Prediction markets allow users to trade on the outcomes of various events. These platforms began scaling across the United States last year and now track a broad spectrum of events—from geopolitical developments to winners of entertainment awards. But what exactly are prediction markets?

As an economics researcher with long-standing expertise in market mechanisms and incentive structures, my answer is simple: prediction markets are fundamentally just ordinary markets. Markets are foundational tools for resource allocation, directing goods and services to those who need them most. In this process, markets also exhibit information aggregation capabilities: the clearing of supply and demand integrates all participants’ private information into price signals.

Prediction market platforms and their associated instruments directly leverage this information aggregation power to forecast the likely outcomes of future events. Platforms introduce specific assets tied to particular events, where holders receive payouts if the predefined outcome occurs. Users trade these assets based on their own assessments of event probabilities. For years, many organizations have used prediction markets to uncover tacit knowledge held by employees—assessing whether key products will launch on schedule. Researchers have also employed these tools to evaluate the reproducibility of experimental results. Today, numerous media outlets collaborate with prediction markets, harnessing collective intelligence to supplement frontline reporting and traditional journalism, enriching content from multiple dimensions.

Prediction markets aggregate individual forecasts about the future from all participants and synthesize them into a trading market that quantifies the probability of various events. Betting on event outcomes on such markets follows the same logic as predicting stock prices on equity markets or trading oil prices in commodities markets. The difference lies in the underlying asset: while oil prices are influenced by multiple complex factors, prediction market contracts only generate payouts when a specified event occurs.

When oil prices rise, we can infer that demand exceeds supply—but we may not know why: Is it due to concerns over escalating tensions in the Middle East, or the emergence of new applications for oil? Prediction markets enable the creation of separate tradable contracts for each distinct possibility, allowing for precise disaggregation of predictions. For example, a market could be created for “whether the Strait of Hormuz will remain open at a specified time.” The contract’s payout rule might stipulate: $1 per contract if the event occurs. As users continuously trade, the market price becomes a probabilistic indicator reflecting the collective judgment of all traders on the likelihood of the event.

The mechanism works as follows: suppose the current price of each contract is $0.50, indicating the market assesses a 50% chance of the event occurring. If you believe the probability of passage exceeds 50%, say reaching 67%, you can buy the contract. If your assessment proves correct, the contract you bought at $0.50 yields a $0.67 return. This purchase pushes up the market price and the implied probability, signaling that other traders believe the event was previously underestimated. Conversely, if someone believes the current price is too high, they may sell short or sell outright, driving down the market’s estimated probability.

Compared to other forecasting methods, well-functioning prediction markets offer significant advantages. First, they directly output quantified probabilities—a core strength. Public opinion polls and surveys only capture the proportion of opinions; inferring event probabilities requires additional statistical analysis to link sample data to broader populations. Moreover, poll results are typically static snapshots at a given moment, whereas prediction markets continuously update as new participants enter and new information emerges.

Even more crucially, prediction markets come equipped with intrinsic incentive and constraint mechanisms. Both buyers and sellers risk real capital, and incorrect judgments result in financial loss. This compels participants to rigorously assess their available information and focus on areas where they possess genuine expertise. Conversely, the desire to profit from information and professional insight motivates individuals to conduct thorough research and uncover relevant evidence. A notable example occurred ahead of the 2024 U.S. election, when prediction market participants employed unconventional survey techniques to gather information inaccessible to traditional polling firms.

Finally, prediction markets boast remarkably broad coverage. In theory, traders with insights into the oil industry can express their views through long or short positions on crude oil contracts. Yet in practice, countless events lack reliable pre-forecasting mechanisms via mainstream commodity or equity markets—precisely where prediction markets excel. For instance, recent prediction markets have introduced contracts evaluating the performance of various AI models across different tasks. Such niche trends are nearly invisible in traditional commodity markets. Anyone can create and fund a prediction market to address such specialized questions.

Prediction markets are not new. Early prototypes date back to 16th-century Europe, where people used them to predict the next pope. Modern prediction markets integrate knowledge from economics, statistics, market design, and computer science. In the 1980s, Charles Plott and S. N. R. S. Shyam pioneered the formal academic framework for this system. Shortly thereafter, the world’s first modern prediction market—the Iowa Electronic Market—launched. Leveraging internet technology, this model aggregated fragmented information globally, enabling continuous growth and maturation.

However, unlocking the full potential of prediction markets still faces significant challenges. First, infrastructure-level issues remain unresolved: how to determine final event outcomes and achieve consensus; how to ensure market transparency and transaction traceability; and how to implement scalable dispute resolution mechanisms when contract payouts are contested or subject to manipulation.

Second, there are critical challenges in market design. First, participants must include those with access to core information. If all traders are uninformed, market prices become meaningless. Conversely, if informed actors avoid participation, forecasts will be biased. I proposed this idea as early as 2016: both the Brexit referendum and Donald Trump’s first presidential election were underestimated by prediction markets because participants failed to recognize the rising tide of populism.

Additionally, insider trading poses serious risks—especially when traders possess not only privileged information but also the ability to influence the event itself. Imagine if insiders attending a papal conclave placed bets on “the next pope” in a prediction market, using inside knowledge to front-run trades—or even manipulated the election outcome to benefit their own positions. Once participants perceive the market as rigged, they will exit en masse, leading to systemic collapse.

Another risk: malicious actors may deliberately manipulate prediction market prices to shape public perception of event probabilities. In this case, prediction markets transform from tools aggregating diverse viewpoints into instruments for influencing public opinion. For example, a political campaign could use campaign funds to artificially inflate the market probability of its candidate winning, creating a false impression of dominance. However, prediction markets possess self-correcting properties: when prices deviate significantly from reasonable levels, arbitrageurs step in to bet against the mispricing, thereby restoring equilibrium.

These challenges underscore the need for improved rules governing participant eligibility, contract design, and overall operational standards. Yet if industry practitioners can systematically address these issues, prediction markets will ultimately become indispensable tools for human foresight and navigating uncertainty.

By: Scott Duke Kominers, Research Partner at a16z crypto, Translated by: Chopper, Foresight News