EXCLUSIVE: A new study finds that political prediction market traders are not representative of all voters, but election market prices are more accurate than leading poll-based methods and not politically biased. The study found: ✅ In its first month of listing election contracts in 2024, Kalshi's prices beat every poll-based forecast (Silver Bulletin, FiveThirtyEight, The Economist) in the national race, all seven battleground states, and 19 Senate races ✅ A trader who believed the poll-based models and traded against Kalshi's national price would have lost 92 to 100 cents per dollar. ✅ No poll-based forecast would have been profitable to trade against Kalshi in any group of presidential contests. 💰 @Kalshi's 2024 election traders were 86% male and half were under 30. 🏠 Their home zip codes voted 28% for Trump. A market that priced its traders' politics would have leaned toward Harris, not Trump. Instead, 60% of the money went to Republican positions, and Kalshi priced Trump as the favorite through the final month. 🔢 The study used data from @Kalshi in 2024, @PredictIt in 2016, and New York betting markets from 1884 to 1940. ✍️The study was authored by @HarryDCrane, Professor of Statistics at @RutgersU and member of @CFTC Innovation Advisory Committee & @KolemanStrumpf, Burchfield Presidential Chair of Political Economy, Department of Economics, @WakeForest University. Story via @Rotowire Below