I mentioned this on a different post - the biggest problem with prediction markets is not the gambling or dumb people losing money. Its the fact that it gives very powerful people a vehicle to make lobsided bets on outcomes they control.
A small example of this would be NFL / NBA Refs fixing playoff games with a bad call or two. This actually happened 20 years ago, an NBA ref went to prison over being bribed just $2000 per game.
The much worse example is the fact that you can make 100-1 odds on whether the US airstrikes Iran today... or How many times Pam Bondi says the word "China" in a press conference.
Somebody poor grunt who chose to earn a living by laboring (which has proven to be much less effective than being born with money) will be putting fuel in the bombers and thinking "I could just make an anonymous bet..."
The irony that operational national security would be greatly improved if only they maintained a well staffed and resourced government kitchen for the Pentagon, but won't for many silly reasons. Oh no, lots of people would have to sit on idle standby many times, or gov't employees would get free meals.
1- Making a bet with privileged information.
2- Creating the event and making the bet.
2 would be a war crime, 1 would be a probabilistic leak.
Trump claimed they didn't want to pass through congress because they leak, and there were no leaks about the event. But if any personnel made a polymarket bet, that would constitute a leak. It wasn't acted upon, but if personnel continues to leak information in this manner, it is possible that an adversary will eventually listen to this signal, and that it was just ignored because it is too fresh.
This analysis would also make it clear why it would be immoral to participate in such markets as a civilian. Because if it is your country you might be compensating an insider for information, benefitting the enemy. And if you are not, you might be harming the enemy, but you would be an unlawful belligerent.
This gets into a philosophical point about what a prediction market actually is. If it's a device for anonymously aggregating fragmented group information into a coherent accurate prediction, the lopsided bets are a feature; the only point of the market is the price signal, and the lopsided bets true up the price.
But most of us understand that prediction markets aren't that, no matter what Robin Hansen said when he was helping invent the modern incarnation of things like Polymarket and Kalshi. They're gambling venues, and we have "Nevada Gaming Commission"-style concerns about fairness. To me, the next logical step is to say that they should be heavily regulated, but in the era of DraftKings, that seems off the table.
> But alarms were raised Thursday morning, hours before the royal appearance, when a run of bets for brown started coming in, displacing light blue as the favorite.
> "Nobody was backing brown at all and suddenly everyone wanted in on it," Paddy Power (search), owner of the eponymous chain of betting shops that inaugurated the hat bet 10 years ago, told The Times.
> Power's odds on brown went from 12-1, to 2-1, to even and finally to 8-11 before he yanked the bet at 11:30 a.m., 2½ hours before the Queen was due to show.
> "Someone must have been in the know. We laid 50 pounds at 20-1 and 200 pounds at 10-1 and some smaller bets," David Hood, spokesman for rival betting chain William Hill (search), told the Daily Telegraph.
> ...
> When Elizabeth II finally made her appearance, she was indeed wearing a brown hat with cream trim.
> "Somebody has made a tidy sum," sniffed Hood.
> Both he and Power, who estimated his firm lost about 10,000 pounds, or $18,000, suspected palace insiders.
That's the actual point. Everyone else is there to make money gambling, but the whole premise is to incentivize people with secret information to share it anonymously with the public, and take a reward for doing it.
All without traceability or secret drops or whatever.
> it gives very powerful people a vehicle to make lobsided bets on outcomes they control
I'm sceptical that prediction markets uniquely enable this. Like, if you want to bet on U.S. airstrikes in the short term, you could always buy oil options (or short exposed companies). If you're in for the long term, you're buying something that benefits from cheaper gas, e.g. an additives company.
You are not wrong, and I should clarify I also have a big problem with the current state of legal insider trading of elected officials, but this polymarket problem is much more extreme. You can get a guaranteed 100-1 payout by blowing up some random people on the other side of the planet. Way worse than making even 2-5x on a leveraged futures bet with insider info. In that example, the victim is usually just other rich people.
All of these things are much more subject to the problem that effects policy generally: the law of unintended consequences. Betting on the policy, rather than an intended/expected longer-term outcome that is easily derailed by intervening events outside of your direct control is much more direct (plus, if you are corrupt enough to bet on policy you control, that policy is probably already seeking a longer-term aim that serves your existing financial interests, so the ability to bet on the policy itself makes the corruption more attractive by providing a more immediate and certain payoff on top of the longer-term, less certain one.)
Prediction markets don't uniquely enable it, but they make it far more effective and easy.
Insider trading is illegal. And for trades that aren't technically insider trading, often having some information ahead of time isn't as useful as it seems. Markets are known to react unpredictably to news; sometimes they move the opposite way from what you'd think, especially over the mid-long term, and there are many other influences on the price.
With a prediction market though, if you know what'll happen in the world, you know exactly what you'll win in the market.
Plenty of more fun dynamics. For example, in some cases it becomes a way for voting for decisions one otherwise wouldn't control. If a person in position to make a decision doesn't really care about any choice in particular, seeing the prediction market lean one way would incentivize them to choose the opposite, making a short sale immediately before.
It also makes sense for the people voting: by betting against the outcome they want, they end up either a) paying for getting things their way, or b) getting consolation payoff if the decision makers pick the undesired choice.
> During the 2024 general election campaign, allegations were made that illicit bets were placed by political party members and police officers, some of whom may have had insider knowledge of the date of the general election before Rishi Sunak, the Prime Minister at the time, publicly announced when it would be held.
> ...
> In April 2025, the Gambling Commission charged 15 people with offences under Section 42 of the Gambling Act 2005, including Russell George, Tony Lee, Nick Mason, Laura Saunders, and Craig Williams. Trials are not expected to begin until September 2027 or January 2028.
> the biggest problem with prediction markets is not the gambling or dumb people losing money. Its the fact that it gives very powerful people a vehicle to make lobsided bets on outcomes they control
This is quickly becoming the point of them, at least insofar as they are enjoying an extremely favorable regulatory environment courtesy of the Trump crew.
well, at least for really odd ones - like the china example - the liquidity is (probably) going to be really low. you need people buying both sides to make money.
But for big events/talked about stuff/etc ofc this is not true.
In the context of legislating prediction markets or not, sports is not a concern at all.
Whether it's a net positive or negative for important shit like war and corruption, we'll see, but if it helps in the important stuff, but damages sports, sorry bud.
I'm getting some really skeezy ads for prediction markets on TikTok at the moment, the message is effectively "hey, are you broke? earn $50+/day on Kalshi!"
I have occasionally tried checking Polymarket and Kalshi to get an idea of the general political/cultural/technological consensus on various issues that are difficult to research otherwise, eg. "what are the chances that the Senate changes hands in the 2026 midterms?" People have thought about it enough to wager a million dollars and the consensus is at about 1/3. I have this abstract prediction market in my head, each bet placed by some statistically average person with diverse experiences and exposures from my own bubble, who carefully considers their information and puts their two cents into the pot, and I assume that by adding all our ideas together we form some sort of combined intelligence which is more insightful and reliable together than any individual pundit could be.
And then I go back to the home page, and see all the rabid sports fans, and realize that these bets are not being placed by deep thinkers.
I'm a little confused by the "Yes" versus "No" asymmetry.
For example, one of the top trending ~~bets~~ markets right now is on whether Miami or Indiana will win the NCAA football championship tonight. You can either take "Yes" on Indiana at 74c, or "No" at 27c, or you can take "Yes" on Miami at 27c or "No" at 74c. Or, there's another potential outcome - you can also bet on a tie at 10c yes/91c no.
Is this research suggesting that an optimistic Miami fan can somehow get a better return by buying "No" on Indiana than a "Yes" on Miami?
Why is Kalshi structured with these yes vs. no options for all outcomes?
Part of this perceived arbitrage is the fee structure. Kalshi has a weird transaction cost structure but taking advantage of that 1c arb probably costs you 2c in fees to Kalshi, so nobody does it.
> Why is Kalshi structured with these yes vs. no options for all outcomes?
it's basically how they do margin. otherwise you wouldn't be able to sell / post asks without already having a long position. for kalshi, it's actually one single security in the background they just present it as two order books (but really it's one). for polymarket, they are two distinct products that trade separately, and technically could have arbitrage between them. although in practice they're normally priced correctly to sum to 1 (or 1.01)
10c yes seems really high for a tie. NCAA rules don't allow for ties in football. I know prediction markets have very long shot bets but I would expect that to be closer to 1c
Edit: it looks like the tie market is only for if the game is tied at halftime, which makes much more sense
I hope we manage to leverage prediction markets to actually achieve goals rather than just making a casino out of it.
For instance, if you spot malware in a commit you could bet heavily against it being merged, and that would attract the maintainers' attention, and they'll see what you see and not merge it, and you get paid for the code review--that money would come from whoever bet that it would get merged, which you could require be the author of the malware. I haven't worked it out entirely but it seems that there are opportunities to build games that reward dilligence and transparency and penalize deception and spam.
Why not just bet heavily against and then inform maintainers? By just betting on it instead it makes you look like you, or someone you know planted the malware
The analysis is interesting, but I think it ignores a few factors:
1. The article mentions the bid/ask spread for contracts, but I believe that Kalshi also has its own fee structure. Small edges (an expected loss of 0.57¢ on a 1¢ contract implies an expected gain of 0.43¢ on a 99¢ contract, or a 5.75ppt edge) can be easily eaten by even small fees, and liquidity provision is all about small edges.
2. The article ignores the time value of money, and contracts take time to resolve. If a contract won't resolve for six months and the risk-free rate is 5%, then buying a "sure thing" over 97.5¢ is a loss net of otherwise earnable interest.
3. Long shots offer greater implied leverage to bettors, making them more attractive. This is still (sometimes) an exploitable mispricing, but it's closer to the well-understood "bet against beta" factor.
(Edit to add) Also, I think their explanation of the non-returns on finance is lacking:
> Why is Finance efficient? The likely explanation is participant selection; financial questions attract traders who think in probabilities and expected values rather than fans betting on their favorite team or partisans betting on a preferred candidate. The questions themselves are dry ("Will the S&P close above 6000?"), which filters out emotional bettors.
Financial contracts are the ones that are most perfectly hedges with existing markets. "Will the S&P close above X?" is a binary option, after all, so it's comparatively easy for a market-maker to almost perfectly offset their Kalshi positions with opposite positions in traditional markets.
on point 1, an important thing to know is that these markets have a non-linear fee structure where the rate is higher near 0.5 and lower near tail prices
True, but from the pdf it seems like the fee charged of market makers is 1.75¢ × P × (1-P) per contract. Near P=0 that's approximately 1.75% of the notional amount invested, but near P=1 that's approximately 1.75% of the potential gain.
As I read it, the implication is that a market maker in the high-P regime needs to still have an expected edge of 1.75% to profit net of fees, which means that the 'maker return' table in this article is net negative after fees for all categories save for entertainment, media, and world events.
Fees are also waved if a market maker hits a certain monthly quota. With the recent adoption of “professional” market makers on the platform, I’m sure they can get around such fees.
I will also add to the 2nd point that some of these platforms due give fixed interest to positions in unresolved markets.
Something that appears to be missing: Certain events attract "advertising" types of bets. E.g. There is value in making a candidate appear to be a leader, so dedicating dollars to swinging the market is more of a form of advertising than an intelligent bet.
So it would be interesting to measure the inefficiencies of various bets vs the total market value in that bet.
e: Although full disclosure, I did not pick apart the entire paper. Maybe it's buried in there.
super interesting, re: spending money to move the line is just another form of non-profit-seeking "consumption."
i didn't filter for manipulation specifically, but i did find that politics was actually one of the most efficient categories (only ~1% maker/taker gap), suggesting the market absorbs those flows pretty well.
> but i did find that politics was actually one of the most efficient categories (only ~1% maker/taker gap)
I confess I'm surprised by that result in particular. I realize your results are for Kalshi, but ISTR some reports from the presidential elections on Polymarket.
But more generally: When you say there is "only a ~1% maker/taker gap", is that weighted by the size of the bets? or is it averaged over the number of bets placed?
How do prediction markets account for interest rates? I feel like I should be willing to pay no more than ~96 cents for a contract that will definitely resolve to a dollar in a year. Who puts up the other 4 cents?
Interest on open positions. Polymarket pays about 4.00% annualized holding rewards on eligible markets/positions (not all). Kalshi pays about 3.25% APY on cash plus open positions (collateral).
The usual thing is that the market ends up around $0.95 for things like that, if the actors are all solid investors. It only takes one overly enthusiastic yes buyer to break that ceiling, the smart money won't "correct" it down to $0.95
There's another idea, which is make contacts that pay out in shares of an ETF, but I haven't seen this idea put into practice
This article lacks even the most basic understanding of probability and statistics. Slot machines "93 cents on the dollar" return is a statistical certainty of 7% loss. You are playing a repeated game which by the law of large numbers will converge to the 93% probability.
In prediction markets if the markets are fully efficiently priced, in the absence of transaction costs you WILL get 100% back in the long run.
Slots are also unskilled games, prediction markets clearly some participants have a clear market edge, thus not efficiently priced.
> Takers pay a structural premium for affirmative "YES" outcomes while Makers capture an "Optimism Tax" simply by selling into this biased flow.
It's still operating like a casino in that there's a "house edge" that comes from taking bets. Unlike a casino, there is nothing stopping the average person from market making, which is why it doesn't make sense this structural inequality exists.
i understand how probability works. the "93 cents" vs "43 cents" comparison is looking at realized historical data, not theoretical odds. if the markets were efficiently priced, you would get 100% back. the entire point of the paper is showing that, historically, they aren't efficiently priced (longshots return ~43%), and explaining who captures that inefficiency.
> In prediction markets if the markets are fully efficiently priced, in the absence of transaction costs you WILL get 100% back in the long run.
This is basically equivalent to the observation that, in a perfectly efficient market, no entity can ever make a profit.
And yet, in the real world, entities make profits all the time. In fact, they make wild, unimaginable, world-changing, history-altering profits. This is a tacit admission that our markets aren't even remotely efficient, and that includes predictions markets. Efficient, rational markets are the exception, not the rule.
In a perfectly efficient market all entries can make the same profit on a given investment at the same level of risk and time horizon. There’s nothing inefficient about a market having a risk premium etc.
If you're making nonzero profit that means that it's feasible for anyone else (literally anyone else, assuming zero barriers to entry, which we do assume for an efficient market) to make slightly less profit by selling the same product at a lower price, which iteratively pushes all profits towards zero. An efficient market also assumes perfect information, which includes information of future events, so talking about risk/uncertainty is already out of the question. If that sounds absurd, then yes, that's the point: our assumptions about what it takes in order to achieve an efficient market approaches the absurd. Which isn't to say that markets aren't often useful, especially compared to the alternatives, but rather that appeals to rationality don't survive contact with the enemy.
You’re forgetting the economy is finite. You can’t infinity add new participants with infinite product to sell.
Instead in an efficient market everyone is already occupied making X ROI and gives as much up by entering a new market as they gain.
Put another way, if you already own a sock with 10% ROI, you can sell it and buy a sock with 10% ROI but the transaction is pointless so it doesn’t occur.
> An efficient market also assumes perfect information, which includes information of future events, so talking about risk/uncertainty is already out of the question.
Perfect information means something different here. In Chess both players have perfect information of the game state, they don’t know the future. Poker has randomness and imperfect information but there’s other games with randomness and perfect information.
Free markets aren't even an exception. They're an abstract construction that exists to make economic analysis scientific by removing all confounding variables from the equation. I'd be extremely surprised to find one example where the conditions required of a free market truly existed.
If people knew more about economics than just whatever is being parroted as 'economics' in mainstream media they would know that there's a variety of types of markets that happen in the real world and none of them are the abstraction of a free market that allows econ 201 students to compare what happens when you introduce trade between a country that produces 4 apples for 3$ each and a country that produces 5 oranges for 4$ each.
dataset: 72.1m trades and $18.26b volume on kalshi (2021-2025)
core findings:
longshot bias: well documented longshot bias is present on kalshi. low probability contracts are systematically overpriced. contracts trading at 5 cents only win 4.18% of the time.
wealth transfer: liquidity takers lose money (-1.12% excess return) while liquidity makers earn it (+1.12%).
optimism tax: the losses are driven by a preference for "yes" outcomes. buying "yes" at 1 cent has a -41% expected value. buying "no" at 1 cent has a +23% expected value.
category variation: finance markets are efficient (0.17% maker-taker gap) while high-engagement categories like media and world events are inefficient (>7% gap).
mechanism: makers do not win by out-forecasting takers. they win by passively selling "yes" contracts to optimistic bettors
> Optimism tax: the losses are driven by a preference for "yes" outcomes. buying "yes" at 1 cent has a -41% expected value. buying "no" at 1 cent has a +23% expected value.
This is interesting and makes a statement about positive or negative orientation in human psychology. Also, couldn’t the bets just be worded in the double negative instead of the affirmative to influence the optimism bet?
This reminds me of the old scheme where if you just bet against ND football you'd make money because ND fans were so rabid that the "ND is good" positions became overpriced.
Yes, in the study they pinpointed this beautifully:
“A fan betting on their team to win the championship is not calculating expected value; they are purchasing hope.”
To me this is all the more reason to get regulatory gatekeeping out of the financial markets
If the odds in some financial products are worse than gambling while everyone can access gambling, then people should stop making a distinction under the guise of protecting investors
it just drives investors to actual gambling because they cant get the exposure they were already looking for
> it just drives investors to actual gambling because they cant get the exposure they were already looking for
This argument gets trotted out by Wall Street every decade or so, usually under the guise of "democratising" some piece of finance. It's almost always bunk.
Most investment capital is looking for safe returns. It's not competing with gambling. Even within the high-risk end of finance, the game is in turning that high risk into above-market but predictable returns through portfolio mechanics. (Fuckups aside, you can't generally portfolio mechanic your way out of the negative expectated value of a lottery ticket.)
More simply: the notion that we need to increase risk and profitiabilty for intermediaries in investments to keep people from gamblig is a false economy. Gamblers are seeking a different thrill from what financial markets are designed to provide. To the degree we have a problem, it's in letting our markets look more like casinos.
> exposure they were already looking for
Broadly speaking, if you want exposure to the economy you're investing. If you want exposure to a number that goes up, you're gambling. This is an overly-simplistic delineation. But it works for first-order estimates.
The same financial products are used in both gambling and smart investing. The canonical example here being options. And the restrictions on what the public can and cannot invest in are complete bullshit. You can't buy shares in a series A startup because that is deemed to be too risky for anyone who is not an "accredited" investor ("accredited" here literally means rich). But anyone who wants to can bet on sports, go to a casino, or buy a 2x levered VIX ETF.
You're misunderstanding the dynamics here. Modern prediction markets are 90% sports gambling by volume. The trick is that, by positioning themselves as general financial markets and accepting the corresponding regulatory gatekeeping, they're exempt from the often much stricter regulations that states put on normal sports gambling apps.
I wonder how much of the activity on prediction markets these days is competing LLM scripts? I would guess the overlap in prediction market punters and AI boomers is high.
A small example of this would be NFL / NBA Refs fixing playoff games with a bad call or two. This actually happened 20 years ago, an NBA ref went to prison over being bribed just $2000 per game.
The much worse example is the fact that you can make 100-1 odds on whether the US airstrikes Iran today... or How many times Pam Bondi says the word "China" in a press conference.
Somebody poor grunt who chose to earn a living by laboring (which has proven to be much less effective than being born with money) will be putting fuel in the bombers and thinking "I could just make an anonymous bet..."
It's a national security issue.
We saw this with the Venezuela attack. A flurry of trading and someone made $400,000 for placing a bet mere hours before the "surprise" attack. https://www.pbs.org/newshour/nation/a-400000-payout-after-ma...
1- Making a bet with privileged information. 2- Creating the event and making the bet.
2 would be a war crime, 1 would be a probabilistic leak.
Trump claimed they didn't want to pass through congress because they leak, and there were no leaks about the event. But if any personnel made a polymarket bet, that would constitute a leak. It wasn't acted upon, but if personnel continues to leak information in this manner, it is possible that an adversary will eventually listen to this signal, and that it was just ignored because it is too fresh.
This analysis would also make it clear why it would be immoral to participate in such markets as a civilian. Because if it is your country you might be compensating an insider for information, benefitting the enemy. And if you are not, you might be harming the enemy, but you would be an unlawful belligerent.
But most of us understand that prediction markets aren't that, no matter what Robin Hansen said when he was helping invent the modern incarnation of things like Polymarket and Kalshi. They're gambling venues, and we have "Nevada Gaming Commission"-style concerns about fairness. To me, the next logical step is to say that they should be heavily regulated, but in the era of DraftKings, that seems off the table.
A classic example is the color of the Queen's hat at Royal Ascot.
https://www.upi.com/Odd_News/2008/06/20/Bets-placed-on-queen...
https://news.williamhill.com/horse-racing/queens-hat-betting...
And the relevant one from 2005 - https://www.foxnews.com/story/hat-trick-upsets-british-booki...
> But alarms were raised Thursday morning, hours before the royal appearance, when a run of bets for brown started coming in, displacing light blue as the favorite.
> "Nobody was backing brown at all and suddenly everyone wanted in on it," Paddy Power (search), owner of the eponymous chain of betting shops that inaugurated the hat bet 10 years ago, told The Times.
> Power's odds on brown went from 12-1, to 2-1, to even and finally to 8-11 before he yanked the bet at 11:30 a.m., 2½ hours before the Queen was due to show.
> "Someone must have been in the know. We laid 50 pounds at 20-1 and 200 pounds at 10-1 and some smaller bets," David Hood, spokesman for rival betting chain William Hill (search), told the Daily Telegraph.
> ...
> When Elizabeth II finally made her appearance, she was indeed wearing a brown hat with cream trim.
> "Somebody has made a tidy sum," sniffed Hood.
> Both he and Power, who estimated his firm lost about 10,000 pounds, or $18,000, suspected palace insiders.
All without traceability or secret drops or whatever.
POSIWID
I'm sceptical that prediction markets uniquely enable this. Like, if you want to bet on U.S. airstrikes in the short term, you could always buy oil options (or short exposed companies). If you're in for the long term, you're buying something that benefits from cheaper gas, e.g. an additives company.
Insider trading is illegal. And for trades that aren't technically insider trading, often having some information ahead of time isn't as useful as it seems. Markets are known to react unpredictably to news; sometimes they move the opposite way from what you'd think, especially over the mid-long term, and there are many other influences on the price.
With a prediction market though, if you know what'll happen in the world, you know exactly what you'll win in the market.
It also makes sense for the people voting: by betting against the outcome they want, they end up either a) paying for getting things their way, or b) getting consolation payoff if the decision makers pick the undesired choice.
if you're not the person-in-complete-power, your bet is really likely to be 'rigged' against you
I'd rather play dice or buy lotteries
> During the 2024 general election campaign, allegations were made that illicit bets were placed by political party members and police officers, some of whom may have had insider knowledge of the date of the general election before Rishi Sunak, the Prime Minister at the time, publicly announced when it would be held.
> ...
> In April 2025, the Gambling Commission charged 15 people with offences under Section 42 of the Gambling Act 2005, including Russell George, Tony Lee, Nick Mason, Laura Saunders, and Craig Williams. Trials are not expected to begin until September 2027 or January 2028.
This is quickly becoming the point of them, at least insofar as they are enjoying an extremely favorable regulatory environment courtesy of the Trump crew.
OK, and? The market is just paying them to make information about their decisions public.
But for big events/talked about stuff/etc ofc this is not true.
In the context of legislating prediction markets or not, sports is not a concern at all.
Whether it's a net positive or negative for important shit like war and corruption, we'll see, but if it helps in the important stuff, but damages sports, sorry bud.
I have to say I was this huge fan of the idea and I didn’t anticipate it would happen like this.
And then I go back to the home page, and see all the rabid sports fans, and realize that these bets are not being placed by deep thinkers.
For example, one of the top trending ~~bets~~ markets right now is on whether Miami or Indiana will win the NCAA football championship tonight. You can either take "Yes" on Indiana at 74c, or "No" at 27c, or you can take "Yes" on Miami at 27c or "No" at 74c. Or, there's another potential outcome - you can also bet on a tie at 10c yes/91c no.
Is this research suggesting that an optimistic Miami fan can somehow get a better return by buying "No" on Indiana than a "Yes" on Miami?
Why is Kalshi structured with these yes vs. no options for all outcomes?
it's basically how they do margin. otherwise you wouldn't be able to sell / post asks without already having a long position. for kalshi, it's actually one single security in the background they just present it as two order books (but really it's one). for polymarket, they are two distinct products that trade separately, and technically could have arbitrage between them. although in practice they're normally priced correctly to sum to 1 (or 1.01)
Edit: it looks like the tie market is only for if the game is tied at halftime, which makes much more sense
For instance, if you spot malware in a commit you could bet heavily against it being merged, and that would attract the maintainers' attention, and they'll see what you see and not merge it, and you get paid for the code review--that money would come from whoever bet that it would get merged, which you could require be the author of the malware. I haven't worked it out entirely but it seems that there are opportunities to build games that reward dilligence and transparency and penalize deception and spam.
1. The article mentions the bid/ask spread for contracts, but I believe that Kalshi also has its own fee structure. Small edges (an expected loss of 0.57¢ on a 1¢ contract implies an expected gain of 0.43¢ on a 99¢ contract, or a 5.75ppt edge) can be easily eaten by even small fees, and liquidity provision is all about small edges.
2. The article ignores the time value of money, and contracts take time to resolve. If a contract won't resolve for six months and the risk-free rate is 5%, then buying a "sure thing" over 97.5¢ is a loss net of otherwise earnable interest.
3. Long shots offer greater implied leverage to bettors, making them more attractive. This is still (sometimes) an exploitable mispricing, but it's closer to the well-understood "bet against beta" factor.
(Edit to add) Also, I think their explanation of the non-returns on finance is lacking:
> Why is Finance efficient? The likely explanation is participant selection; financial questions attract traders who think in probabilities and expected values rather than fans betting on their favorite team or partisans betting on a preferred candidate. The questions themselves are dry ("Will the S&P close above 6000?"), which filters out emotional bettors.
Financial contracts are the ones that are most perfectly hedges with existing markets. "Will the S&P close above X?" is a binary option, after all, so it's comparatively easy for a market-maker to almost perfectly offset their Kalshi positions with opposite positions in traditional markets.
As I read it, the implication is that a market maker in the high-P regime needs to still have an expected edge of 1.75% to profit net of fees, which means that the 'maker return' table in this article is net negative after fees for all categories save for entertainment, media, and world events.
I will also add to the 2nd point that some of these platforms due give fixed interest to positions in unresolved markets.
So it would be interesting to measure the inefficiencies of various bets vs the total market value in that bet.
e: Although full disclosure, I did not pick apart the entire paper. Maybe it's buried in there.
i didn't filter for manipulation specifically, but i did find that politics was actually one of the most efficient categories (only ~1% maker/taker gap), suggesting the market absorbs those flows pretty well.
I confess I'm surprised by that result in particular. I realize your results are for Kalshi, but ISTR some reports from the presidential elections on Polymarket.
But more generally: When you say there is "only a ~1% maker/taker gap", is that weighted by the size of the bets? or is it averaged over the number of bets placed?
In any case: Thanks for a very interesting paper!
I'm glad you enjoyed the paper :)
Edit to add that on non eligible markets your theory is correct, for example: https://polymarket.com/event/will-jesus-christ-return-before...
There's another idea, which is make contacts that pay out in shares of an ETF, but I haven't seen this idea put into practice
In prediction markets if the markets are fully efficiently priced, in the absence of transaction costs you WILL get 100% back in the long run.
Slots are also unskilled games, prediction markets clearly some participants have a clear market edge, thus not efficiently priced.
> Takers pay a structural premium for affirmative "YES" outcomes while Makers capture an "Optimism Tax" simply by selling into this biased flow.
It's still operating like a casino in that there's a "house edge" that comes from taking bets. Unlike a casino, there is nothing stopping the average person from market making, which is why it doesn't make sense this structural inequality exists.
This is basically equivalent to the observation that, in a perfectly efficient market, no entity can ever make a profit.
And yet, in the real world, entities make profits all the time. In fact, they make wild, unimaginable, world-changing, history-altering profits. This is a tacit admission that our markets aren't even remotely efficient, and that includes predictions markets. Efficient, rational markets are the exception, not the rule.
In a perfectly efficient market all entries can make the same profit on a given investment at the same level of risk and time horizon. There’s nothing inefficient about a market having a risk premium etc.
Instead in an efficient market everyone is already occupied making X ROI and gives as much up by entering a new market as they gain.
Put another way, if you already own a sock with 10% ROI, you can sell it and buy a sock with 10% ROI but the transaction is pointless so it doesn’t occur.
> An efficient market also assumes perfect information, which includes information of future events, so talking about risk/uncertainty is already out of the question.
Perfect information means something different here. In Chess both players have perfect information of the game state, they don’t know the future. Poker has randomness and imperfect information but there’s other games with randomness and perfect information.
If people knew more about economics than just whatever is being parroted as 'economics' in mainstream media they would know that there's a variety of types of markets that happen in the real world and none of them are the abstraction of a free market that allows econ 201 students to compare what happens when you introduce trade between a country that produces 4 apples for 3$ each and a country that produces 5 oranges for 4$ each.
dataset: 72.1m trades and $18.26b volume on kalshi (2021-2025)
core findings:
longshot bias: well documented longshot bias is present on kalshi. low probability contracts are systematically overpriced. contracts trading at 5 cents only win 4.18% of the time.
wealth transfer: liquidity takers lose money (-1.12% excess return) while liquidity makers earn it (+1.12%).
optimism tax: the losses are driven by a preference for "yes" outcomes. buying "yes" at 1 cent has a -41% expected value. buying "no" at 1 cent has a +23% expected value.
category variation: finance markets are efficient (0.17% maker-taker gap) while high-engagement categories like media and world events are inefficient (>7% gap).
mechanism: makers do not win by out-forecasting takers. they win by passively selling "yes" contracts to optimistic bettors
This is interesting and makes a statement about positive or negative orientation in human psychology. Also, couldn’t the bets just be worded in the double negative instead of the affirmative to influence the optimism bet?
If the odds in some financial products are worse than gambling while everyone can access gambling, then people should stop making a distinction under the guise of protecting investors
it just drives investors to actual gambling because they cant get the exposure they were already looking for
This argument gets trotted out by Wall Street every decade or so, usually under the guise of "democratising" some piece of finance. It's almost always bunk.
Most investment capital is looking for safe returns. It's not competing with gambling. Even within the high-risk end of finance, the game is in turning that high risk into above-market but predictable returns through portfolio mechanics. (Fuckups aside, you can't generally portfolio mechanic your way out of the negative expectated value of a lottery ticket.)
More simply: the notion that we need to increase risk and profitiabilty for intermediaries in investments to keep people from gamblig is a false economy. Gamblers are seeking a different thrill from what financial markets are designed to provide. To the degree we have a problem, it's in letting our markets look more like casinos.
> exposure they were already looking for
Broadly speaking, if you want exposure to the economy you're investing. If you want exposure to a number that goes up, you're gambling. This is an overly-simplistic delineation. But it works for first-order estimates.