The capital ratio and the interest rate spread: A panel threshold regression approach

Abstract

Using annual observations of insured US commercial banks, this paper investigates the possible asymmetric impacts of capital ratios on the interest rate spread where the asymmetry depends on the magnitude and also the channel the bank used to increase the capital ratio. To this end, we employ a panel threshold regression approach with one and two threshold variables, where the change in the capital ratio itself and the main components of change in the capital ratio, namely, change in the capital or change in risk-weighted assets have been used as the threshold variables. The results show a significant threshold effect. In the single threshold models, banks that correspond to regimes with a higher increase in the capital ratio, a higher contribution of capital to the increase in the capital ratio, and a higher contribution of risk-weighted assets to the increase in the capital ratio also show a higher impact of capital ratios on the interest rate spread compared to their counterparts. In the panel threshold regression model with two threshold variables and four regimes, we find similar results. Our baseline results also hold stable using the panel smooth transition regression approach as an alternative modeling approach, using alternative proxies as threshold variables and performing the main tests in different bank size categories.