Wall Street Quants Discover Edge in Polymarket Earnings Forecasts
Every financial quarter, Wall Street's armies of analysts laboriously construct financial models, scrutinize alternative data streams, and compete for executive access in their quest to predict corporate earnings. Emerging research now suggests that anonymous bettors on prediction market platform Polymarket might be outperforming these traditional analysts, offering a potentially superior forecasting tool for investors.
Quantitative Research Reveals Striking Accuracy
A comprehensive report from brokerage Wolfe Research has uncovered compelling evidence about the predictive power of Polymarket users. The study found that when Polymarket participants bet that companies would miss earnings estimates, those firms actually missed their targets 44 percent of the time. This figure more than doubles the historical benchmark of 18 percent for such misses.
Even more impressive, when Polymarket bettors expressed high confidence that a company would exceed earnings estimates, that prediction came true 90 percent of the time, significantly surpassing the 81 percent norm observed in traditional analysis.
"The accuracy is possibly due to crowdsourcing," explained Yin Luo, who leads quantitative research at Wolfe Research. "In this case, investors betting on Polymarket earnings releases are likely to be much more diverse than consensus earnings, which are based only on sell-side analysts."
Academic Research Confirms the Trend
This finding represents the latest indication that prediction markets could evolve into valuable information sources for investors, potentially emerging as rivals to traditional sell-side analysts whose primary function involves earnings forecasting. A working paper from researchers at London Business School and Yale University, updated in early April, concludes that these nascent platforms demonstrate high accuracy, incorporate new information more rapidly than analysts, and avoid certain biases inherent in Wall Street estimates.
The academic researchers propose several factors contributing to prediction markets' accuracy. Users risk their own capital when placing bets, creating stronger incentives for accurate predictions. The study also uncovered evidence suggesting that those who wager on earnings markets possess unusual sophistication, and insider trading may occasionally influence outcomes.
The Mechanics of Earnings Prediction Markets
Since September, Polymarket has enabled users to place wagers on corporate earnings through yes-or-no contracts concerning whether specific large stocks will beat consensus estimates. To compare the accuracy of these bets against Wall Street predictions, Wolfe Research examined approximately 430 earnings releases covered by Polymarket, representing roughly one-quarter of such events for Russell 1000 companies during the study period.
Predicting whether a company will exceed earnings forecasts presents considerable complexity. Most stocks routinely beat estimates, partly because corporate executives often guide expectations downward to generate positive surprises when actual results are announced.
Current Market Landscape and Future Potential
Despite the promising research findings, earnings-linked event contracts currently represent only a minor fraction of activity on platforms like Polymarket and its primary competitor, Kalshi. In the most recent week, earnings markets generated just $795,315 in volume on Polymarket, accounting for a mere 0.03 percent of total platform activity, according to user-compiled data from Dune Analytics.
Nevertheless, the research strengthens the argument that these novel derivatives could eventually play a significant role on Wall Street, even though sports betting currently dominates prediction market volume. Exchange operators and financial institutions are investing substantially in prediction market platforms, though the sector remains in its early developmental stages.
"The signal they generate will offer an increasingly rich and high-frequency lens through which to study information aggregation, belief formation, and the pricing of uncertainty across virtually every domain of event space that moves the market," the Wolfe researchers wrote about prediction markets' potential.
