The Impact of COVID-19 Pandemic on Real Customers’ Credit Scoring in tha banks of the Islamic Republic of Iran (Case study: Mellat Bank Branches in Yazd province)

The provision of facilities plays an important role in the banking industry. The credit risk is increased by customers' requests, and using a method to manage this risk is essential. During recent years, the customers’ credit scoring has been emphasized by most banks because of the increasing volume of deferred and due date claims. The importance of this issue is compounded by the prevalence of the Covid-19 virus, which affected the economies of all countries, especially the incomes of non-governmental businesses. In this paper, at first the impact of the prevalence of the virus on possibility of paying customers' installments is studied based on McNemar's Test. Then, the real customers’ credit scoring is performed for the Mellat Bank branches in Yazd province, and the effect of the Covid-19 virus is investigated on the significance of different variables, using the Logit regression. In this study, a random sample of 1200 real customers have been selected from the customers that have received facilities from the Mellat Bank branches in Yazd province from 1396 to 1399. The samples are divided into two equal parts; before and after the outbreak of the virus. The data on 16 selected variables are collected for both good and bad loan applicants. The results show that the prevalence of the virus has a significant effect on customers' default, so that it has increased sharply. Also, during the virus outbreak, the customer’s job type, marital status, and spouse's monthly income significantly impacted on the customers’ default with coefficient 3.3, -1.73 and 1.47 respectively, but these variables were not significant before the outbreak. Before the outbreak, the effective variables on customers’ default were the number of loan installments, collateral, credit history, and average account balance with coefficient 3.03, 2.82, -2.73 and 2.39, while the job type, monthly income and collateral were more effective after the outbreak with coefficient 3.3, 3.09 and 2.8.