Data-Driven Decision Making in Auction Markets

This dissertation consists of three essays that examine the promises of data-driven decision making in the design and operationalization of complex auction markets. In the first essay, we derive a structural econometric model to understand the effect of auction design parameters on sellers' revenues. In addition, we develop a dynamic optimization approach which makes use of the rich structural properties identified from empirical data to guide auctioneers in setting these parameters in real-time. In the second essay, we focus on bidding strategies across different market channels and examine the interactions between different strategies and auction design parameters. In the third essay, we investigate the effect of information revelation policy on price dynamics and market performance. This research offers important implications to both theory and practice of decision-making in information-rich and time-critical markets. From the theoretical perspective, this is, to our best knowledge, the first research that systematically examines the interplay of different informational and strategic factors in dynamic, multi-channel auction markets. In particular, it sheds light on real-time decision support in complex markets and thus contributes to the nascent literature on smart markets. From the managerial perspective, our research shows that advanced data analytics tools have great potential in facilitating decision-making in complex, real-world business environments.