Kalshi and Polymarket recorded a record total trading volume of $50.6 billion in July. However, this figure is almost by nature misleading.
Six weeks before American and Israeli aircraft killed Iran's supreme leader on February 28, a group of new, previously inactive crypto wallets began buying Polymarket contracts predicting "Khamenei out of power by March." In the hours preceding the attack itself, six accounts generated total profits of $1.2 million from bets placed with almost absolute timing accuracy relative to the strike. By the time the operation concluded, $529 million had been bet on the strike timing and another $150 million on Khamenei's removal — with the majority of bets resolving in favor of traders before the Pentagon released a single word. No diplomat, no briefing, and no news agency anticipated this market. This week, as Kalshi and Polymarket announced a total monthly trading volume exceeding $50 billion, a new wave of commentary questioned whether prediction markets have evolved into a faster, cleaner signal than official diplomatic statements. It is the right question. However, most of those asking it are going to get the wrong answer, notes Cointribune.
The $50.6 billion and the misleading picture
Kalshi and Polymarket recorded a record total trading volume of $50.6 billion in July. However, this figure is almost by nature misleading. Approximately $20 billion came solely from contracts related to the World Cup, while sports betting and bets on crypto prices — rather than geopolitics — drove both platforms to new historical highs. The genuinely political and geopolitical segment is smaller, but it is growing faster and carries significantly greater importance. Polymarket alone handled over $2 billion in war-related bets during the worst weeks of the Iran conflict in early March. Both platforms now host live, continuously re-priced contracts concerning the fluctuating ceasefire process in Ukraine — including a contract, until this week, on whether Russia would even respond to a Ukrainian proposal to end attacks on civilian targets in the Black Sea. This is not a new debate. In 2003, the research arm of the Pentagon itself attempted to create a government version of this exact model, the Policy Analysis Market. Congress abolished it within a week of its launch after senators labeled it a "terror gambling parlor." Twenty-three years later, the public is conducting this experiment anyway — just without anyone inside the government capable of turning it off.
The argument for markets is stronger than skeptics admit
Let us begin with the defense of markets, because it is stronger than most critics admit. The internal journal of the Central Intelligence Agency itself argued in 2006 that prediction markets could outperform conventional analysis. It cited the success of the Iowa Electronic Markets in predicting presidential elections, orange juice futures outperforming official forecasts by the National Weather Service, and an internal experiment at HP Labs where prediction markets among employees beat official sales forecasts roughly three out of four times. The mechanism is simple. Markets aggregate scattered information that no single analyst, no matter how high-ranking, possesses in its entirety. The Iran bets prior to the strike look, at first glance, like this exact mechanism functioning at war speed: money moving based on information before it becomes news.
When "prediction" looks more like a leak
However, if we examine more closely what kind of information this was, the phrasing that "markets beat the CIA" begins to fall apart. Companies specializing in blockchain analysis traced the profits before the strike to a small number of wallets with no prior trading history before mid-January. These wallets were almost exclusively concentrated in contracts tied to the specific date and target of the attack. This is not the classic picture of scattered public information converging into a price. It is the picture of a leak — or an extraordinarily successful estimate by someone close to the operational planning — monetized through an anonymous, liquid, and legal financial tool. A member of Congress has already demanded answers from Polymarket regarding a single $500,000 bet on the Iran attack. The market did not outperform US intelligence in analysis. At best, it gave someone with inside information — American, Israeli, or Iranian — a faster and harder-to-trace way to monetize it than anything available in the past. This represents a truly significant development for open-source intelligence. However, it is not the same as claiming that "the market knows more than the analysts." Confusing the two is precisely the error currently spreading through the international relations debate.
The contract trap
The second problem is more mundane and, for analysts, far more dangerous: contract phrasing. A widely circulated Polymarket contract asking whether Russia and Ukraine would achieve a ceasefire before the end of 2026 resolved as "yes" months ago. The market deemed the condition fulfilled due to a 32-hour Orthodox Easter truce in April, which both sides violated within hours and which collapsed back into full-scale conflict by the time the exchange paid out. Anyone reading that resolved contract as proof that the war had ended or was heading toward an end would have made a specific and serious error. Kyiv and Moscow have since gone through at least two more ceasefires and collapses, while this week Ukraine awaits a response to a far more limited proposal concerning only civilian targets in the Black Sea. The Iran contracts face the opposite problem. A $345 million Polymarket contract for a "permanent peace agreement" remains blocked due to disputes among traders over whether a temporary US-Iran agreement satisfies the condition. The reason is that the contract wording was never tied to a specific, verifiable event. The price does not represent a probability for the event you think counts, unless the contract itself precisely defines what that event is. And most do not.
Neither ignoring them nor blindly trusting the price
Combining these two problems, the honest stance is neither "ignore these markets" nor "trust the price." Prediction markets are best read as real-time leak and sentiment detectors rather than oracle machines. They are genuinely useful for identifying when someone with access to information moves early. However, they are completely useless — and often actively misleading — when used on their own as an assessment of probability for what is actually going to happen. This is precisely the distinction that the political system never allowed to mature. The Policy Analysis Market of 2003 was not shut down because its statistical foundation was wrong. CIA analysts themselves still refer positively to the research behind it. It was abolished because official betting on state violence is politically toxic. The same reaction is visible today in congressional letters regarding bets on Iran and in the delayed rollout by the SEC of ETFs tied to prediction markets. The result is that a genuinely useful analytical tool has so far existed only in an unregulated, marginally gambling-adjacent format where World Cup fans, major political traders, and potentially informants bet together in the same order book. And there is no institutional mechanism capable of separating the signal from the noise.
Three possible scenarios for the coming year
Base scenario – approximately 55% probability
The current pattern continues without significant correction. International relations analysts, journalists, and even some government officials increasingly use Polymarket and Kalshi prices in briefings and memos as a supplementary indicator, because a price feels more concrete than a qualified diplomatic report. Occasionally, embarrassing misinterpretations will surface — such as the Ukraine ceasefire episode resolved as "yes" — which will be mocked on social media and subsequently quietly corrected without altering core practices. Trading volume will continue to grow, sports betting will remain dominant in total numbers, and geopolitical contracts will stay a noisy minority attracting attention disproportionate to the percentage of money they move.
Negative scenario
A market based on the price of a specific event will influence an actual political or military decision — an attack, a negotiating position, a sanctions decision — and it will later emerge that the decision was influenced by a small, possibly coordinated group of informed or manipulative traders rather than genuine collective wisdom. The ensuing scandal, amplified by a Congress already sensitized by the Iran betting case, will trigger the exact type of political backlash that led to the shutdown of the Policy Analysis Market in 2003. Restrictions or a total ban could be imposed on geopolitical contracts and war-related contracts, distinct from the pressure campaign already underway regarding sports betting. The useful signal — the detection of informed money — would be regulated along with the noise, because no one had built the tools to separate them in time.
Positive scenario
A specific and credible entity — a quantitative research firm, an open-source intelligence organization, or a university lab building on wallet-tracking work currently done informally by companies like Bubblemaps — could transform the leak-detection method into a repeatable and reliable product. It would identify anomalous, recently activated wallets concentrating bets on strictly defined and verifiable contracts, separating them from the usual political and sports money flow. This would convert today's fragmented journalistic practice into the analytical tool envisioned by the 2006 CIA study, without requiring government agency funding or triggering a political battle that leads to its cancellation.
The real story is not "markets versus intelligence agencies"
The interesting story was never that "prediction markets are better than intelligence agencies." That phrasing flatters both sides and tests neither. The real finding is more limited and far more useful: These markets occasionally move prior to public disclosure because someone holding real information trades on it. All this occurs within a much larger volume of sports betting and vaguely phrased contracts, which can mislead anyone reading the price as a probability. The real objective, therefore, is not learning to trust the number. It is learning to distinguish between these two categories.
The next test comes in the Black Sea
It is worth watching what happens to the contract price regarding the Ukraine ceasefire proposal in the Black Sea over the coming days. Kyiv submitted the proposal on August 13, and Moscow has yet to respond. If the market moves sharply before any official Russian statement, it will serve as one of the cleanest live tests for this methodology. It will reveal whether these platforms are detecting a real signal — or merely repricing based on rumor.
www.bankingnews.gr
Σχόλια αναγνωστών