Event Contract Arbitrage: Test the Whole Payoff

Event Contract Arbitrage Rule Matching Comes First

Write the precise conditions for each leg before comparing prices. Include the underlying event, observation date, threshold, reporting source, publication version and treatment of ambiguous outcomes. If two contracts are supposed to offset, explain why one pays in every state where the other does not. Similar titles can help discover candidates, but they do not establish that relationship.

For a fictional example, Contract A pays Yes if the first published monthly index exceeds 7.0. Contract B pays Yes if the latest revised value available three weeks later exceeds 7.0. Buying Yes in A and No in B may appear to cover opposite sides of the same question. It does not. The contracts refer to different observations of the index, so the positions can both lose or both win.

Suppose the first value is 6.9 and the later revised value is 7.1. Yes in A loses because the first value did not exceed 7.0. No in B also loses because the revised value did exceed 7.0. Reverse those two observations, and both positions win. The pair has exposure to revisions rather than a constant combined payout. A price difference between the contracts may reflect that difference in meaning.

Current Polymarket resolution documentation directs attention to the market's rules and resolution source. The analytical task is to turn those specific terms into a payoff statement. The older Polymarket and Kalshi arbitrage bot discussion is relevant to cross-venue research, but no cross-venue pairing should proceed merely because a familiar event appears on both services.

State the Payout Assumptions in Ordinary Language

Consider a different fictional pair whose rules have been verified as compatible. Contract C's Yes pays one dollar if the specified event occurs and zero otherwise. Contract D's No pays one dollar if that same event does not occur and zero otherwise. Assume for this example that their exceptional-outcome provisions also preserve the intended complementary payout. One unit of each position then produces one dollar in either ordinary outcome.

Describe the two branches explicitly. If the event occurs, C Yes pays one dollar and D No pays zero. If the event does not occur, C Yes pays zero and D No pays one dollar. The combined ordinary payout is one dollar in both branches. This statement is the core of the candidate. It should remain understandable without referring to a chart or using the word arbitrage as a substitute for the reasoning.

Next identify states outside the ordinary two-branch model. What if the source does not publish? What if the event is cancelled? What if the venues use different rules for a dispute or an invalid market? The analysis must establish what each leg pays in those cases or mark the relationship unresolved. A constant payout in two familiar scenarios is insufficient if a plausible contractual exception produces a different result.

The broader event market trading guide helps distinguish binary questions, mutually exclusive outcome groups and merely related markets. These structures produce different payoff relationships. A collection of contracts under one event heading does not automatically form an exhaustive set, and a collection spanning several venues requires an even more explicit explanation of which states it covers.

Event Contract Arbitrage Costs Use the Whole Quantity

Suppose 1,000 units of C Yes can be purchased at $0.43 each and 1,000 units of D No at $0.54 each. Under the compatible-payoff assumptions, the purchase amounts are $430 and $540, totaling $970. The combined ordinary payout is $1,000. Before fees and other costs, the difference is $30. This is a hypothetical snapshot, not a claim that such prices are available in a real market.

Now assume the charges for the first leg total $6 and the second leg total $9. Total acquisition cost becomes $985, leaving $15 between that cost and the assumed $1,000 payout. Expressed against the $985 committed, the simple return would be about 1.52% if every assumption holds and the stated payout is received. That percentage says nothing about how long the capital remains committed or whether another opportunity can be repeated afterward.

The fees in this example are invented inputs. Actual fee schedules can depend on market, side, price, execution role and other terms. Use the current schedule for each leg, including the treatment of partial executions and any additional charges relevant to funding or redemption. Do not assume that one venue's method can be applied to another. Kalshi's fee schedule is an official starting point for its applicable charges, not a universal event-market formula.

Also verify that the quoted quantity exists at the assumed price. If only 100 units are available at $0.54 and the remaining 900 cost $0.56, the second leg costs $54 plus $504, or $558 before charges. With the first leg at $430 and the same hypothetical $15 combined fees, total cost becomes $1,003. The apparent $15 margin disappears and becomes a $3 shortfall under the simplified payout assumption.

Build an Event Contract Arbitrage Funding Plan

A combined profit calculation does not describe where money must be available. If the legs are on separate services, each account needs enough usable funds for its own purchase and charges before execution. Money expected from one leg after settlement cannot necessarily finance the other leg at entry. The funding plan should identify the required amount in each account, the form in which it must be held and the permissions needed to use it.

In the 1,000-pair example, the first account needs at least the modeled $436 acquisition amount and the second at least $549. A combined balance of $985 is not sufficient if almost all of it sits in the wrong account and cannot be moved before the quotes change. The correct constraint is local availability at each execution point, together with any additional reserve the operating process requires.

Pre-positioning funds can make execution more feasible, but it creates its own commitment. Capital waiting in an account is unavailable for other uses, and account or transfer conditions can change. Record this idle allocation when evaluating the strategy's practical return. A calculation that divides gains only by the brief amount held in completed pairs may ignore the larger pool needed to keep both accounts ready.

Account eligibility and venue context remain prerequisites. Do not assume that access to public prices grants permission to place either leg, or that similarly branded services share accounts and operating rules. The funding plan should reference the actual permitted account contexts. It should never depend on circumventing restrictions or treating a transfer path as proof of legal or contractual eligibility.

One Completed Leg Changes the Problem

A candidate pair is not an executed pair. If C Yes fills while D No becomes unavailable, the account owns a directional position in C. Its payout now depends on the event outcome. The original $15 projected result no longer describes the account because the combination on which that calculation depended was never completed. Record the unfinished state immediately rather than continuing to display the original theoretical margin.

Suppose 1,000 C Yes units fill at $0.43, while only 400 D No units fill at $0.54. Under the simplified compatible rules, 400 units are paired and 600 C Yes units remain unpaired. Before fees, the paired portion costs 400 times $0.97, or $388, against a $400 payout. The unpaired portion costs $258 and can pay either $600 or zero. Its outcome exposure is much larger than the small margin expected from a complete pair.

The response should follow a predefined limit on unpaired quantity, cost and duration. Possible actions include cancelling remaining orders, reducing the unpaired position at available bids or completing some quantity within a newly authorized cost ceiling. There is no universal rule that finishing the hedge at any price is best. The decision must compare the remaining exposures and costs using current evidence.

An event contract arbitrage bot can help implement these controls, but automation does not remove the incomplete-leg state. It should make that state easier to see and reconcile. The older prediction market arbitrage bot guide provides additional context for distinguishing a detected relationship from a fully executed and accounted-for position.

Treat Quote Timing as Part of the Evidence

An apparent discrepancy can result from comparing observations taken at different times. A quote from one venue may precede a public announcement, while the other quote reflects the announcement. The difference is then partly a timing artifact. Preserve source timestamps where available and local receipt times, and define how much temporal mismatch the analysis tolerates before it rejects the comparison.

Even synchronized observations do not guarantee synchronized execution. Network delays, order validation, market pauses and changing depth can intervene after the scan. The purpose of a timing rule is to keep the research record honest and bound avoidable staleness, not to certify that a pair will fill. The executor still needs price limits and a recovery policy for each leg.

Distinguish an order-book quote from a last trade or a displayed reference price. A last trade describes an earlier exchange between other participants. It may have been small or executed before available liquidity moved. A midpoint may not be directly tradable at all. The cost model should use the side of the book that the intended action would actually consume and the depth required for the planned quantity.

The older discussion of a Kalshi versus Polymarket price difference bot is useful for organizing this comparison. The central discipline is to preserve the observation type and time alongside the number. A scanner that reports only a percentage gap makes it difficult to distinguish a real executable candidate from a stale or semantically mismatched comparison.

Settlement Timing Can Separate the Cash Flows

Event contract arbitrage settlement risk includes more than the possibility of an unexpected final interpretation. Even when the intended payouts eventually occur, the two legs can settle at different times. One account may receive proceeds while the other remains unresolved. A position that is economically paired under the model can still create a liquidity problem if the operator expected both cash flows to become available together.

Return to the $985 acquisition example. Suppose the winning leg pays $1,000 after several days, while the losing leg's final record is confirmed later. The final combined result may still match the simplified $15 calculation, but the operator must reconcile both accounts and preserve enough records to establish what happened. If the expected winning leg is delayed instead, none of its projected payout should be treated as spendable merely because the event appears obvious.

Avoid converting a short observed holding period into a confident annualized return. Repetition depends on finding new compatible opportunities, maintaining funding and completing each pair. Settlement delays and idle capital can materially change the relationship between a single trade's percentage and the account's result over a longer period. Report elapsed time and committed capital directly before making any broader inference.

A useful settlement record contains the final rule-based outcome, the quantity held, the amount credited and any remaining unresolved discrepancy. It should connect to the original compatibility analysis. If the final result exposes an overlooked exception, the review must update the contract-pair registry before another candidate using the same assumption is permitted.

Passive Entry Changes Execution Risk

Some participants try to improve a candidate by posting a resting order on one leg. This can reduce the purchase price if another participant accepts the quote, but it changes the sequence of risk. The first leg may fill only when new information has moved the other leg's price. The scanner's earlier paired cost is no longer the relevant comparison at the moment the resting order executes.

A resting quote therefore needs an active link to the remaining-leg conditions. The operator can define a price or evidence change that triggers withdrawal, subject to the venue's actual cancellation behavior. While cancellation is pending, the order may still create exposure. The system should reserve for that possibility and should not assume the quote disappeared merely because a local strategy stopped displaying it.

An event contract market maker faces similar inventory questions when other participants choose when to trade. The lesson for a paired strategy is to evaluate the completed position distribution, not just the attractive limit price. Receiving a favorable-looking fill can be bad news if the complementary position has become sufficiently more expensive or unavailable.

This does not mean passive execution is inherently unsuitable. It means its value must be tested using fill timing, subsequent executable prices and the cost of unfinished cases. A research report that assumes every resting order fills at its quoted price while the other leg remains unchanged is assigning favorable execution to both sides without evidence that the combination was available.

Separate a Price Gap From a Capacity Limit

A candidate can satisfy the payoff test and still be too large for the operating process. Suppose the account can tolerate at most 100 unpaired units during execution, while the scanner recommends a 1,000-unit pair. The fact that the full pair has a small projected gain does not authorize a first leg that could leave 1,000 directional units. The executor must reduce the intended size, use a verified supported method that respects the limit, or decline the candidate.

Splitting the order into smaller stages can bound exposure, but it also changes the economics. Later stages may encounter different prices or fees, and each stage requires its own reconciliation. A successful first stage does not guarantee that the remaining quantity remains attractive. Recalculate the candidate using current depth instead of assuming the initial snapshot applies to the entire sequence.

This distinction belongs in the research record as well as the trading system. Report the theoretical quantity visible in the books separately from the quantity the account's controls permit. A scan that repeatedly finds large gaps beyond operational capacity may still provide useful market observations, but those observations should not be counted as captured opportunities. The account's feasible set is defined by both market conditions and its own bounded ability to manage unfinished work.

Keep a Candidate Ledger and an Incident Ledger

The candidate ledger records what the analysis saw: contract pair, rule version, quote times, depth, proposed size, fees and the calculated payout relationship. Include rejected candidates and the reason for rejection. This shows how often an apparent discrepancy survives semantic and execution checks. It also prevents a few completed favorable cases from standing in for the full workload of evaluating many unsuitable pairs.

The incident ledger records what the account experienced: submitted orders, accepted quantities, partial fills, cancellations, unknown responses and final reconciliation. Link every incident to the candidate that initiated it. If an order response is uncertain, preserve the attempt identifier and investigate it before creating another order. A retry with a new identity can expand exposure rather than repair the original request.

Review the ledgers together. A candidate may have been correctly priced but poorly executed. Another may have executed exactly as planned while relying on a mistaken contract equivalence. Those are different failures. Corrective work should target the relevant layer: compatibility review, fee estimation, quote synchronization, reservation policy or recovery logic. A generic label such as bad arbitrage conceals the information needed to improve the process.

The analysis is complete only when it can explain both the intended combined payout and the account's actual path to it. A small quoted gap is the beginning of an investigation. Compatible terms, executable quantities, local funding and reconciled settlement determine whether the proposed combination has the economics attributed to it. Maintaining those distinctions is more useful than treating the word arbitrage as a promise about the final result.