A risk-neutral approach to the RAROC method of loan pricing using account-level data

Purpose This paper has used account-level data of corporate and retail borrowers, assessed their credit risk through the risk-neutral principle and examined its implication on loan pricing. Design/methodology/approach It derives the capital charge and credit risk-premium for expected and unexpected losses through a risk-neutral approach. It estimates the risk-adjusted return on capital as the pricing principle for loans. Using GMM regression, the article has assessed the determinants of risk-based pricing. Findings It has been found that risk-premium is not reflected in the current loan pricing policy as per Basel II norms. However, the GMM estimation on RAROC can price risk premium and probability of default, LGD, risk weight, bank beta and capital adequacy, which are the prime determinants of loan pricing. The average RAROC for retail loans is more than that of corporate loans despite the same level of risk capital requirement for both categories of loans. The robustness tests indicate that the RAROC method of loan pricing and its determinants are consistent against the time and type of borrowers. Research limitations/implications The RAROC method of pricing effectively assesses the inherent risk associated with loans. Though the empirical findings are confined to the sample bank, the model can be used for any bank implementing the Basel principle of risk and capital assessments. Practical implications The article has developed and validated the model for estimating RAROC, as per Basel II guidelines, for loan pricing that any bank can use. Social implications It has developed the risk-based loan pricing model for retail and corporate borrowers. It has significant practical utility for banks to manage their risk, reduce their losses and productively utilise the public deposits for societal developments. Originality/value The article empirically validated the risk-neutral pricing principle using a unique 1,520 retail and corporate borrowers dataset.