Approach to Constructing Malmquist Indices

In this study, we utilise a new, non-parametric efficiency measurement approach which combines the semi-oriented radial measure data envelopment analysis (SORM-SBM-DEA) approach for dealing with negative data (Emrouznejad et al., 2010) with the slacks-based efficiency measures of Tone (2001, 2002) to analyse productivity changes for Indonesian banks over the period Quarter I 2003 to Quarter II 2007. Having constructed the Malmquist indices, using data provided by Bank Indonesia (the Indonesian central bank), for the banking industry and different bank types (i.e., listed and Islamic) and groupings, we then decomposed the industry’s Malmquist into its technical efficiency change and frontier shift components. Finally, we analysed the banks ’ risk management performance, using Simar and Wilson’s (2007) truncated regression approach, before assessing its impact on productivity growth. The first part of the Malmquist analysis showed that average productivity changes for the Indonesian banking industry tended to be driven, over the sample period, by technological progress rather than by frontier shift, although a relatively stable patter