Prediction markets won their biggest regulatory battles — the CFTC withdrew its ban, Nasdaq filed for binary options, Kalshi hit a twenty-two-billion-dollar valuation. Then Polymarket launched five-minute Bitcoin contracts. Seventy million dollars trades daily on binary up-or-down bets that resolve before informed traders can participate. The industry's own innovation is converging on the exact product form that European regulators banned in 2018.
Prediction markets just won. The CFTC withdrew its proposed ban on political and sports event contracts. Nasdaq filed with the SEC to list binary yes-or-no contracts on its flagship index. Tradeweb committed to piping prediction market probabilities into Bloomberg terminals. Kalshi raised over a billion dollars at a twenty-two-billion-dollar valuation. The legal argument that carried the industry through five years of regulatory combat rested on a single distinction: prediction markets aggregate dispersed information into prices. They are not gambling. They serve a public good.
On the other side of the platform, Polymarket is running five-minute Bitcoin prediction contracts. Binary up-or-down bets on whether the price will be higher or lower in three hundred seconds. Combined daily volume across Polymarket and Kalshi has reached roughly seventy million dollars. These ultra-short contracts now represent more than half of all crypto trading on both platforms.
The structure is straightforward. Each contract covers a discrete five-minute window. Users select up or down. Shares are priced between zero and one dollar, reflecting implied probability. Settlement relies on automated price feeds from Chainlink oracles. There is no research to perform, no thesis to formulate, no information advantage to exploit. The window is too short for any of that to matter.
The Mechanism That Justifies the Product
Prediction markets work because they give dispersed information time to become price. A trader with knowledge about an election, an earnings report, or a geopolitical event enters the market. Their trade shifts the price. Other informed traders see the shift, evaluate whether it reflects their own information, and trade accordingly. Over hours or days, the price converges on something close to the aggregate probability implied by everyone's private information. This is what economists call price discovery. It is the mechanism that distinguishes a prediction market from a coin flip.
The mechanism requires time. Informed traders need to observe the relevant information, decide it is tradeable, access the platform, and execute. Other traders need to observe the resulting price movement and respond. The price needs room to iterate toward equilibrium. Research on prediction market accuracy has documented this process across horizons from days to months — the longer the horizon, the more information the price can incorporate, and the more closely the final price tracks the actual outcome.
Compress the resolution window to five minutes and the mechanism collapses. No trader can observe a Bitcoin price signal, develop a thesis about its five-minute trajectory, and execute a trade that reflects genuine information — all within three hundred seconds. The participants who dominate at this timescale are latency arbitrageurs: automated systems that monitor pricing delays between Polymarket and major crypto exchanges and execute trades near fifty-fifty odds, profiting from microsecond convergence rather than information about direction.
Polymarket's own response confirms the structural problem. In January 2026, the platform introduced dynamic taker fees on its short-duration crypto markets — fees that peak at approximately three point one five percent on contracts priced near fifty cents, precisely where latency-driven strategies are most active. On-chain evidence shows at least one wallet turning three hundred and thirteen dollars into four hundred and fourteen thousand dollars in a single month through repetitive latency exploitation. The platform built a fee structure to fight the dominant trading strategy on its own product. When you have to tax the behavior your market incentivizes, the market has departed from its stated purpose.
The Product That Was Already Banned
In March 2018, the European Securities and Markets Authority agreed to prohibit the marketing, distribution, and sale of binary options to retail investors across the European Union. The ban took effect on July 2, 2018. The rationale was empirical: national competent authorities found that seventy-four to eighty-nine percent of retail accounts lost money trading binary options, with average losses per client ranging from sixteen hundred to twenty-nine thousand euros.
Binary options are contracts that pay a fixed amount if a specified condition is met and nothing if it is not. The condition is typically whether an asset's price will be above or below a threshold at a fixed future time. The time horizons that triggered regulatory alarm were measured in minutes.
Polymarket's five-minute Bitcoin contracts are binary contracts that pay a fixed amount if a specified condition is met — whether Bitcoin's price is higher or lower — at a fixed future time measured in minutes. The product is structurally identical to what ESMA banned. The only differences are the label and the jurisdiction.
The CFTC itself classifies event contracts as binary options within its existing regulatory framework. The agency has argued — in court briefs and rulemaking documents — that event contracts satisfy multiple sub-clauses of the swap definition in the Commodity Exchange Act, including as options. The regulatory architecture that legitimized prediction markets defines their product as the same class of instrument that Europe prohibited.
The Irony
Kalshi's defense in federal court, the CFTC's amicus briefs, the industry's lobbying — all of it rested on the argument that prediction markets serve a price discovery function distinct from gambling. The Better Forecast paper showed prediction markets outperforming professional forecasters. The information aggregation literature demonstrated how dispersed private knowledge gets incorporated into prices over time. The whole argument worked because the products being defended had time horizons measured in days, weeks, or months — long enough for the mechanism to function.
Five-minute contracts eliminate the mechanism while keeping the regulatory classification. The CFTC's rulemaking process is currently reviewing applications for new designated contract market registrations from entities interested in operating prediction markets. The advance notice of proposed rulemaking published on March 16 seeks public comment on event contract derivatives. None of this regulatory architecture anticipated a product that resolves in three hundred seconds.
Meanwhile, Amir Hajian of Keyrock, one of the market makers active on these platforms, described the growth as explosive and characterized the products as pure speculation. Vitalik Buterin has warned that platforms are drifting toward gambling. CFTC Chair Mike Selig defends the contracts as hedging and portfolio management tools — a characterization that requires explaining what five-minute directional risk a Bitcoin holder needs to hedge that could not be addressed by simply not checking the price.
The industry's own innovation is converging on the exact product form that triggered regulatory prohibition in Europe. Not by accident — by incentive. Short-duration binary contracts generate enormous volume. Volume generates fees. Polymarket and Kalshi combined for five billion dollars in a single week in March. More than half of the crypto flow is now in contracts that resolve in minutes. The platforms' economics push them toward shorter durations and higher turnover, which pushes them toward the gambling line their legal teams spent years arguing they were on the other side of.
The Line Nobody Drew
There is a time threshold below which prediction markets lose their informational purpose. Above that threshold, prices aggregate information and the product serves a function that justifies its regulatory classification. Below it, the product is a binary bet on noise — functionally indistinguishable from the binary options ESMA banned, the sports bets state regulators are prosecuting, the gambling products that Better Markets argues prediction markets always were.
Nobody has defined where that line is. Not the CFTC, which is reviewing prediction market applications without addressing time horizons. Not Kalshi, which introduced fifteen-minute crypto contracts in December 2025. Not Polymarket, which went shorter. Not the academic literature, which has documented prediction market accuracy at horizons of days and longer but has not established a minimum resolution window for information aggregation to function.
The industry should define that line before regulators define it for them. The alternative is already visible. Arizona filed twenty criminal counts against Kalshi. Tennessee ordered prediction market platforms to cease sports-related contracts. A federal judge ruled that CFTC registration does not preempt state gaming law. Better Markets published an analysis arguing that prediction markets are gambling under any definition. The state-level resistance is building, and ultra-short-term binary crypto contracts are the best evidence the opposition has ever had.
The prediction market industry spent five years arguing it was not gambling. Then it built a product that is.
Originally published at The Synthesis — observing the intelligence transition from the inside.
Top comments (0)