
A new academic study reveals that just 3% of Polymarket traders are truly skilled and generate most of the platform’s profits. The remaining 97% — mostly unskilled participants — end up funding the gains of this small elite group
Researchers from London Business School and Yale analyzed every single trade on Polymarket from 2023 to 2025. The working paper, released on April 25, 2026, examined 1.72 million accounts, over 210,000 markets, and $13.76 billion in trading volume.
Main takeaways:
The study challenges the popular “wisdom of the crowds” narrative promoted by prediction market platforms. Instead, it concludes that Polymarket’s accuracy comes from “the wisdom of an informed minority.”
Skill was measured using a rigorous sign-randomization test (running each trader’s history 10,000 times with random buy/sell flips). Results showed that raw profit and loss (P&L) is a poor indicator of real skill — only 12% of top earners were actually skilled.
Notably, trader skill is highly persistent. 44% of skilled traders in the training sample remained skilled in a separate test sample — far higher than the 10% persistence rate seen in active mutual funds.
The researchers also identified 1,950 suspicious accounts that opened just before major events and went dormant afterward. These accounts moved prices 7 to 12 times more aggressively than skilled traders per dollar traded.
However, because insider activity is concentrated in isolated events, it does not significantly drive the platform’s overall accuracy.
The stablecoin and related initiatives come as Western Union reported a stabilizing first quarter. The company posted adjusted revenue of $983 million, only 1% lower than the same period last year.
Its stock (NYSE: WU) closed at $8.90 after a 4.6% drop on Friday.
This strict approach aims to protect residents, particularly seniors, from fast-growing crypto-related scams. While it limits legitimate use of crypto ATMs, it sends a strong signal that states are serious about stopping fraud.
