Imperfect Information, Lending Standards and Capital Requirements

Abstract

We propose a quantitative model of lending standards with two reasons for inefficient credit: lenders' moral hazard from deposit insurance or government guarantees, and imperfect information about the persistence of asset price growth, which generates incorrect but rational beliefs in the lenders. We calibrate the model to match recent credit boom-bust episodes. Then we study which patterns of real estate price growth and banks' beliefs could serve as early warning indicators of a crisis. Finally, we propose a Value-at-Risk (VaR) rule to implement the capital requirements. The VaR framework ensures that the probability of banks not having enough equity to cover their losses is maintained at a certain level. Capital requirements should be state-contingent and lean against lenders' beliefs by tightening after periods of asset price growth. However, the relationship between asset price growth and financial risk is not monotone and this should be integrated in the setting of the capital requirements and early warning indicators.