Presentation

Shri K.R. Kamath, Chairman, IBA and Chairman, Punjab National Bank; Smt. V.R. Iyer, Chairperson, Bank of India; CMDs of banks, delegates to the BANCON 2013; members of the print and electronic media; my other colleagues from the banking fraternity; ladies and gentlemen! At the outset let me thank the Indian Banks’ Association and the Bank of India for giving me this opportunity to share my thoughts with you from the platform of BANCON. This annual event is the most awaited event on the calendar of bankers as it provides them an opportunity for a free and frank deliberation on contemporaneous issues that the banking community is facing and what the future portends for them. In that sense, the theme chosen for this year’s BANCON - “Bank of the future, gearing up to meet the emerging environment” aptly sumps up the purpose of organising this event. I observe that the ‘who is who’ from the world of banking and finance have spoken on this theme and all of you have a pretty clear vision of what the banks of the future should look like and what you need to do to meet the emerging challenges.

All of you who know me closely know that, very often, I hold a contrarian view of things. At the present juncture, as a bank supervisor, I am much more concerned about the “future of the banks” than the “banks of the future”. Arguably, credit administration and the asset quality of banks is the most significant factor that affects the future of the banks and in my address today, I intend to highlight the concerns surrounding these areas. The purpose of my address today is not to find fault but to encourage introspection and self-criticism, to identify where we have gone wrong in the past and what corrective steps need to be initiated to undo some of the past mistakes and steps that need to be taken to move ahead. To this end, I intend to begin with looking back at historic trends in asset quality and credit management of banks and then use the ‘learnings’ to suggest a comprehensive solution to tackle the issues.

Introduction

1. In a bank dominated economy such as India, the asset quality of the banking system has important implications for the stability of the overall financial system. The general perception about a bank’s health is greatly determined by the level of non-performing advances (NPAs) held in its books. It is, therefore, not surprising that the current spurt in NPAs in banks has attracted a lot of attention.

2. The business of banking essentially involves intermediation – acceptance of deposits and channelling those deposits into lending activities - and credit risk is a direct fallout of this intermediation process. Certain amount of default and impairment of assets are likely to show up in the normal course of banking business and hence, credit risk management assumes a critical role in ensuring that such impairment is contained to a minimum.

3. Needless to say, when banking was simple, credit risk management was also straightforward. Lending decisions were made on impressionistic basis as the banks knew their borrowers and their businesses quite closely and hence they did not appreciate a need to collect and process elaborate information/data for supporting their credit decision making framework. Over time, as banking activities diversified and became more complex and the products became more sophisticated, risks also increased and became more complex. Although, the risks from intermediation became more transmitive and contagious, the evolution in the credit risk management failed to keep pace. The advanced credit risk management necessitated a granular analysis of the risks that the banks were being exposed to; however, they failed to appreciate these requirements and relied on a primitive management information system. To my mind, the failure of the banks to collect and analyse granular data/information on various elements of credit risk is one of the major reasons why the banks failed to foresee the impending problems. This is one theme that would continuously resonate all through my address today.

4. Against this backdrop, let me present my analysis of how Indian banks have dealt with credit risk over the last two decades, the present state of asset quality of Indian banks and the various factors that have contributed to the present situation. The analysis is based on a detailed study of data/information compiled by RBI from the banks. I would like to thank the banks and their data management teams for providing us extensive data and working closely with my team for facilitating the analysis that I have presented before you today.

5. Credit risk management is interlinked with the regulatory framework and I shall attempt to explore how the regulation at various stages has contributed to the present situation, how market forces have captured the regulation and exacerbated the problem. I would also try to dispel some common myths surrounding the lending to priority sector, small and micro enterprises by providing hard facts. Finally, I will list out the way forward for the regulators, policy makers, banks and bank customers.

Evolution of NPA regulation in India

6. The NPA trends, if studied carefully, can be observed to mirror the changing regulatory landscape. Let me first trace briefly the evolution of NPA regulation in India.

7. Until mid-eighties, the management of NPAs in India was left to the banks and the auditors. As the need for fine tuning regulatory structures to deal with the changing risk-profile of banking was felt, in 1985, the first-ever system of classification of assets for the Indian banking system was introduced. This system, called the ‘Health Code’ system, involved classification of advances into eight categories ranging from 1 (Satisfactory) to 8 (Bad and Doubtful Debts). A significant change in this evolution process in regulatory instructions, however, came in April 1992 with the introduction of prudential norms on income recognition, asset classification and mathematical methods for the computation of provisioning requirements. A graded norm for NPA recognition was brought-in, beginning with a four quarter norm for classification of advances as non-performing. With the introduction of 90-day norm for classification of NPAs in 2001, the NPA guidelines were brought at par with international standards.

8. Even as the NPA classification norms were being gradually tightened to bring them at par with international standards, RBI also introduced guidelines on “restructuring of advances” during the early 1990s. The guidelines required that standard assets, where the terms of the loan agreement regarding interest and principal had been renegotiated or rescheduled after commencement of production, be classified as sub-standard. In 2001, the instructions were further strengthened to clarify the asset classification treatment of restructured accounts prior to commencement of production as well.

9. The classification of advances as per the newly introduced “prudential norms” enabled a proper assessment of the extent level of non-performing assets in the Indian banking system for the first time. The initial figures for the NPAs in the system were quite high and hence, created sufficient incentive for the regulators and the banks alike to bring them down to manageable levels. Over time, as the banks introduced improvements in credit risk management systems and processes, the headline NPA ratios declined appreciably (Table 1). Empirical evidence, thus, indicates that increased stringency in regulation facilitated reduction of NPAs.

10. After the early 2000s, the pace of introduction/ tightening of regulatory reforms slowed. Regulatory norms were not further tightened during the “good” pre-crisis years which resulted in lowered credit standards and increased delinquencies. On the contrary, the norms were relaxed in some instances post the crisis. Some instances of flip-flop in regulation were also witnessed e.g. changes in norms with regard to usage of floating provisions, relaxation in norms for provisioning coverage ratio, delay in the introduction of dynamic provisioning coverage. Further, the 2008 special dispensation that permitted restructured accounts to be classified as standard has led to skewed incentives for banks/ borrowers to resort to restructuring to classify assets, even the un-viable ones, in the standard category and report lower NPA levels.

11. Generally, there is a tendency to ask for relaxation in NPA regulations whenever NPA levels rise. However, the empirical evidence suggests the such relaxations lead to the opposite result. Past experience highlights that stringency in regulation pre-empts any slackening in credit appraisal and monitoring standards by the banks and forces them to be on a constant vigil against impending delinquencies in their asset portfolio. I would mention that the banks should, on the contrary, welcome stricter prudential norms as they aim to become banks of the future.

Trends in asset quality

(i) Trends in NPA ratios

Let us me now turn to the trends in the asset quality of the banking sector. As I mentioned earlier, the introduction of prudential norms resulted in a significant rise in the NPA levels of banks during the early 1990s. Gross NPAs as a percentage of gross advances stood at 19.1 per cent as on March 1994. Thereafter, the ratio progressively declined during the period when the structural reforms were implemented and also through the “boom” years leading up to the global financial crisis. The historical trend in Gross and Net NPA ratios is indicated in Annex 1. The reduction in NPA levels could be attributed to several factors, including:

• Introduction of prudential norms for asset quality and other regulatory initiatives in the 1990s encouraged improved risk management in banks contributing to improvement in asset quality;

• As interest rates were falling, banks garnered substantial treasury profits which were utilized for writing off NPA accounts;

• Overall good performance of the economy and concomitant rise in credit growth;

• Abundant liquidity conditions;

• Increased restructuring, etc.

12. Post the onset of the Global Financial Crisis in 2008, the NPA ratios started increasing, indicating a marked deterioration in asset quality of the banking system. A closer scrutiny of the asset quality, however, reveals considerable divergence between the performances of various bank groups.

13. Initially, the NPA levels of public sector banks (PSBs) and other bank groups displayed a divergent trend. While the Gross NPA (GNPA) ratio for PSBs stood at 21 per cent in 1994; for new private sector banks and foreign banks, the same was between 1 to 2 per cent. Till 2003, while the GNPA ratio for PSBs declined gradually, it increased for other bank groups. During 2003-06, the NPA ratios across all bank groups showed a secular declining trend. Early on in the crisis (during 2007-09), the NPA ratios decoupled - the GNPA ratio of new private sector banks and foreign banks increased sharply while they continued to decline in case of PSBs. In fact, foreign banks witnessed the highest spurt in NPAs during 2009. The trend, however, reversed after 2009, when NPAs rose significantly for PSBs, while it declined for other bank groups.

14. This divergent trend clearly indicates that the ability to manage asset quality across banks varies markedly and, in the post crisis years in particular, the concerns on asset quality are largely confined to the PSBs.

15. PSBs share a disproportionate and increasing burden in case of NPAs among the bank groups (i.e. share in gross NPAs as compared to share in advances). The share of PSBs in gross NPAs has increased over the last decade and particularly since 2009 (Table 2). NPAs accounted for 85 per cent of the NPAs of the banking system in 2013 as compared to 75 per cent in 2003. During this period, the PSB’s share in total bank credit increased only marginally, from 74 per cent to 76 per cent. This is in sharp contrast to the performance of the other segments of the banking system, especially the new private sector banks, whose share in NPAs has fallen from over 14 per cent in 2003 to 8 per cent in 2013.

(ii) Beyond the Headline NPA numbers

16. Any analysis based on the headline NPA numbers i.e., the gross and net NPA ratios, have limitations. These ratios do not reveal the actual dimension of problems of asset quality of banks. A 360-degree view of trends in asset quality requires the capture and analysis of more granular data – activity and segment wise - about the various aspects of NPA management viz., slippages, write offs, recoveries, upgradation, restructuring, etc.

17. Till late 90s, the banks captured only basic data about asset quality i.e. gross and net NPA ratios. It is only following the regulatory nudge that the banks started collecting and reporting data on flow of NPAs (fresh accretions and recoveries) from 2001 onwards. Further, it is only recently i.e. since 2009, that collection of more granular data on asset quality like segmental NPAs, write-offs, recoveries etc., has started, primarily due to regulatory impetus. Unfortunately, this initiative in banks was limited to collecting this data for the purpose of reporting to the regulator. Few efforts were made to analyse and use this data internally for improving the quality of internal credit management.

18. Over the next few minutes, I will present an analysis of trends in asset quality and credit management of banks based on this granular data. I will then attempt to draw some inferences from this analysis about the quality of credit management in banks and how it has contributed to the recent trends in asset quality.

(iii) Trends in Slippages and Recovery

19. The flow of NPAs over the last decade shows that while the reduction in NPAs was to the tune of Rs. 4.9 trillion between 2001 and 2013, the accretions to NPAs were close to Rs. 6.3 trillion. If the entire period is split into two - 2001-2007 and 2008-2013, some interesting differences emerge. While during the first period, accretions to NPAs and reduction in NPAs are largely matched (Rs. 1.6 trillion of accretions as compared to about Rs. 1.7 trillion of reductions), during the later period the position altered dramatically with accretions to NPAs far exceeding the reductions (about Rs. 4.7 trillion of accretions as compared to Rs. 3.2 trillion of reductions) (Table 3).

20. Slippages, i.e. fresh accretion to NPAs during the year, provide a better metric to assess the credit management system in banks. While the slippages reduced during early 2000s, they started rising significantly since 2006-07. Growth rate of slippages which was negative till 2005-06, turned sharply positive in 2006-07 reaching a peak of 51.6 per cent in 2011-12. The net slippages (slippages net of recovery during the year) also showed the same trend. Contrary to the popular notion that the rising NPAs are a fallout of the global Financial crisis, the data suggests that the credit administration in the banks had started weakening and the asset quality had started deteriorating even before the onset of the crisis. Further, slippages exceeded reduction in NPAs, especially post crisis, as the ratio of reduction in NPAs to slippages fell dramatically – from about 105.3 per cent in 2001-07 to 70.8 per cent in 2007-13. Year wise figures of slippages and reduction from 2001 is indicated in Annex 2.

21. Simultaneously with increased slippages, recovery efforts of the banks suffered compounding the asset quality concerns as evidenced by the increasing trend in the ratio of slippages to recovery and upgradation since 2006-07 and the relative large share of write offs in total reduction in NPAs of banks.The ratio of slippages to recovery and upgradation represents the extent to which banks have been able to reduce their NPAs through recovery efforts. The ratio for the banking sector as a whole deteriorated from a low of 125.4 per cent in 2005-06 to 264.1 per cent during 2009-10 and remained elevated at 257.0 per cent in 2012-13. Recovery performance also varied widely across banks. This is evident from the divergent slippage and net slippage ratio in different bank groups indicated in Annex 3.

(iv) Trends in Recovery & Write-offs

22. There is evidence of increased use of write offs by banks to reduce NPAs, which is a pointer to weaknesses in credit management. Write offs were initially introduced as a tool for banks to manage their tax liabilities on impaired assets. However, they subsequently emerged as a tool for banks to manage their reported gross NPA numbers. Write offs, in fact, contributed significantly to the reduction in the quantum of gross NPAs (in some years, write offs accounted for nearly 50 per cent of reduction) as compared to actual recoveries and upgradations. Write offs as a percentage of terminal reduction (reduction on account of write-offs and actual recovery alone) has consistently been above 50 per cent mark (Annex 4).

23. These practices clearly engender moral hazard issues as they reduce the banks’ drive to improve recovery efforts. They also result in leakages in the recovery process. This is evidenced by the fact that, on an average, less than 10 per cent of the total amount written off (including the technical write off) is recovered (Table 4).

24. Further bearing testimony to the poor recovery efforts of the banking system are the trends in the ratio of upgradation to slippages. Even in the ‘good’ times, the ratio was never more than 20 per cent – a clear sign of poor standards of credit and recovery administration as well as a certain amount of apathy on the part of banks in expending efforts to revive accounts. The spirit of good credit management is to revive genuine problem accounts and not to retain them as NPAs for an eventual write- off.

25. Masked in the overall recovery statistics that I have so far discussed, are considerable divergences in trends across different bank groups. The new private sector banks and foreign banks recorded higher slippage ratio (gross and net)² immediately after crisis but were able to arrest the increasing trend in slippages through focused attention on credit risk management including exit strategies. In recent years, the ratio has risen sharply for PSBs. This indicates that new private sector banks and foreign banks were able to manage their asset quality better than PSBs as they were quick in identifying NPAs, while PSBs resorted to retrospective restructuring to report lower NPAs initially. This practice eventually tipped the scale against the PSBs and thus, in hindsight, regulatory guidelines also contributed to the decline in the asset quality of the PSBs.

26. The bank group wise trends in slippages are further re-enforced when the trends in slippages and fresh restructuring are examined. The ratio of slippages and fresh restructuring to advances rose sharply for the PSBs post crisis – from 5.2 per cent in March 2009 to 7.1 per cent in March 2013. The ratio reduced for foreign banks and new private sector banks and stood at a much more robust 1.8 per cent in March 2013.

27. It is, thus, clear that the weaknesses in credit and recovery administration that existed prior to the crisis, especially in the case of PSBs, were not dealt with in a timely manner. The crisis only exacerbated the problem.

28. It is also evident from the above trends that the gross NPA numbers, by themselves, are not a major cause for concern. Viewed in conjunction with the trends in restructuring of advances, however, there are evident and growing concerns about the asset quality of banks in India. In order to better understand the underlying issues, let me now turn to an analysis of the trends in restructuring of advances.

(v) Trends in restructuring

31. Restructured accounts have garnered wide attention in recent times as the quantum of restructured accounts has increased sharply post-crisis. Concerns have emerged from various quarters about the quantum of “forborne” assets and their potential impact on the asset quality of banks. If we look back to the early part of the 2000s, the growth rate in restructured accounts has showed a mixed trend. From 2003 to 2008, the restructured advances ratio was much lower than the NPA ratio. There was, however, a sharp uptick in growth rate in 2008-09 due to the one time dispensation with regard to asset classification granted to borrowers by the Reserve Bank with the onset of the crisis. Thereafter, the growth rate of restructured accounts has remained relatively high with the ratio of restructured accounts to standard advances showing a secular increase and the restructured advances ratio remained higher than the GNPA ratio. As on March 31 2013, the reported GNPA ratio for the banking system was 3.4 per cent. However, the ratio of gross NPAs and restructured accounts to advances was much higher at 9.2 per cent.

32. It is, indeed, appreciated that difficult times warrant some degree of forbearance, some hand-holding to performing and other-wise viable units in tiding over temporary difficulties. There is, however, sufficient evidence, anecdotal and otherwise, that restructuring is often used by banks for ‘evergreening’ problem accounts to keep the reported NPA levels low. Additionally, there are many instances of:

• unviable units taking recourse to restructuring leading to a situation where deserving and viable units get overlooked, especially in the small-ticket segment; and

• promoters not bringing in/ or being compelled to bring in sufficient equity in the project especially in the restructured accounts, which reduces their downside risks in the project.

33. As in the case of the overall NPA ratios and recovery performance, there are significant bank group wise differences in restructuring. Share of different bank groups in gross advances, NPAs and the restructured standard advances of the banking system since 2002 is given in Annex 5. In fact, the divergence in the asset quality of different bank groups is even more pronounced if restructured standard assets and cumulative write offs are considered along with NPAs, than if gross NPAs are looked at in isolation. To further analyse this trend, I would like to use a ratio namely “impaired assets ratio,” which is a ratio of gross NPAs, restructured accounts and cumulative write offs to total advances and which, I believe, is a more robust indicator of asset quality of banks. Between 2009 and 2013, this impaired assets ratio rose sharply from 6.8 per cent to 12.1 per cent in the case of PSBs. In contrast, the ratio fell for new private sector banks and foreign banks and stood at 5.3 per cent and 6.4 per cent respectively in March 2013 (Table 5).

34. The relative share of gross NPAs, restructured accounts and cumulative write offs in total impaired assets of different bank segments (Annex 6) further re-emphasizes the point I had made earlier – that banks which are quick in identifying and recognising NPAs are able to better manage their impaired assets. In March 2013, gross NPAs constituted 29 per cent of impaired assets in the case of PSBs while restructured accounts and write offs constituted 52 per cent and 20 per cent, respectively. In contrast, the shares stood at 36 per cent, 22 per cent and 42 per cent for new private sector banks and at 45 per cent, 2 per cent and 53 per cent for foreign banks (Table 6). These figures clearly illustrate that in case of PSBs, restructuring was used extensively as a tool for NPA management. Such masking of NPA accounts reduces incentives for improved credit management, including recovering efforts.

35. The problems with regard to asset quality seem to be concentrated in the old generation banks -the PSBs and the old private banks and I would urge the management of these banks to be sensitive about this trend and be more willing to recognize the problem at the initial stages so that an early resolution of the NPA problem can be found.

36. It is, thus, clear that a clear picture of asset quality across different bank groups can emerge only if restructured accounts and write offs are reckoned. As I mentioned earlier, write offs create adverse incentives in the system and hinder recovery efforts in the bank. In the case of restructured accounts, there is a clear need for a relook at the entire gamut of issues – identification of accounts as restructured, determination of the sacrifice, if any, associated, the regulatory treatment of accounts classified as restructured, etc. I will return to these issues a little later in my address.

(vi) Trends in asset quality across segments

37. Any analysis of asset quality of the banking system is incomplete without a thorough investigation of the trends in the different segments of the economy. The general expectation is that the impairments in large borrower segments should be lower as banks devote more resources for the administration of large ticket advances as compared to administration of small advances.

38. An analysis based solely on gross NPA ratios would tend to prove the above hypothesis. Gross NPAs numbers suggest that asset quality is a greater concern in respect of the agricultural advances as compared to that of industrial advances. However, if we look at the ‘impaired assets ratio’, it shows that the asset quality is of a higher concern in the industries segment as compared to the agriculture/ retail segments and has deteriorated more significantly for this segment after the crisis. A bank group-wise data on ‘impaired assets ratio’ across various segments in given in Annex 7.

39. Among the big-ticket advances, infrastructure finance has been affected significantly. Inadequate commercialisation of projects due to various regulatory, administrative and legal constraints and absence or insufficiency of user charges in many projects have put strain on the banks. That infrastructure projects in our country face such constraints is well known. Yet, it is deplorable that banks had not factored in such issues in their credit appraisal or ensured that contingency arrangements are in place for eventualities like inability to procure coal, gas or to acquire land for road projects.

40. Advances to the medium and large segments account for about 50 per cent of total bank advances and of total NPAs indicating that the deterioration in asset quality is driven by the medium and large enterprises. This is also reflected in the ‘impaired assets ratio’ which stands at 14.8 per cent for medium and large industries as compared to 10.6 per cent for micro and small enterprises as on March 2013.

41. Even in cases of restructuring, there seems to be a distinct bias towards the large-ticket borrowers than the weaker segments of the economy (i.e. micro and small enterprises, agriculture and priority sectors). Statistics on restructured advances shows that the medium and large segments account for over 90 per cent of restructured accounts while the share of micro and small segments keeps dwindling over the years (Annex 8). Further, the large ticket accounts hold a major share in CDR cases. This runs contrary to the spirit of the allowing the facility of restructuring as economic downturn is more likely to impact the small borrowers who are more vulnerable to business cycles and hence may require restructuring to tide over the temporary problems, as compared to large borrowers. The ground reality is that advances to smaller borrowers, with genuine needs get overlooked and slip into NPA which enables building of a perception that the quantum of non-performing assets is more in the case of small borrowers and hence promotes a rush towards large-ticket advances, ignoring the basic fact that the lower NPAs amongst larger borrowers is primarily on account of extensive restructuring/ write offs of such accounts. In their scramble to lend to large borrowers, therefore, banks lose out on a massive business opportunity in terms of the small borrowers.

42. One of the common myths about asset quality of banks is that directed lending has significantly contributed to the rising NPAs of banks, as the gross NPA ratio of credit to priority sector is higher than that of the non-priority sector. However, if the impaired assets ratio is considered, then the asset quality of credit to non-priority sector is a greater concern than in the case of the priority sector. This ratio in case of the non-priority sector has also deteriorated to a greater extent in the post crisis period.

43. To summarise, I would say that the perception that agricultural advances or priority sector lending carry more credit risk than the non-priority sector is entirely misplaced and needs to undergo a change. The smaller borrowers are per se not a cause of stress to the banks; rather it is a bias against them that turns them into weak accounts.

Study Conclusions & other issues

44. So far, we have explored the trends and various facets of the asset quality of banks. Let me walk you through the broad ‘learnings’ that we can draw from our analysis.

(a) Primitive Information Systems - At the outset I had highlighted how the evolution of information systems had not kept pace with the changing banking landscape. Improvements in information systems were not in keeping with the increase in asset size of banks and the increasing complexities in credit management. The lack of granular data on slippages, early indications of deterioration in asset quality, segment wise trends, etc., hampered timely detection of problem accounts and weakened banks’ credit risk management capabilities. As a result, banks failed to identify the reversal in trends in asset quality in the pre-crisis period. As you would all appreciate, a robust database containing granular information is an imperative for better credit management and hence as part of the Top Management of banks, you must drive your officers towards building up this database and also use the information for the early identification of problem accounts, trends in asset quality, etc.

(b) Higher NPAs not only from the GDP slowdown – Undoubtedly,themacro-economic environment has a significant influence on the asset quality of banks. The recent deterioration in the asset quality is almost concomitant with the deceleration in the country’s growth rate, re-emphasizing the oft quoted argument about the cyclicality of asset quality of banks and its linkages with the macroeconomic performance of the economy. However, a closer look at the trends in the growth rate of gross NPAs and in the slippages of Indian banks indicate that the seeds of recent deterioration in asset quality had been sown well ahead of the slowdown in economic growth, i.e. in 2006-08 when most of the asset quality indicators reached their most robust levels. Therefore, the current decline in asset quality cannot be solely attributed to the recent decline in the country’s macroeconomic performance. Besides, the deceleration in GDP growth rate, there are other factors/causal relationships which are responsible for the current state of the asset quality of the banking system. I will now touch upon some of these factors in depth.

(c) Lax credit management - The analysis reveals that impairment in assets were an off shoot of the deficiencies in credit management that had crept in during the pre-crisis “good years”. Banks with higher credit growth in 2004-08 ended up with higher growth in NPAs during 2008-13 period. A bank wise analysis of credit and NPA growth indicates that the compounded annual growth rate (CAGR) of NPAs during the period 2008-13 was highest in case of banks whose CAGR of credit was also higher during 2004-09. Thus, it was during the pre-crisis years that deficiencies in credit appraisal crept in, credit monitoring was neglected and recovery efforts slowed. Let me elaborate on some of these points with some specific illustrations about the deficiencies in credit appraisal.

• Evidence suggests that the banks were not taking adequate cognisance of the build-up of leverage while sanctioning or renewing limits. In fact, banks’ credit appraisal processes failed to differentiate between promoter’s debt and equity and over time, promoters’ equity contribution significantly declined and leverage increased. In particular, there has been a significant increase in the indebtedness of large business groups in recent years. A study of ten large corporate groups by Credit Suisse has revealed that the share of these ten groups in total banking sector credit more than doubled between 2007 and 2013 even while, the overall debt of these groups rose 6 times (from under Rs. one trillion to over Rs. six trillion).

• Ironically, the banks were found to be lending more to sectors that had high impairments, pointing to possible lacunae in credit appraisal standards. For example, while the CAGR of credit for the period 2009-2012 for the banking sector was 19 per cent; the segments like iron and steel, infrastructure, power and telecom witnessed much higher credit growth despite the impaired assets ratio for these segments being significantly higher.

• Indian corporates, operating in India and abroad, have been increasingly accessing international debt markets to raise capital. While this is presumably being done to take advantage of the low interest rate in the international markets, in an environment of fluid exchange rate markets, corporates run the risk of incurring losses from adverse movement in exchange rates for their un-hedged exposures. The un-hedged exposures and an eventual increase in interest rates could put pressure on the corporates and eventually spill-over to their lenders.

• Banks are not conducting adequate contingency planning, especially for mitigating project risks. While on the one hand, the stress tests/ simulation models run by banks do not employ extreme stress scenarios; the extent to which the borrower can withstand it and the haircut to be imposed by banks in such cases; on the other hand they also do not factor in possible eventualities like failure of gas based power projects to ensure supply of gas, failure to complete land acquisition for highway projects, failure to acquire environmental clearances etc. Banks definitely need to factor in such adverse situations and have a back-up/contingency plan in their appraisal process.

• Restructuring was extended to companies that were facing larger problems of over-leverage and inadequate profitability pointing to possible lack of due diligence in assessing viability while restructuring.

• Further, companies with dwindling debt repayment capacity were raising more and more debt from the system. These trends pose concerns for the asset quality of banks – on one hand, the ability of corporates to service debt was falling and on the other hand, with increased borrowings their exposure to interest rate risk was rising.

A pointer to the potential deterioration in appraisal and credit management standards in banks is the increasing incidence of advance related frauds, especially large value frauds (over Rs. 500 million) in recent years. Over 64 per cent of the total amount involved in fraud cases reported till March 2013 were advance related. My assessment of the situation is that, in such cases more often than not, it is the failure of the credit appraisal and post disbursement supervision processes which results in accounts being classified as fraud. There appears to be a tendency on part of the sanctioning authorities to brandish their own credit appraisal failures as a fraud perpetrated by the borrower. The moral hazard risks associated with identifying business failures as frauds is that the lacunae in credit appraisal do not surface and fixing of staff accountability becomes a casualty.

(d) Sub-optimal credit management in PSBs – Ouranalysis reveals that the weaknesses in credit management, which I have dwelled upon at length, were significantly more pronounced in PSBs. In fact, the wide divergences in the asset quality of PSBs and new private sector and foreign banks bears testimony to the fact that the economic downturn is not the sole reason for recent deterioration in the asset quality of banks as all bank groups were not affected to the same extent.

PSBs suffer from some structural deficiencies related to the management and governance arrangements. Instances of lax credit management (credit appraisal, credit supervision, etc.)and poor governance and management standards which, though persisting even before the crisis, were not dealt with in time and eventually impacted much more emphatically than was anticipated.

The private sector and foreign banks, on the other hand, were quick to identify the early threats posed by the slowdown and effectively managed them. This is evidenced by a sharp rise in the slippage ratios of private sector and foreign banks in the immediate aftermath of the crisis, but the deterioration was quickly arrested through improved credit management systems and more concerted recovery efforts. In contrast, PSBs failed to respond equally swiftly and have suffered far more significant deterioration in asset quality post crisis. Failure to acknowledge the problem as also inability to look for an early exit route at the incipient stage, used very successfully by some private sector and foreign banks, have likely exacerbated losses for PSBs.

How resilient is our banking system?

45. While I have so far dwelt on the problems that characterise the credit management in Indian banking system, let me count some positives. The first and foremost comforting factor is that we have thrived through much more challenging times in the past in so far as the NPA ratios are concerned. In fact, in March 1994, the NPA levels were much higher than the present level. The aggregate banking system GNPA ratio was 19 per cent in March 1994 and for PSBs was 21 per cent. Against that benchmark, the current position of NPAs in the banking sector is not alarming. In March 2013, the GNPA ratio was 3.4 per cent for banking system and 3.6 per cent for PSBs. Even if we were to consider the impaired assets ratio, for the system as a whole it is much lower at 10.6 per cent while for the PSBs it stands at 12.1 per cent.

46. Another silver lining on the cloudy horizon is provided by the strong capital position of Indian banks. Our stress testing of the banks’ asset portfolio (by applying different static credit shocks, including shocks to banks’ restructured accounts) as of June 2013, has revealed that the system level CRAR would remain above the required minimum of 9 per cent. However, only under severe shock of 150 per cent on NPA, the Core CRAR would go down to 6 per cent.

47. The relatively lower level of provision coverage ratio (PCR) of banks in India as compared to their global peers is a weak spot in an otherwise fairly resilient Indian banking system. The sad part is that this ratio has seen a declining trend in recent years. A study conducted in the Reserve Bank (covered in the Financial Stability Report of June 2013) assessed that, under stressed macroeconomic conditions, the current provision coverage of banks in India may not be sufficient to cover expected losses. The PCR presents a more dismal picture when restructured standard advances are also considered, as it stood at just 30 per cent in March 2013 down from 35 per cent in March 2009. Hence, it is essential the banks increase their provision cover from current levels and strengthen their balance sheets.

48. Thus, on the whole while the banks will be able to withstand the present deterioration in asset quality, the rise in slippages and the quantum of restructuring coupled with low PCR levels is a source of worry and over a period of time the situation could pose grater concerns especially if timely corrective actions are not taken.

Recommendations and Way Ahead

49. Let me now turn to the way forward for policy makers and banks on the concerns regarding asset quality I have listed thus far. The recommendations that I am making can be broadly classified into two categories- one set for the short run and one set for the long-run. In the short run, the objective should be to address the issues with the existing stock of impaired assets in a time bound manner. The long run measures will have to deal with issues that are more deep-seated and hence, would require collaborative efforts from all the stakeholders. These involve improvements in systems and processes, in regulatory framework and legal infrastructure, etc.

(i) Short-term measures

Review of NPAs / restructured advances

50. In the short run, an immediate review of the accounts classified as NPAs and those which have been ‘restructured’ needs to be taken up. It would be critical to assess the viability or otherwise of these accounts on a case-to-case basis. The accounts which are found viable would need to be supported through additional finance/other means in order to ensure turnaround while those found unviable should be taken up for prompt recovery/resolution, so that the loss given default of banks could be minimized.

51. In accounts where restructuring is contemplated, it is essential that the decision is taken within a pre-stipulated time-frame. Further, it needs to be ensured that promoters assume their share of losses without resorting to further borrowing from the banking system or employ structured products for their contribution towards equity. If need be, new promoters should be brought on board to ensure that the down-side risks are equitably shared.

52. As I mentioned earlier, all unviable accounts should be put under time-bound asset recovery drive, sale of assets to ARCs or under the SARFAESI Act, to protect the loss in economic value of the assets. Another option which needs to be explored in such cases is management/ownership restructuring and permitting banks to takeover units where promoters’ equity is low or non-existent (and hence the promoters have little interest in rehabilitating the unit); running the unit through an intermediary or agent and then disposing off the unit when it is sufficiently rehabilitated.

53. I strongly feel that the mechanism of restructuring of small accounts needs a special mention in terms of corrective action needed. SMEs and other segments covered under the priority sector are an important segment of the economy as they provide employment to a large number of people. It is important that the viable businesses under these segments facing temporary problems are not discriminated against. The entire approach to restructuring has to be reoriented to show more compassion to the small customers.

(ii) Long Term Measures

Improved credit risk management

54. The first line of defense against deterioration in asset quality of banks is the bank’s own credit risk management. If the asset quality profile of the banking system needs to be improved on a sustained basis, then sound and robust credit risk management system needs to be in place in the banks. Credit risk management comprises of different facets like credit appraisal, credit monitoring, efficient system of fixing accountability, etc. Let me touch upon the improvements required in each of these areas.

(a) Enhance Credit Appraisal – My assessment is that the banks’ credit appraisal standards have slackened remarkably over the last few years. To arrest the rapid deterioration in their asset quality, the banks need to focus on certain aspects more rigorously than is currently the case. Let me list a few of them:

• Group leverage: Trends in the leverage of the group to which the borrowal firm belongs; the growth in overall indebtedness of the group, particularly borrowings from the banking sector, over the last two/ three years must be closely examined. It is important that the leverage at the holding company level is also factored in the appraisal.

• Source/structure of equity capital: One of the most crucial aspects in project finance is the source and structure of equity contributed by the promoters. As I mentioned a little earlier, focused attention needs to be given to this area. Banks need to ensure that the promoters’ contribution is funded through equity and not debt so that the promoter has sufficient skin in the game. Attention must be paid to the source of equity as, very often, the project is funded through structured borrowings by the promoters at the holding company level within the larger group and which is down streamed as equity in the subsidiaries. There is also a need to examine the borrowers’ ability to raise additional equity to tide over adverse situations.

• Complex project structure - If the project structure is complex/ ring-fenced (as in a SPV structure), then the bank should take extra precautions while assessing such projects and, if necessary, prescribe additional equity requirements to guard its interests.

• External constraints – Often projects get delayed or do not reach sufficient commercialisation due to factors beyond the control of promoters/ banks viz. regulatory, political and legal constraints like environment and other clearances, etc. As banks are taking increasingly larger exposures to infrastructure projects, they must conduct necessary sensitivity analysis and contingency planning while appraising the projects and build adequate safeguards against such external factors.

(b) Need for quicker decision making - Timely decision making is crucial to the success of any project. The appraisal, sanction and disbursement decisions should be timely and fast. Similarly, additional limits/ concessions, if needed, should be sanctioned quickly in order to enable the business unit to smoothly tide over the problem.

(c) Strengthen Credit Monitoring - An effective appraisal needs to be backed by efficient monitoring to ensure that the account remains standard and performing. In this regard, developing an early warning mechanism and a comprehensive MIS assumes an important role to enable regular viability assessment.

(d) Enforce Accountability - No amount of rules and regulations would aid in addressing the issues at hand, if their enforceability is weak. Thus, accountability for lapses in credit risk management is as important as the management of credit risk itself. The present system is biased against lower rung officials as, in more cases than not, it is this section of staff that is held accountable for accounts becoming non-performing despite the loan sanction having been done at senior levels or even at Board level. The officials at junior level should be brought under the ambit of probe for accountability only if there is a clear cut case of deficiency at the monitoring/implementation level. In other words, the sanctioning authority must also share the burden of responsibility in case of failed or improper credit monitoring. People ask me whether it is practical for the Board/Top Management to monitor loans. My answer to that is if you wish to enjoy the authority you must also be willing to accept the responsibility. You may bring in a kind of mechanism which allows you to monitor the performance of loans that you have sanctioned. If you can’t do that, please abdicate this responsibility in favour of the branch managers/ lower level officials. The accountability framework should also encompass instances of delayed decision making or inaction especially when the delay results in disruption in a unit’s working or in its failure due to delayed sanction/ disbursal/ renewal/ restructuring. Accountability also needs to be fixed on CAs, certified valuers, lawyers, etc. who collude with the borrowers and submit false certifications to facilitate borrowers in availing bank finances.

(e) Address corporate governance issues in PSBs – As I had mentioned earlier, there are some inherent structural weaknesses in the corporate governance arrangements in PSBs. These will need to be addressed through a slew of measures such as explicit ‘fit and proper’ criteria for appointment of top executives, instituting a system of an open, market-wide search for Chairman/ other top executives, review of the role of RBI / government nominee on the Board of PSBs, fixation of tenure of top executives of banks so as to ensure a sufficiently long tenure, market based remuneration for top executives, review of the incentive structure for the executives, greater accountability of the executives and of the Board, etc.

Improved information systems

55. As I have said earlier, information systems form the bedrock of credit risk management. Unless the risks are identified, it will not be possible to manage them. Any robust management system necessarily hinges on a strong information base. While the banks have made significant progress in this area, a lot still remains to be done. Let me illustratively talk of some of the further work needed in this area:

• Efforts should be made to capture more exhaustive and granular data. For instance, an MIS for capturing common exposure across banks needs to be developed. Recently, RBI has initiated efforts in this direction.

• The information system must enable timely detection of problem accounts, flag early signs of delinquencies; facilitate timely information to management on these aspects, etc.

• An appropriate coordinating mechanism across departments within a bank and across banks must be developed.

• And, most importantly, the information system of banks should be integrated into decision making, capital planning, business strategies, and reviewing achievements.

Strong Regulatory Framework

56. The onus on addressing the asset quality deterioration does not rest on banks alone and the policy makers also need to take a re-look at the extant regulatory framework for stressed assets. The review will need to address a few hard questions - Are extant instructions about the classification of accounts as impaired too stringent and prohibitive? Are viable units being classified as NPAs? Is the banks’ reluctance to extend support to accounts classified as non-performing accelerating the transformation of viable units facing temporary cash flow problems towards non-viability?

57. A thorough relook at the overall regulatory approach in the country is warranted. Regulations should be facilitative, practical and commensurate with the risks involved. At the same time, we must appreciate that it’s a tight-rope walk to balance the needs to ensure that viable accounts are supported and that banks are dis-incentivized from ever-greening the advances portfolio.

58. As I hinted earlier, one of the key areas where regulatory prescription needs realignment is that of restructuring. Global practices warrant that restructured accounts should be treated as non-performing. This is a practice which needs to be adopted in India as well keeping in view the extant realties. It is also suggested that simple rephrasing, rescheduling, refinancing, extension of tenure of loans should not be treated as restructured. Similarly, an account should not be treated as restructured if there is a no sacrifice involved on the part of the bank and the bank is satisfied about the realizable value of the collateral. Further, the accounts in respect of which interest is being serviced regularly should not be classified as an NPA. In particular, if the borrower continues to pay a rate of interest above the base rate of the bank, then a mere reduction in the rate of interest being charged from the borrower should not be a sufficient reason for classification of the account as restructured/non-performing. The discretion with regard to treatment of an account as restructured or not should be left to the Board of the bank, based on broad regulatory principles.

59. Another area where reform is needed is technical write-off. The practice of technical write offs of NPAs engenders moral hazard and leakages in the recovery process and needs to be dispensed with. It is much better to pay tax than to distort the system with technical write-off. Further, as I said earlier, there is a need to increase provisioning requirements in line with international norms to ensure the resilience of the banking system

60. The prevalent practices of restructuring accounts to treat them as standard and of technically writing off accounts are not uniformly used across all borrower segments. These practices result in the position of asset quality of banks being mis-reported. Such mis-reporting, in turn, leads to mispricing of risks and inefficient allocation of resources impairing the overall efficiency of the banking system. Hence, my recommendations that all restructured accounts be treated as non-performing in line with global best practices and that the practice of technical write offs be dispensed with. If, however, these regulatory dispensations have to continue, then it needs to be ensured that these dispensations are made applicable to all segments of bank borrowers. Rules may need to be set out in this regard or, at the very least, some broad principles formulated.

61. Across the world, regulatory frameworks are built on one of two approaches – rules based or principles based approach. In India, though we do not have a declared approach to the regulatory framework, the extant framework is built on an approach which is a blend of rules based and principles based approaches. Overall, the approach appears more discretionary. Although flexibility in regulatory prescription is sometimes coveted, arbitrariness in regulation would lead to the build up of ‘fault lines’ in the system. A uniform approach – either principle or rule based regulation – needs to be adopted and consistently followed so as to bring the necessary stability in credit risk management practices across banks and eliminate ad-hoc implementation processes.

Reforming legal & institutional structures

62. A facilitative financial sector infrastructure including robust a legal and recovery framework and an effective credit information systems can play a critical role in ensuring that the asset quality of banks remains healthy. On one end of the spectrum are mechanisms such as the Corporate Debt Restructuring (CDRs) which are aimed at rehabilitating viable but temporarily distressed corporate advances. On the other end are mechanisms such as the Debt Recovery Tribunals (DRTs) and the SARFAESI Act which provide a legal framework for dealing with unviable borrowings. In addition there are Asset Reconstruction Companies (ARCs) which facilitate improved credit risk management by banks by sale of problem loans and creation of a secondary market for such loans. It is, however, sad to note that these frameworks are beset with their own set of problems. Let us critically examine these structures one-by-one.

(a) Corporate Debt Restructuring (CDR) mechanism - There has been major increase in CDR references in the recent period and an exposure-wise breakup of CDR referred cases shows that big-ticket accounts had a dominant share.

There is enough evidence, to suggest that the provisions of the CDR mechanism have not been used very judiciously or effectively. While the debtors and creditors seek the benefits of restructuring, they tend to avoid the painful sacrifices in terms of provisioning and promoters’ sacrifice. Such circumvention of norms not only camouflage the weakness in the credit portfolio of banks but also weaken their defense against expected losses. The inherent credit weaknesses of such accounts are further aggravated due to lower stake of the promoters. I would go on to add that the availability of standing regulatory forbearance in the matter of CDR has prompted banks to avoid using other means of credit management judiciously.

To my mind, the CDR mechanism needs a thorough overhaul in terms of process, structure, administration and governance. Illustratively,

• Viability of the unit must be established, without any cause for reasonable doubt, prior to restructuring.

• Restructuring or exit must be approved within a definite timeframe.

• Promoters’ stake in the project should be ensured upfront through core equity infusion (and not quasi/ debt)

• An independent oversight and monitoring system should be in place, especially for large CDR cases, to ensure that the account does not slip up post restructuring. The CDR cell and the banks concerned should coordinate in this aspect.

(b) Debt Recovery Tribunals (DRTs) & other legal provisions - There are concerns about the functioning of DRTs and the enforceability of legal provisions for recovery of loans. The legal system for recovery seems to be defaulter friendly. Inordinate delays hamper effective resolution even as the time lag erodes the value of recoverable assets. Concrete measures viz., setting up of more DRTs and DRATs, full-fledged computerization of processes, digitization of records, enforcing strict adherence to timelines, etc. would need to be taken to make these mechanisms more effective The recent amendments to the DRT Act vide the Enforcement of Security Interest and Recovery of Debts Laws (Amendment) Act, 2012 (which has been brought into effect from January 15, 2013) are a step forward in this regard. In this context, I will urge the bank management to launch vigorous follow up of the suit filed accounts and hold periodic meetings and interactions with the DRT Chairs to resolve the issues.

(c) Asset Reconstruction Companies (ARCs) - The Reserve Bank had encouraged the setting up of ARCs and put in place a regulatory framework for the same. However, the ARCs have not really taken off. Despite the accelerated increase in NPAs in banks, the incremental flow from banks to ARCs has significantly reduced from Rs 130 billion in 2008 to Rs 60 billion in 2012³ There is a need to review the functioning of the ARCs with a view to revitalizing them. Regulatory measures such as clarifying / standardizing the NPA auction process, valuation norms, etc., could be contemplated.

(d) Credit Information Companies (CICs) - The first credit information bureau was set up in India about a decade ago with a view to assist in credit risk management of banks and financial institutions at various stages across their customer’s life cycle. However, the use of data from the CICs has remained restricted to the retail segment and that too mostly by the private and foreign banks.

There is a need to leverage CICs not only for sanctioning purposes but also for monitoring and recovery. At the monitoring stage, CICs can furnish a 360 degree view of customers, thereby enabling portfolio risk assessment and early warning triggers to identify and address delinquency. At the recovery stage, CICs can help entities in reaching their not so easily contactable base, thereby improving recoveries. Let me list some steps to enable broad-based use of CICs:

• Improve the quality of data captured by the credit bureaus and ensure that the data submitted by credit institutions is complete and timely.

• Enlarge the information database of CICs by bringing more and more customer segments under their fold.

• Collate specific information like details of mortgaged assets, previous defaults, recoveries, write-offs, involvement in frauds or other criminal activities, etc.

• Share data with other CICs to enable a holistic view of the credit profile of the borrower.

• Build elaborate databases on additional information not directly linked to the credit information which can be used by the lenders to study trend in an industrial sectors, geographic areas and also issue triggers/warnings to lenders on likely credit events.

Key messages and measures

63. Let me conclude my address by revisiting the key takeaways for all of us:

1. The present level of stressed asset as an outcome is not a large concern in itself but the extant processes, systems and structure of creation/ restructuring of stressed assets are a problem. The existing level of NPAs is manageable but if corrective actions to arrest the slide in NPA are not initiated quickly, the stability of financial system could be at risk.

2. Economic slowdown and global meltdown are not the primary reason for creation of stressed assets. In fact, the deficiencies in credit and recovery administration in the system involving banks, borrowers, policy makers, regulators and legal system have led to the present state of affairs.

3. Credit quality has a high positive correlation with the prudential norms and regulations prescribed by RBI. Lax, soft and flip-flop approach to regulation may have contributed to creation of NPAs and stressed assets in the system by not creating sufficient incentives for the banking system to recognize NPAs and make provisions for stressed assets.

4. Level of leverage of corporate borrowers, high credit growth, lack of equity and instances of advances related fraud cases are also highly correlated. These are the first order derivative of inadequate credit appraisal and recovery management.

5. Appraisal standards are lax for bigger loans both at the time of sanction as also restructuring while the banks tend to be very stringent for smaller borrowers. Also, restructuring and write off processes are highly biased towards bigger loans as compared to smaller loans, even though, the data suggests that credit risk is higher for big-ticket advances than small borrower. Banks need to be more sympathetic to small borrowers and be more stringent in their approach to big borrowers.

6. Banks following the process of recognizing NPAs quickly and more aggressively are able to better manage their asset quality

7. High pace of credit growth in certain segments has resulted in lower credit quality in those segments in subsequent periods

8. Overall standard and quality of credit management and recovery management is very poor.

9. Time is of essence and timely action based on information needs to be taken everywhere.

Conclusion

64. To sum up, the issues that I have highlighted today require proper attention and concerted efforts by all stakeholders - regulators, policy makers, banks and bank customers. The first and foremost lesson for all of us is that regulatory forbearance is not a panacea for the ills that afflict the credit management environment in the banking sector. Sweeping the problems under the carpet is not a solution. We must be willing to face the situation, learn from our past mistakes and move forward. For this journey to be successful, the banks would need to strengthen their credit appraisal standards and lay greater focus on the quantum of equity brought in by the promoters, the sources of equity and plan for contingency situations in case of large infrastructure projects. The appraisal and approval process for restructuring proposals needs to be improved while ensuring that the benefits of restructuring are also extended to the smaller borrowers. The CDR Mechanism has been grossly mis-utilized and needs a thorough overhaul including setting up of an independent oversight body for approval of proposals as well as for monitoring. For the reasons that I have mentioned earlier in my address, regulatory guidelines on restructuring and technical write-off should be reviewed and these measures should be phased out for ensuring a stronger credit administration and recovery environment. Some concrete steps like careful examination of existing NPAs for determining further course of action - rehabilitation or recovery, would need to be taken in the short term. Accounts identified as viable will need to be quickly rehabilitated with support from the bank and the borrower including through infusion of new equity from new promoters. Above all, a robust accountability mechanism for all levels of hierarchy in the stakeholders – banks, borrowers, regulators, policy makers and the government at large – will have to be put in place to ensure that banks’ asset quality improves on a sustained basis.

65. I would conclude by reiterating that my address today is not aimed at finding fault but an attempt to initiate a debate and encourage some soul-searching in the banking community about the past mistakes and what we need to do differently going forward. I believe the Top Management of banks present here, borrowers, regulators, policy makers and the Government functionaries, would reflect on the key messages that I have attempted to put across this afternoon and make whole-hearted efforts to ensure that our banks emerge stronger out of the present crisis and are able to become the banks of the future. A quick and determined action is the need of the hour and I wish you all the very best in this endeavour. I once again than IBA and Bank of India for providing me this opportunity to share my thoughts with all of you.

Thank you!

Annex 1

Annex 2

Annex 3

Annex 4

Annex 5

Annex 6

Annex 7

Segment wise Impaired Asset Ratio⁴ (in percent)

Annex 8

¹ Address delivered by Dr. K. C. Chakrabarty, Deputy Governor, Reserve Bank of India at BANCON 2013 on November 16, 2013 in Mumbai. Assistance provided by Ms. Dimple Bhandia and Ms. Vinitha is gratefully acknowledged.

² Gross slippage ratio = fresh accretion to NPAs during the year to standard advances at the beginning of the year Net slippage ratio = Slippage ratio net of recoveries

³ Source : Association of ARCs in India

⁴(GNPAs + Rest Std Adv+Cum W/O) /(Gross Advs + Cum W/O)