Not just a gambling platform, but also a dark horse in financial report forecasting: Polymarket challenges the position of Wall Street analysts.

In the battlefield of financial forecast predictions, a quiet revolution is underway. Every quarter, large numbers of Wall Street analysts build financial models, explore alternative data, and compete to communicate with company executives, all in the pursuit of accurately predicting corporate financial performance. However, a recent study has shown that players making anonymous bets on Polymarket may be more accurate in predicting company earnings than Wall Street analysts. Why are prediction platforms more accurate than analysts? A report from brokerage firm Wolfe Research indicates that when Polymarket users bet that a company's earnings will fall short of expectations, their accuracy rate is as high as 44%, more than double the historical benchmark of 18%; and when traders are highly confident that a company's performance will exceed expectations, the accuracy of this prediction can reach 90%, higher than the industry average of 81%. Yin Luo, head of quantitative research at Wolfe, said, "This high accuracy may stem from a crowdsourcing model. Investors betting on financial forecasting on the Polymarket platform may be more diverse than the market consensus expectations based on sell-side analyst forecasts." This result once again highlights the potential for prediction platforms to become important sources of information for investors, and even potential competitors to sell-side analysts - whose core job is profit forecasting. A recent working paper updated in early April by researchers from the London School of Economics and Yale University pointed out that these emerging platforms have high accuracy in forecasting, can integrate new information faster than analysts, and can also avoid inherent biases in Wall Street profit forecasts. The researchers believe that the high accuracy of prediction platforms is due to the fact that participants are betting with real money; they also found that users participating in profit forecasts are exceptionally professional. The researchers also said that insider trading may be one of the factors affecting the accuracy of predictions. Despite being small in size, the potential of these platforms has caught the attention of Wall Street capital Regardless of the driving factors behind it, these two studies highlight the immense potential of Event Contracts tied to profits. While currently, the trading volume of such contracts on platforms like Polymarket and its main competitor Kalshi is still relatively small, research has confirmed their value. This further supports the view of the platforms: although sports betting still accounts for the majority of their trading volume, these new financial derivatives will eventually play an important role on Wall Street. Predicting whether a company's profits will exceed expectations can be quite complex. First, most companies' profits tend to exceed expectations, in part because company executives often deliberately lower expectations in order to achieve performance surprises. Since September last year, Polymarket has allowed users to bet on whether the profits of certain large stocks will exceed expectations in the form of "yes/no" contracts. To compare the accuracy of these bets with Wall Street expectations, Wolfe studied about 430 financial reports covered by Polymarket, which accounted for about one fourth of the total number of financial reports of companies in the Russell 1000 index during the same period. The Wolfe research team wrote about prediction platforms: "The signals they generate will provide an increasingly rich and high-frequency perspective. With this perspective, we can almost delve into all areas of events that drive market volatility, studying information aggregation, belief formation, and uncertainty pricing mechanisms." Despite exchanges and other financial institutions pouring investments into prediction platforms, the field is still in its early stages of development. According to data compiled by users on Dune Analytics, the trading volume of Polymarket's financial forecast platform in the past week was only $795,315, accounting for 0.03% of the platform's total trading volume. Vinesh Jha, founder of alternative data platform ExtractAlpha, believes that Polymarket can serve as a supplementary source of information. However, it may be too early to widely apply it to the trading systems of quantitative fund managers. ExtractAlpha also offers profit prediction bets. Jha said, "The market is more concerned with the specific profit numbers and how the market will react, rather than just determining whether performance will exceed expectations. It's too early to draw conclusions, and the related data is still too thin."