Estimating Loss Given Default

Overview

The loss given default is a measure of how much value is lost when a borrower defaults on her loan.

Whenver a default occurs, the bank should then keep a record of the loan payments that are made after the default has occurred. Note, default is a subjective judgment. It does not mean that the borrower can no longer make any payments, just that they are unlikely to be able to fulfill the obligations of the loan. The borrower may be able to make partial payments, and the loan may have some collateral which can be sold to cover the remaining value of the loan. All payments, including the value recovered from collateral should be recorded along with the date of receival of payment.

Then the loss given default for a given default is the value of the loan at the time of default minus the present value (at the time of default) of the recovered value (from payments and collateral)

Calculating LGD

For each loan that has defaulted in the dataset, subtract the present value of the recoveries from the current value of the load at default. This is the LGD for that loan.

Next, calculate an average value for the LGD. In addition, you can calculate a standard deviation. If you are using a simplified risk model such as BAsel you only need the mean, however, you can assume a distribution for the LGD (probably a beta distribution) which can be fit wiht the mean and standard deviation.