Prediction market

A prediction market is a financial or information market in which participants trade contracts whose values depend on the outcome of future events. These events may concern elections, economic indicators, financial markets, sporting competitions, geopolitical developments, technological milestones and other uncertain outcomes.

Prediction markets aggregate participants' expectations through market prices. In a binary prediction market, a contract typically pays a predetermined amount if a specified event occurs and nothing if it does not. The market price can therefore be interpreted as an implied probability of the event, subject to factors such as liquidity, transaction costs, fees, risk preferences and market design.

Prediction markets may operate through centralized platforms, exchanges or decentralized systems based on blockchain technology. Their legal status varies across jurisdictions, particularly when contracts resemble financial derivatives, gambling products or event-based wagering.

Structure and operation

Prediction markets allow participants to buy and sell contracts linked to clearly defined future outcomes. Each contract specifies the event, the conditions for resolution, the settlement rules and the payout associated with each possible result.

For example, a binary contract might pay 1 monetary unit if an event occurs and 0 if it does not. If the contract trades at 0.65 monetary units, its price may be interpreted as a market-implied probability of approximately 65%, assuming the contract's payout and settlement conditions are understood.

A simplified representation is:

$M a r k e t P r i c e \approx I m p l i e d P r o b a b i l i t y \times P a y o u t$

The relationship is approximate because market prices can also reflect risk premiums, trading frictions, liquidity constraints and differences between participants' beliefs.

Types of prediction markets

Prediction markets can be classified according to the events they cover, their contract structures and their trading mechanisms.

Binary markets

Binary prediction markets offer contracts based on two possible outcomes, commonly represented as Yes and No. A contract pays a predetermined amount if its specified condition is satisfied and a different amount, often zero, otherwise.

If a contract pays 1 unit when an event occurs and trades at price $p$, its implied probability can be approximated by:

$P (E) \approx p$

This interpretation assumes a unit payout, no additional settlement costs and market conditions under which the price reasonably reflects the expected payoff.

Categorical markets

Categorical prediction markets offer contracts linked to several mutually exclusive outcomes, such as the winner of an election or the destination of an economic indicator within specified ranges.

When contracts pay one unit for the correct outcome, their prices may be interpreted as implied probabilities for the respective outcomes. Under idealized conditions, the probabilities across all exhaustive outcomes sum to one.

$\sum_{i = 1}^{n} P (E_{i}) = 1$

Scalar markets

Scalar prediction markets concern numerical outcomes, such as inflation rates, economic growth, commodity prices or the date of a future event. Their contracts can be designed to reflect different values or ranges of a continuous variable.

Conditional markets

Conditional prediction markets estimate the likelihood of an event under a specified condition. For example, a market might estimate the probability of a policy being adopted conditional on a particular party winning an election.

Such markets can help distinguish expectations about an event from expectations about the circumstances associated with it.

Applications

Prediction markets have been used in several fields to collect expectations and improve forecasting.

Politics and elections

Political prediction markets allow participants to trade contracts associated with election results, candidate nominations, legislative outcomes and other political events.

Market prices can provide frequently updated estimates of electoral outcomes. However, they are not equivalent to opinion polls, which measure responses from sampled populations, and can be affected by trading volume, participant composition and market structure.

Economics and finance

Prediction markets can be used to forecast economic indicators, central bank decisions, inflation, interest rates and other economic events. They can complement conventional forecasting methods by incorporating information from participants with different sources of knowledge.

Their use in finance overlaps with certain forms of event-based derivatives, although the legal classification of a contract depends on its structure and jurisdiction.

Business and management

Organizations can operate internal prediction markets to estimate project completion dates, product demand, sales targets, supply-chain disruptions and the outcomes of strategic initiatives.

Employees trade using virtual or real incentives, depending on the design of the market. The resulting prices provide a way to aggregate information dispersed across departments and organizational levels.

Science and technology

Prediction markets have been studied as tools for estimating the reproducibility of scientific findings, the success of research projects, technology adoption and the achievement of specified technical milestones.

These applications can help researchers identify areas of uncertainty and compare collective forecasts with expert assessments.

Sports and entertainment

Markets can be organized around sporting results, tournament winners, awards and entertainment outcomes. Depending on the jurisdiction and platform, these products may be classified as prediction markets, betting products or regulated financial contracts.

Market mechanisms

Prediction markets use different mechanisms to determine contract prices and facilitate trading.

Order-book markets

In an order-book market, participants submit buy and sell orders at specified prices. The trading system matches compatible orders according to its rules.

Prices change as participants revise their expectations or respond to new information.

Automated market makers

Some prediction markets use automated market makers (AMMs), which apply predefined mathematical rules to facilitate trading without requiring a direct match between every buyer and seller.

An automated market maker may adjust contract prices in response to the quantity and direction of trades. The mechanism's design affects liquidity, transaction costs and the relationship between prices and implied probabilities.

Contract resolution

Every prediction market requires rules for determining the final outcome of a contract. These rules specify the relevant event, authoritative information sources, deadlines and procedures for resolving ambiguous or disputed outcomes.

A simplified process is:

$E v e n t D e f i n i t i o n \to T r a d i n g \to O u t c o m e V e r i f i c a t i o n \to S e t t l e m e n t$

Clear resolution rules are important because disputes about the meaning of an event can undermine confidence in the market.

Information aggregation

Prediction markets are designed to aggregate information held by participants with different knowledge, analytical methods and expectations. Trading provides incentives for participants to act on their beliefs, while market prices summarize the resulting supply and demand for contracts.

A simplified representation is:

$D i s p e r s e d I n f o r m a t i o n \to T r a d i n g D e c i s i o n s \to M a r k e t P r i c e \to F o r e c a s t$

The accuracy of the resulting forecast depends on factors including participant expertise, market liquidity, access to information, incentives and the design of the contracts.

Accuracy and limitations

Prediction markets can produce useful forecasts, but their prices should not be treated as guaranteed outcomes or objective probabilities under all circumstances.

Important limitations include:

Liquidity constraints: Thinly traded contracts may not reflect the views of a broad range of informed participants.
Market manipulation: Participants with substantial resources may influence prices, particularly in markets with limited liquidity.
Information asymmetry: Some participants may possess better information than others, and access to reliable information may be uneven.
Participant bias: Traders may share systematic assumptions or political, social and financial biases.
Risk preferences: Prices can reflect risk aversion, funding constraints and incentives in addition to beliefs about an event.
Ambiguous outcomes: Poorly defined resolution criteria can create disputes about contract settlement.
Regulatory uncertainty: Different jurisdictions may classify comparable contracts differently.
Rare events: Markets may struggle to estimate events for which historical data and informed participation are limited.

Prediction-market forecasts are therefore best considered alongside other evidence, including statistical models, expert judgment, opinion polls and historical data.

Regulation

The legal treatment of prediction markets varies across jurisdictions. Depending on the contract design and applicable law, a platform may be subject to rules governing derivatives, commodities, securities, gambling, consumer protection, anti-money-laundering requirements or financial intermediaries.

Some jurisdictions permit specified event contracts under financial-market regulation, while others restrict or prohibit particular forms of event-based trading.

Decentralized prediction markets introduce additional questions concerning jurisdiction, contractual enforceability, consumer protection, the identity of operators and the reliability of mechanisms used to verify outcomes.