PROGRESS,
CONSTRAINTS AND LIMITATIONS OF FINANCIAL SECTOR REFORMS IN
THE LEAST DEVELOPED
COUNTRIES*
Martin
BROWNBRIDGE and Samuel K GAYI
1.
Introduction
Many developing countries implemented financial sector
reforms, as part of broader market oriented economic reforms since the late
1980s. This paper evaluates the achievements, limitations and constraints of
financial sector reforms implemented in eight Least Developed Countries (LDCs):[i]
Madagascar, Malawi, Tanzania, Uganda and Zambia in sub-Saharan Africa (SSA),
and Bangladesh, Laos and Nepal in Asia. All are low-income countries
characterised by shallow and undiversified financial sectors, which have
experienced considerable financial fragility. The research is based on an
UNCTAD project undertaken during 1995-97.
The
financial sector reform programmes implemented in the eight LDCs entailed
financial liberalisation and institutional reforms to systems of prudential
regulation and supervision and distressed public sector banks. The reforms
began in the mid- to late- 1980s in some countries, and in the early 1990s in
others. In none of the countries were the reform programmes complete in the
mid-1990s. The banking system dominates the financial sectors of these LDCs and
has been the major focus of the reforms, although some of the countries have
begun reforms intended to develop capital markets and encourage the growth of
non bank financial institutions (NBFIs) such as leasing and finance companies.
This paper concentrates on the impact of the reforms on the banking system.
The
objectives of the reforms are to build more efficient, robust and deeper
financial systems, which can support the growth of private sector enterprise.
Efficiency entails two components: improved credit allocation (i.e., credit
allocated to borrowers with higher expected returns for given levels of risk)
and more, or higher quality, financial services for a given level of inputs
(e.g., bank staff). Improved credit allocation could be derived from reduced
government intervention in directing credit or setting interest rates so that
banks would have more freedom to allocate credit according to commercial
criteria. The second component of efficiency could be brought about through
increased competition, with competition resulting from liberalised entry and/or
removal of regulations, which restrict competition, such as interest rate
controls. But this assumes that financial markets will be competitive and not
oligopolistic.
More
robust financial systems are those less vulnerable to financial fragility.
Financial fragility can impose heavy costs on taxpayers and disrupt the real
economy through reduced availability of credit and other services such as
payments (Goldstein and Turner, 1996; Llewelyn, 1997). Financial sector reforms
can contribute to enhancing robustness through three mechanisms; reducing one
of the causes of fragility, government direction of credit to uncreditworthy
borrowers; restructuring distressed banks; and strengthening prudential
regulation and supervision.
The
literature includes some scepticism about the efficacy of financial
liberalisation in low-income countries. Financial liberalisation has not always
brought about the expected benefits: there have been few innovations in
financial markets, competition is limited by oligopoly and liberalisation may
exacerbate urban bias (Chandavarkar, 1992). There is doubt as to whether higher
real interest rates encourage greater financial saving, and thus deepen the
financial system.
Market
failure, arising from informational imperfections, is pervasive in financial
markets, and are especially severe in rural areas (Stiglitz, 1994). Market
failures may prevent liberalisation from improving the efficiency of credit
allocation. In particular, potentially profitable borrowers may be denied
credit because of high informational and transactions costs. In Africa the
severe informational problems afflicting financial markets suggest that even
the long term benefits of liberalised financial systems may be small, while in
the short term, financial liberalisation might actually worsen the efficiency
of intermediation because, lacking information about firms' expected
profitability, banks lend on the basis of collateral values (Collier, 1994).
The
segmentation of financial markets between formal and informal sectors impedes
the efficacy of liberalisation to enhance competition and efficiency (Aryeetey et
al, 1997). The liberalisation of financial markets may lead to financial crisis
unless preceded by macroeconomic stabilisation and prudential reforms (Alawode
and Ikhide, 1997; McKinnon, 1988). Caskey (1992) questions the costs of
financial sector reform programmes, in particular rehabilitating government
banks at government expense and the scarce human resources, which the reforms
absorb.
The
rest of the paper is organised as follows. The next section outlines the main
components of the financial sector reforms implemented in the eight LDCs.
Section 3 assesses the impact of the reforms on financial depth, on competition
in the banking system and on credit allocation, factors related to the
efficiency of financial intermediation. The subsequent section examines the
extent to which the reforms have created more robust financial systems. It
analyses the causes of financial distress and evaluates measures to
rehabilitate government banks and the reforms to the prudential system. A final
section concludes.
2.
What Financial Sector Reforms were Implemented in LDCs?
2.1
Pre-reform policies
The eight LDCs covered in this paper began their reforms
from different starting points for a variety of reasons, but mainly because
they experienced different levels of financial repression or financial sector
inefficiency, which reflected differences in pre-reform financial sector
policies. In addition, some governments
were slow in identifying the problem and acknowledging the need for reforms.
Pre-reform policies
in all eight LDCs included controls over interest rates, and most used a
variety of lending directives, rediscount facilities or special lending schemes
designed to increase the volume of bank credit extended to priority sectors
such as agriculture, often at preferential interest rates. Government ownership
of banks was important in all eight LDCs, but the share of the private sector,
if any, and the role of market forces in financial intermediation varied
considerably between countries.
In
Laos, Madagascar and Tanzania government-owned financial institutions had a
monopoly of formal sector financial markets: private sector banks had been
nationalised and new entry from the private sector was not allowed. Financial
resources were allocated according to administrative directives, and banking
was essentially a form of quasi government financing for state owned
enterprises (SOEs), rather than genuine financial intermediation. Until 1988
Laos had a soviet style banking system comprising a single monobank incorporating
both central and commercial banking functions.
The
other five LDCs had mixed banking systems comprising government and private
sector banks, or joint ventures between government and the private sector,
although except in Zambia, wholly owned private sector banks commanded much
less than half the banking market. In these countries government-owned banks
were expected to pursue a variety of non-commercial objectives, such as lending
to SOEs and small farmers. In Bangladesh, Nepal and Uganda the government banks
also undertook major branch expansion programmes in the rural areas in response
to government directives. The private sector banks were operated along
commercial principles, although they were constrained by the controls imposed
by the central banks on interest rates and where applicable, sectoral credit
directives and controls on the location of branches.
2.2
Components of Reforms
The reforms implemented in the eight LDCs included
liberalisation of financial markets and institutional reforms to the prudential
regulatory framework and government banks. Annex table 1 lists the main
components of the reforms in each country.
In all
eight LDCs interest rates were liberalised. With the exception of Zambia (where
liberalisation was implemented in a short space of time), this took place in a
phased manner lasting several years, which began with administered interest
rates being raised before partial and then full decontrol was implemented.
Treasury bill (TB) auctions were introduced in Nepal, Uganda, Malawi, Tanzania
and Zambia to allow a role for market forces to influence TB rates and to
facilitate the use of indirect techniques of monetary control. Most of the
LDCs, but not Nepal, liberalised sectoral credit directives.
All the
LDCs have allowed new entry by private sector banks and other financial
institutions, although only in Tanzania, Laos, Madagascar and Malawi were entry
restrictions liberalised by the reforms. In Laos the private banks are
restricted to the capital city. Nepal, Bangladesh, Uganda and Zambia all
licensed private sector banks and financial institutions during the 1980s
before their reform programmes began, at a time when prudential regulations
were weak and hence prudential requirements on new entrants were low. In these
countries the revisions to the banking legislation raised entry barriers in
terms of minimum capital, expertise of promoters, etc, once this legislation
had been enacted in the late 1980s or early 1990s. All the LDCs have allowed
entry by foreign as well as domestic banks, but in Nepal foreign investment
must be in the form of joint ventures with domestic partners with foreign
ownership restricted to 50 percent of the equity.
New
banking legislation, comprising stronger prudential regulations, was enacted in
all the LDCs, and was accompanied by institutional reforms to strengthen the
supervisory capacities of the central banks. All the LDCs implemented measures
to tackle the financial distress in their government-owned banks, including
recapitalisation and reforms to the management and operations of these banks.
These reforms are discussed in greater detail in section 4. In Laos, the
reforms began with the single monobank being split into a central bank and
state-owned commercial banks in 1988.
There were
differences in the timing and sequencing of reforms between the eight LDCs, but
some common features are evident. Most of the countries had implemented
stabilisation programmes, beginning in the mid- or late 1980s, several years
before embarking on the main programme of financial sector reforms.
Stabilisation measures had included raising controlled nominal interest rates,
as noted above, in order to achieve positive, or less negative, real interest
rates. The main programme of financial sector reforms in the eight LDCs began
in the late 1980s or early 1990s. In most of these programmes liberalisation
was begun alongside the institutional reforms to the prudential systems and the
distressed banks, but the institutional reforms, and in particular the reform
of distressed banks, took much longer to implement than the removal of
administrative controls. While most of the latter had been abolished by the
mid-1990s, many of the institutional reforms were still ongoing in 1996/97.
3.
Have Reforms Enhanced the Efficiency of Financial Intermediation?
3.1
Financial Deepening
A major objective of financial sector reforms is to
boost financial depth, and therefore increase the resources available for
financial intermediation. The main channel through which this should occur are
interest rate reforms (raising controlled rates or deregulating interest
rates), which are intended to lead to higher real deposit rates and hence price
incentives for depositors. Greater non-price competition among banks for
deposits might also boost deposit mobilisation.
The
impact of the financial sector reforms on financial depth, as measured by bank
deposits and M2 as percentages of GDP, varied between countries. In each of the
three Asian LDCs, a marked financial deepening took place. In both Bangladesh
and Nepal, bank deposits increased by around eight percentage points of GDP
between 1985 and 1995, while in Laos bank deposits rose by almost seven
percentage points of GDP between 1990 and 1995, albeit from a very small base
(see table 1). In contrast there was little change in financial depth in
Madagascar and a small decline in Malawi (the growth of NBFI deposits, which
are excluded from bank deposit and M2 data, at least partly explain the decline
in Malawi). Tanzania suffered a sharp contraction of financial depth in the
second half of the 1980s but recovered almost half of the fall in the first
half of the 1990s. In Uganda, a small recovery was achieved in the first half
of the 1990s after the collapse in financial depth in the 1980s, but the
financial system remained very shallow. In Zambia the reforms were unable to
prevent continued rapid decline in financial depth which began in the first
half of the 1980s.
The
better performance of the Asian LDCs in enhancing financial depth is probably
attributable to the greater macroeconomic stability in these countries.
Inflation rates were relatively moderate in the three Asian countries, which
has allowed real deposit rates to be generally positive (in Bangladesh and
Laos) or only marginally negative (in Nepal): see table 2. In contrast, the
five African LDCs suffered higher, and more volatile, rates of inflation. In
these conditions it has been difficult to maintain deposit rates at positive
real levels, especially during bouts of very high inflation. Moreover high
inflation made rates of return on current account deposits, which account for a
large share of deposits, steeply negative. Two other factors may also have
contributed to the lack of financial deepening in the African LDCs. First, the
removal of foreign exchange controls allowed residents to purchase and hold
foreign currency legally, while rapid exchange rate depreciation would have
made the holding of foreign currency assets more attractive relative to
domestic currency assets.[ii]
Second, the introduction of TB auctions led to steep rises in TB rates, often
surpassing time deposit rates, and this is likely to have led some of the
larger depositors to substitute TBs for time deposits (Adam, 1995).
3.2
New Entry and Competition
Financial liberalisation is intended to stimulate
greater competition in banking markets through two channels: new entry by
private sector banks to challenge the oligopolistic market position of the
established public sector and/or foreign banks, and the removal of
administrative constraints on competition such as the interest rate controls.
This should in turn lead to an improvement in the quality, and lower cost, of
services offered to the public, as banks compete for business. Competition may
also encourage banks to provide a broader range of financial products in an
attempt to attract business.
In all
the eight LDCs the reforms facilitated new entry, mainly by the private sector,
into banking markets. The number of banks and financial institutions increased,
and the dominant market share of the major banks was eroded, although it
remained large.
In
Laos, eight private sector banks had captured 30 percent of the deposit market
since banking markets were opened to the private sector in 1989. Eight joint
ventures banks had been set up since the mid-1980s in Nepal. There were 20
private sector banks in Bangladesh in the mid-1990s with 28 percent of the
deposit market. Two new private sector banks were set up in Madagascar while
the government sold equity in two of the three government-owned banks to
foreign banks.
Considerable
new entry by domestic private sector banks occurred in both Uganda and Zambia
(although in both countries new entry preceded the start of financial sector
reforms). Since the mid-1980s, eight new banks were set up in Uganda, and 15 in
Zambia, of which six were later closed down by the central bank. The domestic
private sector banks had captured deposit market shares of approximately 15
percent and 20 percent in Uganda and Zambia respectively at the end of 1995.
New entrants have been fewer and more recent in both Malawi and Tanzania,
reflecting a more cautious licensing policy by the authorities. In Tanzania,
five private sector banks have been set up since 1993 while in Malawi two new
banks were set up in the first half of the 1990s. In Malawi, Zambia and Nepal,
the reforms also facilitated the growth of NBFIs, such as leasing companies and
building societies. The NBFIs provided some diversification in lending products
and competition for deposits (usually wholesale deposits) with the commercial
banks.
Increased
competition stimulated some improvements in financial services. Some of the new
entrants introduced longer openings hours, cut queues in banking halls and
provided more personalised services. A number of innovations occurred and new
products were made available: these included credit and debit cards, automated
teller machines (ATMs), interest bearing current accounts, and savings accounts
with cheque books. Cheque clearing has been speeded up. Competition for
deposits increased in the urban areas, with both price and non-price
competition. There is more competition for corporate clients, especially from
the entry of foreign banks, which focus on this sector. The government-owned
banks are making efforts to improve services and to provide services oriented
to customer needs. However the impact of new entrants on the cost, quality and
range of financial services has been limited for a number of reasons.
First,
although the major government and/or foreign banks lost some of their market
share to the new entrants, they still retain a large enough share to exercise a
degree of oligopoly. This has enabled them to maintain large interest rate
spreads, needed to cover the cost of their own inefficiencies and of their
non-performing loans. Second, the slow pace of reform of some of the government
banks retarded improvements in the cost and quality of their services. Third,
with few exceptions,[iii] the new
entrants (both foreign and domestic banks) have avoided the rural areas, hence
what benefits that have occurred have been confined to the urban areas. Some
rural areas are likely to have suffered deterioration in the availability of
financial services as a result of branch closures by government banks. Fourth,
competition has been impeded because banking markets, particularly credit
markets, are segmented. The foreign banks serve large, and especially foreign,
corporate customers, the government-owned banks remain focused on SOEs, or privatised
SOEs, while the domestic private sector banks and NBFIs mainly lend to local
urban-based SMEs and to the informal trade and service sector.
3.3
Impact of the Reforms on Lending
Interest rate liberalisation, together with the removal
of allocative credit directives and the adoption of commercial lending policies
by public sector banks, is intended to enhance the efficiency of credit
allocation, by allowing the price mechanism and the commercial judgement of
bankers to determine credit allocation (Fry, 1988). A crucial premise
underlying liberalisation is that inefficiencies in credit allocation arising
from market imperfections such as imperfect information are less important than
government failures arising from directed credit policies. It is also assumed
that there exists demand for loans from creditworthy borrowers with profitable
investment opportunities, which would be denied credit under a repressed
financial system because administrative controls or the non-commercial lending
policies of public sector banks channel the available credit to less efficient
borrowers, such as loss making SOEs. Hence liberalisation is expected to allow
a reallocation of credit towards the users most capable of generating higher
rates of return to capital. Liberalisation could also reduce the pressure on
banks to accommodate less creditworthy borrowers and therefore lead to an
improvement in the quality of their loan portfolios. (Gelb and Honahan, 1991: passim)
An objective
assessment of whether the financial sector reforms have actually brought about
a more efficient allocation of credit is very difficult. Ideally one would need
data on private and social rates of return earned by borrowers plus a
counterfactual with which to make a comparison. Such data are not available.
Instead, we examine the volume of bank credit to the private sector on the
grounds that the private sector is assumed to use resources more efficiently
than the public sector. Moreover the economic reforms being implemented by all
eight LDCs aim to boost incentives for private investment, hence if the banking
system responds to this change in incentives, it should extend more credit to
the private sector. Other factors, however, may have a bigger influence than
financial sector reforms on the volume of credit extended to the private
sector, not least macroeconomic variables such as the government borrowing
requirements and monetary policy, hence observed changes in private sector
borrowing are not necessarily attributable to financial sector reforms.
Bank
credit to the private sector as a share of GDP has expanded strongly in Nepal
since the mid-1980s and in Laos during the 1990s. In the latter case this was
attributable to the privatisation of SOEs. There was a small increase in
Bangladesh. Zambia, Malawi, Madagascar and Tanzania all suffered sharp declines
in private sector bank borrowing during the 1990s. Uganda registered a small
increase, but from a negligible base. In all four mainland African LDCs,
private sector bank borrowing amounted to less than 10 percent of GDP in the
mid-1990s.
Clearly
the efficacy of the reforms in terms of enabling banks to channel more credit
to the private sector faced major constraints in the African LDCs. The lack of
financial deepening constrained the funds available to banks, and large
government domestic borrowing requirements crowded out the private sector.
Moreover, in conditions of high inflation, as in Zambia, banks were very
reluctant to extend credit to private sector borrowers even when funds were
available, because of the fear that borrowers would not be able to service the
very high nominal interest costs of the loans. High interest rates had
contributed to widespread loan defaults by farmers hit by drought in Zambia in
the early 1990s. Banks instead invested in TBs, a safer alternative.
It is
likely that financial liberalisation has also affected the sectoral allocation
of credit. Banks probably extend less credit to agriculture and to small farmers
in particular because of the removal of lending directives, the cut back of
special lending and refinancing schemes and closure of rural branches, although
some banks fund the larger commercial farmers.[iv]
The share of agriculture in total bank lending has fallen sharply in Malawi
since the mid-1980s and in Bangladesh during the 1990s. Because of the high
administrative costs, informational problems and difficulties in enforcing loan
repayment, lending to small farmers on a purely commercial basis is unlikely to
be viable for the banks.[v]
4.
Have Reforms Led to More Robust Financial Systems?
All the LDCs experienced banking crises involving
financial distress in one or more major banks in the late 1980s and/or early
1990s, with the exception of Malawi where financial distress had been diagnosed
and effectively tackled in the early 1980s prior to the reforms. This section
assesses whether the reforms have made banking systems more robust; i.e., less
vulnerable to financial distress. We start by examining the causes of financial
distress and also discuss whether financial liberalisation may have contributed
to distress. We then assess the efficacy of the remedial measures taken to deal
with distressed banks and the reforms to strengthen prudential regulation and
supervision.
4.1
Causes of Financial Distress
Two distinct types of financial distress afflicted the
banking systems of the LDCs. The most important in terms of the scale of
losses, and the fact that it was common to all LDCs, entailed endemic distress
in government commercial banks; banks which, except in Zambia, accounted for
over 50 percent of the banking market. Most of these banks were insolvent and
required some form of recapitalisation by the government.[vi]
The second, which was important only in Bangladesh, Uganda and Zambia, involved
domestic private sector banks with much smaller shares of the banking market.
Six private sector banks were closed down by the central bank in Zambia in 1995
and 1997, while in Uganda one was closed down and two were restructured in the
mid-1990s. Non-performing loans were at the core of both types of distress, but
the causes of poor loan quality were different.
Honohan's typology
of bank crises distinguishes between (i) epidemics caused by macroeconomic
shocks; (ii) epidemics caused by poor management and microeconomic
deficiencies; and (iii) "endemic crises in government permeated banking
systems" in which banks were subjected to non commercial principles which
undermined their solvency (Honohan, 1997: 2-10). Using this typology the
distress to the government banks in LDCs is an example of (iii), and that to
the private sector banks an example of (ii).
Political
pressure to lend to uncreditworthy borrowers was the main reason why the
government-owned banks incurred substantial levels of non-performing loans (in
Nepal, Tanzania and Uganda non-performing loans accounted for between 60 and 80
percent of the total loans of the government-owned banks). Except in Uganda and
Malawi, the largest share of their bad debts were to state-owned enterprises
(SOEs). Most of the SOEs were not profitable, but because of political
pressure, backed up in some cases by government loan guarantees and collateral
which was to prove of limited value, the government banks had little choice but
to fund this sector. The government-owned banks also incurred bad debts from
lending to farmers, especially in Bangladesh, Malawi, Nepal and Uganda, usually
as part of government schemes aimed at supporting agricultural development, and
to politically connected private sector borrowers. Governments pressured their
banks not to foreclose on borrowers when they defaulted, and in some cases
encouraged default by periodically announcing that loans to farmers would be
rescheduled or written off after droughts or other disasters, as in Bangladesh.
Even where banks did pursue defaulters, weaknesses in the legal systems often
meant that foreclosure was a very difficult and lengthy process.
The
loan portfolios of the government-owned banks were further undermined because
in most cases their own policies: procedures, and capacities to appraise loan
applicants and to monitor and recover loans, were very weak, as were internal
controls. Furthermore, the banks were overstaffed and had overextended their
branch networks; hence operating costs were high.
The
distress afflicting the private sector banks in Bangladesh, Uganda and Zambia
was mainly a result of poor management and fraud. Insider lending was a major
contributor to their bad debts: directors took loans from their own banks and
failed to repay. Loan quality was further impaired by a failure to adequately
diversify loan portfolios and by the adverse selection of many of the banks'
borrowers: the banks lent at high interest rates to borrowers in the least creditworthy
segments of the market where default rates were high. Undercapitalisation
afforded them little protection against distress when problems afflicted their
loan portfolios. Incentives on owners for prudent management were weak because
of undercapitalisation, concentration of ownership in one man or family, and
deficient and poorly enforced prudential legislation (Brownbridge, 1998).
With
few exceptions, the foreign owned banks, including the joint ventures in Nepal,
avoided financial distress in all of the countries where they were allowed to
operate.[vii]
They were much less subject to politically pressured lending than were the
government banks, and most had experienced management, prudent lending policies
which focused on creditworthy corporate customers, and good internal controls
which prevented the type of lending problems which afflicted some of the
domestic private sector banks.
Financial
liberalisation, specifically liberal licensing criteria and the decontrol of
interest rates, may contribute to financial crises if undertaken too abruptly
or if poorly sequenced with reforms to prudential regulation, the real sector
and macroeconomic stabilisation (Alawode and Ikhide, 1997). The rest of this
sub-section considers whether the financial liberalisation, which was
undertaken in the LDCs contributed to the financial distress in their banking
sectors.
Financial
liberalisation was not a significant contributor to the distress afflicting the
government-owned banks. This was caused by politically pressured lending, a
characteristic of the controlled financial markets. Their distress was chronic
and pre-dated financial liberalisation in all the LDCs covered here. In most
cases the government-owned banks were already insolvent and reliant upon
liquidity support from central banks for several years prior to financial
liberalisation. The financial distress in these banks was concealed by poor
accounting practices for years while losses mounted, but the true scale of the
losses only became apparent after the financial sector reform programmes began,
when new loan classification and provisioning rules were introduced, when
limits were placed on liquidity support by the Central Bank, and when external
audits were carried out as a prelude to the restructuring of the banks. Further
losses by government banks were undoubtedly incurred after financial markets
were liberalised, but this was mainly because lending practices did not change
rather than as a result of lending in liberalised markets - the banks continued
to fund loss making SOEs, in a few cases, because of political pressure.
The
link between financial liberalisation and the distress of the private sector
banks in Bangladesh, Uganda and Zambia is more ambiguous. These banks were
licensed at a time when prudential criteria for entry, such as minimum capital
requirements, were low, when other prudential regulations were deficient and
when supervisory capacities were weak,[viii]
although not all banks, which were licensed in this period suffered distress.
By itself, however, the licensing of private sector banks does not amount to
financial liberalisation, especially as licensing procedure was generally not
transparent, and the set of criteria for accepting or rejecting applicants for
bank licenses was neither clearly defined nor transparent. The liberalisation
of interest rates probably made some contribution to the distress in the
private sector banks, at least in Zambia where nominal lending rates rose to
over 100 percent following liberalisation and undoubtedly made lending much
more risky. But the distress resulting from insider lending would have occurred
irrespective of whether interest rates were controlled or liberalised.
Wilful default by some
borrowers because of political interference, or simply weak bank management, is
a major contributor to financial distress in both government-owned and private
sector banks. Experiences of
all the LDCs studied demonstrate that it is often difficult, if not contradictory,
for commercial banks to attain social objectives, (for example, poverty
reduction), as well as operate profitably, or along commercial principles. The
management and work culture have to be improved substantially in financial
institutions in LDCs, especially in the private sector banks, if they are to
survive the competition unleashed by financial liberalisation. Even if capital markets become fully
operational, it may take a while for private sector banks to be able to raise
equity capital on them. The management
of these banks would need to demonstrate its capacity to run banks efficiently
and profitably, given the bad track record of co-operative unions in Tanzania
and Uganda, for example.
4.2
Reforms to Distressed Government-owned Banks
Reforms to distressed government-owned banks were
implemented, and in most cases were still ongoing in 1997, in all the LDCs,
with the aim of creating solvent and commercially viable banks. The scale of
the distress in the government-owned banks necessitated some form of balance
sheet restructuring: with one exception - Zambia National Commercial Bank
(ZANACO) in Zambia - they received some form of recapitalisation from the
government budget. They also undertook, or were planning to undertake, reforms
to their management and operational structures. Rehabilitating these banks has
proved to be a costly, lengthy and difficult process, and only in Malawi has it
been an unambiguous success.[ix]
None of the insolvent government commercial banks was liquidated and only in
Madagascar had any been privatised or otherwise divested at the time of
writing.[x]
The
circumstances in which rehabilitation efforts were, and are, being carried out
are not propitious. Commercial banking and accounting skills are scarce,
especially in those countries which previously had socialist banking systems
without any private sector involvement, government financial resources are very
limited, there has been only limited restructuring of major borrowers in most
countries, and the legal framework for enforcing financial contracts is weak
(Sheng, 1996). Moreover successful rehabilitation may entail significant
economic and political costs in terms of lost jobs, closure of rural branches
and lending programmes for farmers, and the curtailment of credit to loss
making SOEs and to politically influential private sector borrowers.
Unlike
in the other LDCs, the financial problems afflicting the two Malawian
commercial banks (both were joint ventures with the private sector) were
diagnosed at an early stage, at the start of the 1980s, and then tackled
promptly, while the scale of the losses was still relatively manageable. The
banks received a capital injection from the government to compensate for some
of their bad loans, but more importantly for their long term viability, their
largest borrower was itself restructured and restored to profitability,[xi]
while the banks strengthened their lending policies and loan recovery efforts.
Moreover the political pressure to lend to priority sectors, which was a cause
of many of their bad debts, appears to have eased. As a consequence both banks
have avoided any further distress.
In the
other LDCs, there were delays in both recognising the scale of the problems
afflicting the distressed banks and in implementing rehabilitation measures.
Major restructuring programmes to Uganda Commercial Bank (UCB) in Uganda and
the National Bank of Commerce (NBC) in Tanzania were begun in 1992 but were
still ongoing in 1996. Bad debts from both banks were transferred to specially
created debt recovery agencies,[xii]
in return for which the banks received, or were due to receive, government
bonds alongside other forms of recapitalisation, such as the write-off of
liabilities to government. Recapitalisation will place a considerable burden on
government budgets. Both banks had also retrenched more than 50 percent of
staff, closed branches and were revamping credit policies and internal
controls. Despite these efforts, neither bank had yet been restored to solvency
by 1995/96 nor was making operating profits. The government was planning to
privatise UCB, but the sale had not taken place by early 1998.
Attempts
were first made to tackle the distress in the three government-owned banks in
Madagascar in 1988. The banks received interest free government loans, wrote
off bad debts and revalued fixed assets. But these measures were unsuccessful
and in 1992 administrators were appointed to manage two of the government-owned
banks,[xiii]
one of which required a capital injection from the government in 1996. The
evaluations of the two banks, carried out as part of the preparations for
privatisation, were only completed in 1997.
Restructuring
plans for the two government-owned banks in Nepal were formulated in 1992
following diagnostic studies undertaken two years earlier. The government
recapitalised both banks in 1992/93 but by 1996 only the Nepal Bank (NBL), a
joint venture had made any progress in implementing the restructuring plans. It
had cut staff and improved credit policies, but its true financial condition
could not be assessed because it did not publish up-to-date accounts. The other
bank, the Rastriya Banijya Bank (RBB), had made few efforts to implement
reforms, continued to lend heavily to loss-making SOEs and was unable even to
provide the central bank with basic balance sheet data. There appeared to be a
lack of political will to tackle the very severe managerial and operational
problems afflicting this bank.
Progress
had also been limited in Bangladesh. The four nationalised commercial banks
(NCBs) together received around $750 million of new capital from the government
budget, but they were still undercapitalised in the mid 1990s. The NCBs were
undertaking measures to improve their management and accounting systems, but
still faced political instructions to lend to SOEs and politically influential
private sector entrepreneurs. Government had begun to restructure some of the
major SOE debtors and established a financial loan court to facilitate debt
recovery.
Restructuring
of the state banks in Laos began in 1990. They were recapitalised in 1993/94
with bad debts transferred to a debt disposal agency in the central bank and
replaced by government bonds, which restored their solvency. The government
tackled the problems in the banks' major borrowers by implementing a major
privatisation programme under which three quarters of all the SOEs had been
privatised by 1995. However the internal restructuring of the state banks still
had a long way to go: it was constrained by serious shortages of skills and
their management's lack of experience of independent decision making based on
commercial principles.
ZANACO
was alone among the major government banks in the eight LDCs in not having
received a capital injection from the government. It was attempting to restore
solvency through internal efforts to cut costs and recover loans, in some cases
by establishing debt work-outs with SOE borrowers. Loan recovery efforts were
facilitated by the programme of privatisation or liquidation of SOEs implemented
in Zambia. ZANACO was able to recover some of the loans made to SOEs either
from their new owners or from the liquidator.
What
conclusions can be drawn from the experience to date of the LDCs in attempting
to rehabilitate their government banks? The problems have clearly not yet been
resolved, Malawi excepted, and whether or not all these banks will eventually
be successfully rehabilitated is doubtful. Governments have provided
substantial financial assistance to recapitalise the banks and to fund
redundancies, but several banks were still insolvent and making losses in the
mid-1990s. In some countries political opposition held up the necessary
reforms. Reputable private sector banks have shown little interest in taking
over the government banks. There is an obvious danger that recapitalisation
will merely provide the funds for further large losses to be incurred if the
restructuring of the banks' management and operational policies are not
effectively implemented, and if they face renewed political pressure to
undertake lending which is not commercially viable.
Three
notable lessons can be drawn from the success of Malawi in dealing with
financial distress in the early 1980s. First, prompt recognition of the problem
and implementation of remedial action makes restructuring less costly and more
likely to succeed. Second, restructuring or privatisation of major clients or
borrowers (e.g., SOEs) is essential if bad debts are to be recovered and if any
future lending to these borrowers is to be viable. Third, once banks have been
restored to solvency, avoidance of further distress depends upon their being
allowed to operate purely along commercial principles free of political
pressure to lend to priority sectors or favoured borrowers.
Some of
the government banks (e.g., UCB, ZANACO, NBL) have made much greater progress
in implementing the necessary reforms than others, but they still face large
problems. Even after major programmes of retrenchment operating costs are very
high. Most have the advantage of a large current account deposit base, partly
from having a near monopoly of public sector deposits, hence interest costs are
fairly low, but earnings are also low because of the past problems in their
asset portfolios. Although recapitalisation by the government can restore
solvency, these banks have to find new, commercially viable, borrowers to
replace their previous borrowers if they are to generate sustained profits.
Building
up and sustaining a sound loan portfolio is the biggest challenge facing the
government banks. Clearly it will be easier to achieve this if major borrowers
such as the SOEs are themselves restructured and privatised, as has occurred in
Laos and Zambia, but to a lesser extent elsewhere. The banks also need to
develop the skills required to appraise and monitor borrowers in a market
economy, skills, which are in short supply. Most importantly, however, the
management need the independence from political pressure, and the incentives,
to concentrating their lending on borrowers which can be served on a strictly
commercial basis. This will probably require some form of privatisation.
Despite financial liberalisation and
increased competition in financial markets, interest rates were still
prohibitively high because of the large domestic borrowing requirements of
governments which are financed by the auction of Treasury Bills in which most
commercial banks and other financial institutions invested heavily, and in a
few cases, because of excessive increases in money supply. While this has encouraged financial
institutions to finance commercial activities with quick and assured returns,
it has deterred them from providing long term capital to finance the fixed
investment necessary for the development of these LDCs. This points to the need
for governments to exercise more fiscal prudence by bringing their expenditure
under control and by curbing the growth in money supply. Financial institutions will provide long-term
capital only when they are convinced that the economy has stabilised; that is,
in an environment of low inflation and restrained growth in money supply. Equally important is the need for LDC
governments to address the lack of indigenous entrepreneurs and industrialists
capable and willing to take on the risks associated with long term investment.
4.3
Prudential Regulation and Supervision
The financial sector reform programmes have led to
substantial improvements to the regulation and supervision of the financial
system, especially in the mainland African LDCs. Legislation and supervisory
capacities were strengthened, and while deficiencies remain, prudential
supervision is now taken seriously in all the LDCs.
Major
weaknesses in both prudential legislation and supervisory capacities characterised
all the LDCs in the pre-reform period (In Laos no prudential legislation
existed prior to the reforms). Legislative weaknesses included minimum capital
requirements that had not been raised in line with inflation, thus allowing
poorly capitalised banks to be set up, and the omission of restrictions on
imprudent activities, such as large loan exposures or insider lending. No
objective loan classification and provisioning criteria were imposed, allowing
banks to overstate the value of their assets, profits and capital. Central
banks lacked the authority to intervene in distressed banks or sanction
violations of the regulations. Many financial institutions, particularly those,
set up by statute, were not subject to the banking laws. Governments did not accord
a high priority to prudential supervision. Central banks lacked adequate
resources for this task, they did not receive the appropriate data from banks
to enable them to undertake off-site monitoring and much of the supervision
that did take place focused on checking compliance with allocative controls (on
interest rates, sectoral credit directives and foreign exchange) rather than on
prudential requirements.
The
reforms adopted in the LDCs have followed the US model, which involves detailed
prudential regulations monitored and enforced through direct supervision by the
supervisory authorities (the central banks are the supervisory authorities in
all these LDCs). All the LDCs except Laos enacted revised banking legislation
in the late 1980s or early 1990s. In the mainland African LDCs this brought
most (but not all) aspects of their legislation into line with international
best practice, although the capacity to implement such legislation was still
weak. Capital adequacy requirements identical or similar to those of the Basle
Accord were adopted. Minimum capital requirements were raised, and restrictions
were imposed on large loan exposures, insider lending and investment in real
estate and non-banking business. Besides banks, the new legislation covered
some of the NBFIs, which had not been covered under the previous legislation in
some LDCs. Central banks issued directives on loan classification and
provisioning.
The
revised legislation gave some of the central banks greater flexibility to issue
prudential directives. It also gave them a degree of authority to intervene in
the event of bank distress. In Uganda and Zambia central banks used their
authority under the new legislation to take over or close down private sector
banks which were insolvent or otherwise unable to meet prudential requirements,
including banks in both countries with political connections.
Supervisory
capacities, methodologies and procedures were also strengthened, with increased
staffing levels, technical assistance, and training for supervisors. In most
countries, off site reporting was revamped, with central banks demanding more
timely and relevant data from the banks. More regular on site inspections,
focusing on prudential issues, have been carried out.
The
mainland African LDCs, notably Zambia and Uganda, have made more progress in
strengthening regulation and supervision than Madagascar and the Asian
countries. Nepal had revised its legislation but off site reporting was impeded
by a lack of timely data and on site inspections were not conducted regularly.
Some of the provisions in the legislation of Bangladesh were weak by
international standards (e.g., loan classification, although this was being
strengthened in 1995). Weaknesses in the legislation facilitated the insider
lending which jeopardised the solvency of Bangladeshi private sector banks,
none of which were closed down by the central bank. The 1988 banking act in
Madagascar lacked many objective prudential requirements, and it was necessary
to enact a new banking act in 1996. In Laos comprehensive prudential
regulations had not been formulated by 1995 and supervisory skills were
rudimentary, although a programme of training was underway.
Deposit insurance
schemes were set up or planned in Tanzania, Uganda, Zambia and Nepal. This will
make it easier for new entrants without a good reputation to mobilise deposits
from the public. There are sound reasons for explicit schemes to protect small
depositors, they can enable governments to acknowledge the costs of financial
distress in a more transparent manner, but the dangers of moral hazard are well
known and this makes effective prudential supervision even more important
(Garcia, 1996). LDC
governments need to reassure depositors that their deposits are safe through the
enforcement of prudential regulations to be able to detect financial fragility
early, and possibly prevent insolvency. Governments in LDCs also owe a duty to
taxpayers not to use their money to support bad management and inefficiency in
private sector banks and other financial institutions.
Despite
the reforms, all the LDCs still face a number of constraints to the effective
regulation and supervision of the financial system. Supervision is impeded by
human resource constraints. As in many other developing countries, there is a
shortage of qualified professionals, bank supervisors require substantial
training in the specific techniques of bank supervision, and there is strong
competition from the private sector to attract qualified professionals (Caprio,
1996). Because accounting standards are poor, supervisors cannot generally rely
on the accuracy of the accounts produced by banks, or even by their external
auditors. Supervisors will often have to rely on their own efforts to detect
fraudulent practices such as insider lending, and will need to develop skills
in detecting false accounting.
Political
interference is another constraint. Although central banks have gained more
authority to act independently, governments retain a lot of both formal control
(often action by the central bank requires the approval of the Finance
Minister) and informal influence. Governments are often very reluctant to allow
banks to be closed down, even when they are insolvent, for various reasons: the
bank owners may be politically influential (in some cases politicians are bank
directors), and governments fear the political fallout from lost jobs, lost
deposits and reduced access to credit which a bank closure would entail. Hence
central banks often face political pressure to exercise "regulatory
forbearance", which also undermines incentives for prudential management
in banks which expect to be able to draw on political support. Explicit prompt
corrective action rules, detailing the circumstances in which supervisors will
intervene in distressed banks, can provide the supervisors some support to
resist pressure for regulatory forbearance, and also improve incentives for
bank owners to avoid getting into situations where supervisors would intervene
(Glaessner and Mas, 1995). Supervision is likely to be more effective if
central banks, with the necessary skilled personnel, can establish a greater
degree of operational independence from governments.
The
revised prudential legislation in the LDCs incorporates regulations modelled on
those of the industrialised countries, such as the Basle capital adequacy
rules. But these may not be adequate for developing countries where accounting
standards are weaker and economic conditions less stable (Dziobeck, Frecault
and Nieto, 1995). In several respects banks in LDCs may be more vulnerable to
distress than their industrialised country counterparts, which may justify
imposing stricter regulations, such as higher capital adequacy ratios, higher
risk weightings for interbank loans, and a complete ban on insider lending as
is the case in Nepal.
6.
Conclusions
The impact of the financial sector reforms implemented
in the eight LDCs covered in this paper was relatively modest. Progress was constrained
because some of the reforms, such as the restructuring of government banks,
proved very difficult to implement and because economic conditions have not
been conducive to the development of dynamic market oriented financial sectors.
In the
three Asian LDCs the reforms facilitated both financial deepening and an
increase in bank lending to the private sector. In the African LDCs this did
not occur, probably because macroeconomic conditions were much more unstable.
All the LDCs experienced some new entry from private sector banks but the
increase in competition and its impact on efficiency was limited. Banking
markets remained largely oligopolistic and were still dominated in most
countries by inefficient government-owned banks. The new entrants brought some
benefits in terms of improved services and wider access to credit, but bank
failures afflicted private sector banks in some countries.
The
rehabilitation of insolvent government banks was a major component of the
reforms in all the LDCs other than Malawi (where similar problems had been
diagnosed and effective remedies undertaken in the early 1980s). In many cases
there were long delays in implementing the necessary rehabilitation measures,
in part because of the political costs involved. All these banks, except for
ZANACO in Zambia, received some form of recapitalisation by the government, but
several were still insolvent and making losses in the mid 1990s. Only in
Madagascar had any of these banks been privatised.
Reforms
to the prudential system have made substantial progress, especially in the
mainland African LDCs, where prudential legislation has been brought up to the
standards of international best practise and supervisory capacities and
procedures considerably strengthened. Moreover, bank supervisors have
demonstrated a much greater determination to enforce the banking laws and to
intervene in distressed banks.
The
reforms have begun to move what were all, to varying degrees, previously
repressed banking systems towards ones which are predominantly commercially
oriented, with regulation restricted mainly to prudential concerns. The reforms
are still ongoing and future policy reforms will have to address several key
issues, including bank-licensing policy, the future of government-owned banks,
development of NBFIs and financial system regulation.
Further
new entry by banks and NBFIs should be encouraged but licensing policy should
be relatively cautious both to ensure the probity and expertise of new entrants
and to avoid supervisory capacities being overwhelmed by the numbers of
financial institutions needing supervision. The entry of local private sector
banks and NBFIs can widen the range of financial services and access to credit,
especially of SMEs, and stimulate more competition, particularly in retail
banking markets. But their vulnerability to financial distress means that
strong prudential regulation and close supervision is essential, an issue
discussed below.
New
entrants should include reputable foreign banks. While they will serve only
limited sections of the banking markets, foreign banks can improve services,
particularly for corporate customers. Some of the foreign banks also provide
valuable training programmes for bank employees, which is an important
externality. Foreign banks have been much less prone to financial distress than
either government or locally owned private sector banks in the LDCs, hence they
provide some stability and credibility to the banking system. However, they are narrowly focused on a
group of prime borrowers outside of which they undertake very little lending in
LDCs.
Even
with more new entry from the private sector, a commercially oriented banking
system is likely to have a relatively narrow focus. Rural banking and lending
to small farmers is unlikely to be commercially viable because of the high
administrative and information costs involved and the difficulties in enforcing
loan repayment. Commercial banks rarely have the expertise needed for lending
to small farmers, their lending procedures (e.g., the focus on realisable
collateral) are not suitable and some do not have a rural branch network. The
banking system is also unlikely to provide long term finance, especially in the
unstable macroeconomic conditions prevailing in the LDCs.
Consequently
it will be necessary to encourage the growth of different types of NBFIs to
serve the segments of financial markets which are unattractive to the
commercial banks. Leasing companies provide a potentially useful vehicle for
short to medium term asset financing for SMEs. Leasing should be commercially
viable (it is already occurring on a significant scale in Zambia, and to a
lesser extent in Malawi), but the legal framework needs to be conducive to
leasing and it is also essential to ensure that leasing companies are subject
to prudential regulation and supervision if they are to mobilise funds from the
market.
Market
failures are pervasive in rural financial markets. Some form of government
intervention to facilitate credit supply to small farmers could improve social
welfare, although this would not necessarily be so in practise (Besley, 1994).
The key to developing rural financial markets is to find the institutional
arrangements, which can best overcome the specific types of market failures afflicting
these markets.[xiv]
The
type of innovative microfinance organisations whose lending technologies (such
as group lending and intensive loan administration) are designed to cope with
the problems entailed in lending to small scale borrowers without collateral
may provide a viable means of serving rural financial markets, but they are
likely to need some form of public subsidy to cover the very high
administrative costs involved (Hulme and Mosley, 1996). Microfinance
institutions, notably the Grameen Bank, have achieved success in Bangladesh,
and the model has been adopted in Nepal and some African LDCs. Microfinance
organisations could improve the range of services they offer the poor by
placing greater emphasis on the provision of savings facilities (Rutherford,
1998). This would help offset the negative impact on savings facilities of the
closure of unprofitable rural branches by the government and private sector
banks.
The
future of government commercial banks has not been resolved.[xv]
With the exception of those in Malawi, none of the distressed government banks
in the LDCs has yet demonstrated conclusively that it can be commercially
viable on a sustained basis, despite costly and lengthy restructuring
programmes. Some of these banks have made progress in cutting costs and
recovering loans and may have a viable future if they can build a base of
creditworthy borrowers. But little progress has been made in restructuring
other government banks. Governments will have to decide either to close these banks
down or to sell whatever parts of them are saleable to the private sector. The
alternative would be further waste of scarce financial resources and eventually
larger costs to government budgets.
It will
be important to continue strengthening prudential regulation and supervision of
banks, and to extend supervision to NBFIs, especially deposit taking NBFIs. In
controlled financial markets fragility mainly arose from government directed
lending to unbankable borrowers. This source of fragility has been reduced, if
not eliminated, by the reforms. But new sources of financial fragility have
already arisen in liberalised markets, in particular from new entry by private
sector banks and NBFIs, and greater competition for funds and borrowers.
Competition may cut interest rate spreads and other forms of income, such as
commissions on foreign exchange dealing, which have to some extent been able to
protect banks from losses incurred in their loan portfolios. Some of the new
entrants will lack adequate resources and experience of the markets they intend
to serve, and some are likely to engage in fraud. Deposit insurance will allow
new entrants without a reputation for prudent management to more easily
mobilise deposits from the public. Liberalised interest rates and foreign
exchange markets will expose banks and NBFIs to new sources of risk, of which
they have little experience of coping with.
Reforms
to the prudential system should be a process, which benefits from constant monitoring.
Some aspects of the prudential legislation may need to be revised to take
account of the particular circumstances of financial markets in LDCs, for
example, it may be appropriate to impose higher capital adequacy requirements
and to ban all insider lending. The regulations should reinforce incentives for
prudent management on the part of bank owners and managers (Caprio, 1996).
Close supervision, particularly of lending policies and of recent entrants, is
needed to detect problems at an early stage. Central banks should intervene
promptly in distressed banks and NBFIs, and sanction infractions of prudential
regulations, both to limit the scale of losses in distressed banks and to
strengthen incentives for prudent bank management. Central banks need
operational independence from politicians if regulation and supervision are to
be effective.
Financial
sector reforms have no doubt drawn the attention of governments in the eight
LDCs to the advantages of efficient financial systems; for example, the cost to
the public purse of politically connected lending is now widely acknowledged,
despite the limited success in tackling the problem in a few LDCs. While it is difficult to rigorously assess
the outcomes of the reforms in the eight LDCs, in part because reforms have yet
to be completed in most, there is little doubt on-going financial
liberalisation has enhanced the role and use of monetary policy in overall
macroeconomic management. There has been some improvement in the efficiency of
the banking system itself: competition, albeit limited in some cases, has been
introduced, commercial bank lending to SOEs is now based on commercial criteria
in almost all the LDCs, rules governing the operation of financial institutions
have been strengthened, regulatory and supervisory systems of central banks
have been enhanced despite weak capacities, in a few cases, to enforce the new
rules; although much remains to be accomplished in the area of divestiture of
public sector banks by governments.
The
case studies presented in this paper demonstrate that a stable macroeconomic
environment is a sine qua non for the
success of financial liberalisation.
Financial sector reforms are necessary, but the implementation of these
reforms is insufficient to bring about enhanced financial intermediation
through stable and sustainable real positive interest rates. The reforms have to be accompanied by sound
macroeconomic, monetary and fiscal policies designed to attain low and
sustainable rates of inflation. In
addition the financial position of banks, and other financial institutions,
must be strong, and programmes for recapitalising weak banks have to be agreed.
A period of instability in financial
markets appears to be inevitable during the transition to financial deregulation:
the relationships between monetary aggregates, economic activity, interest
rates and prices change in ways that defy easy prediction and therefore policy
prescription. This makes the management
of monetary policy during financial sector reforms extremely difficult,
especially in poor countries such as LDCs plagued with weak institutions and
skill shortages. There is the need to
corroborate monetary policy with effective regulation and supervision of
financial institutions including, improved banking legislation (e.g. stringent
capital adequacy requirements), and adequate enforcement of such legislation
(e.g. effective bank supervision) alongside the introduction of market-based
instruments (open market operations) to attain monetary targets. These difficulties notwithstanding, the
inefficiencies of a monolithic financial system are no longer in doubt (e.g.,
as demonstrated in this paper by the major problems that afflicted, and still
afflicts, the Tanzanian commercial banks); and underscore the need for a more
diversified financial sector and ownership of financial institutions as well as
the need to enhance competition and efficiency in the financial sector.
(Percentage of GDP)
|
|
1985 |
1990 |
1995 |
|
Bangladesh |
|
|
|
|
Bank Deposits |
20.8 |
24.5 |
28.9 |
|
M2 |
24.5 |
28.1 |
33.8 |
|
Lao PDR |
|
|
|
|
Bank Deposits* |
n.a |
4.1 |
10.8 |
|
M2 |
n.a |
7.2 |
13.8 |
|
Nepal |
|
|
|
|
Bank Deposits |
15.9 |
19 |
23.8 |
|
M2 |