Picture a hypothetical Polymarket trader in the summer of 2024, in the weeks before Joe Biden stepped aside. Biden is still on the ticket, but his position is under pressure and Kamala Harris is the obvious replacement if he goes. Imagine two contracts stacked on the same screen. One asks whether Harris will win the presidency. The other asks whether she wins, conditional on becoming the Democratic nominee at all. The prices would not be the same. They cannot be, and the gap between them is where the interesting information lives.
That gap is what a conditional market is for. It isolates one variable and asks a sharper question than a headline contract ever could. These markets have existed in academic experiments for decades, but in the wild they stay rare. On Polymarket they surface only occasionally, more often as a refund clause tucked into an individual market than as a distinct, headline product type. They are worth understanding anyway, because once you grasp the mechanic you start seeing where a conditional would sharpen a question.
A trader who wants to bet on Harris, but only if
Stay with that hypothetical summer-2024 screen, Biden still on the ticket. Our trader thinks Harris beats Trump in a straight fight. Fine. A direct market lets them buy YES on Harris to win the presidency. But suppose they think the odds shift meaningfully depending on whether Biden stays on the ticket, drops out, or is replaced through some other process. A single headline contract cannot express any of that. It smushes every possible path into one number.
A conditional market unsmushes it. It asks: given a specific antecedent (Harris is the nominee), what is the probability of the consequent (Harris wins the general)? The contract only pays out YES if BOTH legs come true. If the antecedent fails, the market voids and stakes are returned. That is the whole mechanic. Two events, joined by an if.
The payoff structure sounds fiddly. In practice it feels closer to a bet with a get-out clause. You are not exposed to the noise around whether Harris becomes the nominee at all; you are only expressing a view about what happens once she is.
Where the probability actually comes from
Think of it as long division for uncertainty. If you know the price of a joint market (A and B both happen) and you know the price of the antecedent alone (A happens), the conditional probability (B given A) is roughly the first divided by the second. That is Bayes' rule in a pub-quiz frock.
On Polymarket, the price you see on a conditional contract already bakes that division in. Traders arbitrage between the standalone market, the joint market, and the conditional one, and prices settle where the maths hangs together. When they do not hang together, arbitrage opportunities open up, and disciplined traders work to close them, though on a thin conditional slice that can take far longer than the textbook implies.
The cleaner way to think about it: a conditional market strips out one source of uncertainty so you can trade the other. If you have a view on the consequent but no view on the antecedent, you can now express it without paying to be wrong about the bit you do not care about.
A short history of a very old idea
Robin Hanson, the economist most associated with prediction markets, was writing about conditional markets in the late 1990s under the banner of "futarchy" and "decision markets." The pitch was ambitious. Governments and companies could use conditional contracts to price the outcome of policies before enacting them. Would GDP be higher if the central bank hikes rates, conditional on the hike happening? A market could answer.
The pitch never fully landed in policymaking. It did land, quietly, in political forecasting. Academic experiments at the University of Iowa's electronic markets in the 2000s used conditional contracts to price nomination-then-election paths. What was novel then still turns up on Polymarket, in dollars and open to retail, but as an occasional structure rather than a routine, at-scale product. More often the conditional logic shows up as a refund clause on a single market than as a headline contract of its own.
That matters because the theoretical apparatus is well developed. Conditional markets are not a Polymarket invention, and they are not really a Polymarket product line either. They are an occasional, real-money application of an idea economists have been kicking around for thirty years.
What conditional markets are not
Worth flagging: a conditional market is not a hedge in the insurance sense. If you buy the conditional and the antecedent fails, you get your stake back. You do not get paid out because your side loses. That distinction matters because retail traders sometimes treat these contracts as if they contain downside protection. They do not. They contain scope narrowing.
They are also not the same as multi-leg parlays on a traditional sportsbook. A parlay pays only if both legs win; a losing leg costs you the stake. A conditional market voids the stake if the antecedent fails. The mechanics are structurally different, and the difference between event contracts and traditional betting products is exactly the sort of thing that gets glossed over in slick platform copy.
And they are not the same as a scalar or range market, where the payoff depends on a numerical outcome landing in a bucket. Conditionals are still binary at settlement. The if just gates when settlement happens at all.
Why the mechanic matters for readers and traders
The boring answer is that conditional markets let you express more precise views. The more interesting answer is that they let you decompose a messy question into its parts, and the parts are usually easier to price than the whole.
Take nomination markets. A single "who wins the presidency" contract bundles two questions: who is the nominee, and who wins the general. If you have a strong read on the second but a shaky read on the first, the headline market forces you to price both. A conditional contract lets you trade only the piece you actually understand.
The same logic applies to a football tournament. A market on "Team X wins the World Cup" bundles the group stage, every knockout round, and the final. A conditional "Team X wins the final, given they reach it" lets a trader express a view about the final matchup without paying for the earlier rounds. It also produces genuinely useful information for other traders, because the conditional probability is often more stable than the headline number.
How to read one without getting confused
Start with the antecedent, always. What has to be true for this contract to settle at all? If that condition fails, everything else is moot and your stake comes back. Once you have the antecedent clear in your head, the consequent is just an ordinary binary market with a narrower scope.
Check the resolution wording carefully. Conditional markets have twice as many resolution edges as normal ones, and platforms occasionally write ambiguous rules about what counts as the antecedent being satisfied. If you are new to this, our guide to how prediction markets decide who wins is worth twenty minutes before you put money down.
Then look at the standalone market alongside the conditional. If the standalone gives, say, a 40% probability to the antecedent, and the conditional trades at 60% for the consequent given the antecedent, you can back out an implied joint probability of roughly 24%. If the joint market itself is trading at 30%, something is off, and somebody is going to arb it.
Where the mechanic breaks down
Thin liquidity is the honest answer. Conditional markets are always sliced off a parent question, and the slice tends to attract less volume than the headline. That means wider spreads, more slippage on any real-sized order, and prices that can drift away from their theoretical relationship with the standalone market for longer than the textbook suggests. If you care about why some contracts have paper-thin order books, that piece covers the mechanics.
Resolution ambiguity is the second failure mode. If the antecedent is defined loosely ("if Trump is the Republican nominee" is easy, "if the Democrats have a competitive primary" is not), the market can end up in messy dispute territory. Polymarket's UMA-based resolution process handles most of these cleanly, but not all, and the cases where it does not are exactly the cases where money is at stake.
The third break is more subtle. Traders with strong priors sometimes let those priors bleed across from the standalone into the conditional, pricing them in ways that do not quite hang together mathematically. When that happens, the discipline of Bayes' rule reasserts itself eventually, but eventually can be weeks away.
Getting started without losing your shirt
Do not touch a conditional market with real money until you have watched one settle. Pick a contract with a near-term antecedent, follow it through both possible outcomes, and see how the resolution actually plays out. That single loop teaches more than any explainer.
Once you understand the flow, size small. Conditional markets reward traders who have a specific, defensible view on the consequent-given-antecedent, not traders looking for excitement. If your reason for entering the trade cannot be stated in one sentence beginning "given that X, I think Y is mispriced because," you probably should not enter it.
iPredicta watches for conditional contracts alongside headline markets across Polymarket, Kalshi and the UK-regulated venues, and flags when the maths between related markets stops hanging together. It is the sort of thing worth watching whether you plan to trade the arb or just want a sharper read on what the crowd actually thinks.
Frequently asked questions
How is a conditional market different from a regular prediction market?
A conditional market only settles if a specified antecedent event happens first; a regular market settles regardless. If the antecedent fails, a conditional contract voids and stakes are returned, whereas a regular market always pays out YES or NO. That structural difference lets traders isolate one variable in a chain of events rather than pricing the whole chain at once. In practice, this means conditional contracts often trade at very different levels from their headline counterparts, because they express a probability that has been stripped of one particular source of uncertainty. They are a scope-narrowing tool, not a hedge.
What happens to my money if the conditional part never occurs?
Your stake is returned; the market voids. If you buy YES on a contract asking whether Team X wins the final given they reach it, and Team X is knocked out in the semis, the antecedent has failed. The market settles as void, and your position closes at cost. This is the single most important thing to understand before trading one. Retail traders sometimes assume a losing antecedent means a losing position; it does not. It means the trade never really happened. You still bear the opportunity cost of tying up capital during the market's life, though, which is not nothing.
Are conditional markets legal to trade in the UK?
Broadly, UK regulation restricts direct access to Polymarket's contracts, including its conditionals, for retail users, and the picture is nuanced enough to be worth reading in full. The FCA and the Gambling Commission each have views on where prediction-market contracts sit, and the answer depends on the platform, the contract type, and how it is structured. Our guide to whether Polymarket is available to UK users walks through the practical position. UK-regulated venues like Smarkets offer event markets, but the specific conditional structures you see on Polymarket are less common on the regulated side.
Can conditional markets be used to make policy decisions?
That is the futarchy pitch, and it has been argued by economists including Robin Hanson since the late 1990s, but real-world adoption in policymaking is thin. The idea is that a government facing a decision could set up conditional markets pricing the outcome under each option, then pick the option with the more favourable market view. In practice, political and institutional resistance has kept the concept mostly theoretical. Some corporate settings have run small internal experiments. Where conditional logic does surface today, it tends to be in forecasting rather than policy, with political nomination-then-election paths the standard illustration, though even there the contracts stay uncommon rather than routine.
Why do conditional markets sometimes trade at prices that don't add up?
Usually thin liquidity, sometimes trader bias, occasionally resolution ambiguity. Conditional contracts are sliced off a parent market and inherit only part of its volume, so bid-ask spreads are wider and prices can drift from their theoretical relationship with the standalone contract. Traders with strong priors on the consequent can also leave the conditional mispriced relative to the joint market for surprisingly long stretches. When the maths stops hanging together, arbitrage opportunities open up, and disciplined traders eventually close them. Eventually is doing some work in that sentence.