Technical Efficiency of Banks in Southeast Asia

National financial systems, and banking sectors in particular, assume increasing importance and fluidity with the progress of economic development and the increase in economic openness. This notwithstanding, attempts to measure and formally monitor the performance of the banking sector have largely been confined to western developed economies. As a result, little concrete empirical information and evidence is available about banking productivity and efficiency in non-western countries. Accordingly, the aim of the present investigation is to start filling the gap left by non-industrialized countries in the empirical literature of efficiency studies of banking. The paper examines the evolution and the contemporary state of bank efficiency in major developing economies of Southeast Asia - Indonesia, Malaysia, Philippines, and Thailand - over the period 2001–2005. During this period, the banking sectors of the sampled countries were involved in a process of restructuring that was often guided or even mandated by the respective governments. For example, the Indonesian government pursued a policy of consolidation, reducing the number of licensed banks by more than 40% (from 238 in 1997 to 134 by the end of 2004) during the sampled period. In Malaysia, the number of banks was reduced by 31%, from 36 in 1997 to 25 in 2004. In the Philippines, consolidation reduced the number of banks from 52 to 44, and in Thailand, from 16 to 12 during the period 1997–2004 (Gosh 2006, pp. 63–65). At the same time, an increase in cross-border mergers, i.e., mergers involving foreign firms, exposed the domestic banking sectors to greater competition from abroad (Deloitte-Touche 2005). The present study measures bank performance by employing Data Envelopment Analysis (DEA). The nature and robustness of the DEA results are evaluated with bootstrapping methods. To our knowledge, these methods have not been applied in the context of developing countries in Asia. Our methodology of estimating efficiency and bootstrapping is explained in the following section. Data issues are discussed in Sect. 5.3, results are presented in Sect. 5.4, and policy implications in Sect. 5.5. Section 5.6 concludes the paper.