Who Wins and Who Loses in Prediction Markets? Evidence from Polymarket
papers.ssrn.com94 points by vcf 4 hours ago
94 points by vcf 4 hours ago
Nice paper, and thanks for releasing the dataset.
The "top-1% winners are patient limit-order liquidity providers, not insiders" finding is interesting, and I'd love to see it extended cross-venue.
I work on tooling that normalizes orderbooks across Polymarket, Kalshi, Limitless, and Smarkets. From that angle, a lot of what looks locally like skilled Polymarket market-making turns out to be cross-venue arbitrage that happens to land on Polymarket. The same underlying question routinely trades 3-8% apart across venues for hours at meaningful depth, and a fast multi-venue stack rests limits on the lagging book at the exact moments the leading book moves. Locally that's indistinguishable from disciplined liquidity provision; cross-venue it's closer to FX triangular arb on the consensus price.
If your timestamps are fine-grained enough, a clean follow-up: for the top 1% of Polymarket profit-takers, what fraction of fills land within N seconds of a same-question move on Kalshi or Limitless? If it's materially above baseline, some of "skill" resolves into "cross-venue infrastructure" — which is also a more durable edge than within-venue alpha, so it could partly explain the weak monthly persistence you observe (the cross-venue gap closes when too many players run the same stack).
This might also be consistent with your insider-trading conclusion rather than against it: an insider on a real-world event has every reason to hit the lowest-friction venue with aggressive market orders (Polymarket: permissionless wallets, no KYC, no withdrawal limits). That's a fundamentally different profile from the patient limit-posting strategy your top bucket runs, so the two populations cleanly separate in the data even if both are present.
We have a grad student working on matching markets across venues. Not a trivial task at scale, but we hope to look at that eventually.
this comment was clearly written by AI. please don't do that.
Not sure why you were downvoted/flagged, because you're right. It is also quite an insightful comment worthy of discussion so I'm a little conflicted.
I don’t see why it’s AI, but even if it is, it’s better than most human comments so the complaint should be downvoted.
Take a look at the user's history, it's more obvious in context. It has a lot of claude-specific tells which are noticeable if you've spent time working with claude. AI-generated comments are against the HN guidelines https://news.ycombinator.com/newsguidelines.html#generated
phrenology
if you're not able to tell that OP's comment was AI slop, then you probably don't have much insight to contribute to the conversation either.
not AI slop, simply a copy paste if the abstract of the paper. journals in our field limit the allowed number of words, so the style can feel « unnatural » even when human-written
Maybe the guidelines should be changed? Something about: don’t complain about comments just because they’re AI.
I'm not complaining "just" because it's AI.
I'm complaining because it's AI, and also slop.
> resolves into "cross-venue infrastructure" — which is also a more durable edge than within-venue alpha
anybody who actually trades knows that on these markets, "cross venue infrastructure" (aka vibe coding some exchange api integrations) is much less important / durable than actual alpha.
Slop aside, do you think it's reasonable to assume a decent fraction of those making consistent profits are arbitrage bots?
arbitrage is a rather overloaded word that people use for all kinds of strategies but yes, I predict most of the most profitable and consistent accounts are not actually attempting to forecast the outcome of these markets from first principles
That sounds plausible. Not a trader so I wouldn't know. Saying at least a little about what's actually wrong with it seems more useful than just saying it's slop, which gives me very little info over just a downvote.
it's pretty easy to write basic trading api connectivity. the hard part is knowing what trades to send
even if we are very charitable and assume the comment refers only to like high-engineering-effort infra for trying to be super competitive on latency, that's still like the opposite of a durable edge, since everybody is looking at it. there's very little "hidden" knowledge and it's mostly a matter of elbow grease and careful engineering.
it's only "better than human comments" if you have no idea what profitable trading looks like. it's a very-very thin mildly convincing veneer over what is fundamentally slop.
I don't think that's surprising because the alternative would be that some people are able to predict the future. Whatever strategy one might figure out that works is long term destined to fail, as other people start using them. The only real way to make money there is by providing liquidity since it's a zero sum game. For the stock market this is not true because it's not zero sum, it grows over time.
There is alternative to being “able to predict the future”, which is “I already know the future” or “I can change the future”
Someone flips a coin and looks at it, what orders are you willing to put in?
The potential for insiders should be represented by a complete loss of liquidity.
And yet, many people bet on things like the duration or contents of press conferences, of pre-taped shows, etc.
There are some bets on prediction markets where the future is either already known or in the control of people who may be participating in the market. For example, when people bet on how long the next presidential briefing will be, it doesn't take a prophet to predict this, anyone who organizes said briefing can control it (at least with a very high probability).
So, the question becomes "what is the preponderence of such bets" and "how many people with control or knowledge of bet outcomes actually participate in the market" - not "can some people see the future of any bet better than others".
"predicting the future" and "correct analysis of all available information" often aren't all that different.
From Schlock Mercenary (quoted from memory, may be inexact):
"You cannot see the future. All we are given is the present."
"Of course. But if you look closely at the present, you can find loose bits of the future just laying around."
Not really. Not all players in prediction markets are rational players. A good chunk of it are there for entertainment, and analyze things incorrectly; you can take the other side of those trades, and you won't need to predict the future.
Deciding that someone else's prediction is wrong is a prediction in and of itself.
Yes, but the alternative (that some people are very good at forecasting) is also plausible. It's also useful to have a good prediction model and timely data sources when providing liquidity. We also find that some of the "biggest losers" also provide liquidity; they just aren't as good at it.
The stock market is arguably zero sum as well, just that directionally betting on the US has generally worked during the golden years of the US economy.
The stock markets of the world aren't a money printer.
The stock market is not in the least zero sum. That's just a fundamental misunderstanding. There's dividends, capital allocation, etc.
They can be in cases where investment lenders don’t have 100% capital requirements, but that’s generally no different from other banks.
There's probably also some hedging going on across accounts that look like directional bets.
We study trading gains and losses on Polymarket, the largest prediction market. Using 588 million trades ($67 billion in volume), we show that the gains are highly concentrated: the top 1% of users capture 76.5% of profits. Successful traders provide liquidity using limit orders that resolve favorably relative to realized outcomes while unsuccessful traders take liquidity using market orders. Monthly performance is weakly persistent, however, this may represent sample selection rather than skill. A detailed analysis of the trading behavior of the most successful accounts suggests that "insider'' trading is unlikely to explain the performance of the largest winners.
Full dataset available at https://huggingface.co/datasets/vgregoire/polymarket-users
insider trading on events probably wouldn't show any trends, right? These are point in time events (they call them markets), but they are finite and short lived. An insider would be a one and done thing, so it would be pretty hard to spot them or trend any sort of month over month insider scheming imo.
Also...
> We study trading gains and losses on Polymarket, the largest prediction market
This is not a natural thing to say and I fucking hate that it's impossible to know anymore if I'm wasting time replying to an AI/bot or not
Not meant to sound like AI, but most academic journals limit abstracts to 100 words, so they rarely feel natural...
I agree: insiders are hard to study because they are finite and short-lived. We're pretty confident there are insiders out there trading on Polymarket; however, our conclusion is that they don't account for a significant fraction of the total trading gains on the platform.
I agree - you're not going to be an insider on a significant proportion of trades and it would be stupid to use the same account for more than a couple.
Insiders are going to be earning large amounts in single trades, either by betting a lot when it's odds-on or a small amount when it's out the odds (for a large return).
I think it's just bad tense, which I think makes it not AI amusingly.
> the top 1% of users capture 76.5% of profits
This seems to be similar to OnlyFans, and the economy at large...
Yes, power laws are everywhere. The exact shape of each distribution varies, however, and little is known empirically about the distribution of trading profits in financial markets.
Yeah if you look at the Boltzmann Wealth Model, where every actor gives away 1 dollar to a random person, and you repeat this, then if you start with an equal wealth distribution, you end up with an exponential wealth distribution. That shows how strong exponential curves are :) A few "lucky" individuals become very wealthy, while the vast majority of people end up with very little or nothing.
The effect is so strong that I'm starting to wonder if we should have laws against power laws, like we have in engineering when we try to make things stable.
I mean, do we want the economy to be stable?
Not in a 'oh the rich don't so they control the media and so we don't' sorta way. But like in a 'lets educate people on the pluses and minuses, debate a while, and then come to an informed conclusion' sorta way.
Like, deep down, does the average person actually want a stable economy? Because it seems to me that there is an even split historically between the folks that want stability and a little patch of land and weekly rhythms, and the folks that just want to drunkenly burn couches in the street every full moon, or some such thing.
Not to be glib here at all. I like, would actually like to know the answer. Sorry if this comes off the cuff seeming.
I have a dumbed down version of this question as variant of the Voight-Kampff test (Bladerunner) that goes like this.
You have 2 choices for how the world is shaped, pick 1:
A. You have a modest but comfortable home, a job that pays you enough so that you have what you need and can afford occasional luxuries (e.g., an annual holiday abroad), have good health insurance, access to education and childcare, etc. Everybody else has the same thing, and because of this you live in communities where the arts flourish because nobody has to worry about becoming homeless or destitute.
B. You live in magnificent mansion, one of dozens you own around the world (accessible via one of your personal Gulfstream jets). You have more money then you could ever spend in a lifetime (even recklessly). Your homes are staffed with obedient servants who cater to your every desire. I mean anything. You own them. Your mansions are on palatial estates with secure walls and guards to keep out the rabble outside -- who fight for scraps and are desperate enough to do any kind of work to keep your factories humming and printing cash.
I wouldn't hesitate to choose A because that's a world I'd love to live in and the world of B horrifies me. I don't say this as virtue signaling, it's my innate reaction.
I think that a significant portion of the population would love to choose B. And in some ways, some already have.
I think most people want to earn more the harder they work, and I think that is fine.
However, power laws basically spoil it because it gives a hard worker an exponential advantage, where they can (and will) use that money against other people who made different life choices.
>The effect is so strong that I'm starting to wonder if we should have laws against power laws
This is literally and unironically communism.
Why? You want to earn exponentially more than other people the harder you work? Instead of just linear?
these apps should load with this pie chart showing your likelihood of ever making a profit based on what they know about you. "YOU WILL LOSE MONEY ON THIS APP". Like the cigarette packs.
This is (possibly what you’re thinking of) a requirement in the EU for CFD trading providers. Providers have to (somewhat prominently) state in all of their ads what percentage of traders loses money using the product.
Interesting, would love to see that applied here. Every third ad is about how I can make money on the weather.
Just look at lotteries.
I don't believe in them because when you consider operational costs, less money comes out of the lottery than goes in, so if everyone simply didn't bet on the lottery, they would have more money than if they bet on it.
But everyone who bets thinks "but what if I win?"
What's the baseline here - in a world where every person is betting randomly X times a month, what would the distribution look like? There'd still be a small percentage that wins most of it, right?
We don't know the exact benchmark, but your insight is correct. We provide a simulation similar to what you have in mind towards the end of the paper, but you can generate almost any distribution you want by fine-tuning a simulation...
To relate it to a more-general economic article that has stuck with me for a while:
> If you simulate this economy, a variant of the yard sale model, you will get a remarkable result: after a large number of transactions, one agent ends up as an “oligarch” holding practically all the wealth of the economy, and the other 999 end up with virtually nothing.
https://www.scientificamerican.com/article/is-inequality-ine...
Wait- why isn't there any conflict of interest statement provided in this paper?
Because it's not required and not common practice in our field at this stage. But none of us (I'm one of the authors) is affiliated with or has a financial interest in any prediction market platform.
Isn't it common practice and required to disclose a conflict of interest? Just not to explicitly say there are none.
Yes, when you submit for publication. In our field, you rarely see one for pre-prints, unless you have one to disclose.
Thanks for the clarification. Given the scrutiny on these platforms, this is timely done. Thanks.
the market wins
Just curious but how are bets arbritated on these website?
Meaning who decides if an outcome was yes or no? Answers to things like "Who will win the next Best Picture Oscar?" are fairly obvious and binary.
Can we make bets whose answers are not binary yes/no?
What about "Will celebraty X and Y break up?"? Does Polymarket go to X and Y to confirm if they broke up or something :D
Interesting. Any (in)famous UMA vote debates? Or interesting Unknown/50-50 outcome resolutions?
Check Matt Levine's excellent "Money Stuff" column. He dissects a few examples (Ayatollah Khamenei, a recent public referendum, maybe more).
First one that came to mind: https://www.theguardian.com/world/2026/jan/07/wager-platform...
In terms of damage to society it's irrelevant who the winners are within the Polymarket system, it matters how much the insiders playing on Polymarket have an effect to the outside world of politics and economics. If Polymarket gambling increases corruption and destructive effects on society it simply has to be regulated or made illegal.
[dead]