Dynamic Pricing and Learning in Prediction Markets /

Saved in:
Bibliographic Details
Author / Creator:Schultz, Adam, author.
Imprint:2017.
Ann Arbor : ProQuest Dissertations & Theses, 2017
Description:1 electronic resource (166 pages)
Language:English
Format: E-Resource Dissertations
Local Note:School code: 0330
URL for this record:http://pi.lib.uchicago.edu/1001/cat/bib/11715128
Hidden Bibliographic Details
Other authors / contributors:University of Chicago. degree granting institution.
ISBN:9780355079319
Notes:Advisors: John R. Birge; N. Bora Keskin Committee members: Baris Ata; Rene Caldentey; Devin Pope.
Dissertation Abstracts International, Volume: 78-12(E), Section: B.
English
Summary:In this dissertation, we explore the nature of dynamic pricing, information aggregation, and bias in prediction markets. We begin with Chapter 1, in which we develop a dynamic control model to analyze how a monopolistic market maker can optimally set prices in a prediction market while learning information about the event outcome. We demonstrate the market maker's optimal policy when facing myopic agents, and prove that a myopic (greedy) policy performs relatively well in this context. We also introduce a setting where a sophisticated agent (i.e., insider trader) can exploit the market maker. We characterize the amount of harm imposed on the market maker by the presence of this strategic agent, and propose a policy the market maker can adapt to mitigate the presence of the strategic agent.
In Chapter 2, we explore how market makers use pricing as a mechanism to aggregate information in a biased prediction market. We collect a novel data set of time series data to study a sports betting market, including Twitter data to control for breaking news events that lead to information changes in the market. After investigating how market makers adjust prices, we present an approach to estimate potential bias in bettors' beliefs about the game outcomes. This model allows us to perform a counterfactual analysis in which we characterize the optimal point spread for the market maker for each game. We use this model to assess market makers' expected profit performance. We demonstrate that market makers' pricing policies do not follow the oft-cited strategy of "balancing" the bet and analyze how market makers benefit from deviations from this policy.
In Chapter 3, we explore how biases evolve over time in prediction markets. In particular, we study NBA point spread betting markets and futures odds betting markets for the NCAA tournament. We conclude that biases appear to persist in these markets over time, and we explore potential reasons for this market behavior.

Similar Items