In this decade, many countries have witnessed an unprecedented number of failures of banks and nonbank financial institutions. 2/ Failed financial institutions, however, represent only a fraction of problem institutions. A much larger proportion is in a state of distress. 3/
Although a distressed financial system may have no obvious adverse consequences on the economy, it will perpetuate the system’s instability and inefficiency, and cause resource misallocation. Without the urgency of a financial crisis, these ill effects tend to be easily overlooked, and corrective measures can be delayed. As a result, the cost and difficulty of returning the financial system to stability may be compounded. The consequences of prolonged financial distress can be as devastating as, if not more than, a financial crisis.
Bangladesh’s state-owned banks have been under pressure since 1983-84. The key causes of financial distress were insufficient autonomy in lending decisions, inadequate supervision, and managerial problems. For a number of years, limited remedial measures have been undertaken. Although important corrective measures have been initiated recently, it is still too early to say whether they will finally lead to a lasting improvement in the situation of the state owned banks. Meanwhile, the financial system remains in a state of distress.
The distinguishing feature of the Bangladesh banking problem is that it is in a latent crisis. It is a continuing problem and not an abrupt run on deposits that creates substantial commotion, requires quick remedial action, and is thus resolved over a relatively short period. Because of this distinguishing feature, it is interesting to investigate the causes and consequences of the Bangladesh banking sector distress, its stability and efficiency, and the costs of long inaction. These are the purposes of this paper.
Bangladesh’s state-owned banks have expanded credit to priority sectors in response to government directives without due regard to quality, often at interest rates below the banks’ cost of funds. This has led to inefficient resource allocation and widespread loan delinquency. It has become increasingly difficult for the banking system to play its essential role of supporting the country’s economic development and adjustment goals.
Recognizing the need to strengthen the financial sector, the Government, in late 1984, appointed a National Commission on Money, Banking, and Credit to undertake a major study of the financial sector; in 1986, the Commission issued its recommendations for financial sector reform. 4/ The problem of loan delinquencies was recognized as a key issue that needed to be addressed. 5/ Restoring the financial strength of banking institutions and introducing more flexibility into the financial system were identified as important strategies to revive the banking sector’s crucial role in financial intermediation, and ensure its stability.
The outline of the paper is as follows: Section II reviews the Bangladesh financial system. Section III assesses its financial strength. Section IV discusses the major causes of the banking distress. Section V discusses the consequences of the current situation and the danger of prolonged inaction. Section VI describes measures that were undertaken to strengthen the banking sector and suggests additional measures. Section VII provides a conclusion.
II. The Bangladesh Financial System
1. Structure of the financial system
The Bangladesh financial system consists of deposit money banks, which dominate the system, and other financial institutions. As of June 1989, there were 25 deposit money banks, comprising four NCBs, ten private banks, seven foreign banks, and four publicly owned specialized development banks which engage in lending to the priority sectors of agriculture and industry. Other financial institutions include a development bank, a number of insurance companies, investment and leasing companies, and cooperative institutions (Statistical Appendix Table 1).
a. Deposit money banks
(1) Nationalized commercial banks (NCBs). As of March 1989, the four NCBs (Sonali, Janata, Agrani, and Rupali Banks) controlled about 64 percent of total deposits and 54 percent of advances. They lend primarily to the agricultural, industrial, and export sectors.
(2) Private and foreign commercial banks. Private banks’ market share has increased significantly over the years, and has become the second largest class of institutions after the NCBs, accounting for about one-quarter of the market as of March 1989. Foreign banks’ market share—about 8 percent of total—has not changed substantially since 1981. Private and foreign banks lend primarily to commerce.
(3) Specialized banks. Bangladesh Krishi Bank (BKB) and Bangladesh Shilpa Bank (BSB) are the most important specialized banks. The Krishi Bank lends to agriculture, and agro-based and cottage industries. The Shilpa Bank makes term lending for medium- and large-scale private industrial projects.
The two banks have a relatively small deposit base and the BKB in particular relies heavily on the refinancing facilities of the Bangladesh Bank (BB). Advances to agriculture provided by these two specialized banks exceed the total amount provided by all the other commercial banks combined and represent close to 70 percent of their portfolio.
b. Other financial institutions
These institutions are relatively unimportant, except for the Bangladesh Shilpa Rin Sangsta Bank (BSRS). The BSRS was supposed to provide financing for larger-scale industrial projects, but its operations have become virtually indistinguishable from that of the BSB. The BSRS, together with the BKB and the BSB, will be referred to as development finance institutions (DFIs).
2. The regulatory framework
The BB is empowered by the Banking Companies Ordinance of 1962 to regulate all scheduled banks, which include NCBs, private and foreign commercial banks, the BKB and the BSB; and by the Bangladesh Bank Order of 1972 to regulate and supervise other financial institutions. In practice, however, the BB regulates, supervises, and inspects mainly commercial banks. The BKB and the BSB adhere to some of the BB’s standards but are not closely supervised.
The BB imposes a number of restrictions on banks’ asset composition to ensure that they conduct business in a prudent manner, and have adequate reserves and liquidity. However, before the adoption of some important supervisory measures at the end of 1989, there were substantial weaknesses in supervision and inspection. This was especially true in the area of financial performance, in particular, capital requirement, and loan classification and provisioning. These weaknesses have contributed to the existing problems in the financial sector.
III. An Assessment of the Banking Institutions’ Financial Positions
1. Nationalized commercial banks (NCBs)
From the NCBs’ financial statements one does not get any indication of serious financial problems. While the NCBs’ return on assets is low and has declined from 0.9 percent in 1984 to 0.4 percent in 1988, they reported a net profit after tax of Tk 2.2 billion during 1984-88 (Table 1). Except for 1986, deposits have grown at a double-digit pace in the past few years. Capital accounts are extremely small, accounting for less than one percent of total assets, but this percentage has increased slightly since 1985.
|(In millions of taka)|
|Loans and bills||55,968||67,293||72,991||78,305||87,423|
|(Annual percentage change)|
|Loans and bills||…||20.2||8.5||7.3||11.6|
|(As a percentage of total assets)|
|Loans and bills||64.6||65.2||62.8||62.6||67.7|
These indicators do not disclose the real financial position of the NCBs, which is largely hidden by the prevailing loan classification, provisioning, interest suspension, and other accounting practices. NCBs would be in financial distress—as defined earlier—if bad loans were properly accounted for. Overdue loans have increased from a recorded 18 percent of loan portfolio in 1985 to about 21 percent in 1986 (excluding Janata Bank whose data on overdues are not available, Table 2). 6/
|(In millions of takas)|
|Accumulated interest suspended||641||668|
|Capital accounts and accumulated provisions/overdue||18.8||19.0|
|Accumulated provisions/Total classified||44.8||54.2|
|Accumulated provisions—loss loans/doubtful loans||55.3||64.2|
While the definition of overdue loans is unclear, and banks have different ways of classifying overdue debts, loans are probably considered delinquent when they are overdue by more than one year. 7/ Overdues would be higher if arrears of a shorter period were included. Overdue loans are underestimated also because most NCB credits are overdrafts and drawn under lines of credit, which make it difficult to determine when borrowing is used to cover interest charges on overdue obligations. In such cases, the determination of overdue debt largely becomes a decision of individual managers, which can be quite arbitrary.
In 1986, Sonali Bank, Agrani Bank and Rupali Bank had an aggregate amount of overdue loans of Tk 11.1 billion. However, only Tk 2.8 billion, or 25 percent of overdues were classified as substandard, doubtful and loss loans. The amount classified as loss loans was Tk 0.2 billion or a mere two percent of their overdues. These ratios are unusually low.
Two factors may have contributed to the small number of loans classified as past due. First, up until now, all loan classifications have been made by BB inspectors and not by banks. 8/ These classifications were updated only when banks were inspected. Of the three banks mentioned above, only one was inspected and its loan classifications updated in 1986. Secondly, the BB seemed to have been lenient in classifying banks’ problem loans. For example, in 1985, the officially classified loans of an NCB accounted for slightly less than 10 percent of its overdues and its loss loans were only 0.2 percent of its overdues.
As of 1986, loan provisions of the three NCBs accounted for about 54 percent of classified loans. A breakdown of the data shows that NCBs have more or less provided for 100 percent and 50 percent of loss and doubtful loans, respectively. However, no provision was made for substandard loans and whatever provisions were made did not reach more than a very low fraction of actual nonperforming loans.
Interest income was grossly overestimated since interest was suspended only on doubtful and loss loans. Interest on substandard loans was suspended only after a two-year period without any interest payments. 9/
In 1986, the three NCBs mentioned above had a combined capital accounts and loan provisions of Tk 2.2 billion which is equivalent to 19 percent of their overdues. The implication is that the capital and reserves of these three NCBs would be completely wiped out, if merely 19 percent of their overdues could not be recovered, requiring the write off of this amount. With such a small cushion, one can infer that the NCBs are de facto insolvent. At the same time they are state-owned and the Government is expected to stand by as needed; therefore, the public does not consider such technical insolvency as having economic relevance. The true economic cost of such “distress” is discussed later.
Although NCBs are de facto insolvent, they show substantial bookkeeping profits 10/ and on that basis pay taxes, dividends and bonuses. These expenses, as well as current expenses are largely paid out of their new deposits. While currently NCBs have ample liquidity, should their deposits start to grow at a slower rate, liquidity problems may emerge.
2. Development finance institutions (DFIs)
The three DFIs—Bangladesh Krishi Bank (BKB), Bangladesh Silpa Bank (BSB), and Bangladesh Silpa Rin Sangstha Bank (BSRS)—have been severely affected by overdue loans, which have increased uninterruptedly since 1982. In one case, as high as 79 percent of portfolio was classified as overdue as of June 1989. The majority of overdues are on account of private loans. However, a large proportion of public sector loans extended by these institutions are overdue, setting up an undesirable example for the private sector.
IV. Causes of Financial Difficulties
Problems of the Bangladesh banking institutions are explained by a multitude of economic, political, legal, and administrative factors. The most important of these factors are discussed below:
1. Distortive and rigid economic policies
a. Preferential lending
The authorities have created various credit schemes to promote priority sectors such as agriculture, exports and small-scale industries. Among these, the agricultural sector is given top priority. Being state-owned banks, the NCBs were expected to extend rural and preferential credits to a variety of priority sectors and schemes under subsidized terms and conditions. The BB then partly or totally refinanced such loans. Based on data for December 1985, about two-thirds of NCBs’ loans were preferential loans.
Although preferential lending was not a strict requirement, the authorities exerted moral suasion to persuade these state-owned institutions to comply with the expectations. They were also pressured to lend to public corporations or privatized public corporations. Prior to 1985, at the beginning of each year, the BB announced its preferential lending and refinancing targets for subsectors. Since 1985, however, these directives for preferential lending have been replaced by lists of favored and discouraged subsectors or activities, but the tradition to lend to existing customers in these subsectors seemed to remain. 11/
Lending in accordance with the authorities’ directives reduced banks’ autonomy and undermined their efforts to maintain good credit approval practices. Banks saw no necessity of screening for good customers—a compelling practice for banks under normal circumstances. Although the authorities provided no guarantees for priority lending, banks seemed to expect that if serious default problems developed the authorities would have to step in eventually, especially if the borrowers were public corporations (whose borrowing accounted for one-third of preferential credits).
A number of priority sectors and schemes also had especially high risk, for example, the agricultural sector, whose borrowers’ repayment ability depends critically on such exogenous factors as weather and world commodity prices. Lending to these high-risk sectors without proper appraisal of customers was bound to bring about a substantial amount of poor-quality assets.
b. Rigid and complex interest rate structure
Before the substantial changes in the interest rate structure in early 1990, Bangladesh had comprehensive controls on the level and structure of interest rates. Interest rates on deposits and loans were differentiated by type and maturity. For the same sector, different rates applied to working capital and term loans. Rates on term loans also varied depending on the debt/equity ratio of the borrowing firms. Interest rates were changed only infrequently with a result that real rates fluctuated with changes in inflation. There were no market-determined interest rates except for the rates on certificates of deposits (which were not attractive due to artificially high time deposit rates) and the interbank rate.
Most real rates have remained positive since 1982 due to the relatively low inflation, and, as a result, saving mobilization has not been discouraged. However, administered interest rates proved to be too rigid and created distortions. The interest rates on long-term deposits were considered too high and private and foreign banks refused to accept such deposits.
Nevertheless, the NCBs were expected to accept all deposits at the administered interest rates. As a result, the NCBs had to bear a higher cost of funds and were stuck with excess liquidity, in part reflecting NCB’s reluctance to lend due to the relatively low lending rates, the potential difficulty in recovering loans, and the new awareness of credit risk. The excess liquidity did not earn the NCBs a market-related yield due to the lack of market-based investment vehicles for liquidity management. This situation had eroded NCBs’ earnings and thus their financial position.
Most preferential loan rates were lower than the highest deposit rates and nonpreferential loan rates despite the fact that the administrative costs of preferential loans, in particular, agricultural and small cottage industrial loans, were higher than other types of loans. This interest rate structure, again, adversely affected the NCBs and the BKB, which extended substantial amount of preferential loans. Although the BB refinanced preferential loans with low refinance rates, the NCBs did not depend heavily on the BB due to their excess liquidity. Therefore, the low refinancing rates did not help reduce banks’ cost of funds to any significant extent.
Interest rates on nonpreferential loans were not high enough to provide a cross subsidy from nonpreferential to preferential loans. Although in March 1987, an interest rate ceiling rather than a set rate was established on loans for commercial imports, the ceiling seemed to be set too low. Consequently, almost all commercial loans were made at the ceiling rate and the intended interest rate flexibility did not materialize.
In addition to eroding NCBs’ earnings, the administered interest rates and the regulatory environment failed to reflect a loan’s real risk and return, and thereby contributed to the build-up of poor-quality assets. The interest rate structure, together with the loan classification, provisioning and accounting procedures, which concealed the bank’s real cost of funds, made it difficult for a bank to evaluate a loan based on its financial or economic costs and returns. This has resulted in resource misallocation. Financing was given to projects whose viability was questionable, to firms and public corporations that were weak financially and managerially, and to sectors with considerable over-capacity.
2. The inadequacy of supervisory framework and inspection
Supervision and inspection were lax in the areas of capital requirements, loan classification and provisioning, and accounting standards. The banking laws of Bangladesh stipulate different capital requirements for banks depending on whether they are incorporated abroad or domestically, and whether they are nationalized or private. Banks incorporated abroad are required to maintain paid-up capital and reserves equivalent to at least 5 percent of demand and time liabilities as of the end of the previous year, or Tk 2 million, whichever is higher. Private banks are required to hold capital and reserves of at least 6.5 percent (reduced from 7.5 percent after pressure from banks) of their demand and time liabilities, or Tk 100 million, whichever is higher; and paid-up capital has to be at least Tk 80 million.
As for nationalized banks, the minimum paid-up capital is expressed as an absolute taka amount. Presently, the paid-up capital is Tk 30 million for the Sonali, Janata, and Agrani banks and Tk 20 million for the Rupali Bank, making a capital-to-deposits ratio of less than one percent for the first three banks, and 2.5 percent for the last bank.
The different capital requirements give an unjustified competitive edge to the favored banks. With the extremely low capital requirement, the NCBs are vulnerable to changes in market and policy environments which may easily make them insolvent.
As was discussed earlier, loans were inadequately classified and provisioned for. In inspecting and classifying a loan, BB inspectors often placed their emphasis on the adequacy of formal security rather than on actual loan performance. Collateral was treated as having substantial value although it was difficult to realize its value through foreclosure in the existing legal environment.
Given the large number of loans, the focus on security was perhaps inevitable, as banks had been reluctant to do their own classifications, and BB inspectors could not be expected to review each loan in sufficient detail. Bank officials had little interest in classifying loans which may lead to increased provisions and reduced profits. Government, too, had resisted increased provisions based on realistic classifications in order to increase the flow of dividend and taxes to the budget.
The BB allowed banks to accrue interest on overdue loans in the first two years, a too lenient standard which enabled banks to overstate their incomes. Also, although banks were supposed to follow uniform accounting practices, in reality there were substantial differences among them. There was a large degree of arbitrariness and discretion of individual managers regarding the classification of loans as nonperforming, and the setting aside of reserves for bad loans. More stringent standards were normally applied to private sector than to public sector loans.
3. Managerial weaknesses
a. Insufficient project appraisal
In the industrial sector, the vast majority of the overdue loans are caused by faulty project appraisal. Reportedly a large number of DFI project loans were appraised merely as a formality either because external funds for projects were available, or because projects had already been approved in principle by the National Investment Board. In some cases projects were appraised based on assumptions about the policy environment or availability of infrastructure that later were substantially changed or did not materialize.
The DFIs have also been inflexible, partly due to national guidelines, in determining and modifying amortization schedules. Often the schedules have been unrealistic and inconsistent with the borrowers’ cash flow.
b. Inefficient loan collection
NCBs have poor debt recovery. Poor loan accounting and inadequate management information systems make it difficult to monitor and collect loans. Except in the case of Agrani Bank, branch-level information on the status of the portfolio and debt recovery is not collected and consolidated at each organizational level.
Interest suspense accounts are kept only on an aggregate basis for the whole bank. Most branches would find it impossible to relate the balance on their aggregate account to the amount of interest suspended on individual accounts. In the case of the BSB, there are a number of departments which produce data on loan balances, collections, and arrears but data are often conflicting.
Management is not sufficiently result-oriented. Bank officers’ financial rewards are not linked to their loan collection performance. In fact, since administrative costs incurred to intensify collection efforts can actually hurt profits, which are the basis of bonuses, there is an incentive not to step up collection efforts as well as not to establish accurate non-accrual procedures. Most banks do not have management processes which allow the managers to review, monitor, reward, or punish the performance of the lower-level officers. The number of officers relative to the number of loans outstanding and the level of training received by branch officers are also generally inadequate.
4. Uncertainties regarding debt repayment
There are uncertainties regarding financial obligations owing to the nationalization and denationalization of certain industries and corporations. Furthermore, the government has frequently promised and/or given concessions to overdue borrowers; as a result, many borrowers hold off payment anticipating further concessions. This can be seen in the case of the Exchange Rate Fluctuation Burden Absorption Scheme (EFAS) introduced by the Government to protect would-be borrowers against foreign exchange risk and to give concessions to existing borrowers in foreign exchange. Since its launching in 1983, the EFAS has been amended several times to include additional concessions. Borrowers who expected to receive additional concessions from the scheme had held off their payments.
Bangladesh is a country prone to floods and other natural disasters. The Government often assists the affected areas with special credit facilities and with debt exemption/forgiveness schemes. Neighboring areas which may be less affected also frequently become beneficiaries. In some cases, there appears to be some ambiguity as to whether such assistance is in the form of grants, or loans that are expected to be repaid.
5. Inadequate legal framework
An inadequate legal and judiciary framework has contributed substantially to the loan delinquency problem. All financial transactions in Bangladesh are subject to common law. The law requires notices to be served in person—a requirement that has made the process susceptible to substantial delay. Judges often have limited experience with commercial cases and have, on occasion, made inappropriate decisions which have set damaging precedents. Judicial staff are severely underpaid, raising susceptibility to external influences. In contrast to the NCBs, the DFIs have extensive powers—through special presidential ordinances—to recover loans through summary legal procedures but the problem of excessively long delays and a drawn-out appeals process remains.
6. Macroeconomic conditions and financial distress
Preferential lending and administered interest rates were part of the economic environment even in the 1970s, long before the surfacing of the banking distress in 1983-84. It is therefore interesting to ask whether the surfacing of the distress was brought about by certain adverse economic conditions.
In 1982, a year before loan performance became a major problem, the real economic growth dipped to a record low of 0.8 percent (Chart 1). However, Bangladesh’s growth performance has always been quite erratic. A year earlier, the economy registered a remarkably high rate of 6.8 percent. More importantly, economic growth picked up again in 1983 and remained relatively stable over the following years but loan performance failed to improve. This points to the weak correlation between loan and economic growth performances. It is, therefore, difficult to conceive that the sharp economic downturn in 1982 contributed to the deterioration of the loan performance in the following years.
After remaining negative for four years, the real lending rate turned positive again in 1983 due to a sharp deceleration of inflation. However, the magnitude was small compared to the level prevailing in the mid-1970s. The positive real interest rate is not likely to have significantly increased the debt servicing burden of borrowers.
The sizable depreciation of the taka prior to 1983, 61 percent in 1975, and 23 percent and 22 percent in 1981 and 1982, respectively, could have accumulated to exert a significant impact on the debt burden of some of the foreign loan borrowers. In fact, some borrowers argued that the depreciation had inflated the principal amount of the private sector’s foreign borrowing by 600-700 percent, constituting the major factor of the build-up in overdues. The study conducted by Sobhan and Ahsan, 12/ however, indicated that for the majority of the 178 samples of loan projects sanctioned by the BSB and BSRS after the Bangladesh independence in 1971, the exchange rate burden has been less than 25 percent of the taka equivalent of the foreign currency loan at the time of repayment. 13/ Hence, the claim that the debt burden of borrowers has substantially increased is not supported.
Sobhan and Ahsan also show that for the majority of the projects, the exchange rate burden is less than a quarter of their overdue amount. Projects that were sanctioned prior to 1971, however, have had a much higher exchange rate burden, although not to the extent claimed by some borrowers. Nevertheless, even in the case of the BSB, whose pre-1971 projects accounted for 16.7 percent of the total projects in the private sector, the overdue amount of pre-1971 projects is only slightly more than one-quarter of its total overdues. Therefore, it would not be accurate to attribute the poor loan performance to the taka depreciation.
A correlation between the economic conditions in this period and the unsatisfactory loan performance, therefore, cannot be established, except perhaps for one factor, that is, the accommodative credit policy in 1983/84 and 1984/85. In these few years, BB’s lending to deposit money banks increased significantly, resulting in a rapid credit expansion of deposit money banks to the private sector, a 59 percent and 40 percent increase in 1983/84 and 1984/85, respectively. These increases were substantially higher than the 20-30 percent range in the preceding two years. This rapid increase, coupled with the weak lending practices, discussed in the last chapter, may have contributed to the growing problem of bad loans.
7. Summary of the causes of financial distress
In summary, there is not a single cause that explains the financial distress situation in Bangladesh. It is clear, however, that without preferential lending under government directive and the complex structure of administered interest rates, the banking distress would not have been half as serious.
These initial problems were compounded by inadequate supervision, and managerial and internal control weaknesses. Capital requirement and loan provision rules were too lenient and were not strictly enforced. Project appraisal and loan collection efforts were also deficient. In addition, the legal framework has proven to be an obstacle even to the modest loan collection efforts which were undertaken.
Chart 1Bangladesh: Major Economic Indicators
Source: IMP; International Financial Statistics.
Government activism and assistance in connection with floods has reduced financial discipline. Uncertainties arose due to nationalization and privatization of certain sectors and businesses. Also, the government did not make clear from the start whether government assistance was in the form of loans or grants. As a result, little provision was made by the beneficiaries for future repayments.
It is worth noting that macroeconomic developments, which played an important role in the banking crisis of many countries, were not a key factor in Bangladesh. Nevertheless, the accommodative credit policies of 1983/84 and 1984/85, together with weak lending practices, did contribute to an increase in the portfolio of nonperforming loans which the Bangladesh banks now carry.
V. Consequences of the Banking Distress
The banking distress is not expected to lead to an erosion of the public’s confidence in the banking system since all problem institutions are state-owned and carry implicit government guarantees. However, it is exerting an adverse impact on both the financial position of these institutions as well as on the economy.
Although a number of long overdue corrective measures have been initiated recently to revitalize the financial sector, the banking situation is not expected to improve soon. 14/ Strong and persistent determination will be needed to carry out the announced measures through the many stages of the reform until banking problems are finally resolved. In the meantime, the costs of the banking distress will go on accumulating until corrective measures start to bear fruits.
The discussion below outlines the economic and social costs of the banking distress in order to highlight the need to redress these problems in a sustained way. Banking distress has an adverse impact on the government budget, weakens these institutions’ ability to compete, affects monetary control, distorts resource allocation and promotes social inequity.
1. Budgetary implications of the insolvency of financial institutions
Given the large amounts of overdue loans and the potential losses, and their current rates of earnings, the NCBs and the DFIs cannot be expected to adequately make provisions for losses by themselves. Although the recent changes in the interest rate policy and preferential credit program can be expected to alleviate these institutions’ losses, they will not be able to make these institutions solvent. The costs of insolvency will, eventually, have to be borne, directly or indirectly, by the Government, as the owner of these institutions.
The Government as major shareholder will be responsible for the needed recapitalization of the NCBs. This recapitalization is needed for two reasons: one to replenish the capital that would be reduced or depleted by the writing-off of bad loans, and second, to increase their financial strength by increasing capital to a certain percentage of their deposits or risk assets. 15/
If the preferential credit program is continued, the Government will have to provide direct interest subsidy to the NCBs and the DFIs to prevent them from incurring further losses. Direct budgetary cost will also be incurred in this case. Indirect budgetary impact occurs if these institutions’ dividend and tax payments to the Government stop when their financial situation deteriorates.
The extent of the budgetary impact, both direct and indirect, depends on the extent of bad loans and loan recovery, the required capital to deposit ratio, and related policy issues. As part of the recent financial reform measures, the Government has decided to (1) increase NCB capital to 5 percent of their deposits after losses are written off; 16/ and (2) to provide interest subsidy to six priority sectors. However, the recent performance of bad loans and loan recoveries has yet to be evaluated. Without the knowledge of this important information, it is difficult to estimate the extent of the budgetary impact.
Yet, according to a World Bank estimate, as of December 1989, if provisions amounting to 12 percent of portfolio were required and the capital was to be raised to the recommended minimum of 5 percent of deposits, the amount of needed NCB recapitalization would have been roughly Tk 15 billion. If this estimate turns out to be correct, the Government intends to raise the required capital through (1) the issue of a special non-negotiable government bond of, say, Tk 13 billion, with interest payable at the treasury bill rate and a balloon repayment of principal at the end of 15 years; (2) conservatively revaluing fixed assets on a one-time basis with an estimated revaluation of Tk 2 billion; and (3) reinvestment of future profits to the extent necessary to maintain capital at 5 percent of deposits thereafter. Currently, the treasury bill rate is 8.5 percent per annum. The Tk 13 billion government bond would, therefore, cost the Government Tk 1.1 billion annually. This is expected to be largely offset on a lagged basis by future NCB income taxes and possible future dividends.
Since NCBs have no serious liquidity problems at present, a large amount of cash infusion may not be necessary. On the other hand, a too small amount of cash infusion can generate too little revenue and take NCBs a long time to become profitable and build up the required level of capital. In the meantime, the weak NCBs will continue to impose certain costs on the economy.
Therefore, in order to restore NCBs’ financial strength in a reasonably short period of time, it is desirable that sufficient cash be infused. This amount depends on the size of an NCB’s existing performing portfolio, its ability to expand its liabilities and earning assets, and its spreads. It is important that the authorities do not deliberately lower cash infusion to hide the real cost of rescue operations for political purposes.
The estimated time for the reclassification process to be completed is now September 1990. The longer the recapitalization is postponed, the larger the losses will be. Failure to swiftly strengthen the NCBs’ financial footing will only defer and aggravate their problems. It is, therefore, crucial that reclassification and recapitalization be carried out as swiftly as possible.
2. Implication for banking competition
Given the large burden of delinquent debts, the NCBs cannot compete well with private and foreign banks. According to a World Bank estimate, bad debts and suspended interests put an extra 2 to 8 percentage points of cost on the NCBs, depending on the types of loans. The NCBs need a larger margin to partly compensate for this extra cost and this larger margin puts them in a disadvantaged competitive position.
If they are to aggressively compete for funds, say, to maintain their market share, the increased funds may merely add incentives to lend to either high-risk new customers or nonperforming existing customers, in order to generate higher income or avoid classifying nonperforming loans. Furthermore, since at least a quarter of NCBs’ portfolio is delinquent and cannot be readily liquidated, the NCBs may need to maintain a higher level of liquid assets as a precaution against withdrawals; hence their motivation to actively compete for borrowers could also be reduced.
Lack of competition from the NCBs, which have over 50 percent market share, provides no incentive for other banks to improve their efficiency, and the financial system in general is likely to be less efficient as a result. An inefficient financial sector cannot fulfill its role of supporting the country’s economic development.
The implementation of interest rate reform and other financial reforms may be more difficult and require close monitoring since it is likely to have a major impact on the inefficient and noncompetitive NCBs. The scope and speed of reform may have to be constrained if the NCBs fail to adjust themselves to a more liberal environment; certain reforms also may not achieve the intended result. For example, an active money market may not develop if sound banks refuse to transact with insolvent banks.
3. Implication for monetary control
The large portfolio of nonperforming loans of the NCBs complicates the setting and monitoring of credit targets. This is because credit that is extended or rolled over to make nonperforming loans “performing” does not lead to productive activities nor consumption. It is hard to distinguish between fresh credit extended for output-generating activities and credit for the rollover of bad debt and the capitalization of interest on such debt. Therefore, it becomes difficult to set credit targets that are consistent with the desirable or attainable levels of output, prices and external balances.
If one assumes that half of the interest payments on NCB overdue loans outstanding as of the end of 1985, 17/ were capitalized in 1986 at an interest rate of 14 percent, 18/ the interest capitalization would have accounted for 8.5 percent of the corresponding NCBs’ credit in 1986. If one further assumes that apart from interest capitalization, 10 percent of overdue loans were amortized with new credits, about 21 percent of 1986 NCB credits would have been used for the servicing and amortization of overdue loans. The more new credits are used for these purposes, the less is available for output-generating activities. This ambiguity of the use of credits reduces the usefulness of setting and monitoring credit targets.
Nonperforming assets can tie up financial institutions’ liquidity, and may change problem banks’ demand for assets; in particular, the demand for liquid assets may increase. Consequently, the relationship among the monetary base, bank asset creation, and economic activities can be volatile and difficult to predict.
Should deposits grow at a slower rate, the banking system may have to increasingly rely on the central bank’s refinance facilities. If this happens at a time when a monetary contraction is called for, the authority may have to face the difficult choice of providing the additional refinancing for the banking system and thereby relaxing monetary control, or reducing its credit and thereby squeezing these institutions’ liquidity.
Similarly, the authorities’ decision to recapitalize or bail out troubled financial institutions can put the monetary authority in a similar dilemma, if the latter has to finance the higher budget deficits. The dilemma could be a serious one if large cash infusions are needed.
4. Implication for resource allocation
With the substantial amount of overdues in their portfolio, banks tend to allocate credit to customers that have overdue debt so that their loans can “perform.” Consequently, new and more dynamic sectors and borrowers are denied credit, or get squeezed when there is a general contraction of credit. This has also led to social inequity in that borrowers that are less creditworthy may get credit allocated whereas creditworthy borrowers fail to receive credit.
The inequity is particularly troublesome insofar as part of the bad loans represent willful defaults. Moreover, the way the burden of bad loans has been and will be eventually shared by various economic agents has further redistributive effects. The EFAS scheme, for example, by allowing a certain percentage of exchange rate burden to be converted into preferential shares or debentures to be held by the lending institutions, has transferred the exchange rate burden from a group of profit-seeking private entrepreneurs to financial institutions. The burden will be borne eventually by tax payers who pay for the Government’s bail-out operations.
VI. Measures to Strengthen the Banking Sector
The Bangladesh National Commission on Money, Banking, and Credit has recommended extensive corrective measures in the areas of economic management, the legal and supervisory frameworks, accounting procedures, and the management of financial institutions. These measures have been supported by the World Bank, the International Monetary Fund, and other international organizations. A few loan recovery programs have been implemented since 1983, but more comprehensive financial reform programs were launched at the end of 1989 and early 1990. This chapter discusses measures that have been undertaken and suggests measures that remain to be implemented to strengthen the financial position of the ailing institutions.
1. Loan recovery programs
The large amount of nonperforming assets poses the most serious problem for the NCBs and the DFIs. It was thought that the increase in the taka value of foreign currency loans, due to the depreciation of taka, was one of the principal factors contributing to the build-up in overdues. In 1983, the Exchange Rate Fluctuation Burden Absorption Scheme (EFAS) was, therefore, introduced to give concessions to borrowers in foreign exchange, as a measure to step up loan recovery. Under the EFAS, a fixed percentage of the change in loan values due to exchange rate fluctuations, determined according to the period in which the project was sanctioned, would be converted to preferential shares or debentures to be held by the lending institutions. 19/
In late 1986, a more comprehensive program of loan recovery was initiated. Recovery targets were established for both agricultural and industrial loans of the NCBs and the DFIs. To support this effort, an action program was adopted. The program included the following provisions: prosecution of willful defaulters; prohibitions on defaulters from holding local elected public office or directorship of banks; limitations on access to new credit except for borrowers in good standing and with valid passbooks for agricultural credit, or certificates of “no objection” for industrial credit; and denial of import licenses to industrial loan defaulters.
To encourage agricultural loan repayment, a temporary program of interest amnesty was introduced in September 1986. Accrued interest and penalties on loans were remitted provided that arrears were eliminated by end-February 1987; the deadline was subsequently extended to end-May 1987. The cost of the program, estimated to be about Tk 4 billion, would be shared equally by the Central Government, the BB, and the lending institutions.
In 1989, as required by the BB, each NCB submitted a debt collection strategy plan for its 100 borrowers with the largest amount of overdues. The plan formulates the timetable for collection, measures for monitoring and reporting on performance, and debt negotiations and rescheduling. Out of the four NCBs’ 100 largest defaulters’ overdues of Tk 5.4 billion, collections for the six months ending June 1989 were Tk 0.3 billion, against a target of Tk 0.8 billion. The collections were equivalent to 38 percent of the target, a level which is far from satisfactory. Still, the collection plan marked a significant step in loan recovery efforts.
On the other hand, two DFIs were fairly successful in their loan recoveries against their recovery targets. However, the amount recovered was still small compared to their overdue loans and therefore they have agreed to make provisions for possible bad debts. The third DFI has not shown satisfactory loan recovery performance and still depends heavily on BB refinancing. The BSRS, which was stopped from extending new loans since 1985 was reorganized in March 1987 and merged with the Industrial Advisory Center of Bangladesh.
2. Legal amendment for loan recovery
The National Commission on Money, Banking, and Credit (NCMBC) has proposed a draft loan recovery law to facilitate a speedy and efficient loan recovery. It recommends the establishment of special loan recovery courts with simplified procedures and appropriate jurisdiction to deal with recovery cases of banks and the DFIs. It also proposes that a Standing Law Committee be established with the aim of monitoring important court decisions, key precedents, and improving information on legal issues related to loan recovery.
The Parliament has recently enacted a Financial Loan Courts Act which enables the establishment of special loan recovery courts in each district. Presently, the Financial Loan Courts have been established. These courts will safeguard against undue postponement of court hearings, requiring a debtor to pay 50 percent of the amount decreed prior to any appeal, and allowing appeal within 30 days only to the Supreme Court.
Also, the Ministry of Law has finalized a draft revised Banking Companies Ordinance, initially prepared by the NCMBC. Apart from setting minimum capital requirements for banks, the draft law stipulates regulations to reduce risk concentration and loan delinquency. These include the introduction of a limitation on loans to a single borrower, a prohibition on defaulters serving as directors in financial institutions, and an authorization to enable the BB and financial institutions to publish lists of defaulters and divulge information on delinquent borrowers. Some of these measures have been undertaken by the BB and the Government without the authority of law and hence may be challenged if the banking laws are not changed accordingly. The draft law is awaiting enactment. Its passage, together with the enacted Financial Loan Courts Act, can be expected to improve financial institutions’ loan recovery performance.
3. Interest rate and credit reforms
As was discussed, administered interest rates and the preferential credit program have been the main factors that contributed to the financial distress. The former created distortions in resource allocation and squeezed banks’ earnings. The latter imposed costs and extra risk on state-owned banks from having to lend to high-risk priority sectors at low administered rates. In November 1989, the authorities introduced significant changes in the nature and structure of interest rates and preferential credit program which took effect January 1990.
Apart from loans to four sectors, administered loan rates have been replaced by a band of floor and ceiling to allow banks more flexibility in setting their interest rates. The BB has also simplified the rate structure by reducing the number of lending categories from 29 to 12 (Table 3).
|As a percentage of disbursements||72.7||50.5||52.0||59.0||114.2|
|As a percentage of loans outstanding||18.8||19.6||24.6||29.1||38.5|
|Disbursements since 1977||26,500.8|
|As a percentage of disbursements||63.7|
As of end-March.
As of end-March.
Under the new preferential credit program, the authorities annually reimburse the banks for the cost of interest rate subsidies for six categories of preferential credit. These are agriculture, jute working capital, jute exports, non-traditional exports, special programs for small-scale industries and other special programs which include rural housing and small-scale farmers’ self-sufficiency program. These credits accounted for 43 percent of NCB loan outstanding as of the end of March 1990.
The subsidy cost is measured as the difference between the shadow market rate for such loans and the interest rate charged to borrowers. The shadow market rate is computed based on banks’ average cost of funds, administrative cost, a portion of bad debt expense, and a modest profit margin. The interest rate charged to borrowers is determined by the BB either as an administered single rate or as a band of floor and ceiling, from which the lending bank can choose a rate within the band. Banks will also get the reimbursement of the subsidy on existing preferential loans. All reimbursement costs will be incorporated in the annual budget.
The recent interest rate and preferential credit reforms mark the first step toward a more flexible and market-oriented financial system. However, the reforms have a few major problems. First, from the way lending rate bands are set, it seems that banks will only have limited flexibility in setting their loan rates within the band, and a number of loans may be charged the maximum interest rate. This is so because for some types of loans, for example, loans for jute and jute goods exports, the maximum rate of return for banks (maximum interest rate plus the rate of subsidy) is only slightly higher than the floor, and much lower than the ceiling interest rates on other categories. 20/ Therefore, if banks are to lend to the former, there is no reason why they would not charge the maximum interest rate for those particular categories, since at this rate, their returns inclusive of subsidy would still be lower than their returns from other sectors. Hence, the intended flexibility for the setting of loan rates may not materialize.
Secondly, by stipulating different maximum interest rates for different categories of loans, the authorities have, in effect, set the stage for bank credit allocations during a period of tight credit. That is, when credit becomes tight, loans with lower maximum interest rates will be squeezed compared to loans with higher maximum interest rates, unless the latter have considerable higher risks. However, it is difficult to conceive, for example, that other commercial loans, with a maximum interest rate of 18 percent carry a higher risk than loans for working capital (other than jute), which have a maximum interest rate of 15 percent. 21/ In this case, when the market becomes tight, credit rationing will work against loans for working capital.
Thirdly, the effect of the new preferential lending program will depend on whether banks are under implicit pressure, however subtle it may be, to lend to priority sectors. If NCBs have the implicit obligation to lend to priority sectors, which still account for about 43 percent of NCB loans, their autonomy and efforts to maintain good credit approval practices would remain undermined. Banks’ financial burden in lending to priority sectors, however, will be alleviated because of the subsidies.
It is desirable, therefore, that interest rates be further liberalized. It is important that the authorities keep a close watch on the interest rate structure so that the interest rate floors and ceilings remain in line with the changing market situation, and real interest rates remain positive at all times. The difference in the interest rate ceilings of various types of loans should be as small as possible, giving banks the flexibility to determine their loan rates according to the risk they perceive. As for preferential credit, it is desirable that banks be explicitly assured of their autonomy in lending to priority sectors, that they can refuse loans to these sectors if the risk is too high, and that priority lending be further reduced in the future.
4. Supervisory regulations
The BB is strengthening its bank supervision function. Recently, it has established two supervision/inspection departments and an implementation department to follow up problems identified during or between inspections.
The most significant change in bank supervision is the new system of loan classification, loan provisioning and the suspension of interest announced in November 1989. Banks are now required to classify their loans at least annually, with the BB Inspection Department reviewing their accuracy and requiring modification where necessary. Loans that are overdue by one year, three years, and five years must be classified as substandard, doubtful and loss loans, respectively. Criteria for classifying overdrafts are also defined in relation to loan performance.
In addition to the 50 and 100 percent respective loan provisioning requirements set on doubtful and loss loans, a 10 percent requirement is set on substandard loans, and an one percent requirement on unclassified loans. On the other hand, the transitory measure introduced in 1985 requiring banks to maintain a general provision of 2.5 percent of their total loan portfolio outstanding at the end of 1986, has been abolished. Under the new system, interest on all classified loans must be suspended. Loan provisioning and interest suspension must be done at the branch level to establish an effective loan monitoring system.
State-owned banks will be recapitalized to facilitate loan provisions and write-offs, and to bring the new capital to 5 percent of deposits. This will take place after loan classification and provisioning exercises are completed.
While the building-up of supervisory expertise and a strong supervisory system is a time-consuming process, the measures undertaken are significant steps in the right direction. Apart from loan classification, provisioning, interest suspension, and limitations on banks’ risktaking through concentration on large loans and insider loans, regulations will also be needed to ensure accurate information disclosure and uniform accounting standards. In general, the BB should ensure that prudential banking practices are being followed. Most importantly, the authorities must rigorously enforce all regulations.
5. Measures to improve the NCBs efficiency and profitability
It is essential and urgent that the NCBs improve their efficiency and profitability to restore viability, through the improvement of their management, organization, credit extension procedures, financial management, accounting and information system.
The Board of Directors of state-owned banks should include more private sector members, particularly individuals with professional experience and expertise in banking, finance, or business. Banks need to establish a system whereby the performance of managers is monitored and evaluated against targets, and rewards and penalties given accordingly to motivate staff. Credit criteria for lending should be strengthened with more emphasis placed on a borrower’s creditworthiness than on his collateral. Monitoring and collection of loans need to be streamlined. Uniform accounting standards for branches and reconciliation of interbranch accounts are necessary. Effective internal auditing should be introduced.
Although the Bangladesh banking distress poses no immediate threat of a financial crisis because the government stands fully behind the banks, there is no doubt that it is exerting an adverse impact on both banks and the economy. The most obvious impact is the direct and indirect budgetary drain to recapitalize problem banks, and to make up for lower dividend and tax payments of these institutions to the government. These costs will increase if the distressed situation continues, since the financing cost of problem banks’ deficits further adds to their losses, resulting in the need for a larger recapitalization. Other less obvious but equally costly impacts of the financial distress on the banking system and the economy include the ill effects on banking competition, monetary control, and resource allocation.
Some long-overdue corrective measures have been initiated recently. This marks a significant step in systematically tackling the financial problems of the banking sector. However, rehabilitating a financial system is a time-consuming process. Strong determination is needed especially when the situation poses no immediate threat of a crisis. It is, therefore, important that the authorities realize that the costs of a prolonged financial distress are large, and remain determined to carry out measures that are needed to restore, as soon as possible, the stability of the country’s financial system.
BaliñoTomas J.T. “The Argentine Banking Crisis of 1980” WP/87/77 (February171987).
VelascoAndres “Liberalization, Crisis, Intervention: The Chilean Financial System 1975-85” WP/88/86 (July211988).
JohnstonR. Barry “Distressed Financial Institutions in Thailand: Structural Weaknesses, Support Operations and Economic Consequences” WP/89/4 (January131989).
Government of the People’s Republic of Bangladesh “Report of the National Commission on Money, Banking, and Credit” June1986.
|Deposit money banks||Other Financial Institutions|
|SonaLi Bank||Arab-Bangladesh Bank||Bank of Credit and Commerce International (Overseas)||Bangladesh Shilpa Bank||Bangladesh Shilpa Rin Sangstha Bank||Grameen Bank Bangladesh Rural Development Board||Shadharan Bima Corporation (Insurance)||National Credit Ltd. Bangladesh Commerce and Investment Co. Ltd.|
|City Bank||Bangladesh Krishi Bank||House Building Finance Corporation||Land Mortgage Cooperative Bank||Juban Bima Corporation (Insurance)|
|Agrani Bank||Grindlays Bank||Saudi-Bangladesh Industrial and Agricultural Investment Co., Ltd.|
|International Finance Investment and Commerce Bank|
|Rupali Bank||American Express Bank||Bangladesh Samabaya Bank||Investment Corporation of Bangladesh||Other Cooperative Banks/Societies||Bangladesh Post Office Savings Scheme and Life Insurance Scheme|
|National Bank||Standard Chartered Bank||Rajshahi Krishi Unnayan Bank||Industrial Promotion and Development Co. of Bangladesh Ltd.|
|Pubali Bank 1/|
|Uttara Bank 1/||United Arab Emirate Bangladesh Investment Company|
|State Bank of India|
|Islamic Bank of Bangladesh Ltd. 2/|
|Banque Indosuez||International Leasing and Development Co. of Bangladesh|
|United Commercial Bank|
|Al-Barakah Bank Ltd. 2/||United Leasing Co., Industrial Development Leasing Co. of Bangladesh, Ltd.|
|Bank of Small Industries and Commerce. Bangladesh Ltd.|
Previously nationalized commercial banks.
Operates according to Islamic banking principles.
Previously nationalized commercial banks.
Operates according to Islamic banking principles.
An earlier version of this paper was presented at the IMF Seminar on Central Banking, November 28-December 9, 1988. I would like to thank Paul Murgatroyd, Sergio Pereira Leite, and V. Sundararajan for their comments. All remaining errors are the author’s responsibility.
A series of IMF working papers examine the experiences of countries with financial crises. See Tomas J. T. Balino, “The Argentine Banking Crisis of 1980,” IMF, WP/87/77, February 17, 1987; Andres Velasco, “Liberalization, Crisis, Intervention: The Chilean Financial System 1975-85,” IMF, WP/88/86, July 21, 1988; and R. Barry Johnston, “Distressed Financial Institutions in Thailand: Structural Weaknesses, Support Operations and Economic Consequences,” IMF, WP/89/4, January 13, 1989.
A financial institution is defined to be in distress if it has more liabilities than assets when the latter are properly evaluated, and this situation is expected to persist if significant corrective actions are not implemented. There may not necessarily be a run or a financial crisis in the banking system, because banks are still able to mobilize sufficient liquidity to meet the withdrawals, or because their deposits are perceived to be guaranteed either explicitly or implicitly.
Government of the People’s Republic of Bangladesh, “Report of the National Commission on Money, Banking, and Credit,” June 1986.
This subject has also received considerable attention in the literature. See, for example, Rehman Sobhan and Ahmad Ahsan, “The Effect of Exchange Rate Depreciation on the Loan Repayment Performance of Private Enterprise in Bangladesh,” The Bangladesh Development Studies, Vol. XV, (No. 4, 1987); and other studies cited therein.
This subject has also received considerable attention in the literature. See, for example, Rehman Sobhan and Ahmad Ahsan, “The Effect of Exchange Rate Depreciation on the Loan Repayment Performance of Private Enterprise in Bangladesh,” The Bangladesh Development Studies, Vol. XV, (No. 4, 1987); and other studies cited therein.
The authorities and banks are currently carrying out a loan classification exercise for more updated and complete data.
In its report, the National Commission on Money, Banking, and Credit estimated NCB overdues based on this definition.
Under the new supervisory system introduced in November 1989, banks will have primary responsibility for classifying their loans at least annually.
Under a new supervisory system, loans more than one year overdue must be classified as substandard, and accrued interests on all classified loans must be suspended.
The bookkeeping profits are so large that NCBs are publicly criticized for too big profits and are under pressure to reduce lending rates.
In January 1990, the nature and methodology of preferential lending were significantly changed. These corrective measures will be discussed later.
The exchange rate burden is defined as the sum of the taka values of foreign currency loan with interest converted by the exchange rates prevailing on each date of repayment less the taka value of the foreign currency loan converted at the exchange rate prevailing at the last disbursement date.
See next section for details.
Preferably, a bank’s capital adequacy should be measured against its risk assets or total assets rather than liabilities. This is so because assets carry default risk, which, if losses are incurred, are to be absorbed by a bank’s capital, whereas liabilities carry liquidity risk, which should be taken care of by proper liquidity management. However, since capital requirements for foreign and private banks are stipulated against banks’ liabilities, the requirement for the NCBs could be stipulated in the same way for symmetry.
The Ministry of Law has finalized a draft Banking Companies Ordinance, which calls for a minimum capital requirement of 6 percent of deposits for all banks. However, the government anticipates revising the NCB requirement to 5 percent at this stage.
Again, excluding Janata Bank, whose data on overdue loans are not available.
The loan rates ranged from 9 percent on traditional export credit to 20 percent on domestic commercial credit. Since the NCBs did not lend substantially to the commercial sector, an interest rate of 14 percent, which was on the low side (but closer to agricultural and industrial loan rates) was chosen.
As discussed earlier, Sobhan and Ahsan’s study shows that the proportion of projects seriously affected by this burden has been very small. Hence, the EFAS scheme, by itself, is not likely to solve the loan recovery problem. To the extent that the EFAS scheme put an end to pressure on politicians for forgiveness, and excuses for not repaying loans, etc., it may have improved loan recovery somewhat.
As of end-March.
The maximum rate of return for jute and jute goods exports is 13.6 percent, compared to, for example, a floor interest rate of 13.0 percent and a ceiling of 17.0 percent in the case of large and medium-scale industry loan (see Table 3).
In calculating the explicit or shadow market loan rates, a “reasonable” bad debt expense is included as a proxy of risks of each category of loans. Generally, only a portion of bad debt expense is regarded as “reasonable.” Thus, NCBs which have high bad debt expenses may still have losses on these loans. It is also questionable whether the derived market rates reflect risks accurately.