How to explain non-performing loans by many corporate governance variables simultaneously? A corporate governance index is built to US commercial banks

This paper aims to combine the principal component analysis and GMM dynamic panel data methods in order to estimate the effect of corporate governance system on non-performing loans. The first method is meant to construct a corporate governance index for US commercial banks. The second one allows us to study the relation between the built index and non-performing loans. The advantage of the combination of these methods is reducing the number of corporate governance variables into a single one and ensuring the consistency of GMM estimates, given that a high number of variables leads to an increase in the number of GMM instruments, which in turns results in biased estimators. Applying these methods to US commercial banks, for a period including the financial crisis years, we find that small banks are characterized by a sound corporate governance system that reduces their non-performing loans. In opposition, the corporate governance fails to protect medium and large US commercial banks from excessive risk-taking that damages their loans’ quality and even leads to enormous losses especially during the global financial crisis.