Relationship lending: A source of support or a means of exploitation?

Abstract

© 2020

Using a dataset from the State Bank of Pakistan containing each and every commercial loan generated in the economy from 2006 to 2013, we find that, on average, a longer relationship length is associated with lower risk premiums but higher collateral requirements. However, further examination paints a far more complex picture. The impact of relationship length on risk premiums and collateral varies substantially with the type of lender, as well as the type of borrower. We argue that conflicting empirical findings on relationship lending are the result of using datasets limited to certain types of borrowers or financial institutions.