Anomalies in Probability Estimates for Event Forecasting on Prediction Markets

Ho Cheung Brian Lee, Jan Stallaert, Ming Fan

Research output: Contribution to journalArticlepeer-review

2 Scopus citations


Innovative forecasting methods using new data sources have been developed to address various problems in operations management, such as demand, sales, and event forecasts. One of the methods for forecasting events consists of prediction markets where participants can take financial positions that may generate returns depending on whether certain events occur or not. Results in experimental psychology and behavioral economics have shown that individuals, including experts, can be subject to judgment bias when making probability estimates for future events. We examine, in this study, whether prediction markets are immune to such bias in estimating event probability. We find that even when there are large numbers of transactions and high volumes of trades, probabilistic fallacies still occur. Moreover, when they occur, they tend to be persistent over a certain period of time, and they tend to happen in situations similar to the ones where individual probabilistic fallacies are reported to occur. Our results have implications for the design of prediction markets and at the same time call for caution when using forecasts generated this way.

Original languageEnglish (US)
Pages (from-to)2077-2095
Number of pages19
JournalProduction and Operations Management
Issue number9
StatePublished - Sep 1 2020

All Science Journal Classification (ASJC) codes

  • Management Science and Operations Research
  • Industrial and Manufacturing Engineering
  • Management of Technology and Innovation


Dive into the research topics of 'Anomalies in Probability Estimates for Event Forecasting on Prediction Markets'. Together they form a unique fingerprint.

Cite this