The Market for Secrets: Who Profits from Wars Before the News Breaks

How prediction markets turned political decisions, corporate secrets, and military operations into tradable assets

Financial markets have always traded the future. When a trader buys an oil futures contract, they are trying to determine what the commodity will be worth months from now. When an investor buys shares, they are assessing the company’s future earnings. Even exchange rates largely reflect expectations about interest rates, inflation, and economic conditions. In recent years, however, trading the future has taken on an entirely different form. The asset is no longer necessarily the price of oil after a war begins, but the war itself; not the reaction of a currency to a politician’s resignation, but the resignation itself; not the stock market’s response to a central bank decision, but the decision itself.

This is the world of prediction markets. These platforms offer contracts tied to the outcomes of real events: elections, government decisions, military operations, corporate announcements, court cases, economic releases, and even the behavior of specific individuals. A participant can buy a contract predicting that a particular event will occur before a certain date. If the prediction is correct, the contract settles at one dollar. If the event does not happen, it becomes worthless.

At first glance, this appears to be little more than a transparent way of trading probabilities. The price of a contract indicates how strongly the market believes in a particular outcome. If a contract is trading at twenty cents, the market is effectively assigning the event a probability of roughly twenty percent. The buyer believes that probability is too low; the seller believes it is too high. Their opposing views meet, and a market price emerges that is supposed to represent the combined knowledge of thousands of independent participants.

That is how prediction markets are usually presented to the public: as a form of collective intelligence capable of assessing the future more accurately than opinion polls, television experts, or official forecasts. The idea makes sense as long as participants operate within roughly the same information environment. One understands economics better, another follows politics closely, a third studies satellite imagery, and a fourth analyzes the statements of government officials. Each participant has a different interpretation, but no one has the final answer.

The problem begins when someone enters the market who is not trying to predict the event because they already know it is going to happen.

When prediction becomes knowledge

In the spring of 2026, U.S. authorities charged a member of the military with allegedly using classified information to trade on Polymarket. According to prosecutors, he had been involved in planning a military operation connected to the capture of Nicolás Maduro in Venezuela and had access to confidential information about its timing and nature. At the same time, contracts were trading on whether U.S. forces would enter Venezuela, whether Maduro would be removed from power, and whether other related events would occur.

While public discussion was filled with rumors, speculation, and contradictory statements, the accused allegedly placed approximately $33,000 on outcomes that would follow from the planned operation. Once the operation had taken place and the contracts had settled, his alleged profit was nearly $410,000. He pleaded not guilty, so these claims remain allegations rather than facts established by a court. Regardless of the eventual verdict, however, the nature of the case reveals a fundamentally new problem.

If the prosecution’s account is correct, the trader did not profit because he had a better understanding of international politics or a more accurate assessment of U.S. intentions. He did not discover a hidden pattern, build a superior model, or notice something the crowd had missed. His possible advantage was access to reality before that reality became public. While everyone else was buying probability, he may have been buying knowledge.

A similar case involved a Google employee. The U.S. regulator accused him of using confidential information about the company’s upcoming Year in Search rankings. Contracts were available on which people and events would appear among the most popular Google searches of the year. For an ordinary trader, this would be a difficult analytical problem involving news coverage, cultural trends, celebrity popularity, and public interest. An employee with access to the final internal data before publication would be in an entirely different position.

According to the regulator, the accused traded more than twenty such contracts with near-perfect accuracy and generated approximately $1.2 million in profits. This case should also not be treated as definitive proof of an individual’s guilt before the legal process is complete. But once again, the structure of the market is clear: thousands of participants were attempting to predict the contents of a future list while one of them may already have seen it.

The situation went even further in the case of former U.S. Congressman George Santos. He traded a contract on whether he would attend the State of the Union address. The market was trying to determine whether Santos would appear at the event, but the primary source of information about Santos’s decision was Santos himself. While trading, he posted statements about his intentions, after which the contract price moved in a direction favorable to his positions. The regulator found the conduct manipulative, ordered him to surrender his profits, imposed a penalty, and temporarily banned him from trading.

This case exposes the most dangerous feature of event markets. A participant may not merely possess confidential information; they may also be able to influence the outcome being traded. The same person becomes the subject of the contract, the owner of a position, the source of public information, and one of the people capable of determining the final result. In a conventional market, a trader tries to anticipate what other people will do. Here, the trader can open a position, alter the information environment, and then personally decide the outcome.

This is no longer a prediction of the future. It is trading one’s own behavior.

The wallets that almost never lose

Individual criminal and regulatory cases might be dismissed as isolated incidents in a young industry. An analysis of trading activity on Polymarket, however, revealed a far more disturbing pattern. Researchers examined wallets that appeared shortly before significant events, quickly placed large bets on unlikely outcomes, collected their winnings, and then became inactive. These wallets were informally labeled “Orcas” because of the speed and precision with which they appeared to strike.

A total of 556 wallets displayed similar behavioral characteristics. Particular attention was given to 152 wallets that traded contracts involving military and defense events. According to the research, these wallets generated approximately eight million dollars in combined profits and achieved an average win rate of 97.2 percent.

Naturally, those figures do not prove that every wallet owner had access to military secrets. A blockchain records transactions and the movement of funds, but it cannot explain where the information behind a trade originated. Some of the results may be attributable to luck, selection effects, arbitrage, copied trades, or the distribution of positions across multiple addresses. Suspicious trading behavior is not, by itself, legal proof of a crime.

Yet statistics of this kind cannot simply be ignored. This was not one trader making several successful predictions, but a collection of anonymous wallets repeatedly buying low-probability military outcomes with a level of accuracy that ordinary forecasting can barely explain. It becomes even more difficult to dismiss the pattern as coincidence after the emergence of an actual criminal case in which a member of the military allegedly used classified information to profit from contracts tied to an upcoming operation.

The most troubling part of the story is not even that a possible insider can make money. The greater danger is that their actions may be visible to the entire world. Polymarket transactions are settled on a blockchain. The identity of a wallet owner may remain hidden, but the position, the size of the trade, and the timing of the entry can still be observed.

Imagine a recently created wallet that has shown almost no previous activity. It suddenly places a large bet on an unlikely military operation. An algorithm designed to detect unusual transactions notices the trade. A larger participant then enters the same market, followed by automated trading systems and other wallets. The contract price begins to rise even though no new information has appeared in public sources.

An ordinary trader sees the movement and concludes that the market “knows something.” A journalist notices the changing probability and begins looking for confirmation. Analysts discuss the unusual activity. Eventually, the same signal may be observed by the intelligence services of the country against which the operation is being prepared. No secret document has been published and no source has been identified, yet the actions of someone attempting to profit from confidential information may effectively become a public leak.

The market begins functioning as an open intelligence terminal. It does not reveal who the source is or whether the information is reliable, but it shows that an unknown participant is prepared to risk a substantial amount of money on a specific outcome. If similar transactions repeatedly precede real military events, these wallets will inevitably be monitored by more than speculators.

This creates a deeply uncomfortable paradox. The transparency of the blockchain helps researchers identify suspicious transactions and allows investigators to reconstruct events after the fact. Yet the same transparency enables trading bots, large market participants, and foreign intelligence analysts to detect a potential insider signal almost immediately. A mechanism designed to provide transparency also becomes a channel through which sensitive information can spread.

The market does not care where information comes from

It would be convenient to blame the entire problem on one particular platform, but that would oversimplify the reality. Polymarket says it monitors suspicious activity, cooperates with authorities, and reports wallets that may be connected to violations. In the case involving the U.S. serviceman, the platform reportedly alerted authorities to the suspicious activity. Its public blockchain also enabled independent researchers to identify patterns that might have remained invisible inside a completely closed system.

Similar cases have also occurred on regulated event markets. One political candidate traded contracts tied to the outcome of his own campaign. A YouTube channel editor was suspected of using advance knowledge of unpublished videos. The platform investigated the cases, confiscated profits, imposed penalties, and suspended the traders.

The problem therefore lies not only in technology, jurisdiction, or the regulatory status of a particular company. It is embedded in the very nature of prediction markets. When the outcome of an event is determined by a small group of people, members of that group inevitably possess an advantage over the rest of the market. When one individual can determine the outcome, that person effectively holds an absolute informational advantage. When the contract concerns a military operation, the circle of potentially informed individuals may include commanders, intelligence officers, technical staff, diplomats, contractors, and government officials.

Prediction markets assign a direct financial value to information that was previously far more difficult to monetize. To exploit political or corporate inside information in the past, a person had to find an appropriate security, estimate how the information would affect its price, and execute the trade through a conventional financial account. Now it may be possible to buy a contract directly on the event itself. There is no need to predict how oil prices will respond to a military strike if a trader can place a position on the strike. There is no need to estimate the effect of a resignation on a currency if the resignation date itself is being traded.

This leads to the harshest conclusion of all: perhaps prediction markets have not broken at all. Perhaps they are functioning exactly as markets naturally function. They reward the participant with superior information and cannot independently determine how that information was obtained. To a trading mechanism, deep analysis, random luck, a leak, a state secret, and a stolen document all look the same. They are simply different forms of informational advantage.

A market has no morality and asks no questions. It does not know whether people will die as a result of the event being traded. It cannot distinguish investigative research from a breach of national security. It sees only orders, volume, price, and position. If one participant knows more than everyone else, the system allows that participant to buy earlier and at a better price.

No market design can therefore solve the insider problem on its own. External restrictions are required: participant identification, conflict-of-interest controls, surveillance of suspicious trading, the freezing of questionable payouts, legal responsibility for the misuse of confidential information, and restrictions on contracts that may create serious security risks. Without such measures, the market will continue to perform its natural function—transferring money from less informed participants to better informed ones without asking where their knowledge came from.

When war itself becomes an asset

Financial markets have always reacted to war. Military conflicts affect oil, gold, currencies, bonds, agricultural commodities, and transportation costs. Producers use futures to hedge risk, investors seek to protect capital, and speculators try to profit from price movements. Traditionally, however, war was the cause of a market movement rather than the asset itself.

Prediction markets erase that distinction. The traded object is no longer the price of oil after a strike but the probability of the strike; not the exchange rate after sanctions are announced but the imposition of the sanctions themselves; not the market’s response to a change in government but the date on which a political leader will be removed. A political decision, military operation, or human tragedy becomes a binary contract with two buttons: “Yes” and “No.”

This does not mean that every buyer of such a contract wants a war to begin. A participant may regard the position as speculation, research, or even a form of hedging. Nevertheless, the existence of a tradable market changes the way events are perceived. The more serious the crisis, the greater the attention. The higher the uncertainty, the more attractive the trade. The closer the possible outcome, the more volume enters the contract.

While some people are trying to survive the consequences of a conflict, others are watching its probability change inside a trading application. For one person, a new attack means the destruction of a home. For another, it means that a contract settles at one dollar. This is where financial logic finally separates itself from human reality.

An especially dangerous conflict of interest arises when a participant can influence the event being traded. A politician may open a position before making a decision. A government employee may trade before a document is released. A corporate employee may act before confidential data becomes public. A member of the military may place a position before an operation known to them. Most people with such access will never misuse it for personal profit. But the infrastructure for doing so already exists, and the potential reward is displayed on a screen in real time.

In the past, exploiting secret information often required an intermediary, a carefully selected related asset, nominee accounts, and a complex attempt to conceal the resulting profit. Today, a smartphone, a cryptocurrency wallet, and a few transactions may be enough. That does not guarantee anonymity or immunity. Blockchains preserve evidence, platforms conduct investigations, and regulators are increasingly treating such behavior as insider trading or market manipulation. The technical barrier, however, has become much lower, while the temptation has moved much closer.

What the price really tells us

To an ordinary trader, a prediction market can appear to be an almost perfect information indicator. If the probability of a military event suddenly rises without an obvious public reason, it is tempting to assume that unknown participants possess reliable information. Sometimes they may. But the price alone cannot explain the origin of the movement.

Behind a large purchase may be a genuine insider, a skilled analyst, an automated system, a manipulator, or simply someone willing to place a large amount of money on an unlikely outcome. Other algorithms detect the trade and begin copying it. The crowd then joins the movement, treating the rising price as confirmation that new information exists. A single questionable transaction can therefore generate an entire market narrative.

Copying a suspiciously successful wallet does not automatically create an advantage. The original participant may have entered at a much better price, distributed positions across several addresses, or hedged the risk on another platform. The trader may close the position as soon as bots begin following it. An outside observer sees only part of the structure and cannot know whether the visible trade is a genuine directional position, bait for copy traders, or one component of a more complex strategy.

The price of a prediction contract should therefore not be treated as an objective probability. It is merely the outcome of transactions between particular participants within a particular level of liquidity. In a deep market, an erroneous order may be absorbed by the surrounding flow. In a small contract, one large position may significantly move the price and create the appearance that important new information has arrived.

The central lesson is that a market that moves ahead of the news is not necessarily more intelligent. It may simply include someone who has already read tomorrow’s headline. Or it may include someone who wants everyone else to believe that they have.

The world really has gone off course

Human beings have always tried to predict the future. They have used religion, astrology, mathematics, statistics, opinion polls, intelligence gathering, and financial models. Now the future has been divided into thousands of short contracts, each with its own price.

A military operation takes place: one dollar. It does not take place: zero. A politician leaves office: one dollar. The politician remains: zero. A strike occurs before a specified date: one dollar. Nothing happens: the contract becomes worthless.

Complex historical processes, human decisions, and real tragedies have been placed inside a simple trading interface. The real problem, however, does not begin when people try to profit from their own forecasts. It begins when a position is opened by someone who can determine the outcome or already knows that the decision has been made.

Prediction markets were designed to extract truth from collective knowledge. But a trading system cannot distinguish knowledge from a stolen secret. It sees only price. When that price begins moving before a military operation, an official announcement, or the release of confidential data, it is impossible to know immediately what has happened. Did the market genuinely predict the future, or did someone begin selling information about an event that had already been decided?

Financial markets once traded the consequences of world events. Now the events themselves have become tradable assets. An insider once tried to determine where the price would move after the news. Today, the movement of the price may reveal that the insider already knows the news.

This cannot be explained solely by technological progress or the appearance of a new financial instrument. It is the result of a much deeper process: the market is gradually acquiring the ability to place a price on everything—a political decision, a state secret, a military operation, and a human life.

Even on things that should never have had a price.