The New New Thing: Prediction Markets Become the Boring Infrastructure of American Finance

On 10 July 2026, Bloomberg reported that Goldman Sachs had barred its employees from trading prediction-market contracts, including positions tied to macroeconomic data releases and to geopolitical outcomes. The Wall Street giant's compliance memo, circulated internally, extended an existing prohibition on personal trading into a market category that did not exist in any meaningful form a decade ago. The same day, Unusual Whales reported that New York state had banned the use of smart glasses inside more than 1,240 state, county, city, town and village courthouses, citing the risk of covert recording and the integrity of proceedings. Two days earlier, the same outlet had flagged investor Michael Burry, of "Big Short" notoriety, warning that prediction markets are exploiting regulatory loopholes and that oversight and taxation are coming.

What is unfolding in the second half of 2026 is the slow institutionalisation of a product that, until very recently, lived in the grey zone between gambling, polling, and derivatives. Prediction markets are platforms on which users buy and sell contracts whose payoffs depend on the outcome of real-world events: who wins an election, whether the Federal Reserve cuts rates, how many inches of rain fall in a given county, whether a particular piece of legislation passes. Two venues have come to dominate. Kalshi, a New York-based exchange regulated by the Commodity Futures Trading Commission, focuses on event contracts in the American regulatory perimeter. Polymarket, a crypto-native platform, settled in 2025 its long-running regulatory brush with the CFTC over unregistered binary options and now operates under a no-action letter covering event contracts. The contracts look like bets. The exchanges insist they are information markets: aggregators of crowd belief, useful to traders, journalists, and policymakers. The distinction is a legal one, and the law is changing under everyone's feet.

The compliance wall rises

Goldman's prohibition is significant less for the bank's profile than for the category it identifies. The internal ban extends to contracts tied to macroeconomic data and to geopolitics, two of the most heavily-traded verticals on the major platforms. According to the Unusual Whales summary of the Bloomberg report, the move reflects a calculation familiar to any large broker-dealer: when a market sits in a regulatory grey zone, the safest posture for staff is to stay out until the lines are drawn. Goldman is not the only firm thinking this way. A growing number of US banks have, since the start of 2026, restricted employee trading in event contracts, often citing the CFTC's own evolving guidance on what constitutes a swap, a security, or a game of chance.

The pressure is coming from multiple directions at once. The CFTC has signalled, through a series of enforcement actions and advisory letters, that event contracts touching elections, war, and assassination are off-limits, while contracts on inflation prints, sports outcomes, and weather events remain permissible under narrower conditions. State regulators have moved in parallel. The New York courthouse smart-glasses ban sits at the seam between two anxieties: the integrity of physical proceedings, and the integrity of the markets on which those proceedings are priced. A juror wearing a Meta or Ray-Ban device could, in theory, stream real-time signals to a market participant. Whether anyone has actually done this, no public filing admits. The rule, in the meantime, is a blanket prohibition: more than 1,240 courts, one rule, no exceptions.

Burry's read

Michael Burry, the California-based investor who made his name on the subprime trade that became a film, registered his view on 11 July via Unusual Whales, arguing that prediction markets are exploiting regulatory loopholes and that oversight and taxation are inevitable. The remark matters because Burry is, by long habit, contrarian on financial innovation. His core argument is that any market with the volume, the visibility, and the political salience of a major Kalshi or Polymarket contract will eventually be treated as a market by tax authorities, the CFTC, and the Internal Revenue Service. The current structure, in which retail users frequently report winnings as gambling income, or not at all, cannot survive a serious audit cycle.

That is the standard endgame for any financial innovation that crosses a threshold of size. The futures market in interest-rate products, the credit-default-swap market, the cryptocurrency market: all spent periods as grey zones, and all eventually attracted rule-makers, tax collectors, and the court system. The question is not whether prediction markets will be regulated, but which regulator lands the file. The CFTC wants it. The Securities and Exchange Commission, after a slow start, has begun to assert concurrent jurisdiction over contracts whose underlyings resemble securities. State attorneys general, in New York, New Jersey, and Illinois, have opened their own inquiries into marketing practices. The market itself has grown large enough to fund the lobbying effort to shape the eventual rule.

The information claim, taken seriously

The case for prediction markets, made most energetically by the platforms themselves, is that they aggregate information more efficiently than polls, expert panels, or analyst notes. A contract priced at 67 cents for "yes" is, on this view, a probability estimate, continuously updated, with real money behind it. The argument has some empirical support. Academic studies over the past five years have shown that, on a restricted set of well-defined events, prediction-market prices forecast outcomes at least as well as aggregated punditry, and on some classes of event noticeably better. The Polymarket contract on the 2024 US presidential election moved sharply inside the final seventy-two hours, in a direction the national polls had not yet caught. Kalshi's inflation prints have, on several occasions, led the Bureau of Labor Statistics release by a full session.

The defence has its limits. On long-tail, low-liquidity contracts, the market is thin enough that a single large position can move the price dramatically, and the "wisdom of crowds" claim breaks down. On contracts touching single-name corporate events, the markets have a poor track record, in part because the relevant information is concentrated in a small set of insiders. On geopolitics, where the underlying event is itself a function of decisions by a small number of actors, the price of the contract tends to reflect the dominant narrative among professional geopolitical analysts, and the marginal trader adds little. The platforms acknowledge these limitations in their risk disclosures, sometimes in language dense enough to require legal counsel. They also note, accurately, that the same critique applies to any forecasting method, including the kind made by pundits on cable news without the cost of being wrong.

What the next eighteen months will decide

The calendar of pending decisions is now dense. The CFTC is expected to publish revised guidance on event contracts by the end of 2026, with a particular focus on contracts whose resolution turns on a single identifiable decision-maker. The SEC's enforcement division has, according to people familiar with the matter, opened a preliminary inquiry into the marketing of certain event contracts to retail investors, an inquiry that could produce its first public action before the end of the year. State-level regulation is the wild card. New York's smart-glasses rule is a tiny piece of a much larger story: the state is also the most likely venue for a serious challenge to the platforms' compliance with state gaming and consumer-protection law.

The stakes are not only legal. The platforms have become, in a short span, a meaningful source of liquidity for several categories of event. Hedge funds, news organisations, and political campaigns have begun to treat prediction-market prices as inputs, in the same way they treat polling averages, betting lines, and analyst forecasts. If the market is constrained, or taxed heavily, the price signal will degrade, and a small but growing category of decision-making will become less informed. If the market is allowed to scale under a clear rule, the price signal will harden, and a wider set of actors will begin to depend on it. The political economy of the question is unusual. The platforms' users are, on balance, young, male, and crypto-curious, a profile with limited influence on bank compliance committees and a great deal of influence on the consumer-facing politics of any crackdown.

The intergenerational backdrop

Underneath the regulatory and the cultural story is a quieter shift in who has the time and the money to make speculative bets at all. Unusual Whales, on 11 July, summarised a Federal Reserve survey showing that the share of American adults under 30 living with parents or relatives had risen to roughly half by 2024, up from 37% in 2019, a one-third increase in five years. The macro detail is not directly about prediction markets, but the connection is plain. A generation with lower wage growth, higher housing costs, and thinner savings has, in aggregate, less to lose by placing small bets on event contracts, and less cushion for the losses. The platforms' growth, in this sense, is partly a story about an asset class whose customers cannot afford the older ones.

It is also, on a longer view, a story about a financial system that is running out of easy frontiers. The deep, liquid, regulated markets are crowded. The new, illiquid, lightly-regulated ones are not. The pattern repeats: a decade of low rates and abundant capital finds its way into whatever category is left, until that category is large enough to attract its own regulators. The smart money has, in this cycle, moved first into prediction markets, into crypto, and into the small handful of private credit and infrastructure deals that pension funds will tolerate. The retail money has followed, often on the same platforms, often with less information. The regulators are now arriving, on the schedule they always arrive on, a few years late and a few billion dollars in volume too late to prevent the market from forming.

What remains uncertain

The sources for this article are thin in two respects that the reader should know. First, none of the public reporting on the Goldman prohibition provides the full text of the internal compliance memo, and the precise scope of the ban is, on the available evidence, a reconstruction by Bloomberg from people familiar with the document. Second, the Burry remarks are summarised rather than directly quoted, which means the exact phrasing of his prediction about oversight and taxation is in the hands of Unusual Whales' summary, not in a primary statement from Burry himself. The CFTC's timetable for revised guidance, the SEC's inquiry, and the outcome of the New York state-level activity are all moving pieces. The most defensible claim this publication can make is that the institutional phase has begun, and that the shape of the eventual rule is the next eighteen months' biggest open question for the platforms and their users.

This piece is a long read on the institutional phase of prediction markets in 2026. The story sits at the intersection of US bank compliance, federal and state regulatory action, and a generational shift in who has the disposable cash to speculate. Monexus framed the question around the slow arrival of the regulators rather than around any single platform's growth, and steered the report away from the cultural noise that tends to attach to the topic.