What is a prediction market

Prediction Markets - Better predictions through the knowledge of many

Can't find an expert who can reliably predict the likelihood of success in launching your new product (or the outcome of the next election)? Then use the knowledge of many! With the help of so-called forecast markets, surprisingly good forecasts can be achieved in numerous situations.


Bundling the knowledge of many people

If the goal of the forecast can be clearly defined, but the underlying information and relationships are difficult to access, forecast markets can be a suitable tool for making the best possible statements about the future. Many situations are characterized by the fact that there is no expert who could assess the multitude of influencing factors and their effects alone, while many different people can very well assess partial aspects. Examples are economic developments (interest rates), political developments (election results), market and industry developments (success of product launches, industry growth, competitor behavior), technological trends (chances of success of a technology in the market) or company-related developments (sales trend of a product, timely implementation of a project) .

A forecast market tries to use the market mechanism to bundle the individual knowledge and experiences of people. The approach is based on the assumption that the information basis for decisions is better if it includes the information status of many people, and that markets are the most efficient form of aggregating individual information. A well-functioning securities market, for example, ensures that the prices of the securities traded reflect all information available about the underlying object. Instead of the development of a company as the basis for the value of a share, in forecast markets the parameter to be forecasted or the event to be forecasted as the basis for the value development of a security.

Application example: sales forecast

For example, if the goal is to forecast the sales of a certain product, then securities can be made available to the participants. A security could be called "70", with analog sales values. Each owner of the security "30-70" will later receive EUR 1 payment if the actual sales are in the range of EUR 30-70 million. The same applies to the other securities. If the turnover is outside the range, the respective securities are worthless.

The market participants - for example sales employees from different regions - can trade these securities over a certain period of time on an Internet-based platform. You will receive an initial supply of securities and cash. At the end of the game, the trading profit made is paid out to the participants. Since the game is played over a longer period of time, participants have the opportunity to study the topic in more detail and to obtain additional information. In addition, they can read the overall assessment of all market participants from the current security price and compare it with their own. A price of 0.35 euros for the security "> 70" means in this case that all market participants add a probability of 35% to the occurrence of a turnover of more than 70 million euros. If a market participant estimates the probability to be 50% , he can place a buy order for the corresponding security for EUR 50. Due to the incentive structure, it is to be expected that the participants will disclose their actual expectations and not give any politically motivated assessments, for example.

The design possibilities of a forecast market are numerous. When designing the securities, different payout mechanisms can be selected, the design of the trading venue, trading hours and playing time offer many options, and the incentive system can also be designed very differently. Careful design is important to ensure that markets function properly. This includes ensuring sufficient liquidity, suppressing attempts at manipulation and avoiding speculative bubbles. Ultimately, of course, the selection of market participants also has a decisive influence on the result.


Handelsblatt forecast exchange

Trade yourself on a forecast market and submit your estimates of gross domestic product, inflation rate and similar economic data.

Wolfers, Zitzewitz: Prediction Markets

The article provides a detailed overview of the application, benefits and the different types of forecast markets.

Author: Achim Sztuka