Saluki Posted February 10 Posted February 10 I get quite a few consulting calls about this topic so I figured it would be interesting to start a post on it here. Anyone use them and have opinions on it? Do you use it for sports betting or financial? What do you think of the other users? Legitimate investors or degenerate gamblers? If anyone has questions on the legal aspects, including the current litigation, feel free to ask and I'll answer what I can without revealing anything privileged.
SharperDingaan Posted February 10 Posted February 10 1 hour ago, Saluki said: I get quite a few consulting calls about this topic so I figured it would be interesting to start a post on it here. Anyone use them and have opinions on it? Do you use it for sports betting or financial? What do you think of the other users? Legitimate investors or degenerate gamblers? If anyone has questions on the legal aspects, including the current litigation, feel free to ask and I'll answer what I can without revealing anything privileged. It's a money line ... so the signal tends to be a little more reasonable. Doesn't mean the prediction will happen, so treat it accordingly. We use it as a confirmation tool; does the market see the same thing, what is the current probability, is it rising or falling, what is the timeline. We already have a hypothesis, this is just a sniff test. Also could have used the option market or looked at forwards ... if they are a possibility. Still new to it, so no real opinion at this point. But ..... like any betting line, it is wide open to manipulation, so you need to think like a bookie. Buy a cheap bet on 'X', sell the same bet later at a higher price, capture the spread SD
Parsad Posted February 10 Posted February 10 If you have any significant knowledge about a specific subject, you may have an advantage to the bookies, but make no error that it is something more than just gambling. If you are an investor/value investor, you buy assets when the odds are greatly in your favor...below liquidation value, below the discounted cash you can take out of the business, etc. That is completely different than what Polymarket and the like offer. One makes sense to me, the other doesn't! Cheers!
RichardGibbons Posted February 10 Posted February 10 Yeah, I think either makes sense. Value investing is buying something for less than it's worth and hoping it re-evaluates to its value. Polymarket is betting on coin flips where you can occasionally risk 40c for a 50% chance of winning $1. It's unclear to me which would be better (though, if those were the actual numbers for the bet and there are lots of such examples every day, it's clear that you'd make way more with Polymarket because that's an expected return of 25% in a matter of days, way above any value investor.)
Eldad Posted February 10 Posted February 10 7 minutes ago, RichardGibbons said: Yeah, I think either makes sense. Value investing is buying something for less than it's worth and hoping it re-evaluates to its value. Polymarket is betting on coin flips where you can occasionally risk 40c for a 50% chance of winning $1. It's unclear to me which would be better (though, if those were the actual numbers for the bet and there are lots of such examples every day, it's clear that you'd make way more with Polymarket because that's an expected return of 25% in a matter of days, way above any value investor.) Sounds good in theory but in practice the investment gets to flip the coin again next quarter with the same odds with very little risk of a permanent loss of capital. The mis-priced coin flip would need a large volume of flips to work that may or may not exist in the future.
RichardGibbons Posted February 10 Posted February 10 26 minutes ago, Eldad said: Sounds good in theory but in practice the investment gets to flip the coin again next quarter with the same odds with very little risk of a permanent loss of capital. The mis-priced coin flip would need a large volume of flips to work that may or may not exist in the future. Agreed, that's why I said "lots of such examples every day". Polymarket is clearly worse if there is only one bet you can identify as mispriced every month. I'm too lazy to want to calculate the math by hand, but Gemini says that the optimal bet size with my numbers is 44% of your bankroll, and you'd need 52 bets to be 99% confident that your bankroll would gain over 25%. So, to completely obliterate a value investor's with those odds, you'd need about one such bet a week for a year. If you're not trying to obliterate the value investor, and don't bet as much as Kelly suggests (instead bet 10-15% of bankroll), then apparently it's 22 bets.
Libs Posted February 10 Posted February 10 Someone made a killing on the Maduro raid. It's being investigated; possibly a government insider took advantage. Point being, a lot of the bets might be stacked against you. The person taking the other side of the bet might know a lot more about the subject than you do.
RichardGibbons Posted February 11 Posted February 11 (edited) Yep, there are two holes in my argument. The first is that my numbers are way too extreme. The edge I propose is way too high, betting 40c to win $1 in a coin flip. That's the average actual advantage I'm suggesting the bettor might have, across all the bets they make, all of which have uncertainties. That is a super-high edge. The other big flaw is that the liquidity of these markets isn't even close to infinite, and it wouldn't be long before you start moving the market. Polymarket had $55M in bets on who would win the Superbowl--the biggest sporting event of the year in America. I imagine it's pretty easy to find places where you can deploy $55M as a value investor without eliminating the investment opportunity as a result of your buying. I suspect all the deep markets on Polymarket are based on arbitrage. Edited February 11 by RichardGibbons
Saluki Posted February 11 Author Posted February 11 There are a few practical/philosphical features that affect predition market pricing besides the depth of the market liquidity. As Sir Francis Galton observed, when you have large numbers of people guessing something (like how many jelly beans in a jar) you get a pretty accurate number because some people guess too high and some too low, and the errors on either side are random and cancel each other out so you get price discovery. But if people know each others guesses (like the stock market), it can get out of whack because of feedback loops. Prediction Markets are like that, kinda, but they also have more people with asymmetric information (like the guy betting on the Venezuela invasion). In a random poll, the guesses are random but they also don't have skin in the game. Someone putting up money would have a higher conviction than a passerby who has nothing to lose. So the people betting are more informed, and the people with asymmetric information are willing to bet heavier until the odds move to reflect the true probability and it no longer makes sense to make bets. The problem is that they are allowing bets on things that don't serve a useful purpose like other types of derivatives: hedging and price discovery.
SharperDingaan Posted February 11 Posted February 11 (edited) There is only collective wisdom (net of error cancellation), if there are large numbers of completely independent participants; in most cases, that is anything but the case. It is a simple thing for a few individuals to inflate the days market volume to make it appear that there is a lot of interest ... via a series of wash trades. 'Running the box' over extended periods of time, is a very common practice in the promotion of junior mining companies Simple to test. Compare the forward price curve of commodity X to the actuals that occurred ... if the wisdom is 'good', the difference between actual and predicted price should be at its lowest, at the time line predicted. It almost always, and very reliably isn't ... across multiple commodities, and multiple time frames. Yet for valuation purposes .... securities analysts are routinely required to use the current forward curves (as market estimates) to estimate the future cash flows, that will be discounted to arrive at 'intrinsic value' . Opportunities Prediction floats on the idea that something simple is better than nothing ... hence accuracy is not really a consideration. Whereas the opportunity rests on the failure of the promoter continually being able to convince enough people tomorrow, than white is the 'new black'; ideally via a sudden 'market discontinuity' . Also known as populist propaganda, that some are very good at. What you cannot achieve via hard analytics, you do via the madness of crowds instead. SD Edited February 11 by SharperDingaan
brobro777 Posted February 14 Posted February 14 On 2/12/2026 at 12:31 AM, Saluki said: There are a few practical/philosphical features that affect predition market pricing besides the depth of the market liquidity. As Sir Francis Galton observed, when you have large numbers of people guessing something (like how many jelly beans in a jar) you get a pretty accurate number because some people guess too high and some too low, and the errors on either side are random and cancel each other out so you get price discovery. But if people know each others guesses (like the stock market), it can get out of whack because of feedback loops. Prediction Markets are like that, kinda, but they also have more people with asymmetric information (like the guy betting on the Venezuela invasion). In a random poll, the guesses are random but they also don't have skin in the game. Someone putting up money would have a higher conviction than a passerby who has nothing to lose. So the people betting are more informed, and the people with asymmetric information are willing to bet heavier until the odds move to reflect the true probability and it no longer makes sense to make bets. The problem is that they are allowing bets on things that don't serve a useful purpose like other types of derivatives: hedging and price discovery. yea this is why I don't bet on sports - all these guys have so much better info than me on what the players and teams and coaches and referees and everybody is doing. Like, I don't stand a chance. The last bet that I made was in 2005 in Vegas, McBride vs Tyson. When I made the bet on McBride to win it was like 4:1 but the line kept moving faster toward lower payout for McBride closer to the fight. I stayed away completely from sports betting since then, so I'll always remain a winner vs Vegas!!!
Saluki Posted March 5 Author Posted March 5 In an effort to get my name out there for my consulting work, I was interviewed for this piece on ReadWrite. How insider trading could work on prediction markets I don't think it will lead directly to business, but the backlink to my website will probably help a lot with SEO.
RichardGibbons Posted March 5 Posted March 5 It is pretty interesting that they allow you to bet with such a narrow spread because you actually don't need that much of an edge for it to be profitable. I've started paper sports betting on Polymarket to see if I potentially have an edge, and I've actually won 13 straight bets, at average odds where the market says I have a 47% chance to win. So far, with 118 fake bets, I've risked 57.71 and won 68. Gemini indicates that's significant, but I'm not that convinced when it comes right after 13 straight wins. It feels like extrapolating stock market results in 1999 based on 5-year returns.
Saluki Posted March 5 Author Posted March 5 7 minutes ago, RichardGibbons said: It is pretty interesting that they allow you to bet with such a narrow spread because you actually don't need that much of an edge for it to be profitable. I've started paper sports betting on Polymarket to see if I potentially have an edge, and I've actually won 13 straight bets, at average odds where the market says I have a 47% chance to win. So far, with 118 fake bets, I've risked 57.71 and won 68. Gemini indicates that's significant, but I'm not that convinced when it comes right after 13 straight wins. It feels like extrapolating stock market results in 1999 based on 5-year returns. Prediction markets are peer-to-peer, whereas (most) sportsbooks are structured where the bookie takes the other side of a bet. If you are a market maker (in options or in sports), you want to have a balanced book, so if you have too many people on one side of the trade you can widen the bid/ask (or just one side) until it evens out for the bookie. With a narrower bid/ask spread it doesn't take much specialized knowledge (or other asymmetric info like insider trading) to do a profitable trade. It is a zero sum game though, so while you won a bunch (statistically) there are people on the other side of every trade who lost. I did a few sample trades with Kalshi (through Robinhood) just so that I could see how it works. I lost every trade, which is not surprising since I don't follow sports very much. But I did notice something interesting. For some reason Robinhood lets me trade prediction markets on my phone, but not my desktop, where I prefer to trade. If other people can only trade on their phones, then Kalshi's claim to the courts that there is no way that could geo-fence their market to keep people out who are in states where it's illegal is complete non-sense. Not only can they do it, but it's easier because unlike a desktop VPN, a phone has geolocation data.
Saluki Posted March 12 Author Posted March 12 Prediction markets create insider trading loophole as enforcement lags | S&P Global I got quoted in S&P Global. I'm trying to get the Thanos infinity gauntlet of financial press mentions. Bloomberg, Bloomberg, where art thou?
brobro777 Posted March 13 Posted March 13 Hey I got a question In 2024 I bet using IBKR's Forecast Futures and made money - IBKR issued me 1099-MISC for taxes Are prediction futures going to be treated like 1256 contracts in the future, like ES, NQ, GC, CL and so on? It would be good if it was, for 60% long term capital gains rates and not having to deal with 1099-MISC
Saluki Posted March 13 Author Posted March 13 7 hours ago, brobro777 said: Hey I got a question In 2024 I bet using IBKR's Forecast Futures and made money - IBKR issued me 1099-MISC for taxes Are prediction futures going to be treated like 1256 contracts in the future, like ES, NQ, GC, CL and so on? It would be good if it was, for 60% long term capital gains rates and not having to deal with 1099-MISC That's a great question, and the answer is: probably not. IMHO when Kalshi decided to do sports betting, for some reason they decided to structure them as swaps, not futures contracts and it was a mistake. I don't think their general counsel is very impressive, and from his LinkedIn he spent less than 6 years practicing law, which wouldn't even qualify him to be a partner at a law firm. But it was a risky startup back then and, according to linkedin, he was self-employed for several years before that, so maybe they both took what they could get. When other companies, like IBKR, wanted to get in on the action, they just copied the flawed model that Kalshi used. IBKR has great outside counsel, but they are, ummmm, frugal when it comes to hiring in house staff, and they work them hard, so I don't think they have a lot of spare bandwidth to think about these things. So that is probably why they just did what was easy and cheap and didn't re-invent the wheel. So, futures traded on a designated contract market (DCM), like the CME get the 60/40 treatment. But swaps do not. Many swaps are now being traded on DCMs, but they don't get the 60/40 treatment because they are not futures. Event contracts are usually set up like binary options, which are swaps under the commodity exchange act definitions. You could try to argue that they are traded on a futures exchange and that makes them futures, but it's a weak argument. Futures are traded on an exchange, but that's what makes them legal (off-exchange futures are illegal), not what makes them futures. Futures have common elements like an underlying reference commodity, futurity (settlement or delivery past the spot date), leverage, settlement by offset, initial margin and maintenance margin. Because on a $1 contract, I might pay $0.30 and you pay $0.69 (the exchange keeps the penny), and I get the whole dollar when it settles, there is no mark-to-market and maintenance margin like a futures contract.
brobro777 Posted March 13 Posted March 13 2 hours ago, Saluki said: That's a great question, and the answer is: probably not. IMHO when Kalshi decided to do sports betting, for some reason they decided to structure them as swaps, not futures contracts and it was a mistake. I don't think their general counsel is very impressive, and from his LinkedIn he spent less than 6 years practicing law, which wouldn't even qualify him to be a partner at a law firm. But it was a risky startup back then and, according to linkedin, he was self-employed for several years before that, so maybe they both took what they could get. When other companies, like IBKR, wanted to get in on the action, they just copied the flawed model that Kalshi used. IBKR has great outside counsel, but they are, ummmm, frugal when it comes to hiring in house staff, and they work them hard, so I don't think they have a lot of spare bandwidth to think about these things. So that is probably why they just did what was easy and cheap and didn't re-invent the wheel. So, futures traded on a designated contract market (DCM), like the CME get the 60/40 treatment. But swaps do not. Many swaps are now being traded on DCMs, but they don't get the 60/40 treatment because they are not futures. Event contracts are usually set up like binary options, which are swaps under the commodity exchange act definitions. You could try to argue that they are traded on a futures exchange and that makes them futures, but it's a weak argument. Futures are traded on an exchange, but that's what makes them legal (off-exchange futures are illegal), not what makes them futures. Futures have common elements like an underlying reference commodity, futurity (settlement or delivery past the spot date), leverage, settlement by offset, initial margin and maintenance margin. Because on a $1 contract, I might pay $0.30 and you pay $0.69 (the exchange keeps the penny), and I get the whole dollar when it settles, there is no mark-to-market and maintenance margin like a futures contract. Thanks for the explanation bro, I figured that would be the case. These forecast/prediction futures are created new all the time per event, while these established futures on exchanges like CME have been trading for decades, with the same underlying that are always going to be there. I actually do see positive EV with these bets, especially the political ones, and that's why I did okay in 2024. But it was fairly difficult to deal with the 1099-MISC where IBKR reported the proceeds but not the costs and I had to do all kinds of stuff whereas 1256 mark to market is so easy. Where's a good lawyer when you need one...
Saluki Posted April 2 Author Posted April 2 Does anyone have a subscription to The Business Journals? A reporter there quoted me but didn't give me a heads up or a free copy of the article so I have no idea what he said or if I was quoted correctly. Thanks in advance https://www.bizjournals.com/bizjournals/news/2026/03/19/kalshi-polymarket-gambling-betting-business-2026.html
Recommended Posts
Create an account or sign in to comment
You need to be a member in order to leave a comment
Create an account
Sign up for a new account in our community. It's easy!
Register a new accountSign in
Already have an account? Sign in here.
Sign In Now