The prediction markets never sleep, and over the past 48 hours, they’ve been unusually active, with some dramatic shifts suggesting growing uncertainty—or perhaps surprising clarity—in several hot-button areas. Right now, the highest-volume markets span elections, geopolitics, and tech innovation, particularly on Polymarket, PredictIt, and Metaculus. Each platform brings a distinct flavor, but all are reacting quickly to new information and speculation.
Polymarket remains the clear leader in trading volume, and its election markets are buzzing. The “Will Trump win the 2024 US Presidential Election?” contract leads the way, with over $10 million in volume. As of this morning, the price for “Yes” sits at 54 cents—up from 51 cents just 48 hours ago, following a surge of online interest after a favorable court ruling related to Trump’s immunity case. The bump suggests either increased confidence about his legal path clearing or perhaps that bettors are reading into broader polling movement, even without significant new national numbers released.
Meanwhile, the Biden equivalent market has dropped slightly from 43 cents to 41. This downward drift, while not dramatic, raises eyebrows given the absence of any major scandal or gaffe. Market participants may be responding more to underlying economic sentiment or simply tracking Trump’s media pulse.
On PredictIt, eyes are on a slightly different metric: “Which party will win the 2024 US presidency?” The Republicans are modestly ahead at 52 cents, with Democrats trailing at 48. This gap narrowed sharply on Tuesday from 55 to 52 for Republicans, possibly in response to internal GOP infighting and new fundraising data showing Democrats gaining ground in key swing states.
Comparatively, Metaculus—known for its aggregation of forecaster reasoning rather than monetary bets—is showing Biden with a 41 percent chance to win reelection, aligning closely with Polymarket. But what’s fascinating there is the recent increase in the probability of a "third-party candidate winning at least one electoral vote," which jumped from 3 percent to 8 percent. It doesn’t sound like much, but that’s more than doubling in probability, likely reflecting buzz around RFK Jr. qualifying for ballot access in more states.
The most interesting shift in the past 48 hours is in the "Will AI surpass human performance at all tasks by 2040?" market on Metaculus. This long-term forecast saw a swing from 38 percent to 45, following OpenAI’s recent public demo of a new multi-modal model surpassing GPT-4 in performance benchmarks. The spike hints at how serious forecasters are taking the pace of recent progress, and perhaps that timelines for artificial general intelligence are tightening again.
One emerging trend to watch: the increasing divergence between monetary prediction markets and expertise-driven platforms. While Polymarket prices often reflect sentiment and current media narratives, Metaculus tends to move on technical reports and academic debate. The divergence serves as a reminder—not all predictions speak the same language, and understanding what they're reacting to can be as revealing as the numbers themselves.
続きを読む
一部表示