Prediction Markets & Event Contracts

Table of Contents

    Prediction markets turn questions about future events into tradable contracts. A price of 60 cents for a contract that pays $1 if an event occurs is often described as a 60% probability. That shorthand can be useful, but a reader needs to know what settles the contract, which price is executable, and what costs and market conditions sit behind the quote.

    Economic event contracts make those questions especially concrete. A contract can settle on a published inflation number, a payroll-employment change or a Federal Reserve decision. This guide explains their mechanics and limitations, with hypothetical examples rather than live quotes or venue recommendations. U.S. regulatory context is dated November 14, 2026.

    What Are Prediction Markets?

    A prediction market lets participants trade claims linked to a defined outcome. A common binary event contract pays a fixed amount to the winning side and zero to the losing side at normal settlement. Many use a $1 payout and Yes/No positions, although more complex contracts can have different payout structures. [1]

    The Contract Defines the Event

    “Inflation will be high” is an opinion. “The specified month’s seasonally adjusted CPI-U increase, reported to one decimal place by the named source at the specified time, will be strictly above 0.3%” is a settlement condition. The precise condition determines which side gets paid.

    Traders can buy because they expect an outcome, seek protection against an economic exposure, or expect to sell at a different price before settlement. Those motives affect demand. The price aggregates the trading decisions of participants able and willing to commit capital; it is not a survey giving every person’s opinion equal weight.

    The fixed payoff resembles a cash-or-nothing binary option. The distinguishing task here is evaluating an event’s definition and a market forecast. A stock-option pricing model does not automatically supply a reliable probability for a nontradable economic announcement.

    Contract Price, Payout and Profit

    Separate the purchase price from the settlement payout and the resulting profit. For a fully paid long Yes position, let N be the number of contracts, P the price paid per contract, F the payout per winning contract, and K the total costs of entering and holding to resolution.

    Normal Binary Settlement Profit
    Yes outcome: N × (F − P) − K
    No outcome: −N × P − K
    Assumes no leverage, no interim sale and a valid two-outcome settlement. Other account or contingency terms require separate analysis.

    Buying No expresses the opposite side of the same defined event. In an idealized, frictionless pair, a Yes claim and its exact complementary No claim together deliver F, so their combined value is the present value of F. Actual displayed buying prices can include spreads and costs and need not add to exactly $1.

    A 60-cent economic contract example

    Suppose a hypothetical contract pays $1 if a specified CPI release exceeds its threshold. The current Yes bid is $0.58 and ask is $0.60. You buy 100 contracts at the ask and pay an assumed $2 total fee. Hold to normal resolution, with no further charges, financing cost or interest income. These are invented teaching inputs, not a venue’s fee schedule.

    A $100 Winning Payout Is Not $100 of Profit

    Contract purchase cost = 100 × $0.60 = $60. Including the assumed fee, the outlay is $60 + $2 = $62.

    If Yes wins, profit = $100 − $60 − $2 = $38.

    If No wins, profit = $0 − $60 − $2 = −$62.

    If your own estimated probability of Yes is 65%, expected profit under those assumptions is 0.65 × $38 + 0.35 × (−$62) = $3. That positive expectation depends on your estimate being defensible; either realized outcome remains possible.

    The break-even probability is $62 / $100 = 62%, rather than the 60% suggested by the purchase price alone. In general, with fixed outcome-independent costs and no time-value adjustment:

    Break-Even Probability for a Fully Paid Long
    qbreak-even = (N × P + K) / (N × F)
    This is a cost threshold for the position, not evidence that the event has that probability.

    Before settlement, selling requires a counterparty. In the same unchanged order book, buying at $0.60 and immediately selling at $0.58 loses 100 × ($0.60 − $0.58) = $2 before fees. A limit order controls an acceptable price if filled, but does not guarantee execution. These are the same basic trading execution issues that matter in other markets.

    When Can Prices Be Read as Probabilities?

    For a short-dated $1 binary contract, dividing a representative price by $1 gives a convenient probability-like market estimate. Calling it an unbiased real-world probability requires more. Wolfers and Zitzewitz show that prices can correspond to average beliefs under particular assumptions; wealth and risk preferences can alter that relationship. Average beliefs can themselves be wrong. [2]

    • Spread and quote choice: the last trade, midpoint and executable ask answer different questions. A stale trade can misrepresent current conditions.
    • Fees: the relevant threshold depends on actual order size, execution and charges. For example, Kalshi’s published schedule distinguishes trading and maker fees and allows series-specific treatment; intermediary charges can differ. [3]
    • Time value: a dollar paid later is worth less than a dollar available now when rates are positive. A valuation model discounts the payout before mapping price to a pricing probability.
    • Risk and hedging demand: traders may value protection against a bad economic outcome differently from an otherwise identical speculative bet.
    • Participation and constraints: limited capital, position restrictions and thin liquidity can keep informed traders from fully correcting a questionable price.

    A pricing probability used in a financial valuation model need not equal the actual probability of an event. For events without a replicating traded asset, there may also be no unique model-free risk-neutral probability to recover. Treat the quote as evidence to assess alongside other information, rather than as a mechanically proven chance.

    The broader efficient market hypothesis helps frame information aggregation, but does not certify a particular prediction market. Evaluate a forecasting record across comparable contracts and time horizons. One surprise result does not establish miscalibration: events assigned a low probability can occur.

    How Economic Releases Determine Settlement

    Economic contracts can provide a direct view of expectations about business-cycle indicators. Their usefulness starts with knowing which statistic is being forecast. “CPI,” “jobs” and “the Fed rate” each have several plausible meanings.

    Three economic questions that require different definitions
    Contract topic Definition to check Important distinction
    Inflation Headline or core; CPI-U or another index; monthly or annual change; seasonal adjustment and rounding. A 0.3% monthly headline CPI figure is a different observation from a 3.0% annual rate.
    Payroll employment Named month, total nonfarm payroll change, units, seasonal treatment and release vintage. Payroll jobs are different from the household-survey unemployment rate.
    Policy decision Named FOMC meeting, target-range change or endpoint, action size and cancellation rules. An announced target range is different from the effective overnight rate traded in the market.

    Specify the official source and data vintage

    A data vintage is the version of a statistic available at a particular time. The latest historical database value may differ from the value that settled a contract.

    For a concrete example, Kalshi’s published CPI terms identify the signed, one-month, one-decimal change in seasonally adjusted CPI-U from the Bureau of Labor Statistics. They use a strictly greater-than criterion and exclude revisions made after expiration. That is more specific than simply saying the contract follows inflation. [4] BLS revises seasonally adjusted CPI data as seasonal factors are updated, so a later download may show a different history. [5]

    Kalshi’s published payroll terms similarly exclude revisions after expiration. BLS initially revises monthly payroll estimates in each of the following two months, with additional benchmark and seasonal revisions. A later revised jobs number therefore does not automatically change a completed settlement. Check the actual contract’s cutoff rather than assuming every market uses the first print. [6] [7]

    Read delays, cancellations and disputes before trading

    Trading close, expiration, official publication and payment can occur at different times. A delayed government release, source correction or disputed result may trigger a fallback or review process.

    Do not assume cancellation means a refund. The published FEDDECISION terms, for example, say that cancellation of the target FOMC meeting makes the No change strike resolve Yes and the other markets No. Published payroll terms contain a last-available-month fallback if data are not released by the specified expiration. These are examples of particular terms, not universal market rules. [8] [6]

    Identify who makes the final determination, whether and how a participant can challenge it, the review deadline, and what happens when the named source is unavailable. An appealing headline cannot replace those provisions.

    Reading a Ladder of Economic Thresholds

    A strike ladder lists related contracts at several thresholds. Suppose X is one specified month’s one-decimal CPI change. All three hypothetical contracts below use the same source, vintage, timing and $1 payout. Ignore fees, discounting and spreads, and interpret the prices as one coherent set of probability-like weights.

    Hypothetical CPI thresholds overlap
    Yes condition Price per $1 payout Probability-like weight
    X strictly above 0.2% $0.70 70%
    X strictly above 0.3% $0.40 40%
    X strictly above 0.4% $0.15 15%

    The higher-threshold event is contained within the lower-threshold event. A CPI result above 0.4% satisfies all three conditions. Adding 70% + 40% + 15% = 125% therefore double-counts outcomes; these contracts are not three exclusive alternatives.

    Subtract adjacent cumulative weights to infer a nonoverlapping interval:

    From Nested Thresholds to an Interval
    Pr(a < X ≤ b) = Pr(X > a) − Pr(X > b)
    For identical definitions of X and one coherent probability interpretation.
    Nonoverlapping weights inferred from the hypothetical ladder
    CPI outcome Calculation Weight
    X at or below 0.2% 100% − 70% 30%
    X above 0.2%, at or below 0.3% 70% − 40% 30%
    X above 0.3%, at or below 0.4% 40% − 15% 25%
    X above 0.4% 15% 15%
    Total 30% + 30% + 25% + 15% 100%

    At exactly 0.3%, only the above-0.2% contract wins. Small wording changes such as “at least,” different rounding or a different data cutoff can alter that result. In real markets, subtracting two asks does not establish an executable interval price: both legs’ trading directions, available bids/asks, quantities and fees matter.

    Event Contracts vs Fed Funds Futures and Options

    A policy-decision contract can ask whether the FOMC cuts its target range at a named meeting. CME 30-Day Federal Funds futures instead settle against 100 minus the month’s arithmetic average daily effective federal funds rate. A meeting partway through the month affects only part of that average. [9]

    Different contracts expose the holder to different outcomes
    Instrument Settlement driver Main interpretation limit
    Binary economic event contract Whether a precisely defined release or decision satisfies its condition. The fixed payout does not grow with the severity of the economic outcome.
    Fed funds futures Monthly average daily EFFR. Recovering meeting probabilities requires assumptions about possible rates and timing.
    Asset options around an announcement The underlying asset price and the option’s terms. The asset’s response also depends on expectations, other news and valuation.

    For example, a strangle strategy around an economic announcement expresses a view about asset-price movement. A CPI contract expresses a view about the defined statistic. Correctly forecasting the release does not guarantee correctly forecasting a stock or bond’s response.

    A simplified futures-implied probability example

    Consider a hypothetical 30-day month with EFFR at 4.00% for its first 15 days. Assume only two possibilities for the remaining 15 days: it stays at 4.00%, or a policy cut lowers it to 3.75%. Assume no other rate changes and ignore risk-premium and convexity adjustments.

    Day Weighting Changes the Calculation

    No-cut monthly average = 4.00%.

    Cut monthly average = (15 × 4.00% + 15 × 3.75%) / 30 = 3.875%.

    Suppose the futures index is 96.075, implying a monthly rate of 100 − 96.075 = 3.925%.

    The two-state implied cut weight is (4.00% − 3.925%) / (4.00% − 3.875%) = 0.60, or 60%.

    The 60% follows from the assumed two states and day weights. Additional possible decisions, an unscheduled move or a different relationship between the target range and EFFR would change the inference. The futures index itself is not an event probability.

    Venues and U.S. Regulatory Context

    Separate a platform’s technology and branding from its legal status. A CFTC-registered designated contract market (DCM), a limited research arrangement or simulated forecasting market, and an unregistered foreign platform are different settings. An on-chain settlement mechanism does not itself establish U.S. registration or customer protections.

    For U.S. regulated derivatives, distinguish exchange registration, product self-certification and prior Commission approval. Under 17 CFR 40.2, an exchange can submit its own certification that a product complies with the law. A listing therefore does not necessarily mean the Commission individually approved that contract. [10]

    Checked November 14, 2026: the eCFR text current through November 12 retains 17 CFR 40.11’s prohibition for specified excluded-commodity contracts involving, relating to or referencing terrorism, assassination, war, gaming or unlawful activity, plus provisions for similar activities and a possible 90-day review. How that framework applies to particular products has been contested. [11]

    • Proposals and adopted rules differ. The CFTC withdrew its 2024 event-contract proposal in February 2026 and proposed a replacement 40.11 framework in June. An October proposal would expressly include several event-contract categories in the swap definition. Those proposals must be distinguished from an October 19 effective interim final rule excluding specified licensed casino-style gambling products under stated structural conditions. [15] [16] [19] [20]
    • Court outcomes have differed. On April 6, the Third Circuit affirmed a preliminary injunction protecting Kalshi against New Jersey enforcement concerning sports contracts. On August 28, the Ninth Circuit affirmed dissolution of a preliminary injunction as to sports contracts in Nevada. These dated sports-contract decisions are not universal rulings about economic-release contracts. The Supreme Court docket for New Jersey’s certiorari petition remained pending when checked; a petition is not a decision to hear the case. [12] [13] [14]
    • Agency positions do not end litigation. A CFTC impartial-access interpretation effective November 10 rejects domestic geographic restrictions on otherwise eligible participants. On November 13, the agency announced a suit against Arkansas asserting federal jurisdiction. The interpretation and the agency’s litigation position do not establish every reader’s ability to trade every product. [17] [18]

    Verify the actual legal entity and registration, current product and account eligibility, applicable restrictions and current court orders. This is dated context, not a state-by-state availability opinion. The CFTC cautions that trading with unregistered entities outside the United States may provide little or no protection. [1]

    Limitations and Manipulation Risks

    Moving a quote and changing a settlement outcome are different risks. A large order can move a thin market’s displayed price without improving its forecast. Separately, someone able to influence the underlying event, settlement data or source selection may have an incentive to change which side wins.

    An official economic release provides a specified reference, but publication controls and access to nonpublic information still matter. In its March 2026 staff advisory, CFTC staff emphasized manipulation-resistant contract design, reliable settlement data and safeguards against premature release. This is staff guidance explaining the oversight framework, not a new standalone law. [21]

    Forecasting biases also require care. Research models can produce a favorite-longshot bias, where longshots are priced too high relative to their eventual success frequency, or the reverse under different assumptions. That possibility does not establish a universal correction to apply to today’s quotes. [2]

    A Binary Payoff Can Be an Imperfect Hedge

    A contract that pays the same $1 after either a small or a very large inflation surprise cannot automatically offset a loss that grows with inflation. A loan, business expense or investment portfolio may also respond to a different statistic or time period. This mismatch is basis risk.

    For example, a monthly CPI contract addresses one release. It does not cover every future increase in household expenses or the broader costs of inflation. Add liquidity, source disputes, operational failures and changing access to the analysis. A visible quote does not guarantee that a desired position can be entered, exited or settled on the expected timetable.

    How to Evaluate an Economic Event Contract

    1. Write the condition precisely. Record the statistic, reference month or meeting, units, seasonal adjustment, rounding and strict or inclusive boundaries.
    2. Save the resolution terms. Identify the official source, vintage/cutoff, trading close, expiration and payment timing, plus delay, cancellation and review provisions.
    3. Check access and protections. Verify the entity, registration, account restrictions and applicable current legal position rather than relying on an app’s label.
    4. Price the actual order. Examine bid/ask quantities, likely fills and every applicable fee. Calculate both normal settlement outcomes and the break-even probability.
    5. Challenge the forecast. Compare the same event definition across credible forecasts and related contracts. Investigate inconsistent ladders before treating differences as opportunities.
    6. Check the economic purpose and loss. Separate a forecast from a hedge, assess basis risk, and use only capital whose loss would not disrupt essential needs.

    Common Mistakes

    • Treating 60 cents as a guaranteed 60% objective probability. Quotes reflect trading, constraints and costs as well as beliefs.
    • Confusing the payout with profit. Subtract the purchase price and fees from a winning payout; a losing fully paid long can lose its entire outlay.
    • Using the latest revised data to judge an old settlement. Match the contract’s source and vintage.
    • Adding overlapping threshold probabilities. Convert nested events into disjoint intervals before adding them.
    • Assuming a delay or cancellation produces a refund. The contract may specify another outcome or fallback.
    • Equating a policy decision with a futures-implied monthly rate. Account for day weights and the distinction between the target range and EFFR.
    • Reading registration as a blanket legal or safety assurance. Product design, current orders, eligibility and operational protections still require examination.

    Frequently Asked Questions

    For a short-dated $1 binary payout, it is a useful probability-like shorthand. It is not automatically an unbiased real-world probability. Check which quote is shown, payout timing, fees, liquidity and the exact settlement condition. Trader beliefs and risk preferences can also differ from the actual likelihood.

    Many exchange-traded contracts permit closing transactions before trading ends. Execution still requires available liquidity at an acceptable price, and trading may be paused or access restricted. Selling before resolution realizes the difference between entry and exit prices, less costs, rather than the final binary payout.

    The contract’s vintage and cutoff control. A contract excluding post-expiration revisions is not automatically resettled because a later statistical release changes the historical number. Read the specific terms, including treatment of corrections before expiration and any outcome-review process.

    Threshold events overlap. A result above a high threshold also exceeds lower thresholds. Only mutually exclusive, collectively exhaustive outcomes should sum to 100% in a coherent probability distribution. Bid-ask spreads, fees and differing contract definitions can also complicate comparisons of actual prices.

    No. A binary policy contract can pay on a defined FOMC action at one meeting. Fed funds futures settle on the month’s average daily effective federal funds rate. Inferring meeting probabilities from futures therefore requires assumptions about timing, possible outcomes and the rates associated with those outcomes.

    Registration, product self-certification and prior product approval are separate concepts. Current litigation, contract-specific rules and account eligibility also matter. Check the current position for the actual product and location; the dated regulatory discussion here does not guarantee availability or legal status for an individual reader.

    Sources

    1. CFTC: Understanding Prediction Markets and Event Contracts. Mechanics and customer protections; agency descriptions of nationwide access require the litigation context above.
    2. Wolfers and Zitzewitz: Interpreting Prediction Market Prices as Probabilities, April 2006 working paper. Conditional price/belief mapping and bias.
    3. Kalshi: Fee Schedule, marked effective July 7, 2026. Actual charges vary; the worked fee is hypothetical.
    4. Kalshi: CPI Contract Terms. Statistic, threshold, expiration and contingencies.
    5. BLS: Using Seasonally Adjusted and Unadjusted CPI Data. Seasonal revisions.
    6. Kalshi: PAYROLLS Contract Terms. Release definition, revisions and fallback.
    7. BLS: CES Frequently Asked Questions, modified April 20, 2026. Monthly and benchmark revisions.
    8. Kalshi: FEDDECISION Contract Terms. Defined policy action and meeting cancellation.
    9. CME Group: Using Fed Fund Futures to Trade FOMC Decisions, February 2022 context. Monthly EFFR averaging mechanics.
    10. 17 CFR 40.2: Listing Products by Certification. eCFR current through November 12, 2026.
    11. 17 CFR 40.11: Review of Certain Event Contracts. eCFR current through November 12, 2026.
    12. Third Circuit: KalshiEX v. Flaherty, April 6, 2026. Preliminary-injunction appeal.
    13. Ninth Circuit: KalshiEX v. Assad, August 28, 2026. Sports-contract preliminary injunction and remand.
    14. Supreme Court: Docket 26-299. Pending certiorari petition when checked.
    15. CFTC: Withdrawal of Earlier Proposal and Sports Advisory, February 4, 2026.
    16. CFTC: Proposed Event-Contract Review Framework, June 10, 2026.
    17. CFTC: DCM Impartial Access Requirements, final interpretive guidance effective November 10, 2026, printed pages 72407–72410.
    18. CFTC: Arkansas Lawsuit Announcement, November 13, 2026. Agency litigation position.
    19. CFTC: Proposed Swap-Definition Inclusions, October 19, 2026.
    20. CFTC: Casino-Style Gambling Exclusion, interim final rule effective October 19, 2026.
    21. CFTC Staff Letter 26-08: Prediction Markets Advisory, March 12, 2026. Staff views on contract design and oversight.

    Sources accessed November 14, 2026. Provider documents illustrate specific terms and may change. Examples are hypothetical and exclude taxes; no quoted price is presented as a current market observation.

    Disclaimer

    This article is for educational and informational purposes only and does not constitute investment, legal or tax advice. Event contracts involve loss, liquidity and settlement risks. Rules and availability can change, and the discussion does not establish an individual reader’s eligibility. Consult appropriately qualified professionals about your circumstances.