The Halftime Show Problem
When the information asymmetry is the product, not the bug.
The Seahawks beat the Patriots 29-13. The game was not close. But the most interesting trading of Super Bowl LX did not occur on the game winner market, where $500 million traded on Kalshi alone. It occurred in the halftime markets—specifically, the Bad Bunny prop contracts where someone knew what song was coming before anyone else did.
Two wallets are now the subject of intense scrutiny. One is haeju.eth, which deployed approximately $47,000 exclusively in halftime markets and was correctly positioned for nearly every halftime outcome. The other is anonymous, created less than 24 hours before the game, which placed $32,000 on Bad Bunny's first song and traded nothing else. Estimated insider profits across the halftime markets: $2-3 million.
Combined Super Bowl prediction market volume exceeded $1.3 billion, with Kalshi processing roughly $900 million. The Bad Bunny first song market alone saw $113.5 million in volume. On what was supposed to be the industry's biggest legitimacy event, two wallets demonstrated exactly what critics have warned about for years: entertainment markets are structurally different from political or economic markets, and that structural difference is an invitation to extraction.
The prediction market industry has spent three years arguing it aggregates dispersed information into efficient prices. The halftime show problem reveals the inverse: when the information is not dispersed but concentrated—held by performers, producers, choreographers, stage managers, anyone with access to the setlist—the market does not aggregate. It transfers. From the uninformed to the informed. From retail participants to insiders.
This is not a new problem. But it is now a billion-dollar problem. And the industry's response will determine whether prediction markets mature into regulated financial infrastructure or collapse under the weight of their own contradictions.
The anatomy of an entertainment market
Understanding why halftime markets are different requires understanding what makes information markets work in the first place.
In a well-functioning prediction market—say, a contract on the Federal Reserve's next rate decision—the outcome is genuinely unknown. Fed governors deliberate behind closed doors, but the information that influences their decision is partially observable: employment data, inflation reports, forward guidance, the shape of the yield curve. Traders analyze public signals, form probabilistic beliefs, and trade against each other. The price that emerges reflects the aggregated assessment of thousands of independent observers processing overlapping but non-identical information sets.
This is the Hayekian thesis: markets coordinate dispersed knowledge that no single actor possesses. The price is smarter than any individual participant.
Now consider the halftime show. Bad Bunny's setlist was finalized days or weeks before the game. It was known in full by the performer, his management team, the choreographers, the lighting technicians, the audio engineers, the NFL production staff, and presumably dozens if not hundreds of additional people with access to rehearsals. The information was not dispersed. It was concentrated and asymmetric. The “unknown” outcome was unknown only to market participants. To insiders, it was certain.
The difference is not subtle. In a Fed rate market, informed trading means analyzing information more effectively than others. In a halftime show market, informed trading means possessing information that others cannot access at any level of analytical sophistication.
The structural implication: entertainment markets do not aggregate dispersed information. They create extraction opportunities for anyone with access to backstage.
The wallet evidence
The trading patterns flagged by on-chain analysts exhibit signatures consistent with insider knowledge.
The haeju.eth wallet deployed capital exclusively in halftime-related markets. Not game spreads. Not MVP odds. Not total points. Only halftime. The wallet was “correctly positioned for nearly every halftime outcome,” according to analysis circulating in prediction market communities. The probability of this occurring through pure analytical skill or luck diminishes rapidly as the number of correct positions increases.
The anonymous wallet is more striking. Created less than 24 hours before kickoff. Traded only halftime props. Deposited $32,000. Correctly predicted Bad Bunny's first song. This is the pattern that triggered the Maduro investigation in January: a newly created account, no trading history, single-market focus, large position, immediate correct resolution.
When Polymarket investigated the Maduro trade—where an anonymous account wagered $30,000 hours before Delta Force captured the Venezuelan president and walked away with $400,000—the investigation found no conclusive evidence of manipulation but could not rule it out. The nature of pseudonymous markets makes definitive attribution nearly impossible. You can trace wallets. You cannot prove who controls them.
The estimated $2-3 million in halftime insider profits is small relative to the $1.3 billion in total Super Bowl volume. But the precedent is not small. If entertainment markets are structurally exploitable—if the information asymmetry is inherent to the product rather than incidental to it—then the entire category faces existential questions.
The Kalshi response
Kalshi's response has been institutional signaling rather than enforcement action.
The company announced a partnership with Solidus Labs, a market surveillance firm that specializes in crypto compliance. It appointed Robert DeNault as head of enforcement, a new position. The messaging emphasized proactive investment in integrity infrastructure.
What Kalshi has not done, as of February 9, is initiate any enforcement actions. No wallets have been frozen. No traders have been banned. No referrals to the CFTC have been announced. The pattern flagged by community analysts remains unaddressed beyond the announcement of partnerships and personnel.
This is not necessarily bad faith. Market surveillance takes time. Building robust enforcement infrastructure for a $45-billion annual volume industry is not a weekend project. But the gap between the signaling and the action creates its own loop.
The prediction market industry's legitimacy argument depends on being different from unregulated gambling. Different means oversight. Oversight means enforcement. When suspicious trading patterns emerge on the industry's biggest day and the response is a press release about future capabilities, the difference narrows.
The structural question
Here is the analytical core: what if entertainment markets cannot be fixed?
The halftime show problem is not an edge case. It is a feature of the product category. Super Bowl ad markets—which brand will air a commercial—face the same issue. NBC sold out its inventory at $8 million per 30-second spot. Whether OpenAI or Netflix or Liquid Death will advertise is known to the brand, the agency, the network, and hundreds of people at each organization. Awards show outcomes—Oscars, Grammys, Emmys—are known to accounting firms, venue staff, and presenters before they are known to the public. Reality TV eliminations are known to producers and sometimes contestants. Album release dates are known to labels, distributors, and platform partners.
Every entertainment market involves an outcome determined by a discrete set of actors who know the outcome in advance. The information asymmetry is not a market failure. It is the market.
This creates a trilemma for the industry. Option one: accept that entertainment markets are structurally exploitable and continue offering them, absorbing the integrity damage as a cost of volume. Option two: exit the entertainment category entirely and focus on markets where outcomes are genuinely unknown—economics, climate, geopolitics. Option three: attempt to build surveillance infrastructure so robust that insider trading becomes unprofitable despite the structural disadvantage.
Each option carries costs. Option one erodes the legitimacy narrative that justifies federal classification over state gambling regulation. Option two surrenders the highest-engagement product category. Option three requires enforcement capabilities that the CFTC—operating with one of five commissioners seated and 20 percent fewer staff—cannot provide, meaning platforms must self-police in a zero-fee environment where revenue scales with volume.
The incentive alignment is not encouraging. Kalshi's business model depends on volume. Entertainment markets generate volume. Volume from entertainment markets is partially insider-driven. The insider volume generates fees (for Kalshi, through transaction or withdrawal fees) identically to retail volume. Aggressive enforcement against insiders reduces volume. Reduced volume reduces revenue.
The rational strategy for a volume-maximizing platform is to signal integrity investment while minimizing actual enforcement that would deter high-volume participants. This is not a prediction of what Kalshi will do. It is a description of the incentive structure they operate within.
The Torres precedent
Representative Ritchie Torres introduced the Public Integrity in Financial Prediction Markets Act of 2026 in response to the Maduro trade. The bill extends STOCK Act insider trading prohibitions to event contracts, specifically targeting federal officials, appointees, Executive Branch employees, and congressional staff.
The bill addresses a real problem: government insiders trading on policy decisions they influence. But it does not address the halftime show problem. Bad Bunny's setlist was not determined by federal employees. It was determined by an entertainment production team with no federal nexus.
The gap between the legislation and the structural problem is instructive. Congress can regulate government insiders. It cannot easily regulate entertainment industry insiders without creating a new category of financial regulation that applies to stage managers, choreographers, and backup dancers. The legal infrastructure does not exist.
This leaves the CFTC. And the CFTC, under Chairman Michael Selig, has signaled aggressive support for prediction market expansion rather than aggressive enforcement against prediction market abuse. Selig's January 29 actions—withdrawing the proposed ban on sports and political contracts, signaling intervention in pending court cases on the industry's behalf—suggest a regulatory posture more interested in jurisdiction than policing.
The industry's regulatory champion may not be equipped or inclined to solve the problem the industry's growth has created.
The reflexive loop
The halftime show problem matters beyond its dollar magnitude because of what it reveals about the prediction market industry's reflexive structure.
The industry's growth depends on legitimacy. Legitimacy depends on the claim that these are information-aggregation tools, not gambling platforms. Information aggregation requires information asymmetries to be navigable—traders compete on analysis, not on access to secrets. When entertainment markets reveal that the asymmetries are inherent rather than navigable, the legitimacy claim weakens.
Weakened legitimacy empowers the 36 state attorneys general coalition already opposing federal preemption. It empowers the NFL, which banned prediction market advertising while its games generate billions in prediction market volume. It empowers tribal gaming interests whose compacts are undermined by federally classified gambling alternatives.
The more entertainment market scandals emerge, the stronger the opposition. The stronger the opposition, the greater the threat to the federal classification that enables the industry's growth. The industry's own success—measured in $1.3 billion Super Bowl volume and $113.5 million Bad Bunny markets—creates the conditions for its regulatory unraveling.
This is the Soros pattern. Unstable equilibria persist until they break. The prediction market industry is building volume on a foundation that its own volume undermines. The halftime show problem is not an anomaly. It is a stress test. And the test revealed exactly what the critics predicted: when the information asymmetry is structural, the market does not aggregate truth. It extracts capital.
What happens next
The next 90 days will determine whether this incident is a footnote or a precedent.
If Kalshi's enforcement infrastructure produces results—wallet freezes, CFTC referrals, recovered funds—the industry can point to self-policing as evidence that federal classification is warranted. If the Solidus Labs partnership and Robert DeNault's appointment are followed by visible action, the legitimacy narrative survives.
If nothing happens—if the flagged wallets face no consequences, if the $2-3 million in estimated insider profits remains extracted, if the only response is infrastructure announcements without enforcement outcomes—the February 12 federal hearing, the Ninth Circuit arguments in April, and the probable Supreme Court petition by year's end will all occur against the backdrop of a documented integrity failure that the industry did not remedy.
The reflexive loop will turn. Volume will attract scrutiny. Scrutiny will reveal extraction. Extraction will empower opposition. Opposition will threaten classification. Threatened classification will constrain volume.
The halftime show problem is not about Bad Bunny. It is about whether entertainment markets belong in a prediction market ecosystem at all. The industry's answer to that question will shape everything that follows.