Kalshi’s inflation forecasting edge draws attention in Fed economists’ study

Kalshi outperformed Bloomberg consensus forecasts for headline CPI in a working paper by Fed economists and researchers at Northwestern and Johns Hopkins, according to Webull. The improvement was statistically significant, putting prediction-market pricing in direct comparison with a widely used professional forecasting benchmark.

Results vary by economic indicator

The findings were more evenly matched for other economic releases. According to Webull, the researchers found Kalshi performed roughly on par with Bloomberg consensus on core CPI and unemployment, rather than establishing the same advantage reported for headline CPI.

That distinction matters when interpreting the results: the reported evidence supports an edge on headline inflation, not a uniform lead across all the economic indicators examined.

A strong record immediately before Fed decisions

The paper also found that Kalshi’s rate predictions had a perfect track record on the day before each FOMC meeting since 2022, according to Webull.

The timing is central to that finding. It describes predictions immediately ahead of meetings; the supplied summary does not establish whether the same accuracy held over longer forecasting horizons.

Webull’s account raises the prospect of Kalshi becoming a broader source of macroeconomic expectations comparable to Bloomberg. The reported findings offer evidence of forecasting competitiveness, while leaving that broader role as a possibility rather than an established outcome.

What it means for prediction markets

• Performance is indicator-specific: The statistically significant advantage concerned headline CPI; core CPI and unemployment results were roughly level with Bloomberg consensus.
• Forecast horizons matter: The perfect rate-prediction record applies to the day before each FOMC meeting since 2022, not necessarily to earlier market pricing.
• The evidence has limits: The supplied summary gives no sample sizes or methodological detail, limiting assessment of how broadly the findings apply.