Anand Sinha: Macroprudential policies – Indian experience Address by Mr Anand Sinha, Deputy Governor of the Reserve Bank of India, at the 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, The International Monetary Fund, and The World Bank, Washington DC, 2 June 2011. * * * Inputs from Dr. D P Rath, Muneesh Kapur and Rajinder Kumar are gratefully acknowledged. Introduction 1. Good Morning. It gives me great pleasure to share with you our experience in implementing macroprudential policies in India. The current global financial crisis has brought to fore serious lacunae in the approach to regulation and supervision and put the issue of systemic risk on to the regulatory agenda. A comprehensive definition of systemic risk is, “The risk of disruptions to financial services that is caused by an impairment of all or parts of the financial system, and can have serious negative consequences for the real economy.”1 2. There are two facets to systemic risk. One is in terms of its distribution within the system at any given point in time and another is its evolution with time. The cross-sectional dimension is how risk is distributed within the system at any given point in time. Systemic risk in this dimension arises due to the inter-connectedness of institutions, balance sheet entanglements, common exposures and, sometimes, even common business models of financial institutions. The time dimension on the other hand deals with how aggregate risks in the financial system evolve over time – the procyclicality issues in the financial system. The dynamics of the financial system and the macroeconomy interact with each other increasing the amplitude of booms and busts. The larger is the boom, the larger is the bust and larger is the damage to the economy. Systemic risk management and macro prudential policy 3. The set of policies which deal with managing the downside of systemic risk is known as macro prudential policy. Macroprudential policies primarily use prudential tools to limit systemic risk and thereby minimize disruptions in the provision of key financial services that can have serious consequences for the economy by (i) dampening the buildup of financial imbalances; (ii) building defenses that contain the speed and sharpness of subsequent downswings and their effects on the economy; and (iii) identifying and addressing common exposures, risk concentrations, linkages and inter-dependencies that are sources of contagion and spillover risks that may jeopardize the functioning of the system as a whole2. While the third objective of macroprudential policy [(iii) above] is concerned with the cross- sectional dimension, the first two objectives [(i) and (ii) above] are concerned with the procyclicality issues. The second objective of building defenses, i.e., increasing the resilience of the financial system is viewed as a narrow objective and is attained by build-up of buffers during boom times which can be used when risks materialize during busts. The first objective of dampening the buildup of financial imbalances is considered a broader objective and is essentially “leaning against the wind” aspect during the boom phase for dampening the credit 1 IMF-BIS-FSB(2009). 2 IMF(2011). BIS central bankers’ speeches 1 and asset price boom. The buildup of buffers should achieve this objective by affecting the cost of credit, though evidence is not unequivocal in this regard. A more ambitious interpretation of the first objective would be moderation of credit supply through both booms and busts i.e. ensuring stable credit supply. While the objective of dampening the credit exuberance during boom and, thereby, moderate credit supply looks plausible, increasing credit supply during busts by leaning against the wind i.e. by releasing buffers, does not seem as plausible because of risk aversion that is likely to set in among banks and other economic agents as well as the market pressure and expectation from banks to maintain high levels of capital when risks are apparently highest. Thus macroprudential policy is likely to have asymmetric impact from “leaning against the wind” during booms and busts. BCBS and FSB are currently involved in developing a range of macroprudential policies to deal with the procyclicality issues as also with systemically important financial institutions and other aspects of systemic risk on account of inter-connectedness and common exposures. 4. Reserve Bank of India (RBI) has been using macroprudential polices, more notably the countercyclical policies, since 2004 as a toolkit for ensuring financial stability though it had used them sporadically even earlier. It would be useful to describe the broad structure of the Indian financial system and the linkages between the monetary policy and financial stability in India, to provide a backdrop for discussing the implementation of polices. Structure of the Indian financial system 5. The Indian financial system is heavily dominated by commercial banks. Within the banking system, public sector banks (majority shareholding held by the Government of India) account for nearly 70 per cent of the banking system assets. 6. RBI regulates banking sector, non-banking financial companies (NBFCs), as also the money, forex and Government securities markets which are dominated by banks. Thus, the interconnectedness channels, both from the institutional and market perspectives, come within the regulatory ambit of RBI. There are separate regulators for capital markets, insurance sector and pension funds. With many Indian banks having expanded into the above mentioned activities through subsidiaries, associates or otherwise, there has been a need for coordination among sectoral regulators which has been ensured through inter- regulatory bodies within the umbrella of a high level committee chaired by the Governor, RBI and with representatives from Ministry of Finance. This institutional arrangement has recently undergone a change with the establishment of the Financial Sector Development Council (FSDC) chaired by the Finance Minister. Role of RBI in maintaining financial stability 7. The Reserve Bank of India Act, 1934 provides a broad legal mandate to RBI to secure monetary stability and generally to operate the currency and credit system of the country to its advantage. In practice this meant the dual objective of growth and price stability, the relative emphasis being dependent on the context. Since 2004, RBI has added financial stability as an additional objective in view of the fast growing size and importance of the Indian financial sector3. It is in this setting that RBI has been using macroprudential framework in both time and cross-sectional dimensions for quite long without christening these policies as macroprudential policies as is the case with some other countries, notably some Asian countries. Operationally, while pursuing multiple objectives, multiple indicators, including growth in credit and money, are used to track the macroeconomic conditions. India being a bank-dominated economy, the bank credit becomes a key monetary policy 3 Y.V.Reddy (2011). 2 BIS central bankers’ speeches transmission channel. Thus, the aggregate bank credit growth has always formed an important variable in the conduct of monetary and countercyclical policies. Elements of macro-prudential framework in India Overview 8. RBI’s countercyclical policies have focused on banks due to the centrality and criticality of the banking system in the Indian economy. In any case, application of counter- cyclical policies to the shadow banking system i.e. the Non-Banking Financial Companies (NBFCs) is extremely challenging. These policies have aimed at increasing the resilience of the banking system. The instruments used have been time varying risk weights and provisioning norms on standard assets for certain specific sectors wherein excessive credit growth, in conjunction with sharp rise in asset prices, has caused apprehension of potential build-up of systemic risk and asset bubbles. In the process, the policies have “leaned” against the wind and have had the desired effect of moderating the credit boom in the specified sectors both through signaling effect and affecting the cost of credit. Evidence, though limited, suggests that the leaning against the wind has been more effective in dampening the lending exuberance in the boom phase than in the downturn in ensuring a stable credit supply. Several measures have been taken to reduce the inter-connectedness among banks on the one hand and between banks and NBFCs on the other, and limits have been placed on common exposures to address the cross-sectional dimension of systemic risk. Objective of counter-cyclical policies 9. The objective of these policies is best stated in the words of Dr. Y.V Reddy, former Governor RBI. “The RBI articulated its approach to countercyclicality in its policies by indicating the criticality of the banking system for large segments of the population and for the economy as a whole. Hence the RBI adopted a precautionary approach to essentially protect the banking system from a “bust” were it to occur for any reason.”4 This was amplified in October 2005 in the Mid-term Review of Annual Policy for the year 2005–06, while increasing provisioning on standard assets across the board (except for SMEs and agriculture): “Traditionally, banks’ loans and advances portfolio is pro-cyclical and tends to grow faster during an expansionary phase and grows slowly during a recessionary phase. During times of expansion and accelerated credit growth, there is a tendency to underestimate the level of inherent risk and the converse holds good during times of recession. This tendency is not effectively addressed by the prudential specific provisioning requirements for the impaired assets since they capture risk ex post but not ex ante. The various options available for reducing the element of pro-cyclicality including, among others, adoption of objective methodologies for dynamic provisioning requirements, as is being done by a few countries, by estimating the requirements over a business cycle rather than a year on the basis of the riskiness of the assets, establishment of a linkage between the prudential capital requirements and through-the-cycle ratings instead of point-in-time ratings and establishment of a flexible loan-to-value (LTV) ratio requirements where the LTV ratio would be directly related to the movement of asset values.” 4 Y. V. Reddy (2011). BIS central bankers’ speeches 3 10. It is apparent that the policy does not specifically mention ensuring stable credit supply though this would certainly be a collateral objective. It is also notable that while other available options i.e., dynamic provisioning, through the cycle ratings for capital purposes and time varying LTV ratios have been discussed, the policy prescription preferred was increasing provisions on standard assets which is somewhat akin to dynamic provisioning though not exactly similar. Moreover, RBI never used cap on LTV ratios till much later in 2010 but in a different context. Methodology 11. Ideally, a sound macro-prudential policy should be based on the determination of the economic cycles, assessment and measurement of the build-up of systemic risk and also the effect of the stance of other public policies like monetary policy, fiscal policy etc., on the risk taking behavior of the financial sector. Since the development of a framework is in infancy, RBI’s methodology has not been based on extensive statistical analysis or modeling or on determination of build-up of asset bubbles. It is largely judgmental based on trends in aggregate credit and sectoral credit growth in the macro-economic settings. For this reason, it has not been rule bound which will require either some model or at least some measurement of systemic risk and its sensitivity to the prudential parameters. While undertaking counter-cyclical measures during the high GDP and high credit growth period of 2004–08, there was no explicit attempt to determine the deviation of the credit to GDP ratio from its long term trend, though the GDP growth and the macro-economic setting were kept in view. Similarly, the possibility and not the absolute proof of asset bubbles was explored in terms of broad indicators and possible threats5. Some evidence from Annual Financial Inspections of banks carried out by RBI, together with market intelligence on possible loosening of underwriting standards due to aggressive lending, was also factored in. Dimensions of RBI’s macro-prudential policies Counter-cyclical policies 12. Implementation of countercyclical capital and provisioning regulations in India during the period from December 2004 to December 2010 is reflected in Table 1 below. 5 Y. V. Reddy (2011). 4 BIS central bankers’ speeches 13. Table 2 indicates the movement in monetary measures as well as the movement in the provisioning norms and risk weights of the specific sectors during the three phases of implementation of the countercyclical policies. 14. The period covered in Table 2 is divided into three distinct phases from monetary policy perspective which correspond to three phases from countercyclical policy perspective. The monetary tightening and easing phase corresponds respectively to increase in sectoral capital and provisioning requirements (build up phase) and easing of these requirements (release phase). The period wise classification from this perspective is: – (i) Build-up phase: September 2004–August 2008, (ii) Release Phase: October 2008–April 2009, and BIS central bankers’ speeches 5 (iii) Re-build-up phase: October 2009 till date. It may be noted that the monetary and countercyclical measures have always been in the same direction i.e. have been complementary so far. Build-up phase: (September 2004–August 2008) 15. During 2004–08, the Indian economy exhibited high real GDP growth, of around 9 per cent per annum. Given the high growth and inflationary pressures, monetary policy was in a tightening mode to contain aggregate demand and inflation. During this period, India also received large capital flows, which were intermediated by the banking sector. High growth created a huge demand for bank credit. While the overall bank credit growth accelerated sharply (to over 30 per cent), credit growth to certain sectors such as real estate accelerated much more sharply (reaching more than 100 per cent, year-on-year, for an extended period April 2005–July 2006 and remained above 50 per cent till later than mid-2007). Concomitantly, asset prices, especially those of real estate, rose sharply. This exposed the banking sector to huge risks. In view of the rapid credit expansion in the period 2003–06, it was explicitly indicated by the RBI in April 2006 that growth of non-food bank credit, including investments in bonds/debentures/shares of public sector undertakings and private corporate sector and commercial paper, would be calibrated to decelerate to around 20 per cent during 2006–07 from a growth of above 30 per cent. Inflationary expectations had also started firming up and as a part of monetary management, the repo rate was increased by 300 basis points in stages to 9 per cent by August 2008 from its level of 6 per cent in September 2004. Further, the Cash Reserve Ratio was also raised by 450 basis points in stages from 4.5 per cent in September 2004 to 9 per cent. In order to protect banks’ balance sheets against such risks, the Reserve Bank tightened prudential norms in the form of provisioning norms and risk weights in specific sectors beginning October 2004 (Table 1). 16. Noticing the steep increase in bank credit to the commercial real estate sector in conjunction with that in the prices of real estate, risk weights for banks’ exposure to commercial real estate were increased from 100 per cent to 125 per cent in July 2005, and further to 150 percent in May 2006. The risk weights on housing loans extended by banks to individuals, were increased from 50 to 75 per cent in December 2004. Subsequently, while the risk weights on smaller size housing loans (priority sector) were reduced from 75 to 50 per cent, the risk weights on larger loans and those with LTV ratio exceeding 75 per cent were increased to 100 per cent. When there was a boom in consumer credit and equities, risk weights for consumer credit and capital market exposures were increased from 100 percent to 125 per cent. The provisions for standard assets were revised upwards progressively in November 2005, May 2006 and January 2007, in view of the continued high credit growth in the real estate sector, personal loans, credit cards receivables, loans and advances qualifying as capital market exposures and loans and advances to the NBFCs. The provisioning requirement for all other loans and advances classified as standard assets, namely, direct advances to the agricultural and small and medium enterprise sectors and all other loans and advances were kept unchanged. 17. The tightening of prudential norms made the credit to targeted sectors costlier thereby moderating the flow of credit to these sectors. There is evidence that moderation in credit flow to these sectors was also in part due to banks becoming cautious in lending to these sectors on the signaling effect of RBI’s perception of build up of sectoral risks. For instance, these measures helped moderate the flow of credit to the commercial real estate sector. The credit growth decelerated to around 50 per cent by 2008 from a very high level of around 150 per cent (Y-o-Y basis) in late 2005 as shown in the graph below. 6 BIS central bankers’ speeches 18. Simultaneously, as indicated earlier, monetary policy was also in a tightening mode to contain demand pressures. Thus, while monetary tightening helped in containing the overall credit growth, prudential norms moderated the credit growth to the specific sectors. Thus, monetary policy and prudential norms complemented each other. That is, we deployed both interest rates and prudential instruments during 2004–08 to ensure both price stability and financial stability. Release phase: 2008–09 (October 2008 to April 2009) 19. The Indian economy was also impacted by the global financial crisis, though a major part of the impact was felt indirectly through channels of trade and cross-border capital flows. In order to mitigate the adverse impact of the global financial crisis on the Indian economy, the Reserve Bank aggressively eased the monetary policy. During this period, prudential norms were also relaxed in a countercyclical fashion, again mainly following a sectoral approach (Tables 1 and 2). The relaxations focused primarily on real estate and NBFC sector as these were the segments which had been most severely hit due to the downturn. In addition to easing of risk weights and provisioning norms for standard assets, RBI’s prudential framework governing restructuring of advances (corporate workouts) was also temporarily modified to facilitate restructuring of greater number of units which had potential viability but had been affected in a most unexpected manner. However, despite the easing of monetary policy and aggressive relaxation in prudential measures in a countercyclical fashion, the credit growth slowed down substantially due to, among other reasons, subdued credit demand and risk aversion among banks as is clear from Table 3. BIS central bankers’ speeches 7 Table 3 Credit to select sectors during September 2008 to September 2009 (Per cent) Source: Reserve Bank of India. 20. The credit growth, overall as well as to the target sectors, decelerated. During October 2008 to September 2009, the credit growth to commercial real estate (CRE) decelerated from 45 per cent to 34 per cent on a year-on-year basis, to NBFCs from 61 per cent to 30 per cent, and to housing from 11 per cent to 6 per cent (Table 3). The total non- food credit during the same period decelerated from 29 per cent to 13 per cent. Re-build phase: October 2009 onwards 21. By late 2009, domestic growth began to recover from the slowdown induced by the global financial crisis. However, while the overall credit growth continued to remain subdued, credit growth to the commercial real estate sector remained high well above the overall credit growth (Table 4). As indicated in para 19, in the wake of the global financial crisis, the Reserve Bank had temporarily modified its prudential guidelines for restructuring of advances. However, the extent of restructured advances to the commercial real estate sector was relatively high. Accordingly, provisioning norms on standard assets were increased for the commercial real estate sector in November 2009. This was the period when the Reserve Bank began exiting from crisis-driven expansionary monetary policy as India was confronted with an upturn in inflation – a rising wholesale price index (WPI) inflation and stubbornly 8 BIS central bankers’ speeches elevated consumer price index (CPI) inflation. The exit began by reversing the immediately reversible unconventional measures such as, restoring of export credit refinance facility to pre-crisis level, and discontinuation of special refinancing facilities extended to scheduled commercial banks, etc. Table 4 Deployment of gross bank credit by major sectors Source: Reserve Bank of India. 22. In December 2009, as the economy had just emerged from the crisis, there were apprehensions about asset quality on account of exuberant lending during the boom phase. Since banks were still making good profits, it was decided to prescribe a Provisioning Coverage Ratio (PCR) of 70 percent of gross non-performing advances, as a macro- prudential measure, with a view to augmenting provisioning buffer in a counter-cyclical manner. Banks had to achieve this by September 2010. 23. PCR was intended to be an interim measure and it was hoped that it would be replaced by a forward-looking counter-cyclical provisioning methodology being developed by the Basel Committee on Banking Supervision (BCBS) and International Accounting Standards Board (IASB) or by a methodology similar to Spanish dynamic provisioning framework we are working on. Since in the absence of a calibrated methodology it would be difficult to allow banks to use the countercyclical provisions built up under PCR freely and there were certain design issues too, it was decided to freeze the PCR with reference to the gross NPA position in banks as on September 30, 2010. The buffer (surplus of provisions over specific provisions) will be allowed to be used by banks for making specific provisions for NPAs during periods of system wide downturn, with the prior approval of RBI. It may be added that the banking system has already exceeded the 70 per cent PCR though some banks have yet to reach that level. BIS central bankers’ speeches 9 24. By November 2010, the Reserve Bank had raised the policy rate by 150–200 basis points on account of concerns about high and stubborn inflation. During this period, residential property prices had risen sharply and had attained the pre-crisis level. Some banks had come out with certain residential housing loan schemes (teaser loans and 10:90 scheme) where RBI had apprehensions about asset quality going forward. Moreover, it was felt that these schemes were creating artificial demand for housing loans which could push up the housing prices further (which had already reached pre-crisis levels) with the potential of putting housing beyond the reach of many. In November 2010, therefore, the Reserve Bank initiated the following measures: (i) the provisioning norm for “standard” teaser housing loans was increased from 0.4 per cent to 2.0 per cent in November 2010; (ii) for the first time a cap on Loan To Value (LTV) ratio was prescribed at 80 per cent in December 2010 for loans above Rs.2 million and at 90 per cent for loans up to Rs.2 million; and (iii) the risk weight for residential housing loans of Rs.7.5 million and above, irrespective of the LTV ratio, was raised to 125 per cent. These measures are, however, intended to serve more a micro- prudential rather than a macro-prudential objective. 25. Concluding observations on countercyclical policies (i) View regarding the implementation of countercyclical policies was based on tracking of various indicators in the economy, notably the general credit growth and the sectoral credit growth. This was complemented with market intelligence and some feedback from the Annual Financial Inspections of banks. No detailed statistical analysis or modelling was used. The decisions were judgmental based on constant monitoring of macroeconomy and were not rule based. (ii) RBI, being the monetary authority as well as the regulator and supervisor of banks, NBFCs and important segments of markets i.e. forex, Govt. Securities and money markets, had the necessary information and overall view of the risks building up in the system. It was, therefore, well placed to operate the countercyclical policies. (iii) Monetary policy and the countercyclical policy were in the same direction (Table 2). Such a coordinated response was facilitated due to RBI’s wide regulatory ambit. If policies are not well coordinated, the costs of implementing such policies may be high. (iv) It was important to deal with sectoral exuberance through countercyclical policies even as monetary policy, while dealing with inflation scenario, dealt with generalised exuberance. Interest rate alone, being a blunt instrument, would not have been able to handle the sectoral exuberance, or else, the cost to the economy would have been higher. (v) Combination of risk weights and provisioning requirements for standard assets were used as countercyclical policies. It would appear, however, that varying the provisioning requirements may have been more effective than varying risk weights in moderating credit flow to the specific sectors. This is because, since the average capital adequacy ratio of banks operating in India has been well above 12 per cent for the last many years (as on December 2010, it was above 14 per cent), risk weights may not always be effective in dampening the growth of credit as banks can continue to finance riskier sectors yielding higher returns by allowing their capital adequacy ratios to fall by a few basis points and still remain much above the regulatory requirements. To the extent higher risk weights translate into increase in interest rates, demand for credit may come down. On the other hand, varying provisioning requirement would be potentially more effective as it would impact the Profit and Loss account of banks to which banks are more sensitive. (vi) The countercyclical policies were able to dampen exuberant credit growth in the targeted sectors. However, their effect was asymmetrical during downturn. Despite aggressive easing of monetary policy and prudential measures in a countercyclical fashion, the credit supply did not increase adequately. The credit growth slowed 10 BIS central bankers’ speeches