Article ID | Journal | Published Year | Pages | File Type |
---|---|---|---|---|
6855502 | Expert Systems with Applications | 2016 | 50 Pages |
Abstract
This paper presents a performance assessment of 128 banks from 23 OECD countries from 2004 to 2013, using different financial criteria that emulate the CAMELS rating system. A robust TOPSIS approach for assessing bank efficiency is also developed and presented. First, alternative variable reduction techniques are employed to extract the major factors within each CAMELS criterion. This is done to mitigate collinearity issues. Then, TOPSIS is used to measure bank performance based upon these factors, equally weighted. A comprehensive analysis based on a weighted linear optimization model for multi-criteria classification is also performed, which detects any discrepancies from the original scores. Lastly, censored quantile regressions are combined with bootstrapped TOPSIS scores to produce a model for predicting the impact of different contextual variables on different efficiency quantiles. Results reveal that the effects of ownership, trend, and origin of the bank may vary with respect to efficiency levels, whether high or low.
Related Topics
Physical Sciences and Engineering
Computer Science
Artificial Intelligence
Authors
Peter Wanke, Md. Abul Kalam Azad, C.P. Barros,