Prediction Markets vs Futures Markets: Information Efficiency in FOMC Rate Expectations

Abstract

Prediction markets have grown as alternative venues for trading on macroeconomic events, while Fed Funds Futures remain the institutional benchmark for monetary policy expectations. Yet there is limited systematic evidence on whether these markets produce different expectations or whether prediction markets contain information beyond futures-based measures. This thesis examines both questions using daily implied probabilities from Polymarket and the CME FedWatch Tool across 26 Federal Open Market Committee (FOMC) meetings from February 2023 to March 2026. The analysis compares the two markets across three dimensions: probability spreads, forecast accuracy, and short-run lead-lag relationships. We find that mean spreads are economically small and not statistically distinguishable from zero under meeting-level inference. Forecast accuracy is statistically indistinguishable across all specifications. Short-run predictive relationships appear in both directions, but neither market consistently leads the other, and incremental predictability varies across outcome categories. The thesis contributes a systematic empirical comparison using a four-outcome framework covering hold, hike, moderate-cut, and large-cut scenarios. Overall, the findings suggest that Polymarket provides a comparable aggregate signal to Fed Funds Futures over the sample period, but that the two markets are not perfect real-time substitutes.