In February 2026, a headline on Iran and the Strait of Hormuz is enough to move oil, gold, and the stock market within hours. Reuters described a ‘geopolitical premium’ on Brent estimated at between $7 and $10 per barrel, linked to uncertainty over a possible failure of diplomacy and the possibility of US or Israeli military action.

So far, nothing new: markets have always reacted to crises. What is new is something else: in 2026 geopolitics is not only interpreted by analysts, but is directly priced by platforms that turn political events into tradable probabilities. The best known of these is Polymarket.

Polymarket was founded in 2020 by entrepreneur Shayne Coplan, with the aim of revolutionising the betting and forecasting industry through blockchain technology. The platform has raised over $70 million in funding from prominent investors such as Peter Thiel, Vitalik Buterin and others.

The platform is a ‘prediction market’: a place where you do not bet against a bookmaker, but trade contracts linked to real events (elections, economic decisions, international crises) as if they were assets. If ‘Yes’ costs $0.75, the market is implicitly saying: ‘75% probability’.

This idea – aggregating dispersed beliefs and turning them into a number – is moving out of the niche. In January 2026, Dow Jones (News Corp) signed an agreement to integrate Polymarket data into its editorial products, from the Wall Street Journal to Barron’s and MarketWatch, with dedicated modules also on market expectations (e.g. profit calendars). And already in 2024 there were signs of ‘institutionalisation’ of the phenomenon with the integration of Polymarket shares in financial information environments.

The Iran case: likelihood of escalation as a public thermometer

At the height of tensions in February 2026, Polymarket hosts a constellation of markets on Iran (strikes ‘by date’, escalation, etc.). The ‘Iran predictions’ section indicates that the most active market assigns a high probability to ‘by the end of the year’ strike scenarios, with odds updated in real time.

This is the crucial point: the platform makes visible and exchangeable a geopolitical expectation that, until yesterday, remained confined to intelligence reports, think tanks or diplomatic desks. And a ‘public’ probability has a performative effect: it circulates on social networks, is mentioned in market commentaries, influences perceptions and coverage (energy, defence, gold), becomes a piece of political discourse.

The risk: mistaking a ‘clean’ probability for a truth

This is where the most interesting critique emerged in 2026, in an academic paper that is circulating widely: Prediction Laundering: The Illusion of Neutrality, Transparency, and Governance in Polymarket (Rohanifar, Ahmed, Sultana, University of Toronto / UIUC).

The central idea is interesting: Polymarket does not just produce a prediction; it produces epistemic authority. And it does so by ‘cleaning up’(laundering) the uncertainty and asymmetries behind an easy-to-consume number.

The authors describe a four-stage cycle:

1. Structural sanitisation: someone decides what is ‘betable’ and in what form (which futures become contracts).
2. Probabilistic flattening: different motivations (conviction, hedging, propaganda, pure trading) end up in a single percentage.
3. Architectural masking: the platform can give the impression of ‘crowd consensus’ even when price movements are driven by a few large capitals (‘whales’).
4. Epistemic hardening: disputes and ambiguities disappear from the final showcase; the percentage as ‘historical fact’ remains.

The number appears neutral, but it is often the product of a socio-technical process that may include strategic manipulation, financial incentives and strong information inequality. In other words, one risks the classic error: treating a probability as a truth, instead of a signal.

If we want to treat Polymarket as an information tool (not as an oracle), three things help:

1. Look at the definition of resolution: what counts as a ‘strike’? what sources? what thresholds?
2. Read the number as ‘trading sentiment’, not intelligence: useful for understanding what the market fears or hopes, not for understanding what will happen.
3. Compare with traditional signals: premium on oil, movements on gold, official communications, reports: prediction markets are a piece of the puzzle, not the puzzle.

Polymarket and the US elections: a case study

The 2024 US presidential election is a significant test case for Polymarket.

Here you see, for example, predictions on the upcoming US elections:

In terms of predictive ability, Polymarket proved more accurate than traditional opinion polls. For example, it promptly picked up on Biden’s drop in support several days before his retirement, adjusting the candidates’ odds accordingly.

This highlights the potential of the forecast aggregation model based on economic incentives. However, some results such as Trump’s victory did not fully reflect the final outcome, indicating room for improvement.

An indirect actor in the new geopolitical ecosystem

In the case of the tensions between Iran and the US/Israel, for example, the trend of strike probabilities began to circulate in parallel with the movements of Brent and gold. Even when they do not directly determine the price of oil, these odds function as a public thermometer of strategic risk: if the probability rises rapidly, it becomes a narrative signal that reinforces the idea of imminent escalation. It is a feedback mechanism: geopolitics feeds probability, probability feeds perception, perception influences financial hedging.

In a multipolar world, where deterrence is also measured in communication and expectations, this is a structural change. Probabilities become a form of ‘informational soft power’: they synthesise complex scenarios into immediately understandable and shareable numbers. They do not replace strategic analysis, but compress it into a language accessible to global markets. And it is precisely this fast, numerical, apparently neutral compression that makes Polymarket an actor, albeit indirectly, in the new geopolitical ecosystem.