At 2:47 AM Eastern on February 28, 2026, the first munitions from Operation Epic Fury struck Tehran. Within minutes, the "US strikes Iran by February 28" contract on Polymarket snapped to $1.00. Six freshly created wallets collected roughly $1 million in combined profits. One trader who'd spent months on the other side of that bet lost $6.5 million before breakfast.

By the time Trump posted "Khamenei, one of the most evil people in History, is dead" on Truth Social that afternoon, the prediction market complex surrounding Iran had processed more than $525 million in total volume — making it the single largest geopolitical event contract in the industry's history. The 2024 presidential election was bigger in aggregate, but no single binary outcome moved this much money this fast.

The Iran strike contracts didn't just predict the war. They may have been used by people who already knew it was coming.

To understand what happened on February 28, you need to understand the scale of what prediction markets built around Iran.

The contract architecture was sprawling. Polymarket hosted rolling date-specific markets — "US strikes Iran by January 31," "...by February 14," "...by February 28," and so on — alongside broader contracts on regime change, Khamenei's removal, ceasefire timelines, and whether Iran would retaliate against US military assets. On Polymarket alone, strike-date contracts generated over $525 million in cumulative volume since December 2025. Kalshi ran parallel contracts on a separate set of Iranian leadership and regime stability markets, capturing over $85 million in volume on those questions.

The February 28 contract was the epicenter. It drew nearly $90 million in volume by resolution day. For weeks, the odds had oscillated between 20 and 55 cents, tracking every diplomatic signal, naval deployment, and failed negotiation. When the Oman nuclear talks collapsed on February 26, the contract ticked up. When a $5 million "No" position appeared — suggesting to many that a well-connected whale believed the carrier group deployment was leverage, not prelude — it ticked back down.

Then, in the final hours before the strike, a cluster of small, new wallets began buying "Yes" at prices between 10 and 20 cents. They didn't hedge. They didn't diversify. They bought one thing, and they were right.

Blockchain analytics firm Bubblemaps was the first to map the suspicious wallets, publishing a visual cluster analysis showing six accounts funded through similar paths — all created in February, all with zero prior trading history.

The Block independently reviewed all six Polymarket profiles and confirmed a combined net profit of $989,191.

The largest wallet bought 560,680 "Yes" shares at approximately 10.8 cents each, putting roughly $61,000 at risk. When the contract resolved at $1.00, the position paid out close to $560,000 — a net profit of $494,375 and an 821% return on investment in under 72 hours.

A second wallet purchased approximately 150,000 shares at 20 cents, turning roughly $30,000 into a six-figure payout. A sixth wallet, listed simply as "Anon," made a single trade: 55,556 shares at 18 cents, earning $45,556 in net profit — a 456% return. All six wallets now show zero positions. They're fully exited. Clean.

Outside the suspected insider cluster, other wallets profited through what appears to be conviction rather than information asymmetry. A trader named Vivaldi007 joined Polymarket on February 8 and repeatedly bet on a strike across multiple date contracts. He absorbed loss after loss as earlier deadlines passed without incident. When Operation Epic Fury launched, his surviving positions turned green. Total profit: $385,000.

A wallet called "Roeyha2026" appeared just 11 hours before placing a $50,000 bet on "US strikes Iran by March 1." That position is currently showing $96,800 in unrealized gains.

The most spectacular casualty was a trader operating under the handle "anoin123."

In January and February, anoin123 had built one of the most impressive track records on Polymarket. The strategy was elegant: systematically sell "Yes" shares on Iran strike contracts — effectively betting that each deadline would pass without military action. For weeks, it worked. Iran contract after Iran contract expired uneventfully. By mid-February, anoin123 had accumulated over $2 million in profit and ranked as the fourth most profitable trader in Polymarket history.

Then February 28 happened.

When the strikes hit, every open "No" position anoin123 held went to zero simultaneously. According to Lookonchain, the account lost $6.5 million in a single trading session, swinging from a +$2 million career profit to a -$4.5 million net loss. It is believed to be one of the largest single-day losses any individual has ever sustained on a prediction market platform.

The anoin123 story illustrates a structural reality of geopolitical prediction markets: selling "No" on strike contracts is profitable right up until it isn't. The strategy collects steady premium — analogous to selling options — until a tail event collapses the entire position. The risk-reward is asymmetric in the worst possible direction.

The Iran wallets aren't the first time prediction markets have produced suspiciously well-timed trades around major geopolitical events.

On January 3, 2026, US forces captured Venezuelan President Nicolas Maduro in a nighttime raid in Caracas. Hours before the operation became public, three newly created Polymarket wallets placed concentrated bets on Maduro's removal. One account, created just a week prior with activity limited to Venezuela-related contracts, turned a $32,000 position into $436,000 when the capture was confirmed. Combined, the three wallets netted $630,484.

Lookonchain flagged the wallets. NPR, Fortune, and PBS investigated. Rep. Ritchie Torres introduced the Public Integrity in Financial Prediction Markets Act of 2026, which would bar federal officials and political appointees from trading on prediction market platforms. Chainalysis traced the Maduro wallet to US-based crypto exchanges — suggesting the trader wasn't even trying to hide. No charges have been filed.

In December 2025, a cluster of Polymarket accounts correctly predicted 22 of 23 Google Year in Search results, netting approximately $1 million. A Meta engineer publicly accused the accounts of insider trading, though no formal investigation was disclosed.

In late February 2026 — just days before the Iran strikes — blockchain investigator ZachXBT announced he was probing a crypto platform. Polymarket immediately spun up a contract on which platform would be named. Twelve wallets piled into the correct answer (Axiom) before the public reveal, collectively profiting over $1 million. Insiders were insider-trading on a contract about insider trading.

The pattern is consistent: new wallets, concentrated positions, narrow focus, no prior history, and precise timing that strains the boundaries of coincidence.

The insider trading problem looks different depending on which platform you're watching, and the structural differences matter.

Polymarket is built on blockchain. Trades settle on-chain, wallets are pseudonymous, and the platform operates as a decentralized prediction market. It returned to the US market in late 2025 after receiving CFTC clearance, following a 2022 settlement that had barred US operations. Its CEO, Shayne Coplan, has publicly argued that informed traders are a net positive — telling CBS News that insiders "having an edge on the market is a good thing" because it accelerates price discovery. The platform claims it can self-police through internal audits. There is no public record of Polymarket banning or fining a user for insider trading.

Kalshi operates as a CFTC-regulated designated contract market — the same regulatory classification that governs derivatives exchanges trading corn futures and crude oil options. It requires identity verification, explicitly bans insider trading in its rulebook, and has built an internal surveillance team. On February 25, 2026 — three days before the Iran strikes — Kalshi publicly disclosed its first enforcement actions: a MrBeast video editor named Artem Kaptur was fined $20,000 and suspended for two years after placing $4,000 in trades on content he had inside knowledge of. A former California gubernatorial candidate was separately banned for betting on his own race. Kalshi reported both cases to the CFTC and said it has opened over 200 investigations in the past year.

The contrast is stark. Kalshi caught a guy making $5,000 betting on YouTube videos. Meanwhile, nearly $1 million flowed to six anonymous wallets on Polymarket hours before one of the most consequential military operations of the decade.

The CFTC is the primary regulator for prediction markets in the United States — a role it inherited somewhat awkwardly, since the agency was originally designed to oversee agricultural commodity futures. It has roughly 115 enforcement employees, compared to the SEC's roughly 1,400. Yet prediction market volume now dwarfs many traditional commodity markets.

CFTC Chairman Mike Selig responded to the Kalshi enforcement actions by stating: "If you attempt to engage in manipulation, fraud, or insider trading, we will find you and take action." He called exchanges the "first line of defense." The agency issued a formal advisory noting that insider trading on event contracts may violate federal law.

But the enforcement gap is enormous. Kalshi's regulated framework gives the CFTC clear jurisdiction. Polymarket's decentralized, blockchain-native architecture — with pseudonymous wallets funded through crypto — creates a much harder enforcement target. Even if the CFTC identifies suspicious activity, connecting an anonymous wallet to a real person requires cooperation from crypto exchanges, blockchain analytics firms, and potentially foreign jurisdictions.

The Torres bill (Public Integrity in Financial Prediction Markets Act of 2026) would address one narrow slice of the problem — barring government employees and political appointees from trading. But it wouldn't touch the broader question of whether anyone with material nonpublic information, regardless of their employment, should be able to trade anonymously on geopolitical outcomes worth hundreds of millions of dollars.

There's also a deeper philosophical tension. Traditional insider trading law requires showing harm — someone was deprived of value. As University of Pennsylvania professor Daniel Taylor noted in the Maduro context: "How would the US government be harmed by someone trading on advanced warning of the Maduro operation?" If you can't show a victim, prosecution becomes extremely difficult.

The libertarian argument, articulated by Cato Institute's Alex Nowrasteh, is that insider trading on prediction markets is a "social good" — it forces information into prices faster, which benefits everyone watching the market for signals. Polymarket's Coplan has echoed this view. The counterargument, pushed by Torres and other Democratic lawmakers, is that when government officials can profit from their own policy decisions, the incentive structure becomes genuinely dangerous. When hundreds of millions of dollars ride on whether bombs fall, the line between information and action starts to blur.

The Iran contract complex isn't finished. Khamenei is confirmed dead, but the cascade of derivative markets is just beginning. Ceasefire contracts, regime change timelines, IRGC succession, whether Mojtaba Khamenei (the Supreme Leader's son, reportedly surviving the strike) consolidates power — all of these are actively trading with significant volume.

The insider trading question isn't going away either. If anything, the Iran episode will accelerate regulatory scrutiny. Bloomberg reported that Polymarket's Iran bets have drawn formal notice from regulators. The CFTC's enforcement advisory, issued just days before the strike, now reads like foreshadowing.

For traders, the lesson from February 28 is brutally simple: in geopolitical prediction markets, the information asymmetry isn't between bulls and bears. It's between people who know what's going to happen and everyone else. Six wallets knew. anoin123 didn't. The difference was $7.5 million.

Hammerhead Intelligence tracks prediction market flow, whale activity, and geopolitical risk every weekday at 6:15 AM EST. Weekend deep dives go deeper. This is what we do — we follow the money so you don't have to.