Do Prediction Markets Forecast Cryptocurrency Volatility? Evidence from Kalshi Macro Contracts

2026-04-01

AI Summary

This study provides the first systematic evaluation of Kalshi macro prediction markets’ ability to forecast realized volatility in cryptocurrency markets. Leveraging contract prices related to monetary policy, recession risk, and inflation expectations, the analysis employs time-series regressions, Clark–West tests, MSFE ratios, Benjamini–Hochberg multiple-testing corrections, and orthogonalization techniques. The findings reveal that Federal Reserve rate repricing significantly predicts Bitcoin volatility, recession risk signals exhibit robust out-of-sample predictive power, and CPI repricing forecasts volatility across multiple altcoins. These results demonstrate that macro prediction markets offer a statistically significant and informationally distinct channel—complementary to conventional financial indicators—for forecasting crypto-asset volatility, thereby opening a novel avenue for volatility modeling in digital asset markets.

Abstract

Daily probability changes in Kalshi macro prediction markets forecast cryptocurrency realized volatility through two distinct channels. The monetary policy channel, measured by Fed rate repricing on KXFED contracts, predicts Bitcoin volatility in sample with t = 3.63 and p < 0.001 but exhibits regime dependence tied to the 2024-2025 rate-cutting cycle. The recession risk signal from KXRECSSNBER proves more stable out of sample, delivering an MSFE ratio of 0.979 with Clark-West p = 0.020. The inflation channel, measured by CPI repricing on KXCPI contracts, predicts altcoin volatility for Ethereum, Solana, Cardano, and Chainlink with t-statistics ranging from -2.1 to -3.4 and out-of-sample gains for Ethereum at MSFE = 0.959 with p = 0.010 and Solana at p = 0.048. Both the Bitcoin--Fed-dovish and Chainlink--CPI specifications survive Benjamini-Hochberg correction at q = 0.05. Orthogonalization and baseline comparisons against Fed Funds futures, Treasury yields, and the Deribit implied volatility index confirm that these signals carry information not embedded in conventional financial instruments. The sample covers ten Kalshi event series and six cryptocurrency assets over January 2023 to March 2026.

Problem

• prediction markets
• cryptocurrency volatility
• Kalshi
• forecasting
• macro contracts

Innovation

• prediction markets
• cryptocurrency volatility
• Kalshi macro contracts
• orthogonalization
• out-of-sample forecasting