Creditor information registries and relationship lending

  • Jérémie Bertrand
  • , Paul Olivier Klein

Research output: Journal PublicationArticlepeer-review

7 Citations (Scopus)

Abstract

Banks rely on information to assess the creditworthiness of borrowers. They can secure this information in two ways: on the one hand, they can access public information on the firm from credit databases. On the other hand, they can build a relationship with the firm and secure private, more precise, information on the firm's prospects. In this paper, we investigate what happens to the collection of private information, when banks have access to a broader scope of public information. We argue that banks can either focus on public information and reduce their collection of private information (which is costly), or see the collection of private information as strategic, giving them an advantage when allocating credit. To settle this question, we employ an empirical approach and use a firm-level survey of 2292 firms in seven European countries in 2009. We find that when the coverage of credit registries increases, banks invest less in long-term relationship with their clients. Hence, when public information is widely available, banks do collect less private information on borrowers. This substitution only holds for firms with public information. We investigate if this reduction in the collection of private information alters firms’ access to credit, and we do not find a negative effect. The results are robust to alternative specifications and robustness tests.

Original languageEnglish
Article number105966
JournalInternational Review of Law and Economics
Volume65
DOIs
Publication statusPublished - Mar 2021
Externally publishedYes

Free Keywords

  • Creditor information
  • Information asymmetries
  • Relationship lending

ASJC Scopus subject areas

  • Finance
  • Economics and Econometrics
  • Law

Fingerprint

Dive into the research topics of 'Creditor information registries and relationship lending'. Together they form a unique fingerprint.

Cite this