- Prediction markets act as a leading indicator for inflation data, a new report finds.
- Traders are incorporating Polymarket and Kalshi forecasts into their models.
- Prediction markets' accuracy is debated, however.
Prediction markets have become a leading indicator for pivotal economic data — and traders are monitoring them to get ahead in the market.
Crypto market maker Keyrock made that argument in a December 16 report that explored the impact of fast-growing platforms like Kalshi and Polymarket.
The report looked at Kalshi’s ‘Inflation in 2025’ market, which lets users place bets on where the 12-month percent change in the Consumer Price Index will fall.
“Rather than a lagging statistical update, Kalshi represents a real-time consensus on future inflation, giving traders an informational edge to anticipate macro turns before they are captured in official forecasts,” the report said.
Traders get ahead
With prediction markets offering potentially market moving insights, traders are incorporating their forecasts into their models.
“Prediction markets are already becoming a valuable data signal,” Nass Diba, co-founder of Worm, a Solana-based prediction market, said in the Keyrock report. “Hedge funds monitor these markets to get ahead of sentiment shifts.”
The Keyrock report compared Kalshi’s prediction market data to Nowcast, the Federal Reserve Bank of Cleveland’s inflation estimate index.
It found that ahead of the April 10 CPI release, Kalshi’s implied inflation forecast climbed from roughly 3.05% to 3.58%, signalling that traders were already positioning for firmer inflation.
Nowcast, by comparison, remained anchored near 3.9% until the data hit, then abruptly dropped to 1.5% once the model absorbed the new inputs.
“This demonstrates how prediction markets synthesise dispersed information faster than backward-fitted models, making them valuable for institutional desks trading TIPS breakevens, inflation swaps, or front-end rates,” the report said.
Accurate forecasts?
It’s not the first reported instance of prediction markets providing more accurate forecasts of future events than conventional sources.
During the 2024 US presidential election, Polymarket gained mainstream notoriety when bettors gave Republican candidate Donald Trump higher odds than most other sources to win the US presidency.
A historical analysis conducted by data scientist Alex McCullough showed Polymarket previously predicted the outcome of world events one month out with 90% accuracy.
Other studies have found prediction markets to be less accurate. A study conducted by researchers at Vanderbilt University found Polymarket got only 67% of markets right, while Kalshi hit 78%, and PredictIt scored 93% accuracy.
Still, prediction markets’ ability to reflect the opinions of crowds, coupled with their broad range of sometimes contentious and "disrespectful" bets, have made them popular with both retail bettors and professional traders alike.
Since early 2024, monthly notional volume has surged from under $100 million to over $13 billion, a 130x increase that places them among the fastest growing sectors in the world, according to the Keyrock report.
Some researchers say analyses of Polymarket have overestimated the platform’s trading volume, however.
The prediction market model of binary yes-no outcomes can also be a double-edged sword.
It works well when there are tens of thousands of bettors putting millions of dollars on the line. But when liquidity dries up it’s a different story.
“Traditional models assume continuous price movements, not discrete yes-or-no outcomes,” Duncan Hennes, managing director at accounting giant KPMG, said in the report.
“When paired with thinner liquidity, this leads to wider spreads, sharper volatility, and more complex margin dynamics.”
Tim Craig is DL News’ Edinburgh-based DeFi Correspondent. Reach out with tips at tim@dlnews.com.
