Assessing the performance of Canadian credit unions using a three-stage network bootstrap DEA

We use a novel three-stage network data envelopment analysis (DEA) model (based on production, intermediation, and revenue generation operations) with bootstrapping to evaluate the performance of 14 of the largest Canadian credit unions for the period 2007-2017 and the impact of various events on this performance. For each analysis, we contrast the results of the network DEA with those of a black box DEA. We show that the former provides more insightful information regarding the sources of the inefficiencies. We first found that while overall, the credit unions showed high-efficiency ratios, there is room for improvement, especially for the production sub-process. Moreover, the efficiency of individual credit unions is not consistent across the three different stages. Through the years 2007-2017, the credit union system exhibits a relatively sharp decline in its efficiency, mainly due to managerial issues at the revenue generation stage. Our analyses show that the various stages of Canadian credit union operations have been affected by the 2007-2009 financial crisis, the low policy interest rates that occurred in the following years, and the fact that in Canada, the federal government has eliminated the discount on the federal tax rate. The credit unions can improve their performance at the different stages by exploring Fintech Solutions to reduce their operating costs, seeking a better mix of loans and securities investments, and improving their interest and saving rate settings.