在 2024 年 10 月上诉法院裁定 Kalshi 与选举挂钩的合约不属于赌博之后，预测市场迅速爆发，由 Kalshi 和 Polymarket 领军。月交易量从 2025 年 1 月的 $1.2 billion 升至一年后的逾 $20 billion，名义价值接近 $30 billion；Kalshi 在 2026 年 3 月以 $22 billion 估值募得 $1 billion，而 Polymarket 则寻求以隐含 $15 billion 的估值融资。但产品组合与 2008 年的愿景明显分歧：Kalshi 交易中超过 80% 与体育相关，而这些平台也贩售提及市场及其他类似赌博的合约。斯坦福一项对 4,160 万笔 Kalshi 交易的研究发现，做市商依赖大量乐观者的供给，单一标的合约有 63.4% 的时间对他们有利结算，而更广泛的合约看起来则稍微公平一些。
主要的政策争议在于，这些产品究竟是避险工具、预测工具，还是伪装成赌场。证据显示，预测市场既可能准确，也有许多使用者会亏钱：Bloomberg 发现，自去年年初以来，超过 100,000 个 Polymarket 帐户至少亏损 $1,000，且在机器人攫取大部分利润后，总损失达 $131 million；多伦多大学研究指出，自 2022 年以来，约 69% 的 Polymarket 交易者亏了钱，而哥伦比亚大学与海法大学估计，在 2 年内的 93,000 个市场中出现约 $143 million 的异常获利。Fed 表示，Kalshi 隐含的通膨与利率预测可与既有基准相当，但学者警告存在「prediction laundering」、内线优势，以及类似赌场的危害。监管机构与法院仍在划定界线，第三巡回法院的一项裁决偏向 Kalshi，而最高法院很可能在 1 到 2 年内作出决定；同时，Bloomberg Intelligence 认为，到 2030 年年交易量可能超过 $1 trillion，而 2026 年约为 $300 billion，高于 2025 年的 $51 billion。
Prediction markets were once championed by 22 economists in a 2008 Science manifesto, including Kenneth Arrow, Paul Milgrom, Thomas Schelling, Robert Shiller, and Vernon Smith, who argued that markets with stakes capped at $2,000, or $3,000 in today’s dollars, could improve forecasts and should be freed from unnecessary regulation. The idea had long pedigree, with Justin Wolfers’ 2004 paper finding betting markets beat pre-election polls in presidential races, and advocates hoped these markets would mainly support public-interest forecasting rather than gambling. For years, US gambling laws and CFTC restrictions kept the industry small, while university-run exchanges survived only on limited trading.
After an October 2024 appeals court ruling that Kalshi’s election-linked contracts were not gambling, prediction markets exploded, led by Kalshi and Polymarket. Monthly trading volume climbed from $1.2 billion in January 2025 to more than $20 billion a year later, with face value near $30 billion; Kalshi raised $1 billion at a $22 billion valuation in March 2026, while Polymarket sought funding at an implied $15 billion value. But the product mix diverged sharply from the 2008 vision: more than 80% of Kalshi trading is sports-related, and the platforms also sell mention markets and other betting-like contracts. A Stanford study of 41.6 million Kalshi trades found market makers depend on a large supply of optimists, with single-name contracts settling in their favor 63.4% of the time, while broader contracts look somewhat fairer.
The main policy dispute is whether these products are hedging tools, forecasting tools, or casinos in disguise. Evidence suggests both that prediction markets can be accurate and that many users lose money: Bloomberg found more than 100,000 Polymarket accounts lost at least $1,000 since the start of last year, with total losses of $131 million after bots captured much of the profit; University of Toronto research says around 69% of Polymarket traders have lost money since 2022, and Columbia and the University of Haifa estimated about $143 million in anomalous profit across 93,000 markets in 2 years. The Fed says Kalshi-implied inflation and rate forecasts are comparable to established benchmarks, but scholars warn of “prediction laundering,” insider advantages, and casino-like harms. Regulators and courts are still drawing the line, with a Third Circuit ruling favoring Kalshi and a likely Supreme Court decision in 1 to 2 years, while Bloomberg Intelligence sees annual volume potentially above $1 trillion by 2030 and about $300 billion in 2026, up from $51 billion in 2025.