Building Credit Histories with Competing Lenders

This paper models credit histories as a way of aggregating information among various potential lenders, and is the first one to explicitly model how borrowers may affect this information aggregation through sequential borrowing. We analyze a dynamic economy with multiple competing lenders, who have heterogeneous private information about a consumer’s creditworthiness. We explore how this private information is aggregated through lending that take place over multiple stages. There are two key forces at play. On the one hand, acquiring a loan at an early stage serves as a positive signal—it allows the borrower to convey to other lenders the existence of a positively informed lender (advancing that early loan)—thereby convincing other lenders to extend further credit in future stages. On the other hand, because further lending dilutes existing loans (by increasing the consumer’s probability of default), the early lender takes this into account by charging a higher interest rate on the early loan, which makes the signaling costly. We demonstrate that despite dilution making early loans costly, borrowers may choose to take on small, early loans to signal their credit-worthiness to other lenders. We interpret this mechanism as building a credit history. We also show that information asymmetries can result in inefficiently large loans (relative to the symmetric information benchmark) extended in equilibrium. Our study leads us to examine features of consumer credit related to those consumers who hold multiple balances within a given loan category, e.g. multiple credit card balances. Very little is known in general about what types of consumers hold multiple balances or what credit terms, such as limits, these consumers face. Beyond providing basic documentation of these properties of consumer credit markets using data from TransUnion, we are more specifically interested in understanding how incumbent creditors adjust their credit terms when consumers initiate credit products with new lenders. One key aspect of the data we plan to exploit is the response of exiting lenders to an individual consumer opening of a new credit line. The dilution channel implies that existing credit cards should tighten credit limits attempting to limit the dilution from the new lender. On the other hand, the information aggregation channel implies the contrary—that incumbent lenders should extend their credit limits in response to new positive information, presumably available to the new lender. Since model parameters, in particular, average credit-worthiness, determine which channel dominates in our model, we are also interested in examining how the response of incumbent lenders varies with consumers’ prior credit histories.