Assessing the productivity growth of Indian banks during the post-deregulation period using non-radial Malmquist productivity index

Abstract

By using an unbalanced panel dataset for the period 1991-1992 through 2007-2008 and employing non-radial Malmquist productivity index approach, this paper examines the evolution of total factor productivity (TFP) change in Indian banks, and identifies dominant source of productivity change. The empirical results indicate that TFP in Indian banking industry followed an uptrend during the post-deregulation period, albeit the trend was modest. Further, the observed productivity growth was almost entirely attributable to technological progress. While the public sector banks made steadier progress in productivity, the de novo private banks outperformed their peers. The empirical findings also reflect the enormous relevance of foreign ownership in Indian banking industry since foreign banks have emerged as the leading technological innovators in the banking system.

Economic slowdown, NPA crisis and productivity behavior of Indian banks

“…It is noteworthy here that the efficiency literature suggests the host of production technologies to construct an efficient frontier (see, for more details, Tulkens and Vanden Eeckaut, 1995;Hu and Liu, 2015). The intertemporal or grand frontier approach considers the availability of bank's future NPA crisis and productivity behavior of Indian banks technologies (as adopted by Bhattacharyya et al, 1997;Ataullah and Le, 2006;Das et al, 2009), contemporaneous/annual technology is subject to the availability of bank's current year technology (as referred by Isik and Hassan, 2002;Pasiouras, 2008;Gulati and Kumar, 2016), and window-based technology, for instance, 3-years window frontier, builds on the bank's technology set from the period t 1 to t 3 (as implemented by Asmild et al, 2004;Sufian, 2007). However, the sequential reference technology is considered as a benchmark because it takes all the past and the current year's observations or technologies of a specified bank in the construction of the best-practice efficient frontier.…”

“…All the financial variables are transformed to constant prices using the implicit price deflator of gross domestic product (GDP) at factor cost (base 2011/12 = 100). To reduce the effect of random noise due to measurement error in input and output variables, we followed Denizer et al (2007) and Gulati and Kumar (2016) and normalized all the desirable inputs and outputs along with undesirable output and non-discretionary input variables of a bank by dividing them by the corresponding number of branches of that bank. Based on Cooper et al (2007), the sample used in this study is sufficiently larger and feasible for the productivity estimation in the directional distance function based non-parametric DEA framework.…”