Description

This paper estimates efficiency measures for the banking system in Chile during 2000-2019. Distinct from previous studies, we use input distance functions, introduce the non-parametric slack-based model, and choose the intermediates approach in the determination of inputs and outputs. Our findings suggest that the Chilean system achieved relatively high efficiency levels although with no significant variations during the sample period. Ownership (state, foreign and public trading) and size had a positive impact on efficiency. Merger and acquisitions processes, on average, seem to have targeted highly efficient banks to improve overall efficiency of the controller institution in the short term. At input level, deposits, operative expenses and capital stand out for reaching the frontier level although with diverging trends. We did not find categorical results in the effects of efficiency on bank variables, such as return or dividend policy. Regarding model selection, we find that it affects efficiency averages, scores and rankings although it did not affect the conclusions on inference tests.