Assessment and Prediction of Credit Risk in the US Banking Sector

This research assesses the credit risk within the U.S. banking sector by examining the relationship between various variables proposed by Merton's (1974) structural model, as well as additional factors such as the price of gold, oil, market volatility, market returns, Libor rate, exchange rates with the Euro and the Chinese Yuan. These variables are analyzed across crisis and non-crisis periods. The primary methodology employed is the automated variable selection algorithm, utilized to identify the optimal regression structure. Moreover, the study employs Hull's (2015) formula to approximate the probability of default. The incorporation of structural breaks is significant in this research, as it aids in identifying the change in coefficients and enhancing the predictive models. Key words: US banking sector, CDS spreads determinants, credit risk