Selected Issues


Selected Issues

State-Owned Commercial Bank: Performance, Issues and Reforms1

A. Introduction

1. Historically, state-owned commercial banks (SOCBs) have played an important role in Bangladesh.2 In addition to other banking activities, they were tasked to focus on promoting financial inclusion. With a large network of branches, the SOCBs serve as a source of credit to borrowers, such as small and medium enterprises and agriculture, who are unable to otherwise access financing from private commercial banks (PCBs). The SOCBs currently face challenges from weak balance sheets and high non-performing loans (NPLs). The Government of Bangladesh (GoB) has had to repeatedly recapitalize the SOCBs, but achieving financial soundness of the SOCBs has proven to be elusive. However, the share of the SOCBs in the banking sector has declined significantly since 1990s, and along with it the systemic risk they pose to financial sector stability and to the economy more generally.

B. Bangladesh’s Experience with State-Owned Banks

2. Bangladesh’s experience with the SOCBs is not unique: the SOCBs have been underperforming in comparison to private banks and foreign-owned banks. Experiences from countries around the world show that state-owned banks are often subject to political influence and interference, and suffer from overstaffing, lower quality of services, lack of innovation, and inefficiency. This entails as both direct costs (recapitalization of these banks), and indirect costs (misallocation of resources as credit is provided to inefficient borrowers that are often unable to repay their loans) to taxpayers.

3. In Bangladesh, the SOCBs’ share in the banking sector has been generally declining. The asset share of the SOCBs in total banking sector assets has declined from about 55 percent in 1993 to around 27 percent in 2017. A sharper decline has been observed in the SOCBs’ share in banking sector deposits. However, most of the decline took place during the late 1990s and the first decade of 2000s. Since 2010, the share of SOCBs in the banking sector has remained broadly stable. While the authorities have been imposing limits on the SOCBs’ lending growth, these limits (which vary by the individual banks) have not, in aggregate, been tight enough to result in a further decline in the SOCBs’ share in the total banking sector. In addition, the share of the SOCBs’ deposits is now almost identical to the share of the SOCBs’ assets, a change from the 1990s when the share of SOCBs’ deposits was almost 10 percentage points higher. Similarly, the share of development finance institutions (DFIs) fell notably, from 12 percent in 2001 to less than 3 percent by 2017. Instead, the PCBs’ share in banking sector has expanded significantly.

Figure 1.
Figure 1.

Banking Sector Structure

Citation: IMF Staff Country Reports 2018, 159; 10.5089/9781484360347.002.A001

Source: Bangladesh Bank and Staff Estimates

4. The decline of SOCBs’ share in the banking sector is not unique to Bangladesh. Government ownership of banks has been generally declining in all regions, including South Asia, though the share of assets of government-owned banks in the financial system varies widely across countries and regions. In some countries, government-owned banks account for more than 50 percent of financial sector assets (India and Sri Lanka), while in others (Bhutan, Maldives, and Indonesia) this is in the 30–50 percent range. In comparison with other South Asian countries, the SOCBs’ share in Bangladesh is relatively low (Figure 2). However, it is much higher compared to East Asia & Pacific (except Indonesia) and Sub-Saharan Africa (Cull et al., 2017).

Figure 2.
Figure 2.

State-Owned Banks’ Asset Share, 2010

(In percent of total banking sector assets)

Citation: IMF Staff Country Reports 2018, 159; 10.5089/9781484360347.002.A001

Sources: R. Cull et al. (2017)Bangladesh Bank for Bangladesh data (Cull shows higher share of 34 percent).

5. Figure 3 provides an overview of the structure of bank lending by industries. The charts on the left show a composition of each group’s lending portfolio by industry (in percent), and the charts on the right show the amount of loans in each industry by bank groups (in taka). Several observations emerge from the data:

  • Loans to agriculture sector represent a major share of total loans by the DFIs, and to a less extent, of the SOCBs. For the PCBs and the foreign-owned commercial banks (FCBs), loans to agriculture are a small percentage of their total loan portfolio. However, the PCBs still represent about one third of total agriculture loans.

  • Industrial sector is the single most important borrower, except for the DFIs, with shares between 40–50 percent. Loans to industry represent a very small and declining part of the DFIs loan portfolio. From the perspective of the borrower, the PCBs are the dominant source of loans.

  • Loans to construction provide a similar picture as industry. However, loans to construction represent a very small share in the FCBs total loan portfolio.

  • Loans to trade and other sectors, such as commerce and services, represent a sizable share of total loans, similar to loans to industry. However, during the past decade, this share has gradually declined for the PCBs and the FCBs. The PCBs loans are the most significant source of borrowing for these sectors.

Figure 3.
Figure 3.

Bank Loans by Industry

Citation: IMF Staff Country Reports 2018, 159; 10.5089/9781484360347.002.A001

Source: Bangladesh Bank and Staff Estimates

C. Comparative Performance of Banks in Bangladesh

6. This section describes the relative performance of different groups of banks. The picture that emerges is quite clear: the SOCBs are in much weaker financial position than either PCBs or FCBs.

7. Figure 4 compares NPLs as a percent of total loans (gross and net of provisions and interest suspense). After peaking at 35 percent in 2000, the NPL ratio in the banking system declined significantly, to around 6 percent in 2011. However, starting in 2012, NPLs have begun to increase, partly reflecting the change in loan classification, and have crossed 10 percent in 2016 and 2017. The SOCBs have been showing much higher NPLs than the PCBs and the FCBs, and the difference has significantly increased over time. In 2017, the SOCBs’ NPL ratio was almost five times as high as those for the PCBs.

Figure 4.
Figure 4.

Non-Performing Loans to Total Loans

Citation: IMF Staff Country Reports 2018, 159; 10.5089/9781484360347.002.A001

Source: Bangladesh Bank1/ Net of provisions and interest suspense

8. The increase in the NPL ratio since 2011 also reflects loan deterioration in all types of banks: in all three groups, the NPL ratio has more than doubled. It is noteworthy that this deterioration took place when the real GDP growth remained within 6–7 percent range. A similar picture emerges from net NPLs. For the banking sector as a whole, they fell from almost 30 percent in 2000 to close to zero in 2011 (and even below zero for the SOCBs), but then began to increase again, mainly due to higher net NPLs in the SOCBs, where provisioning did not keep up with loan deterioration.

9. Not only do the SOCBs have a significantly higher share of NPLs, but their capital adequacy ratio (CAR) is much lower compared to the PCBs and the FCBs (Figure 5). The SOCBs’ CAR had reached close to 11 percent in 2013, above the minimum CAR requirement of 10 percent. Since then, it has been declining despite several rounds of recapitalization by the GoB. The ratio has now fallen below 6 percent. In contrast, the PCBs’ CAR has been relatively stable and above the minimum CAR requirement.

Figure 5.
Figure 5.

Capital to Risk Weighted Assets Ratio

(In percent)

Citation: IMF Staff Country Reports 2018, 159; 10.5089/9781484360347.002.A001

Source: Bangladesh Bank

10. The poor performance of the SOCBs in terms of NPLs and CAR has also led to poor profitability indicators. Both return on assets (ROA) and return on equity (ROE), two standard indicators of profitability, show weak performance in the SOCBs (Figure 6). Since 2014, the SOCBs as a group have been reporting losses, and thus have a negative ROA and ROE. In 2017, the SOCBs losses have accelerated, with ROE reaching a negative 20 percent. In contrast, the PCBs and the FCBs have maintained a relatively stable ROA and ROE in recent years, though with some decline in 2017.

Figure 6.
Figure 6.

Bank Profitability Indicators

Citation: IMF Staff Country Reports 2018, 159; 10.5089/9781484360347.002.A001

Source: Bangladesh Bank

11. Figure 7 provides more insights into the financial performance of the banking sector. In 2012, there was a significant jump in the required provisions, in particular for the SOCBs, where required provisions doubled. This jump reflected a major tightening of the provisioning requirement (see section E for more details).3 However, required provisions continued to grow thereafter, and the pace accelerated in 2016 and 2017. The tighter provisioning requirement led to the decline in profitability in 2012, as banks increased their provisioning.4 While the PCBs maintained actual provisions in line with this requirement, the SOCBs had a large shortfall in provisioning in 2012, but were able to reach full provisioning by 2013–14, before falling behind again in the last three years. Due to the under-provisioning, the actual SOCBs’ profitability has been even weaker than suggested by the ROE and ROA.

Figure 7.
Figure 7.

Bank Provisioning

Citation: IMF Staff Country Reports 2018, 159; 10.5089/9781484360347.002.A001

Source: Bangladesh Bank

12. The weak financial performance of the SOCBs can be partly linked to a relatively low interest rate margin – the difference between the costs of funds (deposits) and the lending rates. While the SOCBs’ average deposit rates have been only slightly lower than the PCBs’ rates, their average lending rates have been lower by about 2 percentage points, resulting in a much lower (by about one third) interest rate margin (Figure 8).

Figure 8.
Figure 8.

Bank Deposit and Lending Rates

Citation: IMF Staff Country Reports 2018, 159; 10.5089/9781484360347.002.A001

Source: Bangladesh Bank

13. Table 1 confirms the positive relationship between banks’ interest margin and their profitability. Given their ownership status and perceived government backing, the SOCBs are able to attract deposits at lower costs than PCBs (including government deposits and deposits by state-owned enterprises), but this does not translate into higher profitability as their lending rates are even lower, compared to PCBs’ rates.5

Table 1.

Bangladesh: Interest Margin and Profitability

(Average value over 2008 to 2017)

article image
Source: Bangladesh Bank

14. The implicit government guarantee allows the SOCBs to operate with high leverage. For a given profit and ROA, a higher leverage ratio (the ratio of banks’ assets to bank’s equity) would result in higher ROE. As already indicated by the relatively low CAR, the SOCBs support their assets by relatively low equity, and their leverage ratio has been consistently above other banks -and progressively so in recent years. In contrast, the FCBs’ leverage ratio has been almost steady at around 5, a conservative ratio.

D. Causes of SOCBs’ Poor Performance

15. The relative under-performance of the SOCBs continued with various challenges. The SOCBs have always had higher NPLs, lower CAR, and lower ROA and ROE. The efforts to strengthen the SOCBs’ performance (see next section) have brought some improvements, but much remains to be done to restore their financial strength when compared to the PCBs and to regulatory standards. There is no one cause for the SOCBs’ poor performance. Rather, it is a combination of factors that reinforce each other:

  • Government mandates. The SOCBs are tasked with performing some public policy roles, extending loans to potentially risky borrowers – sometimes at subsidized rates- and providing some services for government without charge. While these demands may have legitimate public policy objectives, if not fully financed (either by budgetary support or pricing policy allowing cost recovery), they will continue weakening the financial position of the SOCBs.

  • Risk management. There are still shortcomings in the SOCBs’ risk management. Comprehensive risk management policy is lacking and early fraud detection could still be a challenge.6

  • Weak governance. The SOCBs have a history of weak governance and outright fraud which has resulted in large losses.7 Independent, qualified, and reputable professionals are not always present on their boards. The “Fit and Proper Test Criteria” guidelines for bank directors are not yet fully implemented. Decisions are often made against the interest of the bank, with insufficient evaluation of loan requests, connected lending, and insider dealing of client company’s share. Importantly, there were often no consequences for these poor decisions. Not many board members or chairmen were charged.

  • Unconditional recapitalization. Repeatedly, the SOCBs have been recapitalized without sufficiently addressing the underlying problems of undercapitalization. Giving more funds to the SOCBs without resolving the causes of past poor performance will not produce sustained financial improvement.

  • Supervision and regulatory forbearance. While the supervisory and regulatory authority of Bangladesh Bank (BB) over the SOCBs has increased, the implementation has not been sufficiently robust. Without a firm enforcement of regulatory limits, and the possible resolution of banks that are continuously in breach of regulatory requirements, BB’s ability to improve the SOCBs’ performance will be limited.

  • Automation and management information system. The SOCBs have been lagging behind other banks in adopting information technology-based banking services, including electronic banking services, electronic fund transfers, and automated teller machines. This has adversely affected their efficiency and profitability, as well as internal controls and fraud prevention and detection.

  • Human resources. There are shortcomings in the SOCBs human resource (HR) management, including the limited role of HR departments in recruitment and promotions. Limited HR capacity could cause insufficient due diligence in lending decisions, and thus increase the risk of wrong or even fraudulent decisions.

16. An ineffective system of bad loan recovery is also adversely affecting the banking sector. The current system allows some borrowers not to repay, even if they are able to do so. The legal framework does not help banks to recover such loans quickly: courts are overburdened; borrowers can file writ petitions to the money loan courts and delay the whole process; seizing the collateral is very lengthy and time-consuming. There are many writ petitions filed against the Credit Information Bureau of BB not to classify these borrowers as defaulters. Borrowers are not always encouraged to service their debt on time, particularly if they can continue borrowing while their case is pending. All of these contribute to relatively high NPLs in the banking sector.

E. Previous Efforts to Improve SOCBs’ Performance

17. In the past, the authorities have undertaken several steps to improve the performance of the SOCBs with mixed results. While these initiatives have brought some limited improvements, as is evident from the analysis in section C, they failed to realize a decisive turnaround in the SOCBs weak performance.

18. The GoB, with the support of the World Bank, undertook the state bank reform program to improve the financial performance of the SOCBs. In 2007, the SOCBs were corporatized and brought under BB’s regulatory authority, with notional capital increase in goodwill/capital equivalent to the SOCBs accumulated losses. Steps were taken to select the SOCBs’ management competitively, and the SOCBs were expected to operate in the same way as private commercial banks. In addition, the GoB began to set up annual monitorable goals for the SOCBs based on cash recovery of NPLs, limits on new NPLs, computerization of branches, profitability, and disclosure. The SOCBs management and boards were made responsible for achieving these targets.

19. In 2007, BB increased the minimum CAR from 9 to 10 percent (and from 4.5 to 5 percent for tier 1 capital). These steps contributed to higher SOCBs’ CAR (though it remained below the required minimum – see Figure 5).

20. The authorities were also implementing a program of the SOCBs divestment. Three banks were slated for privatization during 2005–07, while Sonali, the largest SOCB with a strong rural presence, was supposed to be restructured and divested over the medium term. However, the privatization of Rupali Bank fell through and the whole divestment plan was eventually suspended. The authorities instead changed their focus on improving the SOCBs’ performance, with the view that stronger performance would improve the odds of successful divestment.

21. In 2012, the loan classification was tightened, to bring it more in line with international practices. This resulted in a jump in required provisions (Figure 7). The revision shown here applies to continuous and demand loans. The definition and delinquency periods were also tightened for fixed term loans.

Table 2.

Bangladesh: Loan Definition and Delinquency Periods

article image
Source: Bangladesh Bank

For some priority sectors, including agriculture and SMEs, delinquency periods are longer before provisions are required.

22. Strengthening the SOCBs performance has also been one of the objectives of the Fund-supported program under the Extended Credit Facility. In 2013, the Bank Companies Act was amended, to give BB full regulatory and supervisory control over the SOCBs. BB was given more authority to take punitive actions against SOCB management.8 In 2015, BB announced that it will place observers for the board of directors in the SOCBs with worsening internal governance. As part of its enhanced supervisory powers, BB has also begun to sign memoranda of understanding (MoU) with the SOCBs with specific quantitative targets, including the reduction of NPLs, limits on the growth of their lending portfolio (ranging from 8 to 20 percent in 2017), and cash recovery from the largest defaulters. BB also completed a special diagnostic examination of the SOCBs in 2013 and formulated a plan to address the identified deficiencies.

F. Policy Recommendations and Conclusions

23. Bangladesh’s growth has been strong in recent years, but this should not mask the potential costs of inefficient SOCBs that account for more than a quarter of the banking system. Large NPLs imply that a large amount of the savings is being wasted by financing loss-making activities, and therefore becomes unavailable for financing productive investments. High NPLs and the need for provisions also increase the cost of credit to good borrowers, further dampening investment and growth. Perceived government guarantees of the SOCBs create an uneven playing field for private banks, while the SOCBs undercapitalization presents a potential fiscal liability.

24. Resolute steps are required to resolve the SOCBs’ weak performance, reflecting empirical evidence and mixed-results from the previous efforts. The following steps could be considered:

In the short to medium term:

  • BB and the GoB should strictly enforce the existing bank regulations and rules. In particular, the current practice of no punitive action for not observing the annual performance agreement with the Ministry of Finance or the failure to meet the goals set out in the MoUs with BB should be reconsidered.

  • In performing its regulatory responsibilities, BB should be free of political pressure.

  • To avoid potential conflict between the MoU’s targets and the GoB’s demand on the SOCBs to pursue non-commercial objectives, there is a need to clearly spell out the SOCBs’ mandate and provide budget support, if necessary, to finance this mandate. To the extent that the SOCBs are not allowed to apply cost recovery pricing to the services and products they are asked to provide, the upfront budget transfers rather than ex-post recapitalization should be considered.

  • Independent management should be put in place in the SOCBs that meet the fit and proper criteria.

  • The SOCBs should be gradually recapitalized to meet the regulatory capital standards.9

In the long-term:

  • The authorities should formulate and implement a long-term plan for the SOCBs and their role in the financial system. To the extent that the GoB needs to continue using the financial system to deliver certain policy objectives, such as providing financial services to under-served groups or areas of the economy, it needs to carefully consider the most effective way to achieve such objectives.

  • The GoB should consider the reduction of the number of the SOCBs by either merging or privatizing some of the existing SOCBs. In the long run, there could be 1–2 SOCBs, financed from the budget, that would be tasked with pursuing specific mandates given by the GoB. Other SOCBs could be divested.

  • Strengthening the financial position of the SOCBs with support of above-mentioned short and medium-term measures could prepare the SOCBs for divestment. The GoB could consider SOCBs restructuring before divestment. Alternatively, prospective strategic investors could undertake this task, though this would come at the cost of a lower sale price of these SOCBs.

25. The GoB has ambitious medium-term growth aspirations, and wants to achieve middle-income status. However, for these ambitions to materialize, an efficient financial system, including a healthy banking sector and well-functioning capital market, are indispensable.


  • Cull, R., M.S.M. Peria and J. Verrier, 2017. “Bank Ownership: Trends and Implications”, IMF Working Paper 17/60, International Monetary Fund, Washington.

    • Search Google Scholar
    • Export Citation
  • Mitra, S. and M. Zephirin, 2014. “State-Owned Financial Institutions – Strict Oversight is Key”, MCM Policy Position Note, International Monetary Fund, Washington.

    • Search Google Scholar
    • Export Citation

Prepared by Jiri Jonas and Muhammad Imam Hussain.


This paper focuses on SOCBs. In addition, there are also two DFIs operating in Bangladesh. The DFIs are government-owned and financed, and hold less than 3 percent of banking sector deposits and 2.5 percent of assets. They specialize in providing loans to targeted sectors, including agriculture. While their performance is in some respects even weaker than that of the SOCBs, their limited and declining size imply that they do not pose a systemic risk.


For past due/overdue loans, the minimum time due has been reduced from 6 months to 3 months, while the provisioning for classified loans remained unchanged at 20, 50 and 100 percent for the sub-standard, doubtful and bad/loss loans.


The SOCBs moved from profits to losses in 2012. The profitability of the PCBs also deteriorated in this year, dragging down the ROA and ROE for the entire banking sector. In addition to the change in provision requirement, the stock market correction in mid-2011 and 2012 could also have had some impact on banks’ profitability.


The FCB deposit rates are even lower than SOCBs rates. One reason could be better FCBs’ service quality and advisory expertise. Also, the multinational companies’ policy is to do transactions with only three foreign banks in Bangladesh.


Bangladesh Institute of Bank Management recently conducted a survey of chief risk officers, which revealed significant shortcoming in using risk management tools, not just in SOCBs, but in the whole banking sector. See–02-2018


Several examples of fraud in the SOCBs are described in the article “From Cancer to Pimple” in the Economist, April 16, 2014.


BB can remove the managing director of the SOCBs. However, BB’s action against any member of the Board of the SOCBs requires a consent of the Finance Ministry.


It is sometimes argued that because of the implicit or explicit backing by the government, deposit-taking SOCBs do not need to maintain a high capital cushion as private banks because there is little risk of depositors’ run. However, applying the same capital requirements to the SOCBs provides more transparency about budgetary support than guarantees, and serves as a break on expansion of their activities. See Mitra and Zephirin, (2014).

Bangladesh: Selected Issues
Author: International Monetary Fund. Asia and Pacific Dept