Essays in Corporate Prediction Markets

Loading...
Thumbnail Image

Date

2017-08-22

Authors

Karimi, Majid

Advisor

Dimitrov, Stanko

Journal Title

Journal ISSN

Volume Title

Publisher

University of Waterloo

Abstract

Personal subjective opinions are one of the most important assets in management. Prediction markets are mechanisms that can be deployed to elicit and aggregate a group of people’s opinions regarding the outcome of future events at any point in time. Prediction markets are exchange-traded markets where security values are tied to the outcome of future events. Prediction markets are systematically designed in a way that their market prices capture the crowd’s consensus about the probability of a future event. Corporations harness internal prediction markets for managerial decision making and business forecasting. Prediction markets are traditionally designed for large and diverse populations, two properties that are not often displayed in corporate settings. Therefore special considerations must be given to prediction markets used in corporations. Our first contribution in this thesis is in addressing the issue of diversity, in the sense of risk preferences, in corporate prediction markets. We study prediction markets in the presence of risk averse or risk seeking agents that have unknown risk preferences. We show that such agents’ behavior is not desirable for the purpose of information aggregation. We then characterize the agents’ behavior with respect to prediction market parameters and offer a systematic method to market organizers that fine tunes market parameters so at to best mitigate the impact of a pool agents’ risk-preferences. Our Second contribution in this thesis is in recommending prediction market mechanisms in different settings. There are many prediction market mechanisms with various advantages and weaknesses. The choice of a market mechanism can heavily affect the market accuracy and in turn, the success of a managerial decision, or a forecast based on prediction markets’ prices. Using trade data from two real-world prediction markets, we study the two main types of prediction markets mechanism and provide the much-needed insight as to what market mechanism to choose in various situations.

Description

Keywords

LC Keywords

Citation