Dynamic network DEA and SFA models for accounting and financial indicators with an analysis of super-efficiency in stochastic frontiers: An efficiency comparison in OECD banking

This paper presents an assessment of 124 OECD banks during 2004-2013 in light of relevant accounting and financial indicators that reflect banking production process and performance. Novel DEA and SFA Dynamic Network Super-efficiency models are developed to handle the underlying relationships among major accounting and financial indicators. Additionally, we develop a novel super-efficiency concept for SFA to compare DEA and SFA on common grounds to the extent possible. Firstly, a relational model encompassing major profit sheet indicators, balance sheet indicators, and net income is presented under a dynamic network structure for both models under the non-parametric and parametric specifications. Subsequently, the dynamic effect of carry-over indicators is incorporated into them, so that efficiency scores can be properly computed for these three sub-structures. Differences in scores and ranks are explored by bootstrapped regressions. It reveals that results generated from the proposed models interact differently with the socio-economic and business-related variables.