RBI WPS (DEPR): 08/2013: Macroprudential Regulation and Bank Performance: Evidence from India - आरबीआई - Reserve Bank of India
RBI WPS (DEPR): 08/2013: Macroprudential Regulation and Bank Performance: Evidence from India
RBI Working Paper Series No. 08 Abstract: 1Employing data on Indian banks for 1992-2012, the article examines the impact of macroprudential measures on bank performance. First, it finds that state-owned banks tend to have lower profitability and soundness than their private counterparts. Next, it tests whether such differentials between state-owned and private banks are driven by macroprudential measures; it finds strong support for this hypothesis. JEL classification: G 21; L 51; P 52 Keywords: banking; macroprudential; capital adequacy; loan classification; provisioning; ownership; India Introduction In recent years, countries have put a lot of emphasis on financial sector reforms as a means to improve the overall functioning of the sector. Such reforms have encompassed a significant gamut of measures, including lowering of statutory reserve requirements, deregulation of interest rates, introduction of measures relating to income recognition, loan classification and provisioning, allowing more liberal entry of foreign banks and diversifying the ownership base of state-owned banks. The evidence emanating from empirical research is admittedly mixed. One set of studies find that financial deregulation leads to an increase in the resilience and performance of the banking sector (Boyd and De Nicolo, 2005; Das and Ghosh, 2006, 2009; Yeyati and Micco, 2007), while others find that the net effect of financial deregulation on the banking sector to be negative (Keeley, 1990; Grifell-Tatje and Lovell, 1996; Wheelock and Wilson, 1999). The existing literature tends to look at each macroprudential measure in isolation, thereby ignoring the effect of these measures in totality on bank performance. For instance, there are studies that examine the impact of removal of interest rate ceilings on the banking sector (Kwan, 2002; Feyzioglu et al., 2009). Several others consider the effect of prudential regulations on bank risk and performance (Matutes and Vives, 2000; Hellmann et al., 2000; Claessens and Laeven, 2004; Agoraki et al., 2011). None of the studies take a holistic view on the different macroprudential measures on bank performance. As Allen and Gale (2004) observe, since the aspects of performance, stability, efficiency and soundness of banks are inter-related, careful consideration of all important prudential measures is important for sound empirical analysis. In this context, the paper investigates how various measures of macroprudential regulation affect the performance of the banking sector. More specifically, we consider the impact of three major dimensions of macroprudential regulation – capital adequacy ratio, provisioning norms and loan classification requirements - on the performance of the Indian banking system. We employ four indicators on which to assess the impact: return on asset (RoA) as the profitability measure, net interest margin (NIM) as the measure of economic efficiency, Z-score as the measure of bank stability and finally, advances growth (Gr_Advances) as a measure of bank business. India provides a compelling case among emerging markets to examine this issue in some detail. First, beginning from the early 1990s, the country has experienced significant liberalization of the banking sector. These liberalisation measures were premised on the objectives of enhancing efficiency, productivity and profitability of banks (Government of India, 1991; 1998). Second, India is one of the largest and fastest growing emerging economies with a gamut of banks across different ownership categories. It would be of interest to examine the impact of different regulatory measures on the performance of banks across different ownership groups. Third, a comprehensive and reliable banking database for an extended time span is available for Indian banks. The time-series and cross-sectional variation in the data makes it amenable to rigorous statistical analysis. Additionally, the time period of the study, beginning 1992, coincides with the inception of economic reforms. As a result, it permits us to clearly ascertain the impact of regulatory reforms on the performance of Indian banks. These findings might provide useful leads to other emerging market banks to examine the impact of relevant measures on bank performance across different ownership groups. The paper combines several strands of literature. The first strand is the effect of macroprudential measures on bank performance. Several papers have analyzed the impact of capital requirements on bank risk and performance variables. Employing a partial adjustment framework, Shrieves and Dahl (1992) uncovered evidence to suggest that regulation was effective in the sense that undercapitalized banks (i.e., with capital ratios of less than 7 per cent) increased their capital ratios by more than 100 basis points per annum as compared to other banks. Studies for non-US banks, including UK (Ediz et al., 1998), Switzerland (Rime, 2001) and India (Ghosh et al., 2003) also provide support to the efficacy of capital regulation. In contrast to these studies, we examine the impact of a whole gamut of macroprudential measures on bank behaviour. To the best of our knowledge, this is one of the earliest studies to systematically study the impact of macroprudential regulations on bank behaviour. Second, the paper is related to the literature on the evolution of the Indian banking sector in the post-deregulation era and on the characterization of the state-owned banks in India (Banerjee et al., 2004; Berger et al., 2008; Gormley, 2010; Zhao, 2010; Cole, 2011). The analysis by Banerjee et al (2004) appears to suggest that Indian state-owned banks do not provide adequate credit to the private sector. Berger et al. (2008) examine relationship lending across bank ownership and finds that state-owned banks to be the main bank for state-owned firms; while foreign banks are less likely to lend to small and opaque firms. Gormley (2010) finds that cherry-picking by foreign banks might lead domestic firms to obtain less credit, because of the drop in domestic lending. Cole (2011) demonstrates that the growth rate of agricultural credit provided by state-owned banks is 5-10 percentage points higher in election years. The present paper complements these findings by focusing on the impact of several prudential measures and comparing the response across bank ownership. Finally, this study belongs to the literature which investigates the within-country effects of changes in regulation (Shrieves and Dahl, 1992; Ediz et al., 1998; Stolz, 2007) and to a wider literature which identifies the effects of regulations based on cross-country analysis (Murinde and Yaseen, 2006; Van Roy, 2008; Cosimano and Hakura, 2011). In contrast to the extant literature, this paper explores the impact of several macroprudential measures on bank performance variables. The results suggest that different macroprudential measures exert a differential impact on bank performance. The remainder of the paper continues as follows. Section 2 provides an overview of Indian financial sector reforms. The relevant literature is covered in Section 3. The data and methods are detailed in Section 4, followed by discussion of the results. The final section concludes. 2. The Indian banking system and regulatory environment The Indian banking system is characterized by a large number of banks with mixed ownership. As at end-2012, the commercial banking segment comprised of 87 banks, including 26 state-owned banks (SOBs), 20 domestic private banks, including seven de novo private and 34 foreign banks. Total bank assets constituted over 90 per cent of GDP in 2011-12. In 1991, on the eve of financial reforms, SOBs share in total banking assets was a little over 90 per cent. Prior to financial reforms beginning 1992, the financial system in India essentially catered to the needs of planned economic development. The Government played an overarching role in every sphere of economic activity. High levels of reserve requirements pre-empted a large proportion of bank deposits. Likewise, a system of administered interest rate regime resulted in low-quality financial intermediation. The availability of concessional credit to selected sectors resulted in cross-subsidization such that the interest rates charged to borrowers were not commensurate with the underlying risks. Likewise, the inflexibilities in branch licensing and rigid management structures impeded the operational independence of banks. The overall consequence was an inefficient allocation of scarce resources. The philosophy underlying the financial reforms was to make the banking system more responsive to changes in the market environment. Accordingly, over a period of time, interest rates have been deregulated, competition has been enhanced and the state-owned banking system has been opened up to private participation. Salient among the measures introduced included: (a) lower statutory reserve requirements; (b) liberalisation of the interest rate regime, on both the deposit and lending sides; (c) allowing liberal entry of foreign banks and permitting the establishment of de novo private banks; and, (d) introduction of a wide gamut of prudential measures, in addition to internationally accepted accounting practices. As a consequence of these measures, the competitive pressures on the banking industry have increased. For example, the five-bank asset concentration ratio has declined from over 0.50 in 1991-92 to less than 0.40 in 2008-09. The banking sector has also become more diversified with an increasing number of private and foreign players (See for instance, Prasad and Ghosh, 2005). Reflecting the efficiency of intermediation, the net interest income has declined from over 3 per cent of total assets to close to 2 per cent (Table 1). Three salient macroprudential measures have characterized the process of financial reforms. The first has been the tightening of capital adequacy norms for banks. Driven by the imperatives of liberalization, the capital-to risk-weighted asset ratio (CRAR) for banks was raised to 8 per cent in 1996 (Table 2). More specifically, while Indian banks with international presence and foreign banks were directed to achieve the stipulated CRAR by 1994, other banks were provided another couple of years to achieve these norms. The capital adequacy norms were further raised to 9 per cent in 2000. Second, in 2000, the Reserve Bank of India (RBI) introduced a provisioning of a minimum of 0.25 per cent on standard loans. This measure was more in response to stimulus from domestic forces. These measures were calibrated during the crisis, wherein the provisions were raised over a period of time, initially to 0.40 per cent and thereafter to a peak of 2 per cent in January 2007 before being subsequently lowered. Finally beginning 1993, the norms for recognizing a loan as non-performing have been gradually rationalized, in line with international best practices. Accordingly, the time period for classifying a loan as “sub-standard” was gradually reduced from the initial 12 months to 3 months (90 days) by 2004. Concomitantly, the period for classifying a loan as “doubtful” was also lowered, from 24 months at the beginning of reforms to 12 months by 2005. We investigate the impact of these macroprudential measures on the performance and soundness of the banking sector.
Akin to Besley and Burgess (2004), we code the macroprudential measures as follows. In case there is an increase (resp., decrease) or a tightening (resp., weakening) of a measure in a given year, it is coded as +1 (resp., -1). Provided there is no change in the measure during the year, it is coded as zero. The raw scores across the sub-categories under each of the macroprudential measures are cumulated to arrive at an aggregate index in a given year. As a result, a value greater than one for a given macroprudential measure in a year would signify a tightening; reverse would be the case in case the value is less than one. The macroprudential measure is deemed neutral in case the value of the index in a year equals zero. A significant body of literature has examined the impact of deregulation on bank behaviour. It is possible to broadly distinguish two strands of literature. The first is primarily theoretical in nature, while the second is more empirical in its scope The theoretical literature has focused on the interrelationship among financial deregulation, market power and risk-taking by banks. Hellmann et al. (2000) contend that capital requirements alone may not be enough to curtail bank risk and additional requirements could be useful to reduce risk in a competitive environment. Diamond and Rajan (2000, 2001) suggest that well-capitalised banks might not be conducive to liquidity creation, simply because higher capital lowers bank weaknesses. More recent research indicates that the relation between deregulation and bank behaviour may not be unambiguous (Boyd et al., 2006). Empirical research into the above models report mixed findings. According to the first strand, the impact of financial deregulation is typically assessed either through a dummy variable (Salas and Saurina, 2003) or simply examining the behaviour of banks during periods of financial deregulation (Das and Ghosh, 2006). The findings indicate that the impact of deregulation on bank behaviour depends, among others, on the state of the banking system and differs significantly across bank ownership. The second strand of the literature focuses on the impact of financial deregulation on bank performance. Cross-country studies (Maudos and Pastor, 2001) report improvements in performance, post-deregulation. However, given the difficulties inherent in cross-country comparisons (Rodrik, 2012), studies have also been conducted at the country-level. At the country level, studies have examined, among others, the performance of banks in the US (Elyasiani and Mehdian, 1995; Wheelock and Wilson, 1999), Norway (Berg et al., 1992), Thailand (Leightner and Lovell, 1998), Korea (Gilbert and Wilson, 1998) and Taiwan (Shyu, 1998). These studies exhibit two major limitations. First, they focus on a catch-all measure, thereby ignoring the impact of specific policy dimensions of deregulation. Second, most studies examine the impact of financial deregulation on efficiency and productivity, neglecting other measures of bank performance, such as profitability and soundness. Partly as a response to these concerns, two sets of studies have emerged. The first examines the impact of specific regulatory reforms on various facets of bank performance. The second set examines the impact of financial deregulation on alternate measures of bank performance. As regards the former, Kwan (2002) focused on the impact of interest rate deregulation in bank performance in Hong Kong. The study observed that interest rate deregulation led to significant decline in bank market values. Focusing on China, Feyzioglu (2009) found that interest rate deregulation would raise the cost of capital, improve the return on savings and allow for more efficient financial intermediation. Yet others have examined the one-off impact of changes in loan classification norms on banks’ credit portfolio (Das and Ghosh, 2007). The second set of studies explore the impact of financial deregulation on alternate measures of bank performance. For instance, several studies investigate the determinants of bank profitability or net interest margins (Demirguc Kunt and Huizinga, 2000; Demirguc Kunt et al., 2004; Maudos and Solis, 2009) and bank stability indicators (Ianotta et al., 2007; Beck et al., 2009). These studies veer around the view that financial deregulation generally has a positive effect on performance. In the Indian case, studies have focused on the impact of financial deregulation on efficiency and productivity as also the impact of specific regulatory measures on bank performance. In an early study on Indian banks for 1986–91, Bhattacharya et al. (1997) found that state-owned banks were the best performing banks in terms of efficiency. Kumbhakar and Sarkar (2003) show evidence to suggest that regulatory reform did not exert any perceptible impact on total factor productivity growth. Das and Ghosh (2006) attribute the high cost inefficiency of banks to the under-utilization of resources. Zhao et al. (2010) reported that, by increasing competition, greater deregulation encouraged banks to increase risk-taking. More recently, Casu et al. (2013) report a positive effect of deregulation on total factor productivity (TFP) for Indian commercial banks covering the period 1992-2004. The study which comes closest to the spirit of the present analysis is IMF (2012). Employing cross-national data on 36 countries for the period 2000-11, the analysis examined the impact of several macroprudential instruments on both financial and real variables.3 The results appear to suggest statistically significant effects for both capital requirements and reserve requirements on credit growth. In terms of real variables, the results point to the fact that limits on loan-to-value (LTV) ratio exerts a non-negligible impact on output growth. Although there are certain similarities between that study and the present one, there are also important differences. First, unlike the IMF (2012) study, the present article focuses on a single country. This enables us to bypass issues of data comparability that often plague cross-country studies (Honohan, 2008). Second, comparing institutional and political characteristics across countries are difficult, given the wide diversity in their evolution, customs and the institutional context. As a result, focusing on a single country enables to bypass these limitations (Rodrik, 2012). Third, the set of macroprudential instruments considered across the two sets of studies is significantly different. And finally, we explore the differential impact of macroprudential variables across bank ownership, an aspect not addressed by IMF (2012). This paper seeks to fill the gap in the literature and to shed light on the evidence in the context of various measures of macroprudential regulation and banking in India during 1992–2012. Bank-wise data on commercial banks spanning the period 1992 through 2012 are culled out from the various issues of Statistical Tables Relating to Banks in India. This publication by Reserve Bank of India provides the annual audited data on the balance sheet and profit and loss accounts of individual banks. The financial year for banks runs from the first day of April of a particular year to the last day of March of the subsequent year. Accordingly, the year 1992 corresponds to the period 1991–92 (April–March) and so on, for the other years. The data has the advantage of being perfectly comparable across banks, with the central bank acting as the regulator of the financial system makes it mandatory for the financial entities to present their balance sheets in the same format and criteria. The sample comprises of all state-owned banks (SOBs), which account for around 75 per cent of total banking assets, 20 domestic private banks (DPBs), including 5 de novo private banks (which became operational after the initiation of reforms), which account for around 15 per cent of banking assets and 16 foreign banks (FBs), accounting for roughly 7 per cent of total banking assets. The excluded private and foreign banks are those which have become operational only recently and therefore, lack a consistent time series of relevant variables. The de novo private banks became operational only since 1996. As a result, the number of reporting banks witnessed a sharp increase thereafter. Subsequently, the banking industry also witnessed some consolidation, both domestic and internationally. We also include a dummy variable for take this aspect on board. As a result, we have an unbalanced panel, with a minimum of 58 banks at the beginning of the sample to a maximum of 64 banks. With an average of 20.2 years of observations per bank, there are a maximum of 1307 bank-years.4 The macroeconomic variables are obtained from the Handbook of Statistics on Indian Economy, a yearly Indian central bank publication which provides time-series information on the macroeconomic variables. Table 3 provides a description of the relevant variables, including the data source and summary statistics. 5. Results and Discussion 5.1 Univariate tests Table 4 reports comparisons of various measures of performance, as indicated earlier. The results indicate a clear divergence across ownership. These differences in most cases appear to be economically important, as well. For example, the average NIM for FBs is 3.4 per cent, which is significantly higher as compared to SOBs and DPBs. Return on asset displays the greatest variation, especially for SOBs. Their average RoA is 0.6 per cent, which is around 50 per cent lower than that obtaining for DPBs and roughly half as compared to FBs. All these differences are statistically significant at 0.01 level. The results on Z-score and credit growth5 are no less striking.6 FBs have statistically significant lower Z-score as well as credit growth as compared to other bank groups. To illustrate, credit growth for FBs is roughly 9 per cent, which is significantly lower as compared to 14.3 per cent growth obtaining for DPBs. Overall, the results in table 3 appear to suggest that FBs have the highest margins and profitability, although their stability and credit growth are the lowest across ownership. The raw correlations in table 5 show a statistically significant association between the prudential measures and bank performance. Without loss of generality, capital adequacy norms appear to exert a pronounced impact on most measures. On the other hand, NIM appears to be unaffected after imposition of macroprudential measures. As expected, bank stability improves while loan growth is reduced after imposition of capital and provisioning norms. These raw correlations however, do not account for bank-specific controls or changes in the economic environment. 5.2 Multivariate regression We control for the factors mentioned above in a multivariate regression framework. Akin to Demirguc Kunt and Huizinga (2000) and Martinez Peria and Mody (2004) and Micco et al. (2007), measures of performance are regressed on a set of controls (X) included with a lag to account for endogeneity. The regression specification for bank s at time t is specified as: ![]() In (1), Perf is the performance measure, which is regressed on a set of lagged bank-level controls (X) and ηt are year fixed effects. All expressions control for the impact of mergers (d_merger). We run the regressions with and without the ownership dummy (OD) to ascertain its impact on bank performance. Finally, ν is the error term. We estimate the impact of explanatory variables on performance by fixed effects panel regressions. This method of estimation provides better estimators than simple OLS when the explanatory variables are correlated with the error term. It is quite straightforward to argue that there is probably significant correlation between unobservable individual characteristics of banks (which are captured by the error term of the OLS regression) and some of the explanatory variables. As a result, employing OLS could render biased coefficients. Using a fixed-effects model can solve the problem of correlation. In the fixed effects specification, the differences across banks are captured by the differences in constant term (Greene, 1993). Throughout, inference is based on standard errors that are clustered at the bank level. The bank-specific variables include (log of) total assets (LTA), bank-wise asset share (SHTA) in a given year, demand deposits (DDEP) and fee income (NINT). Following Berger et al. (2005), we include both LTA and SHTA. The former controls for scale economies and the latter for market power of banks. Among the other variables, DDEP takes into account for banks’ funding structure (SOBs tend to rely more on retail deposits as compared to other bank groups) whereas NINT accounts for banks’ income diversification (foreign banks tend to rely more on non-interest income). 5.3 Results and Discussion In Table 6, across the first two sets of specifications, the coefficient on demand deposits is significant and positive, suggesting that banks with greater retail dependence have higher profitability and margins. In the baseline specification, it is observed that a 10 per cent increase in retail dependence improves RoA by nearly 0.3 per centage points. One reason for this could be the low (or, negligible) cost of such deposits, which enables banks to earn higher margins and profitability on such funds. Banks with higher fee incomes are able to generate higher profitability, as expected. Bigger banks appear to exhibit greater stability. These findings are consistent with Beck et al. (2009) who find that bank size exerts a positive impact on stability. Credit growth is slower for big banks, suggestive of the fact that smaller banks increase credit at a faster pace to gain market share. When we include bank ownership, it is observed that as compared to DPBs, FBs have lower credit growth and stability. The effect is quantitatively important, indicating that the average foreign bank has a credit growth that is 0.08 percentage points lower as compared to an average domestic private bank. Considering that the average credit growth in the sample is 11 per cent, this is a sizeable difference. 5.4 The role of macroprudential regulation The previous discussion indicates that FBs display lower stability and exhibit lower credit growth as compared to DPBs. The analysis does not highlight how specific macroprudential measures impact bank performance. To investigate this further, we check whether macroprudential regulation affects the relationship between ownership and performance by estimating regressions of the following form: ![]() In (2), GDPGR measures the real GDP growth in year t, and PRU (PRU=CRAR, LOAN, PROV) is the particular macroprudential measure; the remaining variables are as in Eq. (1). The main coefficient of interest is α3. This coefficient measures the impact of implementation of a given macroprudential measure on state-owned banks (the main effect of macroprudential norms is controlled through year effects). We control for the interaction between ownership and GDP growth, because state-owned and foreign banks might respond differently to the business cycle as compared to domestic private banks. This would not be a problem if the business cycles were uncorrelated with the prudential measures, although such a correlation cannot be ruled out (Micco et al., 2007).7 Table 7 reports the results for return on asset, interest margins, soundness and credit growth. Take for instance, Column 1. The results suggest that, as compared with DPBs, SOBs have higher profitability during periods of economic expansion, although their profitability is reduced after imposition of capital adequacy norms. The impact of the macroprudential measure is economically meaningful, as well. To see this, consider the differential in profitability of the average SOB and the average DPB in a year in which real GDP grew by 6.7 per cent, the average growth rate in the sample. Ignoring the impact of capital standards, the differential equals 0.011 per cent points [=-0.019+0.067*(0.125)]. Taking on board the impact of capital adequacy norms, the point estimates of Col. 1 yield a difference of 0.023 per cent points [=-0.019+0.067*(0.125)-0.012], an increase of over 100 per cent with respect to the no-capital imposition benchmark. In a similar fashion, in case of both loan classification and provisioning practices (Cols. 2 and 3), the difference in profitability works out to be 85 per cent and 62 per cent, respectively. Similar results are echoed when we focus on interest margins. More specifically, the evidence indicates that interest margins of SOBs tend to be higher during periods of economic expansion and lower after imposition of macroprudential norms. Again, the coefficient on the macroprudential dummy is quite large and indicates that the differential between the interest margin of state-owned and private banks more than quadruples after imposition of capital adequacy norms (assuming 6.7 per cent GDP growth, the two values are -0.003 and -0.015). Similar, although of slightly lower order of magnitude, are in evidence when the provisioning and loan classification norms are considered. This provides evidence that the macroprudential channel is at work: the decline in profitability is driven to an extent by the lower margins. Cols. (7) to (9) focus on bank soundness. The evidence indicates that the soundness of SOBs declines during periods of economic growth, although macroprudential norms exert a salutary impact. More specifically, capital adequacy and provisioning norms improve soundness, whereas loan classification norms lower it. The magnitude of the macroprudential effect in all cases is extremely large. In case of provisioning norms for example, the point estimates indicate that the differential between the soundness of SOBs and DPBs halves after imposition of such norms (with 6.7 per cent GDP growth, the two values are -0.38 and -0.17). The final three columns focus on credit growth. In particular, we find that SOBs lower their loan growth after imposition of capital norms. In Col.10 for example, the differential between credit growth of SOBs and DPBs works out to be over 100 per cent (the two values are equal to -0.003 and -0.007 respectively, assuming 6.7 per cent GDP growth). In other words, the evidence indicates that the state-owned banks are less profitable than private banks and the difference in performance is accentuated after imposition of macroprudential norms. One possible way to interpret these findings could be the following. One of the major focus of macroprudential requirements has been to improve the resilience of the banking system by creating a cushion for expected and unexpected losses. With state-owned banks being major players in the system in terms of both size and banking space (Subbarao, 2013), this would have entailed an improvement in their soundness in the long-run, perhaps at the cost of trading-off short-term profitability considerations. When weighed on a long-term basis, it seems likely that the pros of such regulations outweigh the cons so as to provide resilience and help in nurturing a stable and sustainable financial system through the cycle. 5.5 Additional tests In this section, we examine whether our benchmark specification is able to capture important static, dynamic and selection effects, especially for SOBs. In their study of Argentine banking system during the 1990s, Berger et al. (2005) observed that in studying the interaction between bank performance and ownership, it is important to distinguish between static, dynamic and selection effects. The state-owned banks in India provide an ideal laboratory to examine this issue. As part of the process of financial reforms beginning 1991, the relevant banking Acts were amended to enhance the scope for partial private shareholding in SOBs. Over the period 1994-2012, a significant number of SOBs accessed the equity market, with several of them making a follow-on offer. The total amount raised has been close to Rs.250 billion (≈US$ 5 billion); the government shareholding in the divested banks range from 51-82 per cent. Table 8 highlights some relevant details. We employ three explanatory variables to decompose the causes and effects of changes in ownership on performance. The first variable is STAT, a dummy variable that equals one from the year a bank is privatized. The second variable is SELECT, a dummy that equals one throughout the whole sample for banks that were privatized at some point during the sample period. While SELECT controls for any selection effects associated with privatization, STAT measures the effect of privatization itself. The third variable, DYN, is equal to the number of years since the year of privatization. While STAT is included to capture the static effects of privatization, DYN captures the average yearly performance trend in the wake of privatization. To investigate the impact of the interlinkage between ownership and regulation, we estimate the following specification as given by (3): ![]() where STAT, DYN and SEL are the ownership variable2 as discussed earlier and PRU is the relevant macroprudential measure. We introduce the interaction of each macroprudential measures with the ownership variables in a sequential fashion. All regressions take on board the full set of controls variables including year effects and dummies for mergers, although these are not reported. The results are presented in Table 9. Considering the static effect, we find three significant results. First, whenever significant, the coefficient on NIM is negative and that on Ln (1+Z) is positive. In other words, privatized SOBs tend to have lower margins but higher soundness. Second, macroprudential measures appear to exert a positive impact on performance. Thus, profitability and margins improve after imposition of macroprudential measures, although soundness is adversely impacted. Third, tightening of loan classification norms appears to have exerted a positive impact on their loan growth. Thus, for example, the interaction term STAT*LOAN is positive and statistically significant in the Gr_Advances equation with a point estimate equal to 0.12. Given an average advances growth of 10.5 per cent for SOBs in the sample, this entails a difference in advances growth of roughly 0.01 per cent (=0.12*0.105) between privatized and non-privatized SOBs. The dynamic effect of privatization for SOBs is negative and statistically significant for RoA and NIM, and positive in case of soundness, although it is not statistically significant in the other regressions. This indicates that, by lowering the banks’ lendable resources, provisions adversely affect profitability and margins, although it improves soundness. This finding suggests that, to some degree, the beneficial effects of privatization in the short-run on profits and bank margins are, to an extent, reversed, as time passes, especially in case of provisioning norms. Finally, when we look at selection effects, we find several significant results. The first indicates that banks with low soundness were selected for privatization: the coefficient on SEL is uniformly negative in the soundness equation. The second indicates that for SOBs selected for privatization, macroprudential norms tend to exert the most pronounced impact on soundness, the magnitudes of which are roughly equal: while capital and loan classification norms lower soundness, provisioning norms tend to improve it. Third, macroprudential norms relating to loan classification norms tend to dampen credit growth for banks selected for privatization. And finally, capital and provisioning norms exert an opposite impact on profitability. Most relevant for our purpose is the fact that macroprudential norms tend to exert a discernible impact of performance. It might be argued that some of the control variables employed, such as non-interest income and demand deposits, might be endogenous. To circumvent this possibility, the baseline model is re-estimated after deleting these variables. The results are observed to be similar to those obtained earlier. In addition, acknowledging the importance of liquidity and capital in bank behavior, the model is augmented with measures which proxy for these factors. Our main results remain unaltered after inclusion of these variables. Another issue of relevance is that SOBs are much larger than other banks. This raises the question of whether differences in the effects of regulation between state-owned banks and other banks were driven by the ownership structure, or by economies of scale (especially since bank size was included as a control variable). To examine this further, we re-estimate the baseline specification, weighing each observation by the bank’s share of total assets (Levy-Yeyati and Micco, 2007 for the advantages of this approach). Our main results remain materially unchanged in this case. 6. Summary and conclusions Financial sector reforms in India, undertaken as part of the overall process of reforms since the early 1990s, were aimed at improving the efficiency and productivity of the financial sector. While there have been several studies on bank performance, these papers do not pay adequate attention to the important policy dimensions of prudential deregulation and their impact on bank performance. In this context, the present study employs panel data techniques to examine the impact of three important macroprudential measures - capital adequacy norms, provisioning requirements and tightening of loan classification norms - on the performance of Indian banks since the 1990. We focus on four major firm characteristics: profitability, margins, soundness and credit growth. The analysis indicates that the state-owned banks are less profitable than private banks and the difference in performance is accentuated after imposition of macroprudential norms. These results are quite robust. It is apparent in simple univariate comparisons as well as in multivariate regressions that takes on board several control variables. Summing up, the balance of evidence indicates that different measures of macroprudential regulation exert differential impact on banks across ownership. These divergences could, for example, be the outcome of differences in their business models, product sophistication, customer orientation, risk appetite as well as human and other infrastructural efficiencies. It, therefore, appears important for policymakers to take a holistic view of all prudential measures and their potential impact on the banking system in order to avoid possible pitfalls. Contextually, Rajan (2009) has argued that, in order to ensure that regulations are cycle-proof, it is important that they are premised on 3-Cs: comprehensive, contingent and cost-effective. In other words, by being applied comprehensively to all levered financial firms and being contingent on the overall state of the economy, they would discourage regulatory arbitrage and ensure cost-effectiveness and therefore, be less prone to dilution. @ Department of Economic and Policy Research, Reserve Bank of India, Fort, Mumbai 400001, India. Email. 1I would like to thank, without implicating three anonymous referees for their observations and comments on an earlier draft, which improved the exposition and analysis. Needless to state, the views expressed and the approach pursued reflects the author’s personal views. 2See, for example, Sinha (2011). 3The macroprudential instruments considered were capital requirements, limits on LTV, cap on debt-to-income and reserve requirements. 4In case of Z-score, the average number of observations per year is 20; therefore, the number of bank-years is 1280. For credit growth, owing to the calculation of growth rate, one year of observation is lost from the sample. We have a maximum of 1243 bank-years corresponding to 19.4 years of observation per bank. 5To moderate the influence of outliers, the credit growth variable is calculated as in Dinc (2005). 6Consistent with the literature, risk taking is measured by the Z-score, which is a proxy for insolvency risk (Boyd and Runkle, 1993; Laeven and Levine, 2009). A higher Z-score indicates that the bank is more stable. Because the Z-score is highly skewed, we employ its natural logarithm, which is normally distributed (Laeven and Levine, 2009). 7The correlation matrix of GDP growth and macroprudential measures indicate that these correlations are of low magnitude (statistical significance at conventional levels indicated by asterisk).
References Agoraki, M-E., M.D.Delis and F.Pasiouras (2011). Regulations, competition and bank risk-taking in transition economies. Journal of Financial Stability 7, 38-48. Allen, F., and D. Gale (2004). Competition and financial stability. Journal of Money, Credit and Banking 36, 453-80. Banerjee, A.V., S. Cole and E. Duflo (2003). Bank financing in India. India Policy Forum, Brookings Institution. pp.227-52. Beck, T., H. Hesse, T.Kick and N. van Westernhagen (2009). Bank ownership and stability: Evidence from Germany. Working Paper No. 22. Deutsche Bundesbank, Germany: Bonn. Berg, S. A., F. Forsund and E. Jansen (1992). Malmquist indices of productivity growth during the deregulation of Norwegian banking 1980–1989. Scandinavian Journal of Economics 94, 211–228 Berger, A.N., G. Clarke, R. Cull, L. Klapper, and G. Udell (2005). Corporate governance and bank performance: A joint analysis of the static, selection, and dynamic effects of domestic, foreign, and state ownership. Journal of Banking and Finance 29, 2179-2221. Berger, A.N., L.F.Klapper, M.S.Martinez Peria and R.Zaidi (2008). Bank ownership type and banking relationships. Journal of Financial Intermediation 17, 37-62. Besley, T., and R. Burgess (2004). Can labor regulations hinder economic performance? Evidence from India. Quarterly Journal of Economics 119, 91-134. Bhattacharyya, A., C.A.K. Lovell and P. Sahay (1997). The impact of liberalization on the productive efficiency of Indian commercial banks.European Journal Operational Research 98, 332–45 Boyd, J.H. and D.E. Runkle ;(1993). Size and performance of banking firms: Testing the predictions of theory.Journal of Monetary Economics 31, 47–67. Boyd, J.H. and G. De Nicolo (2005). The theory of bank risk taking and competition revisited.Journal of Finance 60, 1329–43. Boyd, J.H., G. De Nicolo and A.M.Jalal (2006). Bank risk taking and competition revisited: New theory and new evidence. IMF Working Paper 297. IMF: Washington DC. Casu, B., A.Ferrari and T. Zhao (2013). Regulatory reform and productivity change in Indian banking. Review of Economics and Statistics 95, 1066-77. Claessens, S. and L. Laeven (2004). What drives banking competition? Some international evidence. Journal of Money, Credit and Banking .3, 563-83. Cole. S. (2011). Fixing market failures or fixing elections: Agricultural credit in India. American Economic Journal: Applied Economics 1, 219-50. Cosimano. T., and D. Hakura (2011). Bank behaviour in response to Basel III: A cross country analysis. IMF Working Paper 119. IMF: Washington DC. Das, A., and S. Ghosh (2006). Financial deregulation and efficiency: An empirical analysis of Indian banks during the post reform period. Review of Financial Economics 15, 193-221. Das, A., and S. Ghosh (2007). Determinants of credit risk in Indian state-owned banks: An empirical investigation. Economic Issues 12, 27-46. Das, A., and S. Ghosh (2009). Financial deregulation and profit efficiency: A nonparametric analysis of Indian banks. Journal of Economics and Business 61, 509-28. Demirguc-Kunt, A. and H.Huizinga (2000). Financial structure and bank profitability. Working Paper 2430. The World Bank: Washington DC. Demirguc-Kunt, A., L. Laeven and R. Levine (2004). The impact of bank regulations, concentration and institutions on bank margins. Working Paper 3030. The World Bank: Washington DC. Diamond, D., and R.G.Rajan (2000). A theory of bank capital. Journal of Finance 55, 2431-65. Diamond, D., and R.G.Rajan (2001). Liquidity risk, liquidity creation and financial fragility. Journal of Political Economy 109, 287-327. Dinc, I.S. (2005). Politicians and banks: Political influences on government-owned banks in emerging markets. Journal of Financial Economics 77, 453-79. Ediz, S., I.Michael and W.Perraudin (1998). Bank capital dynamics and regulatory policy. In Financial Services at the Crossroads: Capital Regulation in the 21st Century. Federal Reserve Bank of New York Conference Volume. FRB New York; New York, pp.19-40. Elyasiani, E., and S. Mehdian (1995). The comparative efficiency performance of small and large US commercial banks in the pre- and post-deregulation era. Applied Economics 27, 1069–79. Feyzioglu, T., N.Porter and E.Takats (2009). Interest rate liberalization in China. IMF Working Paper 171. IMF: Washington DC. Ghosh, S., D.M.Nachane, A. Narain and S. Sahoo (2003). Capital requirements and bank behaviour: An empirical analysis of Indian public sector banks. Journal of International Development 15, 146-56. Gilbert, R. A., and Wilson, P. W. (1998). Effects of deregulation on the productivity of Korean banks. Journal of Economics and Business 2, 133-55. Gormley, T. (2010). The impact of foreign bank entry in emerging markets: Evidence from India, Journal of Financial Intermediation 19, 26-51. Government of India (1991). Report of the Committee on the Financial System, Government of India: New Delhi. Government of India (1998). Report of the Committee on Banking Sector Reforms, Government of India: New Delhi. Greene, W.H. (1993). Econometric Analysis (2nd Ed), Prentice Hall, Upper Saddle River, NJ. Grifell-Tatje, E., and C.A.K.Lovell (1996). Deregulation and productivity decline: The case of Spanish savings banks. European Economic Review 40, 1281-1303. Hellmann, T.F., K.Murdock and J.E.Stiglitz (2000). Liberalization, moral hazard and prudential regulation: Are capital requirements enough? American Economic Review 90, 146-65. Honohan, P. (2008). Cross country variation in household access to financial services. Journal of Banking and Finance 32, 2493-2500. Ianotta, G., G. Nocera and A. Sironi (2007). Ownership Structure, Risk and Performance in the European Banking Industry. Journal of Banking and Finance 31, pp. 2127-49. International Monetary Fund (2012). The Interaction of Monetary and Macroprudential Policies: Background Paper. IMF: Washington DC. Keeley, M. (1990). Deposit insurance, risk and market power in banking.American Economic Review 80,1183–1200. Kumbhakar, S. C. and S. Sarkar (2003). Deregulation, ownership and productivity growth in the banking industry: Evidence from India, Journal of Money, Credit, and Banking 35, 403-14. Kwan, S.H.(2002). Impact of deposit rate deregulation in Hong Kong on the market value of commercial banks. Journal of Banking and Finance 27, 2231-48. Laeven, L and R.Levine (2009).Bank governance, regulation and risk taking. Journal of Financial Economics 93, 259-75. Leightner, E.J and C.A.K. Lovell (1998).The impact of financial liberalization on the performance of Thai banks. Journal of Economics and Business 2, 115-32. Martinez Peria, M.S. and A. Mody (2004). How foreign participation and market concentration impact bank spreads: Evidence from Latin America. Journal of Money, Credit and Banking 36, 511-37. Matutes, C. and X.Vives (2000). Imperfect competition, risk taking and regulation in banking. European Economic Review 44, 1-34. Maudos, J., and J.M.Pastor (2001). Cost and profit efficiency in banking: An international comparison of Europe, Japan and USA. Applied Economics Letters 8, 383–87. Maudos, J., and L. Solis (2009). The determinants of net interest income in the Mexican banking system: An integrated model. Journal of Banking and Finance 33, 1920-31, Micco, A., U. Panizza and M. Yanez (2007). Bank ownership and performance: Does politics matter? Journal of Banking and Finance 31, 219-41 Murinde, V., and H.Yaseen (2006). The impact of the Basel accord regulations on bank capital and risk behaviour: 3D evidence from MENA region. Working Paper 22, University of Birmingham: UK. Prasad, A., and S. Ghosh (2005). Competition in Indian banking. IMF Working Paper 141. IMF: Washington DC. Rajan, R.G. (2009). The credit crisis and cycle-proof regulation. FRB St. Louis Economic Review, Sept/Oct, 397-402. Reserve Bank of India. Handbook of Statistics on Indian Economy (various years), RBI: Mumbai. Reserve Bank of India. Report on Trend and Progress of Banking in India (various years), RBI: Mumbai. Reserve Bank of India. Report on Currency and Finance (various years), RBI: Mumbai. Reserve Bank of India. Statistical Tables Relating to Banks in India (various years), RBI: Mumbai. Rime, B (2001). Capital requirements and bank behaviour: Empirical evidence for Switzerland. Journal of Banking and Finance 25, 789-805. Rodrik, D. (2012). Why we learn nothing from regressing economic growth on policies. Seoul Journal of Economics 25, 137-51. Salas, J. and V. Saurina (2003). Deregulation, market power and risk-taking in Spanish banks. European Economic Review 47, 1061-75. Shrieves, R. and D.Dahl (1992). The relationship between risk and capital in commercial banks. Journal of Banking and Finance 16, 439-57. Shyu, J. (1998). Deregulation and bank operating efficiency: An empirical study of Taiwan banks. Journal of Emerging Markets, 3, 27–46. Sinha, A. (2011). Macroprudential policies: Indian experience. Address delivered at Eleventh Annual International Seminar on Policy Challenges for the Financial Sector on “Seeing both the Forest and the Trees- Supervising Systemic Risk” co-hosted by The Board of Governors of the Federal Reserve System, International Monetary Fund, and The World Bank at Washington, D.C, June. Available at <www.rbi.org.in/speeches> Stolz, S. (2007). Bank capital and risk-taking. Springer-Verlag: New York. Subbarao, D. (2013). Banking structure in India: Looking Ahead by Looking Back. Speaking notes at the FICCI-IBA Annual Banking Conference in Mumbai, August. Van Roy, P. (2008).Capital requirements and bank behaviour in the early 1990s: Cross-country evidence. International Journal of Central Banking 4, 29-60. Wheelock, D.C. and P.W. Wilson (1999). Technical progress, inefficiency and productivity change in US banking. Journal of Money, Credit and Banking 31, 212–34. Yeyati, E.L., and A. Micco (2003). Banking competition in Latin America. First meeting of the Latin American competition forum, Paris, 7-8 April. Yeyati, E.L., and A. Micco (2007). Concentration and foreign penetration in Latin American banking sectors: Impact on competition and risk. Journal of Banking and Finance 31, 1633-47. Zhao, T., B. Casu and A. Ferrari (2010). The impact of regulatory reforms on cost structure, ownership and competition in Indian banking. Journal of Banking and Finance 34, 246–54. |