Market power in electronic commerce: Three essays

Subhajyoti Bandyopadhyay, Purdue University

Abstract

The advent of the Internet and electronic commerce is expected to bring about several changes in the ways that organizations conduct their business. In the three essays that follow, this thesis looks into some manifestations of market power in electronic commerce. There are two broad areas where organizations wield their market power: in their ability to extract surplus from their interactions with other firms (often called the business-to-business transactions) and in their ability to extract higher profits from consumers vis-à-vis other firms (the business-to-consumer transactions). The first essay looks into the nature of increased competition between sellers in a reverse auction environment that is made possible by online business-to-business exchanges. Given the nature of the demand for industrial products, and the capacity constraints of the sellers, the results show that the sellers resort to a price randomizing equilibrium strategy in responding to reverse auction bids. The second essay builds upon the theoretical construct of the first, and simulates managers in real life who might not resort to advanced game theoretic calculations to decide on their ideal strategy. The behavior of two similar competitors is simulated through artificial software agents who start off by randomizing their bids, but gradually analyze their performance over time to place more weights in selecting those prices that seem to yield higher payoffs. If this learning process is considered to be the analogue of the process by which managers analyze their past actions, it becomes easy to understand how a mixed strategy equilibrium can develop as an emerging behavior without the players actually resorting to game theoretic calculations. The third essay develops some suitable metrics for the measurement of market power in the online retailing business. The electronic nature of an online transaction makes available the “clickstream” data, in the form of individual and aggregate web statistics. We explain a strategic conduct of an online retailer in terms of its manifestation in the clickstream data. Our results paint an accurate picture of the early online retailing industry, and show that the richness of clickstream data can be used to advantage in empirical research in electronic commerce.

Degree

Ph.D.

Advisors

Chaturvedi, Purdue University.

Subject Area

Management

COinS