An event contract resolves when its defined outcome is known. The winning side redeems for full value, the losing side expires worthless, and the rule that decides it is written down before you ever trade.
An event contract settles when the event it is tied to is decided. A contract that resolves yes pays its full value, usually one dollar, and a contract that resolves no expires worthless. The source of truth, the resolution criteria, and the timing are all written into the contract terms before trading opens. On regulated United States venues those terms are filed with the Commodity Futures Trading Commission as part of the exchange listing process. Settlement is the moment the market stops being an opinion and becomes a fact.
While a market is open its price is an estimate. At settlement the question is answered, and each contract is worth either its full value or nothing. Nothing about the price during trading changes that final binary result.
Every market states which data or authority decides the outcome, for example an official statistics release or a published result. That source is set before trading, not chosen after, which is what keeps resolution honest.
If your side resolves yes, each contract redeems for its full face value. If it resolves no, those contracts expire worthless. Your profit or loss is the difference between what you paid and what you received.
If a data source is delayed, revised, or unclear, the rulebook says what happens. Regulated venues publish how ambiguous or contested outcomes are reviewed, so resolution does not come down to a single unchecked decision.
Settlement, sometimes called resolution or expiry, is the point at which a market closes for good and each contract is paid out according to the result. Before that point the price floats between one and ninety nine cents and reads as the market implied probability of the event. After settlement the contract is worth one of two values only. There is no middle ground and no partial credit.
This is the feature that separates an event contract from a simple wager with a bookmaker. The contract has a defined payout, a defined event, and a defined way of measuring whether the event happened. When you understand how a given market settles, you understand the actual product you are holding rather than a vague bet on a feeling.
Every well written market names a source of truth. For a market on an economic figure that source might be an official government release. For a market on a published result it might be the body that publishes that result. The contract terms spell out exactly which release, on which date, and how the number maps onto a yes or no answer.
On exchanges registered with the Commodity Futures Trading Commission, these terms are filed with the regulator as part of how a contract is listed. The Commission expects participants to receive timely and transparent information about a contract, including how the payout works, who decides the outcome, and how that determination is made. Reputable reporting on the larger venues describes the same practice, where each market lists named source agencies in its terms.
Read the resolution criteria before you trade, not after. Two markets that look similar can resolve on different sources or different cutoff times, and that difference can flip a result. The criteria are the contract. The headline question is only a summary.
A market usually has a close time and a settlement time, and they are not the same moment. The close time is when trading stops. The settlement time is when the result is confirmed and contracts are paid. Between the two the outcome may already look obvious, but the contract is not final until the named source reports and the venue applies the rule.
In practice many markets settle within hours of the result becoming known. Some take longer when the venue is waiting on official data that is delayed or subject to revision. The rulebook decides whether to wait, so a market can sit unsettled even when the world already knows the answer.
On a centralized exchange an internal team applies the published rule and confirms the result against the named source. The team follows the contract terms rather than its own judgment, and larger venues describe review committees that can make a binding call when an outcome is genuinely ambiguous.
On decentralized venues an oracle reports the outcome on chain and a dispute window lets participants challenge a result before it is finalized. The mechanics differ, but the goal is the same, which is to tie the payout to a defined and checkable source rather than to anyone's opinion. Always check how a given platform resolves, because the model varies.
Real world data is messy. A figure can be released, then revised. An event can be postponed. A question can turn out to be ambiguous in a way nobody expected. Good contract terms anticipate this and state which value governs, whether a revision counts, and what happens if the event does not occur as defined.
If a market cannot resolve fairly, the rules may void it and return stakes, or settle it on a fallback source. None of this is automatic goodwill. It is written procedure, which is exactly why the terms matter more than the headline. When you cannot find a clear rule for a messy case, treat that as a risk in itself.
Your net result is not just the face value of a winning contract. Trading fees, and on some venues settlement fees, reduce what you keep. The spread you paid to enter also matters, since buying at a worse price lowers your return even when your side wins. Always read the fee schedule for the platform you use, because the structure differs from one venue to the next, and we do not quote a single number that would go stale.
Settlement is also where counterparty and platform risk become concrete. Your payout depends on the venue actually crediting and letting you withdraw it. That is one more reason the venue and its rules matter as much as the result of the event.
The principle of settlement is shared everywhere, which is that payout follows a defined and checkable source rather than anyone's opinion. The machinery that applies it differs by venue, and knowing which model you are trading under tells you who decides, how a dispute is handled, and how long you might wait. Three models cover most of what a reader will meet.
On a centralized exchange registered with the Commodity Futures Trading Commission, each market lists named source agencies in its contract terms, and those terms are filed with the regulator through the exchange self certification process. When the result is known, an internal markets team confirms it against the named source and finalizes the market. Reputable reporting on Kalshi describes exactly this model, with most markets settling within a few hours of the outcome, often around three hours, and no separate settlement fee, per Kalshi's published rules and help center as of 2026.
The team follows the written rule rather than its own judgment. Where an outcome is genuinely ambiguous, the Kalshi rulebook provides for an Outcome Review Committee, a committee of the board that can make a binding determination, per the KalshiEX rulebook as filed with the CFTC. Traders can submit a request to settle through the interface, but that acts as a suggestion rather than a binding instruction, and the markets team makes the final call. The strength of this model is a single accountable operator. The trade off is that you are trusting that operator to apply the rule fairly.
On a decentralized venue the result is reported on chain through an oracle, with a window in which anyone can challenge it. Polymarket uses the UMA optimistic oracle, where anyone can propose an outcome and, per Polymarket's documentation as of 2026, a two hour dispute window opens during which a challenger can post a matching bond. If the proposal survives the window, the market resolves to that outcome.
The escalation path is worth understanding before you rely on it. A first dispute does not trigger a vote, it cancels the proposal and restarts the cycle with a fresh proposal, which prevents a single frivolous challenge from blocking resolution. A second dispute escalates to UMA's Data Verification Mechanism, a vote of token holders that makes the final call. This is not a rare path. Reporting indicates Polymarket logged more than 1,150 disputed markets during 2026, already past its full year 2025 total, and a high volume 2026 market on whether a named company would sell Bitcoin by a stated date was disputed twice and left sitting in the oracle queue. The strength of this model is that no single operator decides. The trade off is that resolution can become slow, adversarial, or politicised when a question is close.
A narrow venue such as PredictIt lists political and election contracts only, and those settle on official outcomes such as a certified election result or a formally announced decision. The model is centralized, with the operator applying the written criteria, but the range of questions is deliberately small, which keeps most resolutions straightforward. As always, the contract terms govern the unusual cases, so read them rather than assuming the headline result will map cleanly.
| Model | Who decides | Dispute path | Typical timing |
|---|---|---|---|
| CFTC registered exchange | Internal markets team, named source agencies in filed terms | Outcome review committee makes a binding call | Often a few hours after the result |
| Decentralized oracle venue | Proposed on chain, confirmed if undisputed | Dispute window, then a token holder vote | Two hours if undisputed, longer if challenged |
| Focused political venue | Operator applies written criteria to official results | Handled under the contract terms | After the official result is certified |
Methodology: compiled by Prediction Market Index from each venue's published rules and documentation and from reputable reporting, as of October 2025. Timings are typical, not guaranteed, and individual markets vary. Confirm the rule for the specific market you trade.
The normal case is clean, the source reports, the rule applies, and the market pays. The cases that catch people are the messy ones, and they are common enough to plan for. Take a revised figure. Suppose a market resolves on an official economic release, the first print lands above the threshold, and a later revision drops it below. Which value governs is decided by the contract terms, not by which number feels fairer. Good terms say in advance whether the initial release or a revision controls, and the careful trader reads that line before trading rather than after a revision moves against them.
Take a postponed event. A contract written on something happening by a stated date has to say what counts if the event slips past that date, whether that resolves the question no, voids the market, or extends the deadline. Take an ambiguous wording, where the headline question seems obvious but the precise definition in the terms decides a borderline result. A market on whether a figure reaches a round number turns on whether the rule reads as at least that number or strictly above it. These are not exotic problems. They are the reason two markets that look identical can pay differently, and the reason we say the criteria are the contract while the headline is only a summary.
Settlement is where a contract pays its full value or nothing, but holding to settlement is a choice, not a requirement. On a venue with a working market you can usually sell a position before the close and take the current price instead of waiting for the binary result. That locks a smaller, certain outcome in place of an uncertain one. If you bought a Yes contract at forty cents and it now trades at seventy cents, selling realises thirty cents of gain per contract without betting the position on the final result.
The catch is that an early exit depends on someone willing to take the other side at a fair price. In a thin market the gap between the buy and sell price widens, so exiting early can cost more than you expect, and a confident looking position can be hard to leave near the close. That is why liquidity and the spread matter as much at the exit as at the entry. Our guides to understanding the spread and liquidity and why it matters explain how to read those costs before you commit.
Before you put money on any market, find four things in its terms. First, the named source of truth, the exact release or authority that decides the result. Second, the cutoff, the date and time the question is measured against and when trading closes. Third, the mapping, how the source value becomes a yes or a no, including any threshold and whether it reads as at least or strictly above. Fourth, the messy case rule, what happens on a revision, a delay, a cancellation, or a tie. If any of the four is missing or unclear, treat that gap as a real risk and size the position accordingly, or pass.
Settlement is also where platform risk becomes concrete, because your payout depends on the venue crediting it and letting you withdraw. A correct call on the event still depends on the venue paying out, which is why the venue and its rules matter as much as your read on the question. For the wider picture see our guides to counterparty and exchange risk and how resolution disputes work.
Understanding how these markets work does not make trading safe. Prediction markets can lose you money, and a confident price can still be wrong. Stake only what you can afford to lose, never to chase a loss, and never on borrowed money. If it stops feeling like a free choice, step back. In the United States you can call or text the helpline on 1-800-GAMBLER or visit ncpgambling.org.
It settles once the named source confirms the outcome, which can be minutes or hours after the result is known, and sometimes longer if the venue is waiting on official or revised data. Trading closes first, then settlement follows when the rule can be applied.
The resolution criteria written into the contract terms decide it. Those terms name the source of truth and how its data maps to a yes or no answer. On exchanges registered with the Commodity Futures Trading Commission, the terms are filed with the regulator before the contract trades.
Generally no, once a market is finalized. Before finalization, revisions, delays, or disputes can affect resolution, which is why the rulebook spells out which value governs and how challenges are handled. Always read those rules for the specific market.
The contract terms decide. A market may be voided with stakes returned, settled on a fallback source, or held until the question can be answered. If you cannot find a clear rule for an unusual case, treat that uncertainty as a real risk.
Often you pay trading fees, and some venues charge settlement fees, so your net return is less than the face value of a winning contract. Check the fee schedule for your platform, since structures differ and we do not publish a single figure that would go out of date.
No. Centralized exchanges apply a published rule through an internal team, while decentralized venues use an oracle and a dispute window. The principle of tying payout to a defined source is shared, but the mechanics vary, so confirm how your venue resolves.
Reviewed by Fredrik Filipsson, Editor, on 9 October 2025. Platform mechanics checked against Kalshi's published rules and help center, Polymarket's resolution documentation, and reputable reporting, all as of October 2025.
The rules change fast. Get the changes that affect you, plain and current, not tips.
Independent. Every claim dated and sourced. No platform pays for its place.