Who Watches the Markets?
Prediction market prices are treated as authoritative facts. The markets that produce those prices are structurally compromised.
On January 19, the Kalshi contract for “Will Spotify advertise during the Super Bowl?” spiked from $0.35 to $0.69. No press release. No leak in Bloomberg or the Ad Age Super Bowl tracker. No public information of any kind. The contract settled back to $0.37 by Friday morning, the spike visible on the chart like a seismograph needle.
Whether a company will run a Super Bowl ad is not like a rushing yards total. It is not genuinely unknown until the game is played. It is known to hundreds of people — employees at the brand, the ad agency, the media buyer, the network sales team. NBC sold out its entire Super Bowl inventory at an average of $8 million per 30-second spot, with five to ten ads exceeding $10 million. At every node of that transaction chain, someone had material non-public information and could have traded it on Kalshi for risk-free profit.
This is not an edge case. It is the default condition of any prediction market contract whose outcome is determined by a private decision rather than a public event. And it is one of at least three structural integrity failures in the markets that CNN, the Wall Street Journal, and Google now treat as authoritative sources of truth.
The Maduro trade
On January 3, hours before U.S. forces captured Venezuelan president Nicolas Maduro in a nighttime raid in Caracas, a brand-new Polymarket account placed $32,537 on the contract “Maduro out by January 31, 2026.” The contract had traded at 5 to 12 cents for months. The account had been created less than a week earlier and funded specifically for this trade. The payout: $436,000 — a 1,200 percent return in hours.
“This particular bet has all the hallmarks of a trade based on inside information. It happened very late, right before the very event they were betting on happened; it was a relatively large amount of money; and it happened in a market that is not really regulated and where there is no transparency.”
Dennis Kelleher, Better Markets
Chainalysis reported the trader was cashing out through U.S. crypto exchanges without apparent concern for anonymity. The reason is structural: insider trading on prediction markets is not explicitly illegal. There is no equivalent of the STOCK Act for event contracts. Representative Ritchie Torres introduced the Public Integrity in Financial Prediction Markets Act nine days after the Maduro capture, extending insider-trading prohibitions to event contracts. The bill has not advanced.
The volume that is not real
A Columbia University study published in November 2025 found that approximately 25 percent of Polymarket's total historical volume is attributable to wash trading — traders buying and selling against themselves to inflate activity metrics. During high-stakes periods, the study found the figure spiked to 60 percent.
The mechanics are straightforward. Polymarket charges zero trading fees and does not require identity verification. A single operator can create dozens of wallets, execute circular trades at sub-penny prices, and generate millions in notional volume. The Columbia researchers identified one cluster of 43,000 wallets that generated roughly $1 million in volume, mostly at fractions of a cent. The suspected motivation: farming a future Polymarket token airdrop. Polymarket has confirmed token plans. Fourteen percent of all Polymarket wallets showed wash trading behavior.
The platforms trading against their customers
The integrity problems extend beyond outside actors. Kalshi operates an in-house market-making arm. The first class-action lawsuit against the company, filed in November 2025, alleges that market makers including Susquehanna International Group benefit from “unique contractual and technological integration” — reduced fees, different position limits, enhanced market access. There are now 19 federal lawsuits against Kalshi at various stages.
Polymarket, meanwhile, is building an internal trading desk that would trade against its own customers. Its $9 billion valuation — based on a raise from ICE, the parent company of the New York Stock Exchange — creates pressure to monetize a platform that generates essentially zero revenue from trading fees. Trading against your own customers is one path to revenue.
The parallel to FTX is structural, not rhetorical. FTX collapsed in part because its affiliated trading firm, Alameda Research, enjoyed privileged access to customer funds and market data. Prediction market platforms that simultaneously operate the exchange and trade on it hold an informational and mechanical advantage that no outside participant can match.
The regulator
The entity responsible for policing all of this is the Commodity Futures Trading Commission. In February 2026, the CFTC has one of five commissioner seats filled. Staff has been cut by approximately 20 percent. The agency has roughly one-eighth the personnel of the SEC, despite Kalshi alone processing over $2 billion in trades in a single week.
New chairman Michael Selig, installed in January 2026, withdrew the proposed rulemaking that sought to ban political and sports event contracts. He withdrew the staff advisory cautioning platforms from offering sports contracts in states with ongoing litigation. He directed staff to draft new rules but provided no timeline. Donald Trump Jr. holds advisory roles at both Polymarket and Kalshi. Truth Social is building its own prediction market through a partnership with Crypto.com's CFTC-registered exchange.
“Fuck no.”
Evan Semet, full-time prediction market trader, when asked by NPR whether the platforms are doing enough to combat insider trading
The corrupted signal
Here is why the integrity failures are not merely a consumer protection problem.
Prediction market prices have been absorbed into the country's informational infrastructure. CNN and CNBC have deals to incorporate Kalshi prices into their coverage. The Wall Street Journal's parent company, Dow Jones, has partnered with Polymarket. The Golden Globes broadcast used Polymarket as its “exclusive prediction market partner,” displaying prices as “real-time, market-driven insights.” Google has integrated prediction market data into search results.
When a prediction market price appears on CNN as an authoritative probability — the likelihood of a policy passing, a candidate winning, a geopolitical event occurring — it does not merely reflect beliefs about the future. It shapes them. Viewers internalize the number. Strategists adjust. Capital moves. The price becomes an input to the event it claims to measure.
This is the reflexive loop Soros identified in currency markets. Prices are not passive outputs of information. They are active inputs to the systems they describe. But Soros assumed the participants in the loop were operating in a market with basic structural integrity — enforced rules against fraud, transparency requirements, regulatory oversight. Remove those assumptions and the loop still operates, but what it propagates is no longer aggregated judgment. It is an artifact of whoever has the most inside information, the most wash-traded volume, or the most structural advantage on the platform — laundered into authority by media organizations that display the number without examining its provenance.
The prediction market is no longer a mirror. It is a projector. And the film has not been checked.
The arithmetic
The industry produced $40 billion in volume in 2025, up 400 percent year-over-year. Projected volume for 2026 is north of $100 billion. Every dollar of that growth increases the number of prices flowing into newsrooms, search engines, and decision-making processes. Every integrity failure in the underlying market contaminates the signal that shapes public perception and, through reflexivity, the outcomes themselves.
The question is not whether prediction markets will face a credibility crisis. The question is whether the integrity infrastructure gets built before the crisis arrives or after. The commercial momentum argues for after. The reflexive structure of these markets — where the price is the intervention — argues that after may be too late. Once the institutions that treat prediction market prices as authoritative stop trusting them, the value proposition collapses. And the collapse will be as reflexive as the rise: distrust reduces liquidity, reduced liquidity degrades price quality, degraded price quality validates the distrust.
The loop runs in both directions. It always does.
Reflexivity covers prediction markets as reflexive systems — where prices shape the reality they claim to measure. Written by Dawson Smith.