Contagion and correlation in empirical models of bank credit risk in Israel

We apply a methodology for analyzing bank credit risk in Israel, which distinguishes between contagion and correlation on the one hand, and risk factors that are macroeconomic, sectoral and idiosyncratic on the other. Credit risk may be correlated because the observed and unobserved drivers of credit risk happen to be correlated, or because they are causally related through contagion. Bank credit risk is measured by the proportion of problem loans in credit sectors of Israel’s banking system. Contagion is malignant and infectious if credit risk in one sector increases credit risk in other sectors. Contagion is benign and immunizing if credit risk in one sector reduces credit risk in other sectors. In some sectors, such as construction, credit risk is highly contagious and malign. On the other hand, credit risk elsewhere immunizes credit risk in the construction sector through benign contagion. According to our results there are two aspects related to the construction sector. First, construction is greatly over-represented in bank credit risk. Second, credit risk in construction is highly contagious relative to other sectors. By contrast, our results suggest that the growth in mortgages is unlikely to be a major problem if monetary policy is normalized: Credit risk among persons is less contagious than among construction companies. In some sectors, such as hospitality, credit risk is not contagious but is highly volatile. Although contagion increases volatility, it makes little difference to the correlation between bank credit risks because benign and malignant contagion offset each other. This systemic methodology may be used by banks for stress-testing and in fulfillment of their obligations under Basel III.