- Kalshi’s CPI markets delivered 40% lower forecast error than the Wall Street consensus.
- During volatile CPI shocks, market forecasts beat economists by up to 67% accuracy.
- When Kalshi diverged from consensus, CPI surprises followed about 80% of the time then.
Kalshi released an internal study claiming its prediction market forecast inflation more accurately than Wall Street. The report compared Kalshi’s CPI markets with consensus estimates from economists over 25 months, from February 2023 through mid-2025. The study examined accuracy, timing, volatility, and surprise outcomes to explain how traders outperformed professionals.
Study Compares CPI Accuracy During Volatile Periods
Kalshi analyzed year-over-year Consumer Price Index forecasts priced on its platform against Wall Street consensus estimates. The study measured forecasts one week before each official CPI release. Over the full period, Kalshi’s market-based estimates showed a 40% lower average error.
However, the performance gap widened during volatile economic periods. When CPI readings deviated sharply from expectations, Kalshi’s forecasts outperformed consensus estimates by as much as 67%. These results focused on episodes labeled as “shock” events within the dataset.
The report also tracked forecast disagreement before CPI releases. When Kalshi’s estimate differed from consensus by more than 0.1 percentage point, surprises became more likely. In those cases, actual CPI outcomes diverged significantly from expectations about 80% of the time.
By comparison, the baseline probability of a surprise stood near 40%. Therefore, disagreement between markets and analysts appeared to signal higher uncertainty. Kalshi attributed this pattern to how markets respond quickly to changing conditions.
The study acknowledged a limited number of large shocks. However, it emphasized consistency across multiple volatile periods. They argued that forecasting advantages emerged when conditions became more difficult.
How Prediction Markets Aggregate Information
Unlike traditional forecasts, prediction markets aggregate individual views through real-money trading. Kalshi and Polymarket allow participants to buy and sell contracts tied to specific outcomes. Prices adjust continuously as traders react to new information.
According to the study, institutional forecasts often rely on similar datasets and models. As a result, consensus estimates may adjust slowly when conditions shift. However, prediction markets reflect inputs from traders using varied information sources.
Notably, traders may incorporate sector trends, alternative data and short-term signals. Because participants risk capital, incorrect predictions carry direct costs. Therefore, incentives align closely with forecasting accuracy.
The study also noted timing differences. Wall Street consensus estimates typically lock several days before releases. In contrast, market prices update in real time until publication. This structure reduces lag during fast-moving economic changes.
External research supports similar findings. Earlier this year, a data scientist reported Polymarket achieved 90% accuracy one month before events. Accuracy reportedly reached 94% shortly before outcomes. However, researchers also cited risks from herd behavior and low liquidity.
Related: Phantom Integrates Kalshi Markets for On-chain Event Trading
Market Growth, Research, and Industry Context
Kalshi’s study arrived amid fast platform growth. Earlier this month, the company raised $1 billion at an $11 billion valuation. This followed a $300 million funding round in October, valuing the firm at $5 billion.
Investors include Sequoia Capital and Andreessen Horowitz. Meanwhile, Kalshi expanded distribution through integration with the Phantom crypto wallet. Coinbase is also preparing to roll out prediction markets powered by Kalshi.
Kalshi also formed the Coalition for Prediction Markets with Crypto.com. International expansion plans target multiple country launches within 18 months. However, regulatory challenges persist in states including Connecticut and Nevada.
Kalshi also launched a formal research initiative alongside the study. Researchers from Harvard, Stanford, Yale, and the University of Chicago are involved. Calls for academic papers and conference registration are now open.
The study found Kalshi matched or beat Wall Street estimates 85% of the time, one week before releases. During volatile periods, the mean absolute error fell by roughly 50%. The report emphasized application rather than replacement.
Market-based forecasts may complement existing tools during uncertainty. Data from prediction markets could support traders, policymakers, and executives preparing for sudden inflation shifts.
Kalshi’s study consolidates evidence across accuracy, volatility, and timing metrics. The findings compare market pricing with professional consensus over defined periods. The data outlines how prediction markets performed during stable and stressed inflation environments.
