Prediction Markets

Can my event be successful?

Which of some trends will be the most successful this year?

How will the market develop?

Prediction market is a research method based on the game theory. It is a competitive betting game, designed to exploit the ability of the collective intelligence to predict future events (the development of markets, of prototypes, the impact of communication campaigns, of packaging, format, the outcome of the next political elections, etc.). See for example ‘The Wisdom of Crowds’, Surowiecki (2004).

Prediction markets are exchange-traded markets in which participants are required to predict the development of markets, situations or future phenomena.

They use a betting mechanism: on one hand the participants are prompted to predict events as exactly as they can in order to win in the market, on the other the researcher benefits from this ability of theirs to obtain interesting predictions that are as correct as possible. The correctness of the predictions obtained through the use of Prediction Markets has been demonstrated by a number of experiments; see for example http://ubplj.org/index.php/jpm/article/view/981 or http://ubplj.org/index.php/jpm/article/view/478.

Functioning just like a real market, the price of each ‘share’ present on the market represents the point of balance of the value ascribed to the ‘share’ and is directly linked to the mechanism through which the participants are compensated: the higher the value of the shares in their portfolio at the end of the market, they more they win. Each ‘share’ represents an element that needs to be assessed and so its purchase implicitly represents a vote in its favour, whereas each sale represents a vote against.

For example: if the objective of the market is to understand which one from among a number of candidates will win the next elections (political prediction market), then by buying shares corresponding to candidate B at the price of 100 units per share, the participant is ‘betting’ on B inasmuch as he predicts that other people will make the same choice, thereby causing B’s price to rise. At the end of the market the participant who correctly predicts the behaviour of the others will have a portfolio value of more than 100 and will therefore have made a ‘profit’. The price that people are willing to pay becomes indicator of the consensus for a particular result of the situation: it aggregates information about the preferences, the opinions and the expectations concerning the market, the event or the future phenomenon. Generally speaking, at the end of the market the participants are rewarded in proportion to their ‘profit’.Having the same mechanism as the market, i.e. price-based, in the Prediction Market if lots of people buy a share, its price rises. If lots of people sell it or do not buy it, its price falls.

The fact that the participants have bought more of a certain share is therefore an indication that the prediction is a good one. Thus the price becomes indicator of the consensus for a particular result of the situation: it aggregates information about the preferences, the opinions and the expectations concerning the market, the event or the future phenomenon.

ADVANTAGES

  • Predicting future events such as the success of ideas / concepts / products / services / candidates, etc.
  • Improving predictions for sales and success.
  • Suggesting, during the course of the prediction market, any possible modifications for testing, in progress, the changes in opinion.
  • Collecting qualitative information about the reasons/dynamics that underpin the preferences by analysing the comments and the answers to specific questions asked during the market.