February and you know we're across every RFQ platform now and you know we're you know a fairly significant market maker on all of them now. How many exchanges is that currently? That is well changes daily but I think it's six currently. So we are on Kelshi both Polly's, SX, crypto, profit X, integrating no view I think at the moment. I think they'll probably be about 12 RFQ platforms live for the NFL. We've started the onboarding for Rothera. So yeah there's several of them going forward at the moment. Draftkins etc. So what have you seen from your early trading through being live three or four months? I mean you shared a chart the other day on LinkedIn showing the amount risked on same game parlies just through the World Cup has gone 5X also in literally in the duration of the World Cup so it seems like these these are taking off pretty big. Yeah so I think there's um you know I guess one thing just for the listeners to kind of get right to briefly explain if you're if you're doing a single market on a prediction market I think you frequently trade particularly if you're a liquidity provider you're in and out in and out trying to capture the spread. In the parli market it's almost exclusively retail coming into back a parli and institutional or semi elite traders are putting up the risk capital for those parlies and it's very capital intensive because you fully you have to fully collateralize every every single parli that you write. So it's not quite trading it's it's you need directional risk management it's retail coming to you retail coming to you because the exchange can't be there I mean it's subtle but the exchange can't be a counterparty to one of their clients. I mean a related firm might be able to be etc but generally they're allowing free and open access which is one of the CFT sees instructions for an exchange to operate so you you are one side of retail traffic rather than going in out in out in out. What if we see significant growth if you look at the liability of parlies on the world series between this person the next that oscillated between 90 and 110 million I think game three the first one in New York was 160 million this is an established sport in America with a Texas team in it and Texas don't have sports books so they're all on kelshi and then you see the World Cup final between Argentina and Spain treated almost 400 million of liability so you know you've forexed the World Series games if you forex again for the NFL you're you're talking five billion dollars of liability on an NFL someday across the RFQs and that's not even the bulk of the market because there are structural problems for the RFQ market it's very hard to do long prices on an RFQ it's very hard to do long dated so those two components of the market for the moment probably stay with draft kings and fangil but they will be so they'll be solved on a prediction market basis soon as well yeah you might upset some Americans calling the NBA finals the World Series the open-air which what's the World Series something else is they're baseball yeah yeah what what is the NBA finals called the NBA finals I think right okay yeah so you were saying I'm a sport I'm a sports expert wait I mean it kind of shows you you're going to be deep in the nuts and bolts of the sport to to trade it successfully yes and I was interested in hold specifically so we kind of have this debate in in the Iles and Crycheck slack this week where Mr Grove was saying that the market sports betting market has shown people will pay 20% thick they will pay 25% on SEPs and they'll keep coming back for more so why would the saying game partly hold for market makers on prediction markets why would they trend down down and down and my view was that it probably would given the competition you know the more it seems like everyone is trying to be a market maker nowadays everyone's trying to plug in and obviously the RFG system you only later the parlay the combo if you're the best price so does that trend down to 8% what's your view here where does this settle um I think what you've got to ultimately do when you think about the ecosystem here is you have to look at the cost of recruitment of of the player and draft Kings and Fandall that let's say broadly run these at roughly 20% and they're probably going to trend down marginally on draft Kings and Fandall as well um they've got a high cost of recruitment of players etc etc if these players are coming through you know a super app on Robinhood or similar um you know then maybe the cost of recruitment if they're just tacking in a prediction market to something they're already trading is lower and then the market makers have paid no recruitment costs really so then they risk managing that and Robinhood tells she whoever you want to consider is putting a margin on top who have not paid the CPA of a sports book as well. And then you can see a world where that competition trends down. But I think that world will be in the short term limited by the fact that you have to fully collateralize every transaction. When Draftkins' fan jewel, you know, Ben MGM writer Pali, nobody asks them to prove that they have the capital to pay that Pali. Now Jason famously said in a investor call a few months ago that they wrote something I figured I think it was 1.5 trillion of Pali risk in the prior year. In fact, I think it was 2.5 trillion. Oh, I'll post it to you. Now, I don't have any concerns that these people don't have the capital to pay these out because they've got risk management. They can't all happen. You know, your risk of capitalizing every single negatively correlated Pali, they can't all pay out. But very interestingly in Q4 2025, Fandrew and Draftkins both warned about 400 million of adverse income because all the NFL favourites were winning. So there is significant volatility here, which means no one, whether it's white on data or Sashqahana or Jane Street, if they enter, would price these down to a very low margin because the volatility is significant. And if the retail person is happy to pay it, there's going to be some sort of, you know, ecosystem that maintains a margin. At the moment, I think you're going to see significant growth, you know, exponential is the wrong word to use, but significant growth and significant capital requirement and significant fragmentation of prediction markets and exchanges and clearing houses. So I think while that happens, merchants will be high. And if they solve some of those problems, you'll see some compression. But I don't think you're going to see compression in the next 12 months or not much. Yeah. And so at current rates, what's for parles, if you said, say Fandrew and Draftkins holding 20% or SGP, roughly prediction markets, maybe a holding 15% on combos, do you think lower higher? You've got to look at the structure of what they're laying. So if you do some estimates on our draft Kings who published their liability, it looks like if you make some estimates, if you go out to the American gambling association, get their parlay handle across the year, et cetera, et cetera, it looks like their average parlay average because the odd parlay that they write at 5,000 to one, skews that number a lot. But if you look at their average parlay, it's probably around 100 to one. If you look at the average parlay on a prediction market at the moment, it's more around 10 to one. So I think if you took all the 10 to one parles out of Fandrew and Draftkins, their margin wouldn't be 20%. You know, it's probably mid teens. If you look on the prediction markets at the moment, the margins are running around, they're running around 80 basis points to 100 basis points. So I'd say they're running closer for the market meter closer to 10% but then Kelshia adding their fees on top, which is on a curve structure. So yeah, I'd say retail's probably paying 5% 2 to 3% lower than Draftkins and Fandrew for like price. Yeah. But the RFQs are generally not capturing the high price parlay for the moment because most of them have an order book that only goes to 0.01 cents. So your maximum is a thousand to one. Yeah. And there's a few things that need to be sorted out to help the product get to equality. And so what one of them is the capital requirements because as I understand it, the minute you need to, as you say, collateralize every single parlay you lay, whereas obviously there are competing legs in many. So they can't all win. But at the minute, you have to be able to pay them all. That's absolutely correct. So imagine football, which is a coin and someone does heads, heads, heads. And you bet them 7 million to a million fair value. And someone on the same three coins does heads, heads, tails, you're going to have to post 7 million for the first person, 7 million for the second. You can only lose six because they could both lose or only one of them could win. So your collateral post to your true liability, it's 10 fold. Especially if you've got good risk management, it's 10 fold. If you've got shit management and you just lay the same four fold for 10 million, obviously there is no negative correlation in in that portfolio of risk. Yeah. And you would again, you would assume that we're trying towards getting rid of that because it doesn't make any sense that you would need to be able to pay out all these things which can't literally come in. Or no. I don't know that the regulators ever going to be in a rush to do that because they in really only be left holding the baby, I guess. So I don't know how fast that will occur. And I don't know who it suits. It doesn't suit the bigger market makers. It suits the smaller market makers, but it's a it's a moat for the bigger market makers. Yeah. Yeah. As you say at the minute, there's probably less competition because because of the capital requirement. Yes. Now what? What the prediction markets might ultimately want is netting because if I can net, then I might be inclined to start to capture the 100, 200, 500, 5,000 to one part of the spectrum. And I might be starting to prepare to write an NFL parlay on Tuesday, rather than waiting till Sunday. But for the moment, the long duration and high price is not really getting captured by a prediction market. Learn how opt moves positionless marketing is changing how I gaming teams operate. Discover how operators using opt moves positionless marketing platform to launch personalized serum campaigns, dynamically changed casino lobbies and bed slips and create engaging gamified experiences. Learn more at optmove.com. Let's talk about moat. As you say, one of the, one of the moats for the big prediction markets is capital. But one of the questions we've had recently, either is what's to stop, you know, big finance head hedge funds, people like Citadel coming in using their existing knowledge and trying to turn their hand to sports and then pricing out white swan. How do you guys think about that internally? Do you think that's a risk? I think it's a risk. You know, I kind of repeat that you really need three things to, to survive here. You need fundamental pricing, ultimately, because particularly as margins have any sort of compression, you have to defend your margin by having better pricing than the market. You need pretty serious technology and you need capital. And I think, you know, Citadel or any other big institutional firm, they have technology for sure when they have money, but the technology is not plug and play to a prediction market if you're doing a financial market. It's plug and play for a single. It's a bit more subtle to go to an RFQ and look at all the iterations of that. You're, and then pricing, I think they're going to buy in pricing because they're going to have to go and get the smaller groups or the bigger groups, you know, like it's not impossible that one of them come in and want to take out whites, one data is pricing. You know, we've seen in the past panic technology sold. We saw all shrimp pricing, I think sold as well. So, you know, you know, the US sportsbooks, when they came in, they went and bought some smaller pricing hedge funds than whites, one data. So those sorts of things could happen to, to catch up quickly. Yeah, it's going to happen. You need tech, you need pricing, you need capital. Institutions have most of the tech and capital, but they don't generally have the pricing and the syndicates have the pricing, but sometimes not the capital in the billion dollar capacity. Yeah. No, one of one of your other businesses is called I bankroll, right? So it feels like you must have some access to capital. Yeah, but, you know, the access to capital, we're just talking about entirely different levels because if you want to let a whale have a hundred grand on a blackjack hand, that's too big for most crypto, because you know, starting up, but, you know, your variance is four, five, 10 million in a day or a week. If you've got a five billion dollar NFL market and on a on a Sunday, and, you know, 13 of the dollar 60, 14 of the dollar 65. it's win out of 17, the market is drawing down a billion dollars. It's a different world of risk. There was a day in the World Cup where the biggest bet in the World Cup on almost a daily basis with the superstar scorer followed by the same team to advance. There was a day where in Bapper in France one, the next game was Halland and Norway, the next game was Argentina and Messi and the next game was a Mexican goal scorer in Mexico. They were the four games in a day. And the market drew down about 15% of its risk in one evening. Drawdown 15% of 5 billion on NFL Sunday. That's a proper reporting event even for some of the big boys. Yeah. So as you said earlier, that CFT sent generally requires free and open access for market makers. And I think you said that's generally been true. Is that true across the board? Is there been a struggle trying to connect to any of these? Look, if I'm struggling to connect to someone, I don't want to call them out because it's not going to make the struggle any easier perhaps. I don't need names. I don't need names. I'm just here. There's some exchanges and making it tricky. I think people will find inventive ways to make that happen. Let's be honest. Most people have in-house market makers at the moment. They're arms length and they're this and there that. Look, you've got to worry slightly about where is the cash flow in those circumstances about whether everyone is on equal terms or not. That's for the regulator to manage. We are possibly going to have arrangements that might be mutually beneficial. You know, if white's one data, 10% of a firm and we market make that firm and we market make it aggressively so it becomes unattractive for other people but we've got equity upside in the parent company, you can understand that it's pretty hard to disentangle those two things from each other. Everyone's technically got the same access but someone with access and equity would behave in a different way. So look, the regulator can sort out that. Yeah, good luck to them. Yeah, it sounds like a fun job to untangle because there are exactly those agreements that you've said where market makers have equity in exchanges and maybe they could market make it unsustainably low margins to keep other people out. Yeah, they might be getting rebated, they might be having high fees with rebates and look, I think in some senses that's let's maybe say it's like a victimless crime maybe because you know, the retail person moves if he has a bad price or a bad experience. So I think you can't close your shop and charge 30% because the person moves to an open shop but I think definitely people will try closed or semi-closed shop environments to maintain margins and I think others will seek to be clearing houses for the world. So I think you'll run into different models. Yeah, and it'll be interesting to see what the CFTC allowed and if you are trying to keep it for one or two market makers, they better be good because they're going to be and you know, they might be picked off, they might be going to have to price everything. So there's downsides to that as well. Yeah, this you know, look, there's upsides and downsides to all of those environments but the technology's going to be bad, have to be better, you can't go down, you can't have a problem. But you know, I think the regulator and the operators are going to have to be careful that they don't end up looking like a sportsbook in another regulatory paradigm. You know, if you're in exchange and you only have one market maker, how are you not draft kings? Like of, of, of, and then obviously they would be entitled to kick off and say, well, hang on, these guys are running a sportsbook under parallel regulation. How is that allowed? Why am I in my situation? I've, I've behaved well here. So yeah, it's a, it's an interesting challenge for the regulator. But you know, from an economic perspective, I think the client is ultimately going to switch and move to the place that's a little bit better and a little bit cheaper. Yeah, yeah, yeah, the client wants maximum uptime and lowest cost and ease of use essentially at the end of the day. Yeah, they want the best utility. That doesn't always mean the best price. Yeah, they, you know, they can fleet price and experience and everything together. So they, they'll go to the best utility and price is only one element of that satisfaction. Yeah, yeah, definitely. The lesson of sportsbook in general that the, some, some smart people think price matters more than it does. Couple more for your Bernard, now that you go one, it, it, it sounds to me a bit like you see this period, this 12 months, this two years as a kind of a gold rush, like make money quit while it's new, while the pricing's not so good. Is that fair to say, or you think there's like long lasting sustainable business? No, I think it's very long lasting and sustainable. I think which part of it is sustainable varies. It's interesting because I'm involved in, we do a lot of block trades for insurance type firms. Football teams want to smooth that cash flow for relegation, whether spoons want to smooth that cash flow, they want to buy an option that if England made the semifinal or final, they pay a big premium. But if they get eliminated in the group stages or the round of 16 or quarter finals, they get a payment because no one comes to the pub once England is thrown out of the euros or the World Cup. Restaurants in New York want to sell, want to hedge that it goes to game seven, game six, game five, ticket brokers want to do the same. There is so many economic impacts of sport that if it becomes normalised as a financial instrument to smooth those cash flows and it becomes easy to do on an exchange or on a prediction market, I think it's an asset class that is entirely untapped at the moment. So I think there's substantial growth there. I think there's substantial growth in general that there is significant prospects that prediction markets that currently face 50 states broadly and they don't seem opposed to globally cloning. I think there can be gigantic winners in that space away from the fragmented patchwork of regulations around sports betting. And America is, you know, don't want Donald Trump listening and denying my ASTAR, but you know, America is a basket case of regulations. You've got sweepstakes, casinos, literally being casinos. You've got pick 'em literally being sportsbooks, but only on props. You've got paramutual wagering, you've got historical horse racing, you've got black market casinos, you've got sportsbooks, you've got prediction markets, you've got fantasy betting. Like, you know, it's such a concoction and patchwork that everyone has to go down the super app route to operate on a state-by-state basis, those subcomponents. So I don't know, you know, but all of that together, I think can become much more formalized in a prediction market and it can be very global and it can become a clearinghouse for massive cash flow requirements that resolve around sports. Hotels matter for sports. If the Spurs get into the playoff or whatever, you know, Mary had a selling 20,000 more hotel rooms in Houston or wherever they're from, you know, they want to hedge that. Sponsors want to hedge TV rights, promotion, relegation. And if you look at the value of sports teams generally going through the roof and the use and the integration of sports outcomes in society, they're all hedging opportunities. Not to mention all the politics and financial elements of prediction markets as well, which is not really my area. So I see a huge upside there to basically make everything a trading pass to be honest. Yeah, yeah. And then finally, talk me through this competition you're running. So four million up for grabs for two trading teams who can prove they're good at trading on Calcium. it talks through this. Yeah, so look, I think we go back to a bit of the earlier discussion and we say, how do the institutional people get into prediction markets? The one thing they're missing is price. The fundamental sports price. And I think the one thing I've learned in life, even if you're ultra successful running a sports betting fund and the risk manager, etc. Every incremental piece of information adds value to your fund. So we were looking at how do we find all of the small shops around the world that we would be good to collaborate with. They're hard to find and it's a fake business. Some is too low sitting in their underwear at home. Some is 10 people. We work in Houston or something. So there's a whole spectrum of, you've got some Tony blooms and some jump trading and other opaque individuals that are at the top there and whites want data. And then you've got a whole plethora and long tail down to, I don't want to put you at the end of the tail, but I know you're pretty good at NFL trading and NFL betting. You've got people in America that specialize in golf. So we were looking at that and saying, what's the problem of all of these guys? One of their problems is they don't have capital, but they have great model. You've got Bill and Ted, they've ripe code of fraud. A lot of them are going to lose half a million dollars and close down, but some of them are going to grow into serious businesses. So our job is to look at all of those, assess them in some dragon then type thing and invest in the ones that we think are good. They don't have capital, they don't have technology, but they might have pricing. So they partner with our capital and our technology and we're basically going to build some sort of fund of funds where we all collaborate together to do better. You know, these markets are so big. I don't see a problem. I'm probably going to run this competition four or five times a year and just use it to find people that want to collaborate and give them significant capital. And you see that a lot, you know, there's a firm called fans are they're running fund of funds now. Millennials been doing it for a long time. So they make capital available to people with ideas. They get not survivor buys, but they get a lot of positive selection in that process. And they you get hold of people when capital is very important to them. And if you're good at selecting the right from the wrong, it's basically like seed or venture capital, but just on a very specific asset class trading sports. All right. So if you're listening and you reckon you have a trading strategy on Calcium prediction markets, that could be scaled up and can make a lot of money. Then the NARD has three million dollars for first place for you to trade with and you will want a cut of that when you make your millions. Right. But thank you very much for your time. That was very insightful. I feel like I've just done a sort of lecture in market making on prediction markets to our listeners. We are covering prediction markets, a lot more in depth at Arles and Critech. We recently launched a prediction market monitor every month. And there is a lot of this sort of thing, a lot of charts, a lot behind the scenes on what's really going on at these companies. So drop me a line on LinkedIn and I'll put you in the direction of that. I will say a lot of that data that comes out of them, I can never pronounce it well, as you can see, it's really good stuff that they're putting out into LinkedIn and their reports as well. And the thing is, there's so much data available at the moment. It's extremely hard to keep up. Some of the prediction markets are on chain where you've got complete visibility, some are more opaque. Literally, every day, someone who's sending to Calcium is going somewhere else, a new person sending to Kelsey, you know, Sesh Hohana announced they're doing 500 million investment. The news flow and the data flow here and the data is the really interesting thing. So much data is visible on prediction markets that is not visible in sportsbooks. It's really a fun area to work in and solve and assess. Yeah, there's a lot going on. A lot of highs at the minute, a lot of people trying to build out these teams and a lot of action. Alright, thank you everyone for listening. Chaps you soon.