A new approach to optimal capital allocation for RORAC maximisation in banks

We introduce a new model for optimal internal capital allocation, which would allow banks to maximize their Return on Risk-Adjusted Capital (RORAC) under regulatory and capital constraints. We extend the single period model of Buch et al., (2011) to a multi-period model and improve its forecasting accuracy by including the debt effect and Bayesian learning innovations. The empirical application shows that our model significantly improves the RORAC of a sample of banks listed in the S&P 500 index.