The Oracle That Proved Me Right

The Paranoidist | Issue #10 By Paul Morin | April 11, 2026

Six weeks ago, in Issue #4 of this newsletter, I wrote a piece called "The Oracle Nobody Is Auditing." Its thesis was that prediction markets were being adopted as decision-making infrastructure without adequate scrutiny of their vulnerability to manipulation, their structural incapacity to price genuine uncertainty, and the absence of any regulatory framework to prevent insider trading. I identified the specific risk: thin liquidity, no surveillance, pseudonymous accounts, and an expanding role in media and institutional decision-making that made these markets an attractive target for anyone with non-public information and the will to exploit it.

I wrote: "Probability that a major decision or market event is materially influenced by manipulated prediction market data within the next 24 months: High."

It took six weeks.

On April 7, hours before President Trump announced a two-week ceasefire with Iran, at least 50 newly created Polymarket accounts placed substantial bets on exactly that outcome. The bets were placed in the hours, and in some cases the minutes, before the announcement. These were the sole bets made through these accounts. The Associated Press broke the story on April 9. By April 10, bipartisan calls for congressional investigations were escalating, the Commodity Futures Trading Commission (CFTC) was under pressure to act, and Senator Richard Blumenthal had written a letter to Polymarket demanding answers.

This was not the first time. It was the third time in six weeks. And behind these three episodes lies a pattern so systematic that a team of researchers has now quantified it: $143 million in anomalous profits, across more than 210,000 suspicious trading pairs, spanning everything from the Iran war to Taylor Swift's engagement to the Nobel Peace Prize.

The oracle nobody was auditing is now leaking national security secrets in real time. This issue dissects how we got here, what the mechanism actually is, and what it means for anyone who sits on a board, manages institutional risk, or makes decisions that depend on the integrity of information.

The ceasefire bets were the most recent episode, but the pattern begins on the night the war started.

On February 28, 2026, hours before the United States and Israel launched strikes on Iran, six newly created Polymarket wallets collectively earned approximately $1.2 million by purchasing "Yes" shares in the "US strikes Iran by February 28?" contract. One account, operating under the handle "Magamyman," placed its first trade seventy-one minutes before the news broke, when markets implied only a 17% probability of a strike. When the market resolved, that single account's profits totaled approximately $553,000. The U.S.-Israeli strike on Iran was one of the most closely guarded military operations in recent history. Someone knew.

In January 2026, an anonymous Polymarket user made $400,000 by betting that Venezuelan leader Nicolas Maduro would be ousted, hours before Maduro was captured. The account was new. The bet was the only trade it made.

And then, on April 7, the pattern repeated at scale. Fifty new accounts. Ceasefire bets placed minutes before the announcement. The president had spent the day escalating, not de-escalating, warning on social media that "a whole civilization will die tonight" if Iran did not meet his deadline. There were no public signals that a ceasefire was imminent. Pakistan's intervention, the channel that produced the deal, was not publicly known. The accounts that bet on the ceasefire had information that was not available to the public, and they used a platform with no identity verification, no insider trading rules, and no regulatory oversight to convert that information into cash.

Representative Ritchie Torres, who sits on the House Financial Services Committee, put the arithmetic plainly: "What is the statistical likelihood that anyone other than an insider trader placing a winning bet 12 minutes before a market-moving presidential announcement? There are two answers: God, or an insider trader."

The individual episodes are alarming. The systemic picture is worse.

In March 2026, researchers Joshua Mitts of Columbia Law School and Moran Ofir of the University of Haifa published a study examining every Polymarket market from February 2024 through February 2026: more than 93,000 distinct markets and nearly 50,000 unique wallet addresses. Their methodology combined five signals (cross-sectional bet size, within-trader bet size, profitability, pre-event timing, and directional concentration) into a composite score measuring the probability that a given wallet-market pair reflects informed trading rather than speculation.

The findings: across 210,718 suspicious wallet-market pairs, flagged traders achieved a 69.9% win rate, a result that exceeds the null distribution of random chance by more than 60 standard deviations. The estimated aggregate anomalous profit: approximately $143 million.

Sixty standard deviations. To put that in context: a result three standard deviations from the mean is unusual. Five standard deviations is the threshold particle physicists use to declare a discovery. Sixty standard deviations is not a statistical anomaly. It is a statistical impossibility under the null hypothesis of fair trading. It is proof, at the level of mathematical certainty, that something other than luck is operating in these markets.

The scope extended far beyond geopolitics. The anomalous trading patterns appeared in markets covering Taylor Swift's engagement, the Nobel Peace Prize, cryptocurrency prices, and dozens of other categories. The prediction market ecosystem is not experiencing occasional insider trading. It is, by the numbers, saturated with it.

If the insider trading pattern represents the conversion of classified information into cash, the F-15 episode represents something different: the conversion of human life into a trading chip.

On April 3, an American F-15E Strike Eagle was shot down over Iranian territory. One crew member was rescued. The other was missing, location unknown, with Iranian forces offering a bounty for the capture of the pilot. While a search-and-rescue operation was actively underway inside hostile territory, Polymarket opened a market allowing users to bet on when the pilots would be confirmed rescued. The options included specific days. The implicit alternative to "rescued" was clear to every participant: captured, or dead.

Up to $930,000 was wagered on a single outcome. More than 63% of bettors wagered against a quick rescue.

Representative Seth Moulton, a Marine Corps veteran who served in Iraq, posted a screenshot and responded: "There is an ongoing search and rescue operation for a missing American service member whose plane was shot down over Iran. Their safety is unknown. They could be your neighbor, a friend, a family member. And people are betting on whether or not they'll be saved. This is DISGUSTING."

He described Polymarket as a "dystopian death market."

Polymarket took the market down, stating it did not meet its integrity standards and should never have been posted. But the company was operating 219 additional war-related markets at the time. Moulton demanded they all be removed. Polymarket did not comply.

This is not a tangent from the insider trading story. It is the same story, viewed from a different angle. The structural conditions that enable insider trading on prediction markets (pseudonymous accounts, no regulatory oversight, no identity verification, offshore jurisdiction) are the same structural conditions that enable betting on whether American service members will survive a combat operation. The platform that cannot prevent someone with classified ceasefire information from converting it into $600,000 also cannot prevent someone from wagering on the life of a pilot whose family is watching the news and waiting. The failures are not separate. They are the same failure, expressed in different currencies.

There is a layer to this story that would be almost too convenient for fiction, but it is documented fact.

Donald Trump Jr. is an investor in Polymarket through his venture capital firm, 1789 Capital. He separately serves as a paid strategic adviser to Kalshi, Polymarket's primary domestic competitor. The Trump family has announced plans to launch Truth Predict, a prediction market platform on Truth Social.

The president's son has a financial stake in the prediction market ecosystem. The president's decisions (military strikes, ceasefires, diplomatic announcements) are the events being traded. The accounts making the most profitable trades appear to possess advance knowledge of those decisions. And the regulatory body responsible for overseeing these markets, the CFTC, has no clear authority, no surveillance infrastructure, and no mandate to investigate.

The White House, to its credit, circulated an internal memo on March 24 warning staff that using non-public information to place wagers on prediction market platforms would violate federal ethics guidelines. The memo's existence confirms the concern. Its necessity confirms the vulnerability. Whether a memo is sufficient to prevent exploitation of one of the most profitable insider trading opportunities in modern financial history is a question the memo does not answer.

Senator Blumenthal framed it in national security terms: "Polymarket has become an illicit market to sell and exploit national security secrets unlike any in history, and by extension a potential honeypot for foreign intelligence services watching for those same suspicious bets and wagers."

That last clause deserves emphasis. A prediction market where insiders regularly trade on non-public government information is not just a vehicle for insider profit. It is a real-time intelligence signal available to any foreign adversary with the technical capability to monitor blockchain transactions. When 50 accounts simultaneously bet on a ceasefire minutes before the announcement, that pattern is visible to anyone watching the blockchain. The accounts are pseudonymous. The blockchain is public. If American insiders are telegraphing presidential decisions through their trading patterns, every intelligence service in the world can watch.

Representative Blake Moore, a Republican from Utah, stated the implication directly: "We don't want to imagine a world where America's adversaries use prediction markets to anticipate our next move."

We do not need to imagine it. That world exists.

In Issue #4, I wrote that I was watching "whether any regulator (CFTC, SEC, or international equivalent) establishes manipulation surveillance standards for prediction markets before a crisis forces the issue."

No regulator did. The crisis came anyway.

The regulatory situation is a jurisdictional void that would be comic if the stakes were not so high. Polymarket was banned from the U.S. in 2022 but continues to operate a crypto-based platform offshore that accounts for most of its activity. It has moved to reenter the domestic market by acquiring a CFTC-licensed exchange, giving it a legal pathway to offer contracts in the United States. The CFTC's jurisdiction over prediction markets is contested, the subject of active litigation, and the agency recently filed lawsuits against multiple states to assert sole authority over these platforms.

Meanwhile, the legal definition of insider trading does not cleanly extend to prediction markets. Representative Torres introduced the Public Integrity in Financial Prediction Markets Act (H.R. 7004), which would prohibit federal officials and staff from trading on prediction market contracts where they possess material non-public information. But as the Mitts and Ofir study notes, the bill would not necessarily have prevented the cases already documented: the Iran strike and Maduro trades appear to have involved national security personnel already covered by separate statutes. The issue is not the absence of laws. The issue is enforcement in a system designed, from the ground up, to be unenforced.

Bipartisan bills are now pending in both the House and the Senate. Congressional Democrats have introduced legislation to bar prediction markets from allowing wagers on elections, war, and government actions. Even Polymarket and Kalshi have acknowledged the need to broaden the definition of insider trading on their platforms. The industry is ahead of the regulator in admitting the problem, which tells you how far behind the regulator is.

The DeepStrategy.ai analytical method requires sorting what was risk from what was uncertainty.

The risk column is, for the second consecutive issue, overwhelming. In Issue #4, published on February 28, this newsletter identified the specific vulnerabilities: thin liquidity, pseudonymous accounts, no surveillance infrastructure, no regulatory framework, and an expanding decision-making role that made prediction markets an attractive target for insider trading. The Columbia University wash trading study had already documented that 25% of Polymarket's volume was artificial. The Polymarket UFO resolution incident had already demonstrated that outcomes could be driven by whale capital rather than genuine forecasting. The Yale liquidity critique had already shown that prices cited by media were, in some cases, "phantom figures" set by zero sellers and 50% bid-ask spreads.

Every element of the crisis that followed was documented, analyzed, and published before it happened. The manipulation vulnerability was not a surprise. It was a forecast.

The uncertainty column is narrow but important. Whether the specific mechanism would be insider trading on military operations (as opposed to market manipulation for strategic narrative purposes, which Issue #4 also identified as a risk). Whether the scale would reach $143 million in anomalous profits in less than two years. Whether the platform would extend to wagering on the survival of individual service members in active combat. These specifics were not predictable. That the system would fail in this general category of ways was.

It is possible that the $143 million figure overstates the problem. The Mitts and Ofir study uses statistical flagging, not direct evidence of insider knowledge. Some fraction of the anomalous profits may reflect genuinely skilled forecasters who are better at reading public signals than the median participant. The study's authors are careful to note that their screening identifies suspicious patterns, not confirmed violations. The true insider trading total could be substantially lower than $143 million, even if it is clearly not zero.

It is also possible that the market is self-correcting faster than I expect. The congressional attention, the Harvard study, the bipartisan legislative momentum, the White House ethics memo, and the public revulsion at the F-15 betting market may collectively produce regulatory frameworks that address the worst vulnerabilities within months rather than years. Prediction markets have genuine informational value in well-defined domains, and it would be a loss if the insider trading crisis destroyed the tool entirely rather than producing the oversight that makes it trustworthy. The outcome where prediction markets emerge regulated, surveilled, and structurally sound is possible, and it would be better for everyone, including the platforms.

And the national security signal problem, while real, may be less exploitable than Blumenthal suggests. The blockchain is public, but interpreting trading patterns in real time, distinguishing signal from noise, and acting on the intelligence before the event occurs is not trivial. Foreign intelligence services have many sources of information about U.S. decision-making, and prediction market patterns may be a marginal addition to an already rich intelligence landscape rather than the decisive leak that changes outcomes. I am not certain of this. But intellectual honesty requires noting it.

The core observation holds regardless. The prediction market ecosystem processed hundreds of millions of dollars in geopolitical wagers during a war in which American service members were dying, with no identity verification, no insider trading enforcement, no regulatory oversight, and documented evidence that accounts with advance knowledge of classified military decisions were converting that knowledge into cash. The system I described in Issue #4 as unaudited is now documented as compromised. Whether the compromise is $143 million or $50 million, whether the regulatory response arrives in months or years, the gap between what prediction markets claim to be (transparent aggregators of dispersed information) and what they have demonstrated themselves to be (unregulated venues where insiders convert non-public information into profit while the public provides the liquidity) is no longer theoretical.

The oracle nobody was auditing has been audited. The results are in. The question now is whether the institutions responsible for oversight will act on what is known, or whether this will be filed alongside the Shahed's capabilities, Ukraine's counter-drone offer, and the Gulf states' warnings: another entry in the growing catalog of failures that were not failures of knowledge, but failures of action.

The institution that consumes the analytical process as preparation for multiple futures has what the forecast cannot provide: adaptability. The institution that does not has what the prediction market delivered: a public ledger of classified decisions, visible to every adversary with a blockchain explorer and the patience to watch.

The Paranoidist publishes weekly, with flash issues when events warrant. If this changed how you think about one thing, consider subscribing. If it didn't, tell me what I'm missing.

Paul Morin is the founder of DeepStrategy.ai, author of Uncertainty: When Risk Is Not Enough (a guide to decision-making when probabilities fail), and publisher of The Paranoidist, BoardroomRadar and ScenarioWatch. He has spent more than three decades in entrepreneurship, finance, risk management, and insurance, which is why he worries about the things that keep other people awake at night.

Researched, written, and edited in collaboration with Claude by Anthropic.