Search

To link to full-text access for this article, visit this link: http://dx.doi.org/10.1016/j.ejor.2006.04.043 Byline: Mei-Hui Wang (a), Tai-Hsin Huang (b) Keywords: Data envelopment analysis; Efficiency persistence; Quasi-fixed inputs; GMM procedure; Markov model Abstract: This article introduces a sequence of four systematic methods to examine the extent to which the economic efficiency of Taiwan's commercial banks persists and to uncover the potential dynamic link between bank performance and various financial indicators. Quasi-fixed inputs are explicitly incorporated in the DEA model to account for possible adjustment costs, regulation, or indivisibilities. Among the four methods, the dynamic panel data model and the Markov model appear to be exploited for the first time in the area of the DEA approach. Evidence is found that bank efficiency exhibits moderate persistence over the sample period, implying that the given sample banks fail to adjust their production techniques in a timely manner. Regulatory authorities and bank managers are suggested to be aware of the level of undesirable non-performing loans due to their close relationship with bank performance. Author Affiliation: (a) Department of Banking and Finance, Tamkang University, Taiwan, ROC (b) Department of Money and Banking, National Chengchi University, Taiwan, ROC Article History: Received 28 September 2004; Accepted 7 April 2006