A 14% bet on another UK prime minister by New Year
A Polymarket contract gives Keir Starmer's government a 14% chance of a leadership change before 31 December. The price is small, but the framing tells you something about how political risk is being priced.
At 16:26 UTC on 20 July 2026, the prediction market Polymarket priced a fresh occupant of 10 Downing Street before midnight on 31 December at 14%. The contract, listed as "Another UK prime minister by the end of the year?", puts Keir Starmer's tenure inside a narrow corridor of market doubt: not a coup, not a rout, but a measurable tail risk that an operator in the City can hedge against for a few pence.
That is the story. Not that Starmer is finished, but that political longevity in Westminster has become something traders can write a price on, and that price is non-zero. The 14% figure is the kind of number that lives in the footnotes of British political journalism, ignored at the writer's peril. It says, plainly, that one in seven informed money thinks the parliamentary arithmetic changes between now and the spring budget.
What the market is actually saying
Prediction markets are not polls and they are not gossip. They are pools of bearer capital with a price discovery problem to solve, and they are usually wrong about personalities but roughly useful about probabilities. A 14% reading on a binary "another PM by year-end" contract means that, on the books, traders are willing to pay roughly 14 pence for every pound they would collect if a change of occupant happened. The implied odds sit somewhere between a long shot and a foregone conclusion.
The contract page itself is the only public artefact of the bet. It does not name a challenger, it does not name a cause. It just registers the question, the deadline and the price. That is also a statement about how Westminster is being read from the outside: as a system whose continuity can no longer be assumed between budgets.
The Westminster context the price sits inside
British prime ministers have not, historically, been priced on prediction exchanges. The job was assumed to last at least one full parliamentary term, and the assumption shaped how the press wrote about every leadership crisis. That assumption has frayed. Liz Truss lasted 49 days in 2022. Boris Johnson resigned under pressure the same year. Rishi Sunak followed. Starmer inherited a parliamentary majority in 2024 that is now a working majority, not a sleeping one, with backbench discontent organised around specific bills rather than ideological factions.
The 14% figure does not forecast any of this. It just registers that the distribution of outcomes has thickened. A trader is no longer pricing the question "will the PM survive?" as a near-100% yes. They are pricing it as a question with a meaningful no.
Why the small number matters more than the big one
The instinct is to treat 14% as marginal and move on. That instinct is wrong for two reasons. The first is mechanical: prediction markets work on liquidity, and a contract traded at any non-trivial price is a contract that someone with real money believes in. The 14% is not a residual rounding error; it is a position.
The second is structural. Political risk in the United Kingdom is now legible in the same vocabulary as credit risk and currency risk, and that legibility changes how it gets priced into everything else. Gilts, the pound, London-listed bank stocks and the cost of UK sovereign insurance all reference the same underlying assumption about continuity. The more that assumption is openly questioned in a transparent market, the more it costs to underwrite.
This is the broader pattern. Political events that used to be absorbed by the press are now absorbed first by markets. The market's first read becomes the press's first read, and the press's first read becomes the government's headache. Polymarket does not cause a leadership crisis. It just makes the crisis legible in advance.
What could move the number
Three plausible catalysts sit inside the contract window. A budget defeat on a whipped vote would force a confidence motion and would, in practice, end a government. A by-election loss in a Labour-held seat at a parliamentary majority of working size would convert internal grumbling into a leadership letter. A scandal that breaks between summer recess and the party conference season would arrive with no parliamentary sitting days to bleed pressure off. None of these are forecast in the 14%. They are simply the kind of event the price is willing to entertain.
A 14% probability on a year-end leadership change does not, on its own, mean any of those events are imminent. It means the market no longer treats them as impossible, and that is enough to change the cost of holding UK political risk into the autumn.
What we do not know
The contract page gives a price and a deadline. It does not give the size of the position, the identity of the traders, or the historical calibration of this specific market against the actual frequency of UK leadership changes. Polymarket's record on political contracts has been mixed: it caught the Brexit and Trump calls early and misread several state-level US races. Whether 14% on a UK prime minister is high or low in absolute terms is a question the market itself does not yet have enough data to settle. Treat the number as a signal about how the question is being asked, not as a verdict on the answer.
How Monexus framed this: the wire reported the contract; this publication treated the price as a reading on the cost of political risk rather than as a forecast of any specific event.