How to Make Money on Polymarket: What the Data Shows

Last updated

Most Polymarket accounts do not make money. A Yale and London Business School study of 1.72 million accounts found that about 3% trade with skill, and Galaxy Research found 69% of retail accounts below break-even. Those who profit have an edge that survives the fees. This page sets out what independent studies and our own tests found, what trading costs, and how people try to earn anyway. It reports; it does not tell you what to do, and you can try copying whales on paper, free before any real money is involved.

Do most people make money on Polymarket?

Mostly no. Three studies, all read on 16 November 2026:

  • Yale and London Business School. Researchers at Yale School of Management and London Business School studied two years of Polymarket trades: 1.72 million accounts, 210,322 markets and $13.76 billion of volume. About 3% of accounts showed skill. The authors found it persistent: 44% of accounts that were skilled in one half of the events were skilled in the other, against about 10% for active mutual funds in their parallel test.
  • Galaxy Research (1 October 2026). On about 2.9 million accounts on polymarket.com, 69.2% of retail accounts were below break-even, and retail accounts as a whole were down $338.9 million. The median account lost about 0.5% of the capital it committed. About 4.1% of accounts were automated; they placed 80.8% of orders and finished $246.8 million up.
  • Pew Research Center (22 July 2026). It followed 11,989 active wallets from 7 May to 19 June 2026. The average trade was $6.50, the typical user placed 46 trades over 10 active days, and 58% gained or lost less than $100.

The three measure different things, over different periods, so their figures do not add up to one answer. Yale and London Business School and Galaxy find that most accounts lose; Pew finds that most users stayed within $100 either way. Yale and London Business School and Galaxy also find a small group that does much better than the rest.

Where do the profits come from?

Mostly from the other traders. A Polymarket share pays $1.00 or nothing, so a gain on one side of a trade is a loss on the other, less fees. The market does not create money; it moves it.

The Yale and London Business School paper puts numbers on who ends up where. Skilled takers plus market makers are under 3.5% of accounts and take over 30% of total gains. In The Block's report of the paper, 67% of accounts are unlucky or unskilled losers, and they absorb the aggregate losses. Luck matters too, which is why the figures below about "top earners" are less encouraging than they sound.

Galaxy found a similar split by habit. 28.1% of specialist accounts were in profit, against 30.4% of generalists; among specialists, 25.1% of sports accounts were in profit and 41.2% of tech and science accounts. Accounts were also more likely to go quiet after a loss: 15.2% were inactive for 30 days after a loss, against 6.1% after a win.

What does trading cost?

Polymarket's fees page, read on 16 November 2026, says makers are never charged and takers pay a fee of shares × rate × price × (1 − price). The rate depends on the market's category, and geopolitical and world-events markets are fee-free. The documentation's own example is 100 shares bought at 60¢, which carry a fee of $0.96 in a politics market, $1.20 in sports and $1.68 in crypto. The fee calculator holds the current rates, so that this page does not repeat them and let them go stale.

The fee is not the only cost. Buying at the ask and selling at the bid gives up the spread on a round trip. And the price itself sets the bar: before fees, the break-even win rate equals the price you pay, 80% for shares bought at 80¢ and 30% for shares bought at 30¢. A 70% win rate is strong at 50¢ and a losing one at 80¢.

Our own week-41 report (4 to 11 October 2026, 99,946 settled positions from a directory of 1,049 wallets) shows how different a high win rate and a good result can be. Of the positions bought at 90 to 98¢, 89.7% were in profit, and the group returned -0.9%. Of those bought under 10¢, 13.5% were in profit, and the group returned +18.4%. Across all positions it was 52.3% in profit and +0.5%. It is in our week-41 report (11 October 2026), which win rate against payoff (18 October 2026) reproduces, and one week of settled positions is a snapshot, not a rule. Copy whales on paper and Polymarket's fee is taken out of every result.

What do Polymarket's rewards pay?

Three of Polymarket's programmes pay for how you trade rather than for being right. We describe them from its documentation, read on 16 November 2026, and give no rate that the sources do not agree on.

  • Liquidity Rewards. Resting limit orders are eligible automatically. Each rewarded market sets a minimum order size and a maximum spread from the midpoint, and the score is quadratic, highest near the adjusted midpoint. The order book is sampled once a minute over a UTC day, and two-sided quotes score higher than one-sided ones. Rewards are paid daily, and amounts under $1 are not paid.
  • Maker Rebates. A share of taker fees is paid back to makers in that market. The share depends on the category: 20% in crypto, 15% in sports, 25% in most others, and none in geopolitics, where there is no fee. Fills in a market whose fee is 0 earn no rebate. It is paid daily in pUSD with a $1 minimum, and the percentage is at Polymarket's sole discretion.
  • Taker Rebate Program. Live from 28 May 2026. Your tier comes from your weighted volume over the last 30 days, from no rebate under $2,000 up to 50% at $10 million and above. Only taker trades that pay a fee count, and the category weights can change.

All three are tied to volume, to resting orders or to being a regular taker. They are a way of earning on the cost side, which is a different job from forecasting, and the documentation sets floors: Liquidity Rewards and Maker Rebates pay nothing under $1, and there is no taker rebate under $2,000 of weighted volume. We do not know what any given account earns from them.

What are the main ways traders try to earn?

One line each.

  • Being right more than the price implies. The source of profit for a forecaster. It is also what the 3% in the Yale and London Business School study seem to do.
  • Favourite-longshot bias. A study of more than 300,000 contracts on Kalshi (a different market) by Burgi, Deng and Whelan found low-priced contracts win less often than their price needs, and high-priced ones earn small positive returns. It is a finding about that market, not a Polymarket rule.
  • Arbitrage. Prices that do not add up, or the same event priced differently elsewhere. One study of Polymarket estimates about $40 million of realised profit; it is the authors' own estimate, and we found no independent replication. Gebele and Matthes find persistent gaps between venues averaging 2 to 4%, which they blame on structural frictions such as different event definitions.
  • Making markets. Resting orders, paid in rebates and rewards as above. On Kalshi, the Burgi, Deng and Whelan study found makers lost an average 9.64% against 31.46% for takers; a smaller loss is still a loss.
  • Specialising. Galaxy's figures above: specialists did slightly worse than generalists overall (28.1% against 30.4% in profit), but some categories had more accounts in profit than others.
  • Automating. The 4.1% of accounts Galaxy marked as automated finished well ahead in aggregate. We do not know how their results were spread among them.

Can you make money copying top traders?

It is the obvious shortcut, and the evidence is not kind to it. The Block's report of the Yale and London Business School paper says that 12% of the top earners overlap with the skilled group, and that about 60% of the "lucky winners" went back to losses on a separate sample. It is fair to infer that following whoever tops a leaderboard follows luck more often than skill, though the paper's "top earners" are its own ranking, not a public one.

We tested this ourselves. It is the forward test on our copy trading page: we picked wallets using only copies that settled before 1 September, then checked the same wallets on copies from September to 2 November, which played no part in the picking. It covered 1,036 wallets and about 92,000 copies, each priced with a markup for slippage and the market's fee, and held to resolution. In all three groups of wallets that we show there, the check result was negative: -2.9%, -6.7% and -10.4% a copy in non-sport markets, and -10.9%, -12.1% and -7.5% in sport. The conclusion on that page is that picking wallets by leaderboard rank did not work.

That is one test, over one window, with its own limits, which the copy trading page describes. It does not say copying cannot work. It says the good-looking record you can see in advance did not predict the later one, and past results do not predict future ones.

How do you test an idea without risking money?

If the idea is copying whales, run it on paper first. Whalidate books virtual copies of whale trades at real Polymarket prices, with Polymarket's fee taken out of each result, so the cost described above shows up in the numbers before any money does. It covers copying whales, not market making, arbitrage or news trading. The paper trading page says what a paper result can and cannot tell you. A good paper result is not a promise, and a bad one costs nothing. Polymarket trading strategies goes through the approaches in detail, and how Polymarket works explains the mechanics behind them.

Whalidate is not affiliated with Polymarket. Nothing here is financial advice.

Questions people ask

Is Polymarket profitable?

For most accounts, the studies we read say no. Galaxy Research found 69.2% of retail accounts below break-even on polymarket.com (1 October 2026), and a Yale and London Business School study found about 3% of accounts trade with skill. A minority does profit. Whether you would be one of them is not something a study can tell you, and past results do not predict future ones.

How much can you make on Polymarket?

There is no typical figure to quote, and we do not promise one. What the data shows is the range: Pew Research Center followed 11,989 active wallets in 2026 and found 58% gained or lost less than $100, while Galaxy Research found the median account lost about 0.5% of the capital it committed. Both are reports of what happened, not forecasts.

Do Polymarket's rewards add up to an income?

Polymarket's documentation describes several programmes; three of them are these. Liquidity Rewards pay for resting limit orders and are paid daily; a day's reward under $1 is not paid. Maker Rebates pay a share of taker fees back to makers in that market, paid daily in pUSD, with a $1 minimum accrued. Polymarket says the rebate percentage is at its sole discretion and can change. The Taker Rebate Program rebates a tier of fees by 30-day volume. We give no rate the sources disagree on, and we cannot say what the programmes add up to for you.

Is copy trading on Polymarket profitable?

In our test, usually not. Wallets picked using only copies settled before 1 September were checked on later copies, and all three of the groups we show on the copy trading page came out negative (from -2.9% to -12.1% per copy, after the market's fee). One test is not the last word, and that page describes what it did and did not cover.

Sources