What Are Prediction Markets? Their Economic Significance

What is a prediction market?

A prediction market is a market for contracts whose value is linked to future events. In English, these markets are commonly called prediction markets.

Participants trade contracts tied to outcomes such as whether a company's next-quarter revenue will exceed a threshold, whether a central bank will raise rates at its next meeting, or whether a product launch will be delayed.

A typical contract pays 1 if the event occurs and 0 if it does not. If that contract trades at 0.65, the market price can be read as an approximate 65 percent view of the event's likelihood.

In this sense, a prediction market is not only a place to bet on the future. It is a mechanism for converting dispersed information about the future into prices.

Polymarket and Kalshi

Recent attention to prediction markets has been driven partly by the growth of platforms such as Polymarket and Kalshi.

Polymarket is known as a prediction market platform covering politics, economics, crypto assets, technology, sports, culture, and other topics. Its own help materials describe it as a platform for trading outcomes of real-world events.

Kalshi is an event-contract exchange operated under U.S. CFTC regulation. The CFTC designated KalshiEX LLC as a Designated Contract Market in 2020, and Kalshi describes itself as a CFTC-regulated exchange for event contracts.

Polymarket has also been part of major regulatory developments. In 2022, the CFTC ordered Polymarket to pay a civil monetary penalty and wind down markets after finding that it had offered off-exchange event-based binary options. Later CFTC filing records list QCX LLC d/b/a Polymarket US as a Designated Contract Market with a July 9, 2025 filing date.

These examples show that prediction markets are no longer treated only as internet wagering or speculation. They increasingly sit at the intersection of financial regulation, market design, information aggregation, and data infrastructure. Legal treatment differs by country, so this article focuses on economic significance rather than legal availability.

Three layers of economic significance

The economic significance of prediction markets can be organized into three layers.

First, prediction markets turn uncertain future events into tradable securities. This is close to the idea of Arrow-Debreu securities in economic theory.

Second, prediction markets aggregate information dispersed across many people. This is often described as collective intelligence.

Third, prediction markets can operate as real-time media for information about the future. Market prices are not only transaction prices; they also communicate how likely society currently judges an event to be.

Taken together, prediction markets are not merely forecast games. They are economic institutions with characteristics of financial markets, information markets, and media.

Prediction markets as Arrow-Debreu-like securities

A useful foundation for understanding prediction markets is the concept of Arrow-Debreu securities.

An Arrow-Debreu security pays only when a specific state of the world occurs. For example, a security might pay 1 if next year's inflation rate exceeds 3 percent and 0 otherwise. It is a claim tied to a particular future state.

Kenneth Arrow analyzed the role of securities in allocating risk and showed that, if securities exist for future states, risk can be distributed more efficiently. From this perspective, prediction-market contracts can be understood as modern state-contingent claims.

In contemporary prediction markets, events such as elections, interest-rate moves, IPO timing, technological milestones, or geopolitical outcomes become tradable contracts.

This matters because uncertainty is usually easy to discuss but difficult to price, trade, hold, or hedge explicitly. Prediction markets convert future states into contracts, making parts of that uncertainty tradable.

The analogy has limits. A complete Arrow-Debreu market would contain securities for all possible future states. Real prediction markets are partial, with limited events, uneven liquidity, and imperfect participation. They should be understood as practical attempts to create state-contingent securities for selected real-world events, not as complete markets.

Prices are probabilities and risk prices

When a contract pays 1 if an event occurs and 0 otherwise, a price of 0.70 is often read as an approximate 70 percent probability.

Economically, that interpretation requires care. Market prices reflect not only beliefs, but also risk preferences, budget constraints, liquidity, transaction costs, hedging demand, and speculative demand.

Charles Manski argued that prediction-market prices should not be read too simply as average beliefs or objective probabilities. Strictly speaking, the price is closer to a state price: the current value of a payment in a particular future state.

With that view, prediction markets do more than answer the question, 'How likely is this event?' They also show the price at which the market is willing to bear the risk associated with that event.

This is why prediction markets differ from surveys. Surveys collect opinions. Prediction markets make participants hold opinions as risk. That is the source of their importance as financial-market mechanisms.

Prediction markets as collective intelligence

The second significance of prediction markets is collective intelligence.

Information in firms and societies is dispersed. Sales teams observe customer changes, developers see signs of project delay, policymakers understand regulatory probabilities, investors track market expectations, and researchers assess the pace of technical progress.

F. A. Hayek argued that knowledge in society is not centrally collected but dispersed across many people. Market prices compress that dispersed knowledge and make it available to others.

Prediction markets apply this price mechanism to future events. In ordinary markets, prices communicate scarcity. In prediction markets, prices communicate expectations about future outcomes.

Experimental and survey work by Plott and Sunder, Wolfers and Zitzewitz, and others shows why prediction markets can serve as information-aggregation mechanisms across politics, economics, corporate forecasting, and other domains.

The mechanism is not a simple majority vote. In a vote, uninformed and informed participants have the same weight. In a prediction market, participants who believe they have stronger information can trade more aggressively and attempt to move prices toward their view.

Prediction markets as media

The third significance is the media function of prediction markets.

Here, media does not only mean newspapers, television, or web publishers. More broadly, it means a mechanism that transmits information and updates people's understanding.

A market price for whether a candidate will win, whether a rate hike will occur, or whether a company will go public becomes news in itself. Unlike a published article, it can update continuously as new information arrives.

Traditional media reports what has happened. Prediction markets report, through price, what appears likely to happen before it occurs.

Media can be shaped by reader demand, editorial policy, advertising incentives, and political priors. Prediction markets have their own distortions, including thin liquidity, biased participation, and manipulation risk. But they also impose a different discipline: participants who trade on bad information can lose money.

That discipline is not perfect. Grossman and Stiglitz showed that perfectly informationally efficient markets are difficult even in theory because information is costly to acquire. Still, prediction markets can act as secondary information media that absorb news, expert analysis, statistics, and model forecasts, then convert them into probabilities and prices.

Meaning in the AI agent era

Prediction markets become even more important in the age of AI agents.

AI agents can read news, statistics, company disclosures, papers, social media, government materials, and internal data, then generate forecasts. In the future, AI agents may also become participants in prediction markets.

If AI can forecast, does that make prediction markets unnecessary? More likely, the opposite is true.

As AI systems multiply, many forecasts will be produced from different models, data sources, and reasoning methods. The central question becomes which forecasts to trust and how strongly to weight them.

Prediction markets offer an institutional answer. AI agents, human experts, operators, and statistical models can express their information through trades, and market prices can become a common measure that integrates those predictions.

Prediction markets can also help evaluate AI agents. An AI output can sound plausible as prose, but a forecast is ultimately judged by real outcomes. A well-calibrated forecasting agent can gain credibility; an agent that repeatedly produces poor forecasts faces losses or lower trust.

New risks would also emerge. If many AI agents rely on the same foundation models, news sources, datasets, or reasoning patterns, markets may amplify homogeneous errors instead of aggregating diverse information. Diversity of models and sources, human oversight, trading limits, auditable logs, and transparent outcome resolution will matter.

Summary

Prediction markets should be understood across three layers.

First, they convert future events into state-contingent contracts similar to Arrow-Debreu securities, making uncertainty partially tradable.

Second, they aggregate dispersed knowledge through prices, sometimes revealing information that meetings or surveys do not surface.

Third, they function as real-time media for the future, showing what market participants currently believe is likely to happen.

Polymarket and Kalshi make these roles visible in a contemporary form. In the AI agent era, prediction markets may become even more important as institutions for integrating, pricing, and auditing forecasts from humans, AI systems, models, and data.

A prediction market is not a mechanism for declaring the future. It is a mechanism for treating future uncertainty as price, probability, and risk.

References and sources

Regulatory and platform sources were checked on June 3, 2026. Academic sources are included for the economic and media-theory framing.