Business news and discussion on TalkFuse.
Prediction markets get Democratic win right, with a few surprises
Prediction markets broadly anticipated Democratic victories in the US midterms, including the party taking control of both chambers of Congress, but they underestimated Democrats in several races. The results tested whether markets can turn betting odds into reliable election forecasts and showed that their overall performance was broadly comparable with opinion polls.
Democrats won Senate seats in Texas and Ohio, while several candidates beat market expectations: Annie Andrews defeated a South Carolina opponent who had been given about a 70 per cent chance of winning, and Gina Hinojosa unseated Texas Governor Greg Abbott. Analytics firm BETINT identified 14 known Democratic upsets over candidates favoured the previous week, among more than 400 races with known results. Kalshi reported more than $2.5 billion in election-market trading this cycle, while critics say speculation, limited liquidity and concentrated bets can distort prices.
In short
- Markets anticipated a Democratic sweep but missed some close races.
- Democrats repeatedly outperformed the odds assigned to them.
- Kalshi reported over $2.5 billion in election-market trading.
Steelman: both sides, in good faith
The strongest fair case each way — we don't pick a winner.
The case for
Prediction markets aggregate the views of people willing to risk money, giving participants an incentive to assess information carefully and update forecasts quickly. Their broad anticipation of Democratic gains, alongside performance comparable with polls, suggests they can provide a useful additional signal, and substantial trading may help capture diverse expectations.
The case against
Market prices are not necessarily reliable probabilities: speculation, limited liquidity and concentrated bets can push them away from the best available estimate. The unexpected results in several races show how markets can miss local dynamics, while large trading totals alone do not establish forecasting accuracy.
Explained simply
Prediction markets are platforms where people trade contracts linked to whether an event will happen, such as a candidate winning an election. The prices imply the likelihood traders collectively assign to each outcome, though they can shift as people place or withdraw bets.
In the United States, elections decide who represents voters in Congress and who leads state governments. Congress has two chambers, the House of Representatives and the Senate; control of either can affect which laws are considered and passed. Political parties therefore compete for seats as well as for governorships.
Election forecasts also come from opinion polls, which ask voters how they plan to vote. Comparing polls with market prices can help assess how well each method reflects likely results, while market prices may be affected by trading patterns as well as information about the candidates.