What are PrediCTion MARTS?

Prediction markets are one of the most interesting intersections of finance, forecasting, game theory, information aggregation, and technology. They’re also becoming a major industry. Prediction marts are marketplaces where people buy and sell contracts whose value depends on whether a future event happens. They're often described as "betting markets for information," because prices aggregate the beliefs of many participants. 

For example, suppose there's a contract that pays $1 if a particular candidate wins an election and $0 otherwise.

The idea is that people who have good information, or are better at interpreting it, have a financial incentive to trade, making the market's price a useful forecast.

How they work

Imagine a market on the question:

"Will Company X release Product Y before December 31?"

A trader who believes the release is likely might buy contracts at $0.40.

If the market later rises to $0.70, the trader can either:

This is similar to trading stocks, except the asset represents the outcome of an event rather than ownership of a company.

Why people pay attention to them

Prediction markets have several strengths:

Research over several decades has found that well-designed prediction markets can be remarkably accurate for many measurable events.

Common topics

Markets exist for questions like:

Some organizations even run internal prediction markets to forecast project completion dates or product success.

Limitations

Prediction markets aren't perfect.

Some challenges include:

An example

Suppose the market asks:

"Will the Federal Reserve cut interest rates by December?"

Market Price                                             Implied Probability

$0.20                                                          ~20%

$0.50                                                          ~50%

          $0.85                                                          ~85%

If stronger economic data is released suggesting rates will stay high, traders may sell these contracts, causing the price to fall.

Are they actually accurate?

The evidence is generally positive, but with caveats.

Prediction markets tend to perform well when:

They can be less reliable for obscure questions with few traders or where legal restrictions limit participation.

A useful way to think about them is that they're real-time, financially incentivized forecasts rather than guarantees. A market trading at 80% doesn't mean an event will definitely happen—it means the collective judgment of participants is that there's roughly a four-in-five chance, given the information available at the time.