Date of Award


Document Type


Degree Name

PhD in Business


Department of Management

First Advisor

Aaron L. Jackson

Second Advisor

Marcus Stewart

Third Advisor

Jeffrey Livingston

Fourth Advisor

Greg Reilly


Prediction markets, introduced roughly 30 years ago, are a way to leverage market pricing for information aggregation. Research has shown the markets provide accurate (often superior) forecasts. To date, research has been predominately of an economic nature. Despite numerous articles on the application of prediction markets in businesses (Hewlett-Packard, Google) the management literature has largely ignored the topic. This dissertation follows a three-paper model to begin to address the perceived gap.

Paper one introduces a new application of the market, namely its use as a determinant of employee compensation packages. Employers often claim to utilize a pay-for-performance model. However, principle-agent problems make it difficult for management to identify the varying performance of employees. In numerous settings, employees may have better information about the contributions of peers than management and may be better able to determine appropriate compensation.

Paper two informs the establishment of a corporate prediction market by drawing on employee motivation literature to identify incentive structures and operating processes that best attract and maintain adequate participation levels. We leverage intrinsic and extrinsic motivation research to examine methods of attracting, and perhaps more importantly, retaining sufficient participation levels. The paper concludes with a series of practical recommendations.

Finally, paper three shares results from a lab-based experiment which tested for the endowment effect within a prediction market. The prediction market run at Hewlett-Packard was set-up such that participants were endowed with some, but not all, securities. It was believed varying portfolios would encourage trade (Plott & Chen, 2002). Research on loss aversion suggests people are reluctant to adjust portfolio holdings they inherit. We leverage lab-based markets to address this perceived disconnect and the broader question of whether an anomaly in the marginal trader behavior would alter the market pricing and volume of trades.

Taken together, the three papers bring the phenomenon to the management literature through discussions of application and limitations. It is hoped that this work will expand the discussion beyond the technical aspects of the markets, and identify opportunities for management researchers and practitioners to leverage prediction markets.