The advantage of expected loss lies in tail risk measurement and sub-additivity.

Expected Loss (EL) represents the average loss level, which is a risk cost and thus a part of the total cost. For example, the amount of losses that can be expected by commercial banks in the normal course of business operations, banks usually set up loan loss reserves to buffer these losses.

According to the provisions of the Basel Accord, the three parameters can be evaluated by applying the internal rating method, and the expected loss can be obtained using the formula EL=EAD×PD×LGD. Financial institutions mostly use big data modeling technology to predict user default probability PD and debt loss rate LGD, thereby obtaining the corresponding risk cost.

There are three main parameters in the calculation of expected loss

1. Exposure At Default (EAD): Risk exposure refers to the credit risk that may be incurred due to customer default Business balance.

2. Probability of default (PD): refers to the possibility of a customer defaulting in a certain period of time in the future. It is generally defined as the probability of becoming a non-performing asset. In credit business, that is, the customer eventually becomes a non-performing asset. The ratio of overdue status to M3+.

3. Loss Given Default (LGD): refers to the ratio of the default loss of the debt to the default risk exposure of the debt after the customer defaults.