ASSESSING THE PERFORMANCE OF INDIAN BANKS IN THE PRESENCE OF NON-PERFORMING ASSETS

Abstract

In recent years, the problem of non-performing assets (NPAs) remains a concerning issue in the Indian banking industry, in general, and in the public sector banking segment, in particular. The banking industry in India observed a perceptible fall in the quality of the assets portfolio of banks in the years immediately following the global financial crisis of 2007-09. The official sources point that this troublesome situation has occurred probably due to procyclical lending behaviour adopted by the banks in India. On the one hand, the weakened financial health of large borrowers worsened the investment climate due to their squeezed industrial demand, which reduced their business profits, and thereby, transformed them to overleveraged borrowers. On the other side, excessive volatility in financial markets has significantly increased the cost of capital for investments in securities, and thus, adversely affected the performance of banks in India. In the policy circles, it is believed that this surging burden of NPAs is ailing the financial health of Indian banks, especially in the post-crisis years. Since these toxic assets demand higher loan provisioning that eats up the profitability of banks and eventually affects their efficiency and productivity levels. Recognizing this growing stress of non-performing assets, it is crucial to know to what extent their formation impacted the efficiency and productivity levels of the Indian banking industry. The present thesis is an attempt in this direction. In particular, the broad objective of the present thesis is to provide an in-depth examination of the evolution and formation of NPAs and their accounting in the evaluation of efficiency and productivity of the Indian banking industry. The key objectives of the thesis are: i) to examine the evolution of NPAs in the Indian banking industry as a whole, and across distinct ownership types (public, private and foreign banking groups); ii) to identify the key regulatory and institutional factors alongside bank-specific, macroeconomic and industry-specific variables determining NPAs, and to quantify the degree of persistence in the occurrence of loan losses in the Indian banking industry; iii) to assess the risk-adjusted efficiency of Indian banks in the presence of non-performing assets using sequential directional distance function approach; iv) to evaluate the risk-adjusted productivity change and its components for Indian banks using the sequential Malmquist-Luenberger productivity index approach; v) to scrutinize the impact of the global financial crisis on efficiency and productivity levels of the Indian banks; vi) to test the phenomenon of dynamic convergence in the efficiency and productivity levels for the industry as a whole and across ownership groups; viii) to explore the impact of contextual factors that may influence the efficiency and productivity of banks. An investigation is done for the most recent and prolonged period from 1998/99 to 2016/17, covering the pre-crisis period (1998/99-2006/07), the crisis period (2007/08-2008/09),and the post-crisis period (2009/10-2016/17). For the estimation purpose, this research work adopts the ‘optimal approach’ suggested by Simper et al. (2017) and provides robust estimates of efficiency and productivity. The bank production model used in this study explicitly incorporates the two key risk management variables: i) equity as a good and quasi-fixed input, and ii) non-performing loans as a bad (undesirable) output. So far, such an in-depth analysis of bank efficiency has received little attention concerning the Indian banking industry, especially in the aftermath of the global financial crisis. The thesis explores the factors responsible for the formation of non-performing assets in the Indian banking industry using a two-step system generalized method of moments (GMM) estimation approach. The empirical findings reveal a significant degree of persistence in default risk, and the observed persistence is higher in the gross NPAs specification relative to what has been observed in the net NPAs specification. Further, the study reports that a lower profitability, more diversification in the banking business, the large size of banks and a higher concentration of banks in lending increase the probability of defaults in India. The results thus extend support to the“bad management” hypothesis in the Indian banking industry. The trade-off between short-term operating expenses and future credit quality lends support to the existence of the “skimping” hypothesis, as suggested by Berger and De Young (1997). Moreover, the inverse link between the concentration and the asset quality of Indian banks gives support to the “concentration-fragility” hypothesis. Another interesting result is that large-sized banks’ extend their credit without proper screening and monitoring of the borrower’s creditworthiness under the presumption of “too-big-to-fail”, and thus take an excessive risk which increases their NPAs. The implementation of prudential regulatory reforms has significantly lowered the formation of NPAs in the Indian banking industry. However, the global financial crisis has had no significant impact on the NPA generation among public sector banks during 2007-2009. In the case of public sector banks, NPAs are more sensitive to internal bank-specific factors, while for private and foreign banks, macroeconomic and industry-related factors play a significant role in determining the formation of NPAs. The outcomes are robust for different panel data estimation models and sub-samples of ownership groups. These findings of the study provide valuable insights into the formation of default risk in the banking system of an emerging market economy. An evaluation of the risk-adjusted efficiency evaluation of Indian banks is done by employing the sequential directional distance function approach. This approach integrates the directional distance function (DDF) approach by Chung et al. (1997) and sequential production technology frontier by Tulkens and Vanden Eeckaut (1995). The results reveal that the banks in India still have the potential to enhance their efficiency by 30.5 percent in terms of maximizing their incomes over the analyzed period. The Indian banks have performed better in terms of their efficiency during the pre-crisis years, while the period 2012/13 to 2016/17, marked with the substantial deterioration in their efficiency levels. Higher inefficiency during the crisis and post-crisis years could be attributed to the uncertain macroeconomic environment and ineffective lending practices adopted by the managers during the credit boom period. Among ownership groups, the foreign banks managed well in achieving higher efficiency, followed by the public sector banks. The second-stage regression analysis is performed using an approach suggested by Banker et al. (2019). The outcomes reveal that a higher level of NPAs, bank capitalization, and enormous credit growth lowers the efficiency of banks in India. The dynamic convergence tests signify the strong evidence of σ-convergence in the efficiency levels for the public and foreign banking groups in India. Finally, the thesis investigates the trends of productivity and its sources, namely, efficiency change and technical change in the Indian banking industry. In particular, we see whether Indian banks withstand the shocks of the global financial crisis of 2007-09 and sustain their productivity levels during the turbulent years. The robust estimates of productivity and its sources are obtained using the sequential Malmquist-Luenberger Productivity Index approach of Oh and Heshmati (2010). This approach rules out the possibility of any spurious technical regress and circumvents the problem of infeasibility in computing mixed period distance functions. The empirical findings reveal that the on an average Indian banking industry experienced a productivity regress of (-)1 percent per annum. The observed productivity regress is completely contributed by efficiency decline during the sampled years. The global financial crisis altered the growth trajectory of the productivity growth of banks, and its adverse effect was more pronounced in the public and foreign banking segments. The post-productivity analysis reveals that a higher level of banks’ profitability, better liquidity flow, and expenses incurred by banks on credit appraisal for risk mitigation enhance productivity growth. Also, the analysis confirms that bank diversification and higher credit growth lower productivity growth in India. A phenomenon of productivity convergence is noted at the industry level, while evidence varies across ownership groups. Several significant policy implications are drawn from this research work. On the basis of research findings, the thesis recommend that i) Indian banks need to enhance the mechanism for early detection of distressed debtors’ by appropriately adopting a technology-enabled risk management information system, ii) any prospective NPAs reduction plan by the regulators should not be based on a “one-size-fits-all” approach; instead, the focus should be on the ownership group-specific differentials to frame a clear resolution mechanism for banks, and iii) Indian banks need to put more emphasis on the utilization of installed and available technology for effective credit risk management and to ultimately enhance their efficiency and productivity levels. Overall, this thesis suggest that the integration of up-to-date technology and a robust risk management framework would assist in containing NPAs to a minimum optimal level, and help in maintaining bank efficiency and productivity gains. With the availability of financial data in the coming years, the future research works can be extended to consider the effects of ongoing reforms, including demonetization of 2016, and bank mergers & acquisitions of 2019, on the bank efficiency and productivity levels in India.