Three percent of the accounts on Polymarket took twenty-seven percent of the money. That is from an academic working paper covering $13.76 billion of trades, which sorted out a small group it calls persistently skilled — accounts that repeatedly pushed prices toward outcomes that then happened. They did it by reading public news faster than everyone else, by arbitraging prices that disagreed across related contracts, and by taking the other side of other people's mistakes.
None of those are secrets. They are the ordinary jobs of a trading desk, being performed on a venue that was sold to the public as the opposite of a trading desk — a place where the crowd knows more than the expert, where a nurse in Ohio with a strong read beats the pundit on television. What the data describes is a few fast accounts collecting from a great many slow ones.
And the platforms are now recruiting the fast ones on purpose. Kalshi and Polymarket are courting Wall Street for deeper liquidity, which brings fee revenue and tighter prices. Theis Jensen, the Yale economist who co-wrote the paper, expects the share of traders with any edge to fall from three percent toward below one, leaving hedge funds. Julie Hoover at Bank of America puts it more plainly: tighter spreads mean fewer mispricings to find. This is the arc of every game that turns into a market, the one online poker ran twenty years ago, and the ending is known — the amateurs stop being competitors and become the reason the professionals show up.
Here is what makes it worth sitting with. The professionals are also what make the prices correct, and correct prices are the actual product. Federal Reserve researchers found Kalshi's headline inflation forecast beat the Bloomberg consensus, with its core inflation and unemployment numbers roughly level with it. The paper's own reading is that traders without an edge may do better under sharper competition, not worse, because a well-calibrated price is one they stop overpaying. The thing that ruins it as a game is the thing that makes it work as an instrument.
One gap stays open. Jensen notes that in thin markets a small order moves the price enough to wipe out an institution's own edge, which keeps big firms out of the quiet corners. Whatever is left of the original promise is in there, on contracts too small for anyone serious to bother with.
A market becomes accurate by making it expensive to be wrong. The people paying for that accuracy are the ones who came because they thought they were right.
The thing that ruins it as a game is the thing that makes it work as an instrument.
