Modelling EA banks default rates with jointly spanned and unspanned interest rates and unspanned BEI rates

Abstract

The paper quantifies the influence of interest rates and inflation rates on default rates of banks. By expanding the work of Duffee (1998), with the unspanned risks as in the work of Joslin, Priebsch, and Singleton (2014), we estimate a multifactor model with unspanned interest rates and inflation rates to test the performance of unspanned variables in the default rate term structure of banks. The model is trained in samples of positive interest rates and evaluated in samples of negative interest rates. we check the robustness of the model by comparing the results with the performance of alternative model specifications. The model reveals that unspanned variables have worse performance than alternative models specifications. The negative effect of interest rates on default rates over longer maturities may lead the EA banks to decrease the loan supply to the real economy. As a consequence EA banks will have a lower net interest margin as the return of assets is lower. This may increase the future probability of default. Thus, the solution for EA banks is on the reach to yield behavior as described by Bruno and Shin (2015). This means that EA banks have to modify the allocation of assets more in favour of riskier and longer maturity securities to obtain higher profitability.