University of Minnesota
University Relations

Minnesota Supercomputing Institute

Log out of MyMSI

Research Abstracts Online
January - December 2011

Main TOC ...... Next Abstract

University of Minnesota Twin Cities
College of Liberal Arts
Department of Economics

PI: Varadarajan V. Chari

Competing for Customers: A Search Model of the Unsecured Credit Market

The current dominant role of the credit card as a form of borrowing and the concurrent trends observed in the past raise a set of important questions. Should we think of credit cards as very different from other forms of unsecured borrowing, like personal loans? What are the quantitative implications of these differences for debt and bankruptcy? Is the new financial instrument conducive to consumers cutting their saving rates and relying more on unsecured borrowing, and hence also default? This project addresses these and other questions, by building a quantitative theory of the unsecured credit market that mimics the structure of the credit card market in the US. This framework departs from the competitive, one-period loan pricing assumptions of most models in three important dimensions, supported by the data. First, the researchers are going to model credit cards as lines of credit, i.e. they are going to give households an option to borrow up to a specified credit limit at a given interest rate. Second, credit cards are going to be long-lasting relationships, in which the terms of the contract are not readjusted for the duration of the relationship (commitment by banks). Third, the researchers endogenize the duration of the relationship by proposing a model of bank competition for customer by direct solicitation in a search and matching environment. They use the parameterized model to investigate the quantitative implications of our novel features for debt and bankruptcy related statistics. They find that the model matches the aggregate U.S. statistics for debt and bankruptcy well, and significantly better than an analogously parameterized standard setup with one period loans. The solution of the model with credit cards requires an evaluation of profit and value functions for all possible combinations of credit lines and interest rates. It is a very computationally intensive task. The researchers have been using MSI resources to solve the model in a parallel environment on around 100 processors.

Group Members

Lukasz Drozd, Department of Finance, Wharton School of Business, University of Pennsylvania, Philadelphia, Pennsylvania
Jaromir Nosal, Department of Economics, Columbia University, New York, New York