Prediction markets hit a wall: Goldman bans staff, New York bans the glasses, Burry sees a reckoning

On 10 July 2026, at 22:31 UTC, Goldman Sachs told its employees that they can no longer trade event contracts tied to macroeconomic data or to geopolitics on prediction-market platforms. The directive, internal but disclosed publicly the same evening, extended a longer-running caution inside the bank into a formal prohibition on a category of instruments that barely existed on a retail trader's screen four years ago.

Twenty-three hours earlier, on 9 July, the New York state court system had moved to physically exclude the platforms from its own rooms: a new rule bars smart glasses and similar recording devices from more than 1,240 state, county, city, town and village courthouses, with the explicit motivation of preventing spectators from wagering in real time on courtroom outcomes. And on 11 July, at 03:58 UTC, Michael Burry, the investor who shorted the US housing market before the 2008 collapse, registered a more sweeping view. Prediction markets, he wrote, are currently exploiting regulatory loopholes and will eventually face oversight and taxation.

Taken together, the three moves describe the first serious institutional pushback against an asset class that has, until this summer, expanded faster than any regulator, employer or courthouse could keep up with.

The bank draws the line

Goldman's internal memo lands in a category that has existed, in its current retail form, only since Kalshi and Polymarket emerged as federally designated exchanges in 2023 and 2024. Event contracts let a trader take a position on a yes-or-no question, such as whether the Federal Reserve will cut rates at its next meeting, whether a given candidate will win a given primary, or whether a kinetic strike will or will not happen by a stated date. The prices, set by an order book, double as crowd-sourced probabilities.

For a sell-side bank, that intersection is professionally radioactive. A trader at Goldman with a personal position on, say, whether US CPI prints below 2.5 per cent next month has, at minimum, the appearance of trading on information gathered at work. The bank's prohibition now extends explicitly to contracts tied to macroeconomic data and to geopolitics, the two categories most likely to interact with the firm's own research flow. The constraint is narrower than a blanket ban on prediction markets, but it captures exactly the contracts that retail traders most want.

The move is also a defensive signal. If one of the largest dealers in US Treasuries and rates instruments were perceived to have an insider-information problem in a parallel retail market, the regulatory conversation would not stay friendly. Goldman is closing the door before anyone knocks.

The courtroom blinks

New York's court rule is, on its face, a technology restriction rather than a financial one, but its logic is identical. A spectator in a Brooklyn courtroom can wear ordinary glasses; what they cannot wear, under the new rule, are smart glasses capable of streaming video out of the room in real time, which a co-conspirator outside could feed into a position on a prediction exchange. The scope of the ban is unusually broad for state-level administrative action: more than 1,240 state, county, city, town and village courts.

That breadth is itself a signal. New York's Unified Court System is one of the largest in the United States, with daily dockets that include felony trials, custody hearings, civil liability disputes and the kind of salacious or politically charged cases that move retail betting volume. The rule's authors are not treating the threat as hypothetical; they are treating it as inevitable.

The deeper read is that prediction markets are starting to import the integrity problems of every other market they touch. In sport, the integrity question is match-fixing. In courts, it is witness coaching and evidence tampering. In macro, it is front-running. The New York rule is the first jurisdictional answer to that last category, but the underlying anxiety is generic.

Burry's timing

Michael Burry's intervention, distributed on social media at 03:58 UTC on 11 July and reported the same morning by financial-news outlets, is the most ideologically pointed of the three. Burry, who is best known for his bet against subprime mortgage securities in the mid-2000s and whose trades were later dramatised in a book and a film, has spent most of the last decade running the Scion Asset Management fund with a small public footprint and a large private following. His decision to publicly frame prediction markets as an exploiting-regulatory-loopholes category is consequential because his credibility on financial-system risk has not been spent.

The argument Burry advances is structural rather than moral. Event contracts are, in the US, generally treated as swaps or as exchange-traded derivatives under the jurisdiction of the Commodity Futures Trading Commission, with carve-outs for political and certain entertainment markets. The carve-outs have grown faster than the rule-making. Kalshi's list of approved contracts has expanded month by month through 2025 and into 2026, and the contracts that retail traders actually want most, those tied to Fed decisions, to war timelines, to high-profile criminal verdicts, are precisely the ones whose settlement sources are the most concentrated and the most manipulable.

Burry's prediction, then, is not that prediction markets will be banned; it is that they will be brought inside the perimeter. Taxation on winnings, position-limit and reporting rules, and disclosure of beneficial ownership are all standard for futures markets and none of them currently apply at retail scale to event contracts. The Goldman memo and the New York rule are, in this framing, the first internalised adaptations inside institutions that expect the formal perimeter to follow.

The counter-narrative

The pushback has not gone uncontested. The platforms themselves argue, with some force, that event contracts are a more efficient aggregator of public information than either expert punditry or commissioned polls, and that the prices they print on contested questions are, on average, better calibrated than the alternatives. There is published academic work supporting that claim for political markets specifically, and a more contested but growing literature on macroeconomic and conflict markets. A blanket tightening of access, in that reading, would impoverish the public information environment at exactly the moment when legacy media institutions are losing authority and budget.

A second, more ideological objection runs through the libertarian strain of US financial commentary. To that constituency, prediction markets are a free-speech technology: the price is the speech, and the state has no business telling a private bank or a state court system what its employees or spectators may do with their information. The Goldman's-restrictions-as-corporate-paternalism frame and the New York's-restrictions-as-courtroom-censorship frame both have a constituency inside the American right that does not normally line up with financial regulators.

A third objection is procedural. The CFTC's existing authority over event contracts has been tested in court, with mixed results, and the legal path from "Burry predicts oversight" to "oversight actually arrives" runs through agency rule-making, congressional action, or both. None of that is fast, and any of it can be litigated.

What the wire says versus what the markets say

The mainstream financial press has been more cautious than enthusiastic about prediction markets for most of 2026, and the Goldman memo is being reported as confirmation of that caution rather than as a fresh revelation. The mainstream political press, by contrast, has been mostly absent on the policy question, treating prediction markets as a curiosity or a meme rather than as a regulated market structure. That gap matters. When a market grows to a size where the largest dealer bank on Wall Street feels it has to police its own employees' access, and a state court system feels it has to police the doors of every courthouse in the jurisdiction, the political press's lag is itself the story.

The retail flow data, to the extent it is observable, does not yet suggest the Goldman and New York moves have changed trading behaviour on the platforms. Contract volumes on macroeconomic and on criminal-trial contracts, the two categories most directly affected, remain at or near their 2026 highs in the days following both announcements. The market has been told that the rules will change. It has not yet changed its behaviour.

Stakes

If Burry's prediction is right, the institutional response that begins with Goldman's internal memo and New York's courtroom rule is the opening move in a multi-year tightening that ends with event contracts looking a lot more like commodity futures: position limits, beneficial-ownership disclosure, settlement-source audits, and a tax treatment that treats winnings as ordinary income rather than as gambling revenue. The platforms that survive that transition will be the ones that build compliance capacity early; the ones that treat 2026 as a permanently permissive year will not.

If Burry is wrong, the regulatory perimeter stays where it is, and prediction markets continue to absorb the informational territory that legacy media and commissioned polling are abandoning. Either outcome leaves the platforms larger in 2030 than they are today. The contested question is whether they will be larger inside or outside the regulated perimeter, and who will own them when the dust settles.

This publication reports on prediction markets as a market-structure story first and a technology story second; the Goldman's-restrictions lead is the durable signal, and the New York rule is the visible enforcement proof.

Wire provenance

This editorial synthesis draws on the following public wire/social posts:

• https://t.me/DailyNation
• https://t.me/CryptoBriefing