Yuriy Nevmyvaka, Katia Sycara, Duane J. Seppi
This paper establishes an analytical foundation for electronic market making. Creating an automated securities dealer is a challenging task with important theoretical and practical implications. Our main interest is a normative automation of the market maker’s activities, as opposed to explanatory modeling of human traders, which was the primary concern of earlier work in this domain. We use a simple class of “non-predictive” trading strategies to highlight the fundamental issues. These strategies have a theoretical foundation behind them and serve as a showcase for the decisions to be addressed: depth of quote, quote positioning, timing of updates, inventory management, and others. We examine the impact of various parameters on the market maker’s performance. Although we conclude that such elementary strategies do not solve the problem completely, we are able to identify the areas that need to be addressed with more advanced tools. We hope that this paper can serve as a first step in rigorous examination of the dealer’s activities, and will be useful in disciplines outside of Finance, such as Agents, Robotics, and E-Commerce.
Published by cs.cmu.edu
on 10/12/2003Electronic Market Making: Initial Investigation