부산대학교 도서관

In this study, we modified the model proposed by Amsler, Prokhorov, and Schmidt (2016, 2017) to consider endogenous inputs and environmental variables in directional input distance and cost functions. Compiling data from Taiwan's banking industry, spanning 2002-2019, we estimated technical, cost, and allocative (in)efficiency measures for sample banks. The maximum likelihood estimation was adopted to obtain the parameter estimates for the above two functions. Empirical results support that the distance function considers endogenous inputs and environmental variables and that the cost function involves endogenous environmental variables. The average value of allocative inefficiency was higher than technical inefficiency, where foreign banks were the most allocatively efficient, followed by private and pan-public banks. The various inefficiency measures tended to deteriorate over time. Since economies of scale and scope prevail in the industry, we recommend that the sample banks expand their operating scale and diversify their financial products to lower the long-run average cost.