Bank Lending Policy, Credit Scoring and Value at Risk

The authors use two equations in the probit model, the first models the banks decision to grant a loan, and the second the probability of a default. Through these equations, the paper confirms that:

• Banks provide loans in a way that is not consistent with default risk minimization;
• Loan size does not affect the default risk associated with the loan;
• The inconsistent lending behavior shown by the banks is either due to inefficient loan policy or some other type of optimizing behavior;
• Value at risk (VaR) is a more suitable measure to consider than a default risk for analyzing risks and returns on a portfolio of loans.

The paper further describes the recent research on statistical methods for credit scoring and the econometric bivariate probit model. A detailed analysis of this model illustrates that VaR, instead of default risks, can help optimize banks lending policy. The paper concludes by stating:

• Calculating VaR can enable financial institutions to evaluate alternative lending policies on the basis of their implied credit risks and loss rates;
• At an aggregate level, analysis of VaR reduces the risk of bankruptcy for financial institutions and financial disturbances to the economy.