Although they’ve been around for the last 20 years or so, prediction markets have begun to make news for their application in business operations. Heralded early on in books like James Surowiecki’s The Wisdom of Crowds, prediction markets are a fascinating alternative to traditional forecasting methods, such Monte Carlo simulation, which extrapolate future events from past patterns. Essentially a betting exchange where participants stake something on the accuracy of the information they offer up, a prediction market is a way of capturing emerging patterns.
Prediction markets can be public or closed private exchanges, as in most business applications. Here’s how it might work: a business sets up an online portal to gather intelligence from its employees on such issues as scheduling or production costs. Each employee has a limited number of points to wager with the information he or she offers, and these points are value-at-risk, which means that an employee is likely to offer only information that is accurate enough to be worth the points.
Why bother to play at all? Darwinian competition. With each winning piece of information, the participant gains collective respect. Maybe he or she advances in rank on a leader board or maybe the company honors its top participants in a ceremony.
While the accuracy of prediction markets is still a topic of some fairly warm debate in applied mathematics, a number of risk analysis services are concentrating their solution portfolios on predictive markets.