Kalshi launches research arm as debut study shows 40% outperformance over Wall Street

Prediction market platform Kalshi launched Kalshi Research today, a new division aimed at supporting academic study of forecasting markets by providing researchers access to its proprietary internal data.

The company operates what it calls the world’s largest repository of regulated prediction market data and aims to position itself as a research hub similar to OpenAI or Anthropic.

The company also announced the first-ever Prediction Market Conference, which will bring together academics, traders, forecasters, and policymakers to explore emerging insights in the field.

Alongside the launch, Kalshi published its first internal study comparing its inflation market forecasts with Wall Street consensus.

The study found Kalshi forecasts outperformed traditional economic forecasts by 40.0% overall, matched or beat the consensus in 85.0% of inflation prints one week out, and delivered stronger accuracy during surprise prints, cutting mean absolute error by 50.0% in volatile periods.

Researchers from Harvard, Stanford, Yale, and the University of Chicago are already involved with the initiative. Kalshi says calls for papers and conference registration are now open to the public.