BANKING

The Banking Conundrum

C P Chandrasekhar, Jayati Ghosh

A
Neo-liberal banking reform was launched in the early s fiscal year 2017–18 drew to a close, the Government
1990s to address the low profitability of the public of India decided to bite the bullet and implement a pro-
posal to “resolve” what was being presented as one of
banking system and the large presence of
the leading challenges then facing the Indian economy: large
non-performing assets. It set itself the objectives of non-performing assets (NPAs) on the books of the banks, espe-
cleaning out NPAs, recapitalising the banks and cially the public sector banks (PSBs). The recapitalisation plan,
modifying banking practices to restore profitability and first announced in October 2017, involved infusing `2.11 lakh
crore of new equity into the PSBs, of which `1,35,000 crore
drastically reduce NPA volumes. This did initially have
would be new money from the government, financed with
some effect. However, while the NPA ratio fell between recapitalisation bonds. Another `18,139 crore was the balance
the early 1990s and the mid-2000s, it has risen sharply due under the `70,000 crore Indradhanush plan initiated in
since then. Moreover, while earlier priority and August 2015 and funded from the government’s budget. The
remaining `57,861 crore was to be mobilised by the banks from
non-priority loans contributed equally to total
the market. The plan was to clean up the books of the banks to
NPAs, more recently, large non-priority loans to the a significant extent, enabling them to adhere to the Reserve
corporate sector account for the bulk of NPAs. An analysis Bank of India’s (RBI) voluntary decision to get banks to meet
of these features reveals that these trends Basel III-type capital adequacy norms by 2019.
If corporate borrowers are let off the hook and losses of the
are indicative of the failure of neo-liberal banking
banks are recapitalised with resources from the budget, then
reform in India. private losses are clearly being socialised, since their burden is
being transferred to those paying direct and indirect taxes today
or in the future. This has been under way for some time now.
Between 2000–01 and 2014–15, budgetary allocations for
recapitalisation of banks totalled `81,200 crore. Much of this
was provided for in recent years, with as much as `58,600
crore (or 72% of the total) announced during just four consec-
utive years ending 2013–14. However, the government seemed
to have lost the appetite for such recapitalisation. Even when it
was seen as unavoidable, allocations from the budget for the
purpose were short of what was promised, and what was
promised was short of what was required. In 2014–15, while
`11,200 crore was allocated for the purpose in the budget,
actual capital infusion into PSBs was just `6,990 crore. Then in
2015–16 there was a revival, despite the initial reduction of
even the budgetary allocation for the purpose to `7,940 crore.
In the course of the year, the government announced a four-
year Indradhanush plan, under which the PSBs would be pro-
vided with new capital worth `70,000 crore, with `25,000
crore being disbursed that financial year and the next, and
`10,000 crore in each of the two subsequent years. In its most
recent avatar, the recapitalisation exercise is the `2.11 lakh
crore plan announced in October 2017.
C P Chandrasekhar ([email protected]) and Jayati Ghosh (jayatijnu@ The recent decision to proceed with the recapitalisation
gmail.com) teach at the Centre for Economic Studies and Planning, plan came after much delay for two reasons. First, over a con-
Jawaharlal Nehru University, New Delhi.
siderable period, alternatives to recognising bad assets, writing
Economic & Political Weekly EPW MARCH 31, 2018 vol lIiI no 13 129
BANKING

them off and taking a hit in the form of reduced profits or losses Unfortunately, the evidence on recovery has not been com-
were being quietly sought by all concerned. Second, once the forting. The rate of recovery of NPAs of scheduled commercial
NPAs were recognised, ways to avoid recapitalisation financed by banks (SCBs) through various channels (Lok Adalats, Debt
the government were being sought, on the grounds that it would Recovery Tribunals and the Securitisation and Reconstruction
derail the government’s successful effort at “fiscal consolidation.” of Financial Assets and Enforcement of Security Interest [SAR-
FAESI] Act, 2002) had fallen from 22% of amounts involved in
NPA Recognition cases referred to these channels and being considered by them
The first set of factors delaying NPA recognition was the result at the end of March 2013 to 9.8% by end-March 2017.1 Overall,
of the nature of the NPAs that were accumulated during the the experience with loan recovery has been disappointing. Not
years of high growth after 2003–04. These unrecognised NPAs only has total NPA reduction been stagnant between 2014–15
consisted of large loans most often delivered by multiple lend- (`1,270 billion) and 2015–16 (`1,280 billion) when the sum of
ers to relatively big corporate entities or groups. In 2001, as declared NPAs was rising, but much of this reduction has been
part of the reform, the government put in place a corporate the result of compromises or write-offs that yield the bank little
debt restructuring (CDR) mechanism, to enable defaulting or nothing. NPA reduction is reported under three heads (actual
firms to devise (in consultation with lenders) turn-around recoveries, “upgradation” or transformation of NPAs into pay-
strategies. This effort at revival of large loans that were under ing assets, and compromises/write-offs). Write-offs involve a
threat of default included measures such as extending the complete loss for the banks.
maturity of the loan, reducing the interest rate charged, con- According to finance ministry figures, the share of write-offs in
verting a part of the loan into equity, providing additional the NPA reduction of the PSBs rose from an already high 41% in
financing, or some combination of these. Once agreement 2014–15 to 46% in 2015–16. PSBs have written off a total of
among creditors, and between them and the debtor, on a turn- `2.46 lakh crore worth of loans over the five years, 2012–13 to
around-cum-debt restructuring package could be arrived at, 2016–17. The ratio of declared profits to write-offs has fallen
modified rules permitted the credit assets concerned to be sharply. In 2012–13, PSU banks wrote off `27,231 crore while
identified as “restructured standard assets” and exempted declaring combined net profit of `45,849 crore. The corre-
from provisioning. sponding figures for 2016–17 were `81,683 crore and `474 crore.2
The process was expected to strengthen firms that were Part of the reason for this is that the government has been
defaulters and allow them to resume normal debt service com- encouraging banks to be lenient when pursuing defaulting
mitments. It also exempted banks from provisioning against firms, unless they are wilful defaulters. Thus, the finance
loans that were bad, which would have resulted in losses and ministry’s Economic Survey 2016-17 argued:
eroded capital. If banks had to make provisions for likely losses Cash flows in the large stressed companies have been deteriorating
on such loans at the first sign of them turning non-performing, over the past few years, to the point where debt reductions of more
the impact this would have on their finances would dissuade than 50 percent will often be needed to restore viability. The only al-
them from undertaking such lending. So the exemption from ternative would be to convert debt to equity, take over the companies,
and then sell them at a loss. (GoI 2017: 85)
provisioning was clearly aimed at encouraging banks not to
withdraw, but keep large credits flowing to the corporate sector, The point implicitly being made here is that it is necessary to
especially to capital-intensive projects with long gestation help get companies back on their feet even at the expense of
lags, as in infrastructure. Thus, the government with its neo- bank balance sheets. The finance ministry’s claim was that
liberal agenda chose to treat restructured loans as “standard this is necessary because the companies cannot share any
assets” and not non-performing ones. Banks were given some blame for their current position. The Survey argued:
leeway when classifying assets, which the RBI later claimed Without doubt, there are cases where debt repayment problems have
was misused. This brought down actual and potential NPAs, been caused by diversion of funds. But the vast bulk of the problem
which were initially concealed. has been caused by unexpected changes in the economic environ-
After a significant period of time, when the gap between ment: timetables, exchange rates, and growth rate assumptions going
wrong. (GoI 2017: 85)
declared NPAs and the “stressed assets” of banks (or the sum of
“restructured standard assets” and NPAss) began to widen, the Such arguments notwithstanding, it is clear that the accu-
RBI decided in the second half of 2015 to undertake an asset mulation of NPAs and the losses resulting from that process are
quality review and impose stricter guidelines for bad loan rec- indications of the failure of neo-liberal banking reform in India.
ognition. The result, soon thereafter, was a sharp spiral in the The case for such reform, advanced by a spate of official com-
ratio of gross NPAs to gross advances. Moreover, provisioning mittees set up over the course of the two decades following
for these NPAs sharply reduced profits in some banks and 1991, was that the social banking heralded by nationalisation
forced losses on others. was unviable. Public banks were unprofitable or not profitable
It could be argued that these losses are temporary and enough and were accumulating large NPAs because of the policy
need not impair the capital base of the banks. Write-offs of of directed credit to the priority sector. That is, it was not the
NPA s are technical, and if, in time, much of the value of the failure of the dominantly publicly-owned banking system to
loans in default can be recovered, the balance sheet of the achieve the goals of bank nationalisation that occasioned the
banks would not be damaged. perceived need for the continuous and sweeping shift in banking
130 MARCH 31, 2018 vol lIiI no 13 EPW Economic & Political Weekly
 BANKING

and financial policies that occurred after the July 1991 balance and a half of reform (Table 1). But once the credit boom began
of payments crisis. In fact, bank nationalisation had succeeded in the mid-2000s, NPAs returned, though initially concealed,
in terms of an expansion of the reach and spread of formal through restructuring.
banking, increased credit provision on an expanded deposit In the event the NPA ratio displays a V-shaped tendency in
base, greater focus on underbanked areas and populations, the years after 1991, falling sharply from close to a quarter of
correction of the extreme skew in bank lending in favour of gross advances in the early 1990s to 2% in 2008–09 and then
industry and big business (with more credit going to agriculture rising to around 13% last year (Table 1). But the problem had
and small industry and business), greater inclusion of the poor begun well before 2008–09 and continued thereafter. The
in the provision of financial services and credit, and restruc- lagged spike in the NPA ratio was the result of the RBI’s man-
turing of the banking infrastructure through measures such as date that NPAs that were being kept hidden under the garb of
the creation of regional rural banks and emphasis on social being “restructured standard assets” had to be reclassified,
banking practices. Indeed, the aims of bank nationalisation with a deadline of March 2017. The PSBs were the location of
meant that the realisation of these objectives, rather than prof- much of those NPAs, accounting for 87% of gross NPAs at the
its, should be the basis for assessing banking performance. Going end of March 2017. So, whatever occurred was also clearly due
by such indices, the performance of the PSBs was outstanding, to acts of commission or omission of the government, which is
as they managed to realise within a decade or a little more, known to influence the behaviour of the PSBs.
what the private banks had failed to deliver even partially in
more than two decades Macroeconomic Shift
Notwithstanding this history and these goals of social banking, Understanding the determinants and the implications of V-
the attempt at policy reversal in the early 1990s focused on the shaped movement in the NPA ratio and the policy responses it
so-called “failure” of nationalisation as reflected in the low calls for, requires placing it in the overall economic context,
profitability of the public banking system, the NPA s resulting which reflects the consequences of fiscal and monetary policy
from directed credit to the priority sector and the poor level of reform, on the one hand, and of domestic and external financial
banking services offered to clients of public banks. Therefore, liberalisation, on the other. Neo-liberal macroeconomic policy
an important component of banking “reform” was an effort to reform is focused on weakening the proactive fiscal policies of
write off NPAs, recapitalise banks, change banking rules, and the state, that expand tax- or debt-financed state spending,
modify banking practices that would restore profitability and and relying more on the monetary policy levers of managing
drastically reduce NPA volumes. The evidence suggests that this liquidity and adjusting policy interest rates, which are expected
was achieved in substantial measure during the first decade to drive private consumption and investment in the desired di-
Table 1: NPA Ratios of Public Sector Banks in India (%) rection. In keeping with this perspective, a central feature of
Gross NPAs to Gross NPAs to Net NPAs to Net Net NPAs to post-reform fiscal policy has been the effort to control the fiscal
Advances Ratio Assets Ratio Advances Ratio Asset Ratio
deficit, which was funded in large part by government borrowing
1992–93 23.1 11.8
from the banking system. That effort has been particularly
1993–94 24.8 10.8
1994–95 19.5 8.7 10.7 4.0
successful since the adoption of the Fiscal Responsibility and
1995–96 18.0 8.2 8.9 3.6 Budget Management (FRBM) Act in 2003. Combined with
1996–97 17.8 7.8 9.2 3.6 monetary and banking reform initiatives that reduced the stat-
1997–98 16.0 7.0 8.2 3.3 utory liquidity ratio, which requires banks to invest in specified
1998–99 15.9 6.7 7.1 3.1 government securities, from a peak of 38.5% of net demand
1999–2000 14.0 6.0 6.9 2.9 and time liabilities (NDTL) to 19.5%, this has forced banks to
2000–01 12.4 5.3 6.3 2.7 shift focus away from government securities as an avenue for
2001–02 11.1 4.9 5.8 2.4 longer term investment.
2002–03 9.4 4.2 4.5 1.9 Simultaneously, post-liberalisation changes made banking
2003–04 7.8 3.5 3.1 1.3
extremely important from the point of view of the financing of
2004–05 5.5 2.7 2.0 1.0
private economic activity. Prior to liberalisation, the under-
2005–06 3.6 2.1 1.3 0.7
2006–07 2.7 1.6 1.1 0.6
standing was that banks could provide long-term funding to
2007–08 2.2 1.3 1.0 0.6 industry and the housing market only to a limited extent. Being
2008–09 2.0 1.2 0.9 0.6 dependent on relatively small depositors who would like to
2009–10 2.2 1.3 1.1 0.7 hold their savings in highly liquid deposits, lending to long-
2010–11 2.4 1.4 1.1 0.7 term, illiquid projects would result in maturity and liquidity
2011–12 3.3 2.0 1.5 1.0 mismatches. So the resulting shortfall in the financing of long-
2012–13 3.6 2.4 2.0 1.3 term investment had to be met by creating specialised finan-
2013–14 4.4 2.9 2.6 1.6 cial institutions with access to more long-term capital directly
2014–15 5.0 3.2 2.9 1.8 from the government or the central bank, or through pre-emp-
2015–16 9.3 6.0 5.7 3.5
tion of a part of the resources of commercial banks.
2016–17 12.5
Source: Handbook of Statistics on the Indian Economy, Reserve Bank of India,
A major change brought about by neo-liberal reform was in
various issues. the provision of development finance. The turn to and emphasis
Economic & Political Weekly EPW MARCH 31, 2018 vol lIiI no 13 131
BANKING

on development banking in the immediate aftermath of Indian between 1981 and 2000, rose to 82% by 2012.3 In this sense
independence was explained by two features characterising too, banking was gaining in prominence rather than shrinking
the economy at that point in time: the inadequate accumulation relative to other markets and institutions after liberalisation.
of own capital in the hand of indigenous industrialists; and the One result of these changes was a transformation of the
absence of a market for long-term finance (such as bond or structure of financing of productive activity, especially industry.
active equity markets), which firms could access to part finance Measured as a ratio to gross domestic product (GDP), the
capital-intensive industrial investment. importance of financial assistance from the erstwhile develop-
Post-independence policy perceived that banks per se could ment finance sector diminished considerably after 2000, partly
not close the gap for long-term finance, because there are limits because the development finance institutions had become
to which banks could be called upon to take on the responsibility banks, and partly because they had been rendered irrelevant.
of financing such investments. Banks attract deposits from On the other hand, the capital market did not emerge as a sub-
many small and medium (besides, of course, large) depositors, stitute for these institutions, with the new capital issues market
who have relatively short savings horizons, would prefer to virtually absent, except for periods of an engineered speculative
abjure income and capital risk, and expect their savings to be boom as in the early 1990s. The two main sources of external
relatively liquid, so that they can be easily drawn as cash. finance for industry seem to have been the banks or the private
Lending to industrial investors making lumpy investments, on placement market, with the latter being the target of foreign
the other hand requires allocating large sums to single bor- investors looking for high and/or quick returns. In sum, banks
rowers, with the loans being risky and substantially illiquid. continued to dominate the financing business in India.
Getting banks to be prime lenders for industrial (and infra- The demand for financing of private capital-intensive projects
structural) investment, therefore, results in significant matu- was strengthened by the widening infrastructure gap that
rity, liquidity and risk mismatches, limiting the role that banks resulted from the self-imposed restrictions on public investment
can play in financing long-term productive investment. Other stemming from fiscal conservatism. The government declared
sources need to be found. that given its fiscal “constraints,” crucial infrastructure invest-
This was the gap that the state-created or promoted devel- ment had to be undertaken either through the private sector or
opment-banking infrastructure sought to close. That infra- through public–private partnerships. Since the private players
structure was created over a relatively long period of time and in such “partnerships” typically relied not on their internal
was populated with multiple institutions, often with very dif- resources but on funds borrowed from PSBs, this placed the
ferent mandates. Funds for the development banks came from onus of finding the finance for such projects partly on the gov-
multiple sources other than the “open market”—the govern- ernment, which owned these banks. So, it was natural that the
ment’s budget, the surpluses of the RBI, and bonds subscribed banks would be under pressure to lend to projects varying
by other financial institutions. Given the reliance on govern- from roads and ports to power and steel.
ment sources and the implicit sovereign guarantee that the
bonds issued by these institutions carried, the cost of capital Consequences of External Liberalisation
was relatively low, facilitating relatively lower cost lending for Coincidentally, the effects of those shifts became operative
long-term purposes. Therefore, until the 1990s, India was an when there was another change triggered inter alia by reform,
exemplary instance of the use of development banking as an with large flow of foreign capital into India after 2003. Liber-
instrument of late industrialisation. alisation did from the start increase inflows into the country,
However, as part of financial liberalisation and based on the but large capital flows, which were substantially in the form of
recommendations of the Narasimham Committee reports, the portfolio capital, were a later development. Until 1993–94, total
all-India development finance institutions, which with budget- net inflows amounted to less than a billion US dollars annually.
ary and central bank support and implicit sovereign guaran- Subsequently, foreign investment flows rose sharply to $4.2
tees were seen as distorting the playing field for commercial billion in 1993–94 and averaged about $6 billion during the
banks, were abolished. Some were allowed to atrophy whereas second half of the 1990s. Then, there were even more signifi-
others like the IDBI and the ICICI were allowed to create com- cant changes. During the first decade of this century such in-
mercial banks, with which the development banking arms flows rose to $15.7 billion in 2003–04, and then to $70.1 billion
were “reverse merged.” The result was that investors in capital in 2009–10, despite a fall in the crisis year 2008–09. There-
intensive projects had to turn to the remaining main source of after, after averaging around $64 billion during 2000–13, the
financing—the commercial banks—for long-term funding. figure fell because of the “taper tantrum” in 2013–14.4 But
So liberalisation involved ending the dichotomy between flows bounced back to $73.6 billion in 2014–15, before falling
banking and development financing, with banks now being to $35 billion the next year. In sum, despite high volatility, the
encouraged to foray into term lending of different kinds. The trend has been one of a sharp increase after 2003.
net result was that in the distribution of financial assets among This increase would not have been possible without the
banks and the financial institutions (such as the cooperative relaxation of sectoral ceilings on foreign shareholding and the
banks, the development financial institutions, the nationalised substantial liberalisation of rules governing investments and
insurance companies and sundry other public institutions), repatriation of profits and capital from India. But liberalisation
the share of the banks, which had declined from 71% to 61% began rather early in the 1990s, whereas the boom in foreign
132 MARCH 31, 2018 vol lIiI no 13 EPW Economic & Political Weekly
 BANKING

Figure 1: Aggregate Deposits of Scheduled Commercial Banks (` billion) Figure 2: Ratio of Outstanding Scheduled Bank Credit to GDP (%)
1,00,000 60

80,000 50

40
60,000
30
40,000
20
20,000
10

0 0
1990–91
1991–92
1992–93
1993–94
1994–95
1995–96
1996–97
1997–98
1998–99
1999–2000
2000–01
2001–02
2002–03
2003–04
2004–05
2005–06
2006–07
2007–08
2008–09
2009–10
2010–11
2011–12
2012–13
2013–14
2014–15
2015–16

1990–91
1991–92
1992–93
1993–94
1994–95
1995–96
1996–97
1997–98
1998–99
1999–2000
2000–01
2001–02
2002–03
2003–04
2004–05
2005–06
2006–07
2007–08
2008–09
2009–10
2010–11
2011–12
2012–13
2013–14
2014–15
2015–16
2016–17
Source: Basic Statistical Returns of Commericial Banks in India, Reserve Bank of India, various issues. Source: Basic Statistical Returns of Commericial Banks in India, Reserve Bank of India, various issues.

investment flows occurred much later. That change provides from slightly more than 9% of total outstanding commercial
reason for distinguishing between two phases in the post- bank credit at the end of March 1996 to close to a quarter of
liberalisation years, with 2003–04 as the break. the total more recently. This was a “natural” diversification,
Obviously, these direct and portfolio flows of foreign capital because they were either loans of short-term maturities that
affect domestic money and asset markets. One counterpart of could also be easily pooled and securitised, or they were loans
the capital inflow surge was an increase in the overhang of that were backed by implicit collateral in terms of the asset
liquidity in the domestic economy. There was a dramatic whose purchase was financed. In fact, housing loans accounted
expansion of the deposit base of banks from `1.93 trillion in for a very large share of the total. Moreover, other than for
1990–91 to `9.6 trillion in 2000–01, `52.1 trillion in 2010–11, educational loans, the rates of default in the retail sector have
and `107.6 trillion in 2016–17 (Figure 1). Since banks do not hitherto not been too high.
have the option of sitting on deposits that they must accept What was less natural was the second direction of change.
and pay interest on, the surge in the deposit base would have Despite the huge increase in credit provision, the share of
forced banks to seek new avenues for investment and lending. credit going to industry stood at around 40% of total bank
While the “flexibility” offered by financial liberalisation credit, not too far below pre-reform levels of about 50%. And
helped in this context, the fact that fiscal reform had after long-term loans to corporates, including for infrastructure,
2003 shrunk the space for parking funds in safe government accounted for a significant share of this lending. The share of
securities was a source of pressure. infrastructure lending in the total advances of SCBs to the indus-
The result was an explosion in credit growth (Figure 2). trial sector rose sharply, from less than 2% at the end of March
While the ratio of scheduled bank credit to GDP stood at 1998 to 16% at the end of March 2004, and as much as 35% at
around 20% through much of the 1980s and 1990s, it rose by the end of March 2015. So even as the volume of bank lending
two-and-a-half times between 2000–01 and 2011–12, to touch to industry rose, the importance of lending to capital-intensive
51%. This increase occurred in a period that includes the high sectors and infrastructure within industry has increased hugely.
growth years between 2003–04 and 2008–09, which makes Sectors like steel, power, roads and ports, and telecommunica-
the rise in the ratio of credit to GDP even more significant. The tions were the most important beneficiaries. For commercial
credit deposit ratio of SCBs as a group (which had fallen from banks, known to prefer lending for short-term purposes, this
60.4% in 1990–91 to 51.7% in 1998–99, despite a substantial turn to lending to infrastructure was a high-risk strategy.
increase in the loanable funds base of banks through periodic What this suggests is that the increase in corporate demand
reductions in the cash reserve ratio (CRR) and statutory liquid- for large loans from the banks, for reasons discussed earlier,
ity ratio (SLR) by the RBI starting in 1992 rose sharply after suited the banks as well, since they were under pressure to
2003–04 to touch 74% in 2006–07 and 78% in 2011–12. lend, given the expansion in their deposit base that resulted
from the foreign capital inflow-generated overhang of liquidity
Bank Lending to Industry and Infrastructure in the system. Further, since the government was interested in
The rapid expansion in credit required an expansion in the facilitating capital-intensive private investment, especially in
universe of borrowers and the level of exposure per borrower, infrastructure, it could be presumed that the financing of such
which implied increased risk. There were also significant projects would be backed by the government in the case of
changes in the sectoral distribution of credit, as banks sought liquidity problems or even default. There appeared to be an
to expand the volume of their lending and their universe of implicit sovereign guarantee.
borrowers. Overall, two sets of sectors gained in share. The The net effect of these multiple factors was a sharp increase
first comprised of retail advances, covering housing loans, in lending to capital-intensive projects, including those in in-
loans for automobile and consumer durable purchases, educa- frastructure, where maturity and liquidity mismatches were
tional loans, and the like. The share of personal loans increased significant. But once this tendency of lending large sums to a
Economic & Political Weekly EPW MARCH 31, 2018 vol lIiI no 13 133
BANKING

single project or business group began, it did not stop with such the practice of treating all restructured assets as standard assets,
projects, but was extended to other areas of corporate lending there was a sharp spike in the share of the non-priority sector
as well. In practice, the failure of these projects/investments in total NPAs from 50% in March 2012 to 77% by March 2016.
to generate the revenues needed to bear the debt service costs The role of big corporate borrowers in this accumulation of
associated with their high debt to equity ratios, led to defaults, bad assets is striking. As of March 2017, large borrowers (with
even in cases where much effort at restructuring was made. exposure of `50 million or more), which were provided 56% of
Thus, underlying the V-shaped movement in the NPA ratio, gross advances, accounted for 87% of the gross NPAs of the
was a post-2003 credit boom and a structural shift in credit SCBs. The corresponding figures for the top 100 borrowers
provision. were 15% and 26%. Post liberalisation, Indian banks were sit-
ting on a pile of debt directed at a few large borrowers, a large
New NPAs share of which was bad.
This obviously had an impact on the nature of the NPA s them- The immediate problem this could cause is a loss of depositor
selves. While the NPA problems of the 1990s stemmed substan- confidence resulting in a run on some banks. This, however,
tially from bad assets arising in priority or non-priority sector was not a danger. In India, inasmuch as banks were dominantly
loans to agriculture and small industry, those after 2003 were publicly-owned, and though some private banks too were
dominated by bad assets arising from large loans to a relatively forced to declare a larger volume of bad loans, the bulk of NPAs
few large corporates, including loans for private investment in were on the books of the PSBs. Given the sovereign backing that
the infrastructure sector. As Table 2 shows, between 1997 and public ownership implied, there was little or no danger that
2003, the non-priority sector (including public sector) accounted these NPAs would in any way disrupt the functioning of these
for around a half or a little more of NPAs in the PSBs. Starting banks. In fact, much higher NPA ratios in the late 1980s and
2006, this share began to decline to 38% in 2008, only to rise early 1990s had no adverse impact. In sum, large-scale recapi-
again to reach earlier levels. One reason for this was the use of talisation was not imperative from a stability point of view.
the CDR scheme, which allowed banks to restructure large However, the officially generated fear of insolvency and the
loans subject to default, through means such as extended argument that meeting Basel guidelines was imperative, was
repayment periods, lowered interest rates, partial conversion used to make recapitalisation an urgent necessity. The finan-
to equity, and additional credit. cial community supported this because it backed their case for
However, it soon became clear that many of these borrowers privatisation of public banks through issue of new equity to
were not in a position to restore normalcy of operations, so ensure recapitalisation. There followed a clamour to revise the
that defaults continued or resumed, forcing the recognition of requirement that PSBs must have at least 52% government
the assets concerned as non-performing. After the RBI insti- ownership of equity, so that additional equity could be sold to
tuted the asset quality review to reclassify assets and reverse private players. The problem, however, was that private inves-
Table 2: Composition of NPAs of PSBs (amount in ` billion) tors were unlikely to buy equity in banks that were burdened
Priority Sector Non-priority Sector Public Sector Total with NPAs. So some way of ridding the banks of their NPA bur-
Amount % Amount % Amount % Amount
den had to be found before they could be “recapitalised.” What
1995 192.08 50.0 178.61 46.5 13.16 3.4 383.85
followed were a series of experiments such as attempted sale
1996 191.06 48.3 190.67 48.2 14.11 3.6 395.84
of bad assets to asset reconstruction companies, and segrega-
1997 207.76 47.7 213.4 49.0 14.61 3.4 435.77
1998 211.84 46.4 231.07 50.6 13.62 3.0 456.53
tion of bad assets in a bad bank and selling some good assets
1999 226.06 43.7 276.08 53.4 14.96 2.9 517.1 and businesses to cover losses. None of them worked, so, the
2000 237.15 44.5 285.24 53.5 10.55 2.0 532.94 government had to finally lead the recapitalisation exercise.
2001 241.56 45.4 273.07 51.4 17.11 3.2 531.74
2002 251.39 44.5 302.51 53.5 11.16 2.0 565.06 New Resolution Framework
2003 249.38 47.2 267.81 50.7 10.87 2.1 528.06 But recapitalisation does not mean the end of the effort at pri-
2004 238.41 47.5 256.98 51.2 6.1 1.2 501.49 vatising public banking. What is surprising is that the policy
2005 215.36 45.2 254.94 53.5 5.92 1.2 476.22 establishment that created the circumstances that led to NPAs,
2006 222.36 53.8 182.79 44.2 8.55 2.1 413.7
with liberalisation and enforced reliance on public bank funding
2007 225.19 58.0 156.03 40.2 7.32 1.9 388.54
for capital intensive projects, and postponing the recognition
2008 248.74 61.5 150.07 37.1 5.74 1.4 404.56
2009 242.01 53.8 205.28 45.6 2.97 0.7 450.26
of NPAs by designing the CDR scheme, all of a sudden turned ag-
2010 304.96 50.9 291.14 48.6 3.14 0.5 599.24 gressive vis-à-vis these same banks. Once the asset quality re-
2011 401.86 53.8 342.35 45.9 2.43 0.3 746.64 view resulted in a spike in NPA ratios and provisioning require-
2012 557.8 47.6 588.26 50.2 26.56 2.3 1172.62 ments, a new “prompt corrective action” (PCA) framework was
2013 672.76 40.9 960.31 58.4 11.55 0.7 1644.61 devised, which placed restrictions on banks as a corrective to
2014 798.99 35.2 1472.35 64.8 1.3 0.1 2272.64 trends indicative of fragility. The PCA framework specifies val-
2015 966.11 34.7 1815.98 65.2 2.59 0.1 2784.68 ues of the capital to risk (weighted) assets ratio (CRAR, ratios
2016 1258.09 23.3 4141.48 76.7 34.82 0.6 5399.57 of core equity to risk weighted assets), net NPA ratios, return on
2017 1609.42 23.5 5237.91 76.5 154.66 2.3 6847.32
Source: Statistical Tables Relating to Banks in India, Table 19, Reserve Bank of India, https://dbie.rbi.
assets values and leverage ratios, which define three levels of
org.in/DBIE/dbie.rbi?site=publications. risk thresholds. A bank breaching any of these thresholds is
134 MARCH 31, 2018 vol lIiI no 13 EPW Economic & Political Weekly
 BANKING

called upon to take corrective action varying from holding back The first is the opposition of the debtors, who would use
on dividend payments, to restrictions on branch expansion, in- every means at their command to prevent liquidation, arguing
creased provisioning, and restrictions on managerial compen- that their default is not the result of errors or failures of the
sation. While these may seem like needed actions, identifica- borrower, but of extraneous circumstances, the burden of
tion of banks as having breached any of these thresholds may which has to be shared by creditors. As noted, the government,
set off developments (such as deposit withdrawals) that weak- which in its official Economic Surveys has described the prob-
en the bank’s position even further. lem as a “twin-deficit” problem (the deficit on the books of bor-
Soon, banks were pushed to opt for the resolution frame- rowers leading to default, on the one hand, and the deficit on
work offered by the Insolvency and Bankruptcy Code (IBC) the books of the lenders, on the other), is sympathetic to this
and the National Company Law Tribunal (NCLT). Having long view, fearing a backlash from business.
delayed the resolution of the problem of stressed assets in the The second is the opposition of those with whom the
banking system, the RBI decided to rely on the IBC as an impor- defaulter has liabilities, but who are not included in the JLF.
tant instrument to address the problem. To do that, the RBI There could also be opposition from other third parties, such
shed its reticence to interfere in the resolution process with as home buyers, as in the case of Jaypee Infratech, who have
support from the government. The latter on its part promul- paid up vast amounts in instalment payments but have not
gated the Banking Regulation (Amendment) Ordinance, 2017, been given possession of the homes they had bought. Defaulting
now passed by Parliament, which introduced new clauses into entities owe money not just to the banks but others, including
the Banking Regulation Act, 1949 permitting the RBI to initiate the tax authorities. To the extent that the IBC favours the banks,
action requiring banks to launch proceedings to resolve bad these “third parties” that would lose out would oppose the resolu-
assets with specifically identified clients. tion. This can delay the process and the results can be messy.
The action has multiple components. To start with, large Third, the JLF members themselves who may want assur-
NPAs that have proved difficult to resolve for a long period of ance that there would be limits on the haircuts they would
time have to be identified. The consortium of banks holding take if liquidation is initiated. The market value of the assets
those assets is given a deadline by which the problem should held by these companies and the strength of the collateral
be resolved. For this, agreement in the Joint Lenders’ Forum needs to be tested, and as other cases such as Kingfisher Air-
(JLF) of 50% of the members involved and 60% of the value lines suggest, there is unlikely to be enough to recover a large
of the loans concerned was adequate. Failing the successful share of the debt and interest due. In 10 cases of resolution under
negotiation of a restructuring solution by the stipulated date, the IBC reported in the Economic Survey 2017–18, the claims of
the banks were required to move the NCLT for initiation of liquida- financial creditors were met in full only in one (Prowess Inter-
tion proceedings. During those proceedings, the incumbent national, for which the claim was extremely small). For the
management was moved out, the creditors were put in control rest, the extent of recovery relative to claims varied from 6%
of the process and an insolvency professional appointed to assist to 58%, with only two recovering more than 50%.
the stakeholders, with definite timelines for resolution or liqui-
dation. A resolution plan had to be in place within 180 days of Government’s Response
referral to the NCLT (with additional 90-day grace period if This failure to recover money lent to top corporates has been
needed). If a plan is not agreed upon within the timeline, then accompanied by an effort to sell off assets to private Asset Recon-
the company will go into liquidation. struction Corporations (ARCs), which could acquire NPAs at a
In a first attempt at implementation of this procedure, the negotiated discount. They make upfront payments of as low as
government notified 12 large NPA accounts in June 2017 for 5% of the sums due, with the balance covered by security receipts
which lending banks were required to file insolvency applica- accepted by the banks from the ARCs, which need to be redeemed
tions. At the end of financial year 2016, the size of debt to the only when the ARCs manage to sell the assets concerned. Thus,
commercial banks of these 12 borrowers varied from `3,802 the ARCs were being contracted to recover a small percentage
crore to `41,843 crore, with seven of them burdened with un- of the total NPA value, with their fee depending on the differ-
serviceable debt of more than `10,000 crore. Their combined ence between the acquisition and sale price. The result has
debt totalled `2,26,400 crore. These accounted for as much as been that when the discount on NPAs sold by banks was sought
a quarter of the total NPAs on the books of the SCBs. to be lessened, the volume of NPAs sold reduced.
Even while these cases were being directed to the NCLT and These “failures” only confirm the suspicion that the govern-
the National Company Law Appellate Tribunal (NCLAT), the ment’s decision to recapitalise banks does not mean a commit-
government had flagged more cases of bad, high value debt, ment to provide state backing for the PSBs. Rather, the effort to
and called for their resolution in six months, failing which,
they too would be considered for reference to the NCLT. But the
process seems to have accelerated with the RBI reportedly issu-
ing instructions for proceedings to be launched against 40 or available at
more borrowers, whose NPAs are large and chronic. Churchgate Book Stall
However, it is becoming clear that the problem is not easily ad- Churchgate Station, Opp: Indian Merchant Chamber
Mumbai 400 020
dressed. There are three kinds of difficulties that the process faces.
Economic & Political Weekly EPW MARCH 31, 2018 vol lIiI no 13 135
BANKING

clean the books of the banks may be aimed at preparing them If no agreement for restructuring could be arrived at between
for market, since banks burdened with bad debt and/or under- the borrower and its lenders, liquidation proceedings against
capitalised are unlikely to command a reasonable price for the borrower were to be launched to recover as much of the
their equity. Once banks are recapitalised, the probability of loan as possible.
raising capital from the market by sale of public bank equity at But, as argued earlier, proceedings at the NCLT suggest that
reasonable prices is high. this effort can at best be a partial solution, since, among other
One problem here is that if the banks concerned are to things, finding assets that can cover the defaulted loans is not
remain “public” with at least 51% of equity owned by the gov- easy. Large write-offs are inevitable. So if the government is to
ernment, the headroom available for stake sale may be limited wash its hand of the bad debt problem, and the likelihood of
because of past disinvestment. Besides private entry, an im- debts going bad remaining high even after recapitalisation,
portant component of the transformation of banking engi- other measures of resolution are needed.
neered by liberalisation was a restructuring of PSB ownership. The FRDI Act defines the resolution mechanisms being
This meant that it was not just weak PSBs that were made can- pushed by the government, as an alternative to recapitalisa-
didates for equity dilution. tion. At the centre of the new scheme is the creation of a new
Early in the liberalisation era, in December 1993, the State independent corporation that would take over the task of reso-
Bank of India, with paid-up capital of `200 crore chose to go in lution of bankruptcy in banks, insurance companies, and iden-
for a public issue of shares worth `274 crore at par, but sold at a tified “systemically important financial institutions” (SIFIs).
premium of `90 per share. The shareholding of the RBI and the The FRDIC will also take over the task of insuring bank depos-
Government of India (together) came down to 66.3%, with the its, compensating depositors up to a specified maximum
remaining 33.7% being held by other entities. That was only amount (at present `1 lakh), in case of bank failure.
the beginning. Out of 26 PSBs (including the 19 nationalised As part of its responsibilities, the corporation is to be man-
banks, the State Bank Group and IDBI Bank), as many as half dated to classify the financial institutions under its jurisdiction
that number had no private shareholding even as late as 2002, under different categories based on risk of failure, varying
and only two had private shareholding in the maximum possi- from “low” and “moderate” (or in whose case the probability of
ble 40%–49% range. But in the decade that followed, dilution failure is marginally or well below acceptable levels), to “mate-
has been rapid, so much so that as many as 14 banks had pri- rial” or “imminent” (implying failure probabilities that are
vate shareholding in the 40%–49% range by end-March 2012. above or substantially above acceptable levels), and finally
Another 10 fell in the 20%–40% private shareholding range. critical (or being on the verge of failure). In cases of financial
Private holdings include foreign ownership of equity in 24 out firms placed under the material or imminent category, the
of the 26, with the extent of such ownership varying from 0.1% Resolution Corporation is to be given the power to: (i) inspect
(State Bank of Mysore) to 17.4% (Punjab National Bank) as at the books to obtain information on assets and liabilities;
end-March 2012. (ii) restrict the activities of the firm concerned; (iii) prohibit or
limit payments of different kinds; and (iv) require submission
Financial Resolution and Deposit Insurance Act of a restoration plan to the regulator and a resolution plan to
There is other evidence that an important element of the gov- the FRDIC, if necessary involving a merger or amalgamation.
ernment’s approach to dealing with the problem of high NPAs In cases identified as critical, the FRDIC will take over their ad-
is to try and socialise PSB losses without the intervention of the ministration, and proceed to transfer their assets and liabilities
budget, through the creation of a new debt resolution mecha- through merger or acquisition or to liquidate the firm with per-
nism and authority. On 10 August 2017 the government tabled mission from the NCLT. To leave no choices open, the law pro-
a new bill in Parliament, with the aim of using its majority to hibits recourse to the courts to stay the proceedings at the
push through a desperate policy initiative in the form of the NCLT or seek alternative routes to resolution. Since liquidation
Financial Resolution and Deposit Insurance (FRDI) Act. The act involves compensating stakeholders according to their designated
seeks to create an ostensibly “independent” FRDI Corporation seniority, depending on the net assets available, any stakeholder
(FRDIC), which would take over the task of resolution of failing can be called upon to accept a “haircut,” including holders of de-
financial firms from the RBI and other regulators. To that end, posits in excess of the maximum specified as insured against loss.
it is to be armed with special and near draconian powers to There are many implications of this act. To start with, while
implement its mandate, and given control of the deposit insur- the independent FRDIC and the concerned regulator will deter-
ance framework currently managed by the Deposit Insurance mine whether a financial firm is to be placed in the material or
and Credit Guarantee Corporation of India. imminent category, the task of working out an acceptable res-
As a first step to address the problem, the government prom- toration or renewal plan rests with the firm under scrutiny. So
ulgated the Banking Regulation (Amendment) Ordinance, 2017, the responsibility of restoring viability is that of the bank,
which introduced new clauses into the Banking Regulation insurance company or SIFI, with the regulation and resolution
Act, 1949. These clauses meant that the government could au- authority retaining the right to determine whether this has
thorise the RBI to take special action to resolve the bad debt managed to reduce the probability of failure.
problem. This would involve forcing banks to launch proceed- Second, since mere categorisation in the material or imminent
ings against identified borrowers to recover their unpaid dues. category will send out a signal, banks so designated can become
136 MARCH 31, 2018 vol lIiI no 13 EPW Economic & Political Weekly
 BANKING

the target of a run, as depositors fearing failure would want to the government is obviously willing to write off capital already
move out their deposits. As a result, instead of resolving the invested, but wants to minimise any additional costs. This
problem of vulnerability to failure, the mechanism may pre- way, the mechanism of socialising private losses is transferred
cipitate failure. out of the budget so that its effects are directly borne by the
Third, the restoration and/or resolution plan, to be acceptable, larger “public.”
may “force” the financial firm to accept amalgamation or It is in this odd way that the contradictions inherent in neo-
merger. This would have implications for parties that are not liberal banking reform are being addressed. It should be obvi-
responsible for the state of the firm, including officers, employees, ous that, if the current financial and economic policies are per-
creditors, and small shareholders. For example, retrenchment sisted with, bank losses would continue to rise, as the return of
or downgrading of the status of employees may follow merger NPAs to the books of the PSBs after 2003 makes clear. If that
and amalgamation. And where resolution requires the pre- cannot be prevented, the only option is to force banks to re-
ferred strategy of “bail-in” of the firm, shareholders, creditors, structure when they cross some threshold level of NPAs in their
and if need be, depositors, would be forced to accept a “haircut” books. The FRDI’s hope is that this can be done without knocking
or loss. The unstated objective of the exercise is to save the on the government’s doors. But if experience worldwide is any
government and the regulator from carrying the costs of a guide, this would only precipitate a different kind of crisis,
bailout of the failing firm. forcing the government to step in. The market cannot resolve
Thus, the tabling of the FRDI Bill is a clear declaration by the its own problems.
government that it sees painful resolution or liquidation as a
way out of addressing the bad debt problem that currently notes
afflicts the banking sector in particular. It also makes clear 1 Figures from, Statistical Tables Relating to Banks in India, Reserve Bank of
India, Table 19, https://dbie.rbi.org.in/DBIE/dbie.rbi?site=publications.
that the finance ministry, the central bank, and the government- 2 Figures quoted in Verma (2017).
sponsored regulators will not carry any of the financial burden 3 Figures are from Database of the Indian Economy, RBI.
4 All figures from the RBI’s database at www.rbi.org.in.
associated with resolution, but rather would transfer financial
and other costs (such as job losses) to the employees, officers References
and shareholders, and even depositors holding deposits in excess GoI (2017): Economic Survey 2016–17, Vol 1, Ministry of Finance, Government of
India.
of the insured amount. Since the problem of potential insol- Verma, Sunny (2017): “Non-performing Assets: Govt-run Banks Write off Re-
vency is at present concentrated in the public banking system, cord Rs 81,683-crore Bad Loans in FY17,” Indian Express, 7 August.

EPW’s Special and Review Issues Published in 2017

Economic & Political Weekly EPW MARCH 31, 2018 vol lIiI no 13 137