Kalshi Prediction Markets Outperform Wall Street on CPI — 40–60% Lower MAE and Strong Early Shock Signals

Kalshi Research compared Kalshi prediction-market implied forecasts to Wall Street consensus for year‑over‑year CPI across 25+ monthly releases from Feb 2023 to mid‑2025. Key findings: Kalshi market‑based forecasts delivered a 40.1% lower mean absolute error (MAE) on average versus analyst consensus across the sample. During inflation “shocks” (surprises >0.1–0.2 percentage points) the MAE advantage widened — about ~50% one week before release and ~56–60% one day before release. When Kalshi’s price and consensus diverged by >0.1pp one week out, the probability of a significant surprise rose to roughly 81–84%, and Kalshi forecasts were more accurate in about 75% of such cases. Authors attribute the edge to participant heterogeneity (wisdom of crowds), stronger monetary incentives in prediction markets (profit/loss alignment vs. reputational herding), real‑time pricing and broader information aggregation. Limitations include a relatively short 25‑month sample and few extreme tail events, which constrain statistical confidence for rare shocks. For traders, the report implies prediction markets like Kalshi offer earlier, more accurate signals of CPI surprises; measurable divergence between market prices and Wall Street consensus can serve as a quantitative early‑warning indicator to adjust macro risk positioning, volatility hedges and rate‑sensitive crypto exposure. Primary keywords: prediction markets, CPI forecast, Kalshi; secondary keywords: market‑based forecasting, Wall Street consensus, real‑time pricing, Phantom integration.