A Bayesian Approach for Measuring Economies of Scale with Application to Large Canadian Banks

Traditionally, the literature has not found economies of scale for the very large banks. Among the reasons for these results are that usually large banks are not the sole focus of analysis and the analyzed banks may be subject to a diverse set of regulatory restrictions and limitations with respect to banking services. Our paper draws upon a panel data set containing information on the relatively large Schedule I Canadian banks. Although small in number, they offer extensive interbranch banking services under one set of regulations. In light of this, we propose a Bayesian methodology for estimating returns to scale. This technique allows for random coefficients and distinct input-allocative coefficients for each bank, and it provides reliable estimates of economies of scale using a panel data set with a small number of very large banks. In conclusion, we do find significant economies of scale.