Polymarket, Kalshi : A Study Reveals the Extent of Losses on Prediction Markets

Prediction markets sell a simple promise. Bet on the future, read the crowd, win thanks to better intuition than others. But the latest data on Polymarket tells a much less flattering story, at the very moment Kalshi and the entire sector are attracting more political and institutional attention.

High volume, very few winners

On Polymarket, the majority lose. The analysis published on April 6, 2026 by Andrey Sergeenkov estimates that 84.1% of traders are in the red. In other words, fewer than one in six users shows a positive profit.

What is most striking is not only this ratio. It is especially the weakness of really significant gains. Out of 2.5 million addresses studied, barely 2% exceeded $1,000 in cumulative profit. This finding is not surprising in an ecosystem sometimes driven by shocking bets, such as those concerning the death of an American pilot. Only 0.32% of addresses have crossed $10,000. And only 840 addresses, or 0.033%, exceeded $100,000, according to Sergeenkov.

The idea of a steady income from these platforms therefore looks more like a showcase than a reality. Sergeenkov shows that only 0.98% of traders earned more than $5,000 over a month. And only 0.26% report an average monthly profit above this threshold. Even among the best, the duration remains short.

Why the gap is widening on Polymarket

The drop does not come out of nowhere. The study links this deterioration to the explosion in the number of users after the American election in November 2024. More noise, more new entrants, more impulsive positions. In this type of market, that is enough to reinforce the advantage of already experienced profiles.

Academic work goes in the same direction. A study posted on SSRN at the end of 2025 already estimated that only about 30% of traders made a positive profit on Polymarket. Another preprint dated January 2026 shows extreme concentration of gains: the top 1% captures 84% of all profits. The market can therefore be informative while remaining brutal for less experienced participants.

These figures still need to be read correctly. Sergeenkov’s study reasons by addresses, not necessarily by people. It also only measures realized profits, which can temporarily classify as losses positions still open. But even with these limits, the trend remains clear: the promise of a level playing field seems highly exaggerated.

This finding comes at the worst time for the sector. In Washington, several lawmakers now want to tighten controls on prediction markets related to war, terrorism, assassinations, or sensitive public decisions.

The BETS OFF Act, introduced in March 2026, targets precisely this type of contracts. The DEATH BETS Act follows the same logic by wanting to more explicitly ban bets on death and war. Meanwhile, the American regulatory debate is not limited to these abuses: on the crypto side, the CLARITY Act remains the other major file, with the ambition to impose a clearer federal framework on digital assets. This contrast is revealing. Washington is seeking both to better regulate financial innovation and to close the door to products deemed most toxic.