Republic of Tajikistan: Financial System Stability Assessment Report
Author: International Monetary Fund. Monetary and Capital Markets Department
Publication Date:
09 Feb 2016
eISBN:
9781498303019
ISBN:
9781498303019
Language:
English
Keywords:
ISCR; CR; national bank; interest rate; banking sector; liquidity lending
Download PDF (1.1 MB)

This paper highlights Tajikistan's macroeconomic environment and gives an overview of the financial sector and its stability, and discusses its regulatory, supervisory, crisis prevention, and management framework. The economy of Tajikistan is entering a downturn, and the banking sector is showing substantial weakness. GDP growth has been on a declining trend. Inflation is also projected to rise further. The external position continues to deteriorate, putting pressure on the somoni and eroding already low external buffers. The financial sector of Tajikistan is dominated by banks, which account for 84 percent of total financial sector assets. Dollarization in the financial sector has been increasing and remains a challenge for foreign exchange risk and credit risk management.

Abstract

This paper highlights Tajikistan's macroeconomic environment and gives an overview of the financial sector and its stability, and discusses its regulatory, supervisory, crisis prevention, and management framework. The economy of Tajikistan is entering a downturn, and the banking sector is showing substantial weakness. GDP growth has been on a declining trend. Inflation is also projected to rise further. The external position continues to deteriorate, putting pressure on the somoni and eroding already low external buffers. The financial sector of Tajikistan is dominated by banks, which account for 84 percent of total financial sector assets. Dollarization in the financial sector has been increasing and remains a challenge for foreign exchange risk and credit risk management.

Macroeconomic Environment and Financial Sector Overview

A. Macroeconomic Background and Outlook

1. Tajikistan’s economy is small and vulnerable to shocks. It relies heavily on imports (65 percent of GDP) and its exports (19 percent of GDP) are narrowly based and largely commodity-driven (aluminum, cotton, and gold). Most of its fuel and a large share of food are imported. Tajikistan depends heavily on exports of labor and the resulting remittance income (equivalent to about 50 percent of GDP in 2013), earned mainly by migrant workers in Russia. The vulnerabilities are exacerbated by a poor business climate; Tajikistan stands last among the seven Caucasus and Central Asia (CCA) countries in terms of the ease of doing business (Figure 1).

Figure 1.
Figure 1.

Tajikistan: Macroeconomic Developments

Citation: IMF Staff Country Reports 2016, 041; 10.5089/9781498303019.002.A001

Sources: NBT; WEO; World Bank Doing Business (2015); and IMF staff calculations.

2. GDP growth has been on a declining trend. Economic growth was 6¾ percent in 2014, supported primarily by the rapid expansion of construction (25 percent). Remittances fell by 8½ percent in 2014 and are estimated to drop by 35 percent in 2015. This will have an adverse impact on the services sector and growth. While rapidly growing economic ties with China may help underpin the outlook, the fallout from the adverse external environment—most notably, in Russia—and the slow pace of structural reforms will weigh heavily on growth in 2015 and beyond.

3. Inflation rebounded from recent lows and is projected to rise further. The increase in the CPI remained moderate at 3 percent and 7 percent during the last two years, owing to a relatively stable exchange rate and favorable international food and fuel prices. However, recent electricity tariff increases, ongoing rapid credit growth, and a pass-through from exchange rate depreciation (see below) are expected to push inflation into double digits in 2015.

4. Monetary policy has been largely accommodative. Base and broad money grew at 13¼ percent and 7 percent, respectively, in y-o-y terms through December 2014, but private credit growth surged by 31½ percent during the same period. Although direct National Bank of Tajikistan (NBT) liquidity lending to banks has been limited, the placement of government deposits with commercial banks has supported the liquidity of several large banks (including those in violation of prudential norms) and fueled already high credit growth. The policy rate—the NBT’s refinance rate—remained negative in real terms until recently. Continued losses and negative capital at the NBT could compromise its ability to conduct sound monetary policy.

5. The external position continues to deteriorate, putting pressure on the somoni (SM) and eroding already low external buffers (Figure 2). Although the official exchange rate depreciated by about 23 percent against the US dollar since December 2013, ahead of Kazakhstan and Kyrgyz Republic (with roughly 17 percent depreciation), this felt short of the devaluation of the Russian ruble (37 percent) for the same period. A relatively strong somoni (which supported high import growth) together with falling remittances and depressed exports (largely of cotton and aluminum) sharply eroded the external current account balance. Absent compensating inflows on the capital account side, the authorities intervened heavily in the foreign exchange market, causing international reserves to fall to one month of imports by end-2014. Inadequate reserves coverage in an environment of high dollarization constrains the effectiveness of systemic liquidity management by the NBT.

Figure 2.
Figure 2.

Tajikistan: Macroeconomic Vulnerabilities

Citation: IMF Staff Country Reports 2016, 041; 10.5089/9781498303019.002.A001

Sources: NBT; WEO; and IMF staff calculations.

6. The fiscal stance remains consistent with debt sustainability but faces risks. Preliminary fiscal data for 2014 show a surplus (excluding the foreign-financed public investment program). However, the construction of the Rogun Hydropower Project (HPP) may stress debt and external sustainability. In addition, the fiscal position may be adversely affected by possible costs of bank recapitalization as well as by external debt service-related budgetary outlays on behalf of two loss-making state-owned enterprises (SOEs).

B. Financial Sector Structure

7. The financial sector is dominated by banks, which account for 84 percent of total financial sector assets (Table 2). Microfinance institutions (MFIs) account for most of the remainder. There is a small insurance sector and practically no capital markets activity or advisory services industry. There are 12 leasing companies, but data on their operations are not available. Recently, a license was issued for the second (private) credit bureau.

Table 2.

Tajikistan: Financial System Structure, 2012–14

article image
Sources: NBT; and World Bankstaff estimates.Notes:

Data on leasing companies is not available.

Commercial banks

8. The banking sector remains small compared to other CCA countries and is concentrated. It reports assets of less than 30 percent and loans of 15 percent of GDP, and consists of 17 commercial banks, including one fully state-owned, one majority state-owned, and seven majority foreign-owned institutions. The six largest banks account for 81 percent of total bank assets, and majority foreign-owned banks account for 10½ percent. One large bank is insolvent and another fails to meet all prudential requirements, jointly accounting for 41 percent of total bank assets.

9. Nonperforming loans (NPLs) have grown rapidly, especially in large banks. The NPL ratio for the six largest banks stood at 28¼ percent, compared with the industry average of 25 percent as of December 2014.1 However, the official numbers mask major weaknesses in some banks. They are under-provisioned, with specific provisions accounting only for 42¾ percent of NPLs,2 which results in an overstatement of banks’ capital positions.

10. While some progress had been made on AIB restructuring, further postponing the final decision is likely to result in a higher fiscal cost. The authorities believe that the recent appointment of a new Chairman will expedite the collection of NPLs and position the bank well for new lending. However, this may delay the needed resolution and restructuring of the bank. Therefore, the authorities should pursue the restructuring of the AIB without delay, consistent with best practices of bank resolution, as also outlined in the joint IMF-WB May 2014 TA report on AIB resolutions options, to reduce system-wide moral hazard and jumpstart lending by the bank. Given the systemic nature of AIB, the preferred options would be recapitalization with conditions or a bridge bank (Box 1).

Options for Resolving Agroinvestbank

AIB is a troubled bank with negative regulatory capital, classified by the NBT as systemically important. The delay in its resolution negatively affects market discipline in the overall system, imposes costs on the real economy, feeds moral hazard of other troubled banks, and could have other adverse spillovers to the rest of the banking system.

There are three options:

(1) Recapitalization with conditions: This would involve writing down existing shareholders, restricting dividends, injection of government capital, and eventually re-privatizing the bank to reputable investors. This would not require legal changes, but would carry fiscal costs of 1¾–2½ percent of GDP. The risk of this approach is that—absent strong safeguards—the AIB’s business model would remain unchanged, the legacy loan portfolio would weigh heavily on its financial performance going forward, exposing the public to the need for a further bail-out.

(2) Bridge bank resolution based on a revised legal framework for bank resolution: This would entail placing the bank under a temporary administration, until the legal framework is amended to allow for stronger resolution tools that would allow the transfer of insured deposits and “good” assets to a bridge bank, and the liquidation of “bad” assets and remaining liabilities by a “bad” bank. This approach would have a lesser fiscal cost, assuming some bailing in of creditors and large depositors, and would leave the new “good” bank with a healthier balance sheet and therefore more likely to be viable.

(3) Bank liquidation under the current framework: This would entail withdrawing the AIB’s license, initiating a court-led liquidation, compensating insured depositors, and covering other creditors by the proceeds of liquidation. There would be no immediate fiscal costs, but the franchise value of AIB would be lost, and the potential for a generalized loss of depositor confidence would be greater.

11. The sector suffers from the legacy of directed lending. Although the extent of directed lending has declined markedly in recent years, the misallocation of credit is likely to have imposed important economic and fiscal costs. While these are difficult to measure precisely, due to data and analytical limitations, illustrative calculations can be constructed based on the literature on the effects of financial deepening on growth (e.g., Beck, Levine, Loayza, 2000). For example, assuming that nearly 50 percent of AIB’s loan book consists of nonperforming loans due to directed lending, this could have reduced annual GDP growth by as much as ¼ percentage points. And while the fiscal costs are also hard to measure, restructuring AIB could require the injection of public funds in an amount equivalent to around 2½ percent of GDP (Box 1).

Insurance sector

12. The insurance sector in Tajikistan is small and underdeveloped. Eighteen companies currently operate in Tajikistan, one of which is foreign owned and two are state owned. It is supervised by the State Insurance Supervisory Service (SISS), which is a unit of the Ministry of Finance (MoF). Discussions are underway to move the supervisory function to the NBT. Life insurance is particularly underdeveloped and amounts to less than 2 percent of total gross insurance premiums. The sector suffers from an absence of actuaries, underwriters, or loss adjustors.

Microfinance institutions

13. The microfinance institutions (MFIs) have experienced sustained growth in recent years and the largest institutions compete directly with banks for SME lending. The sector—which represents nearly 15 percent of the financial sector—comprises three tiers that serve over 400,000 borrowers: 42 micro-credit depository organizations (MDOs), 42 micro-credit organizations (MCOs), and 36 micro-credit funds (MCFs). The sector is also concentrated, with the five largest MFIs accounting for around 70 percent of total assets, which compete with banks in the SME segment. One microfinance organization acts exclusively as a wholesaler to smaller institutions.3 MFIs also outperform banks on the number of loans, portfolio quality, and profitability. The NPL ratio in the microfinance sector was around 3¼ percent, while their return on assets reached 6¼ percent by September 2014. However, the sector faces substantial funding constraints, especially in local currency.

Leasing

14. The leasing sector is small and unregulated, and the absence of market monitoring or oversight makes analysis difficult. Banks and MFIs can provide leasing on their books, but the size of their financial leasing transactions is very small. Approximately, 90 percent of leasing transactions correspond to financial leasing, 10 percent relate to leaseback transactions, while operational leasing appears to be nonexistent.

Capital markets

15. The capital market is limited to an emerging government debt market in addition to the NBT’s open market operations. The government securities market is comprised of 91-day treasury bills (T-bills). In 2014, there were five auctions with only two auction participants: the state-owned Amonatbank and the Deposit Insurance Fund (DIF). The NBT’s open market operations are more vibrant and include weekly auctions and instruments with maturities of 7, 14, 28, and 56 days. Most recently, the authorities have established a closed joint stock company “Securities Market of Tajikistan” (100 percent owned by the NBT), with a mandate to develop a secondary market for government securities and to establish a new stock exchange. In addition, Tajiksodirotbank has registered a company under the name “Dushanbe Stock Exchange.” However, there are no market activities because the company has not yet been issued a license to operate as a stock exchange.

C. Progress with 2007 FSAP Recommendations

16. Tajikistan has made some progress in implementing the 2007 FSAP recommendations. The authorities at least partially implemented many of the key recommendations of the 2007 FSAP. In the banking sector, the surveillance framework for monitoring the build-up of credit risk has been strengthened and a wide range of corrective measures were introduced. Governance and autonomy of the NBT were enhanced and some funds were transferred to recapitalize the NBT. The bank regulatory and supervisory framework was strengthened. In the nonbank financial sector, a program of government securities issuance was launched; new laws on Deposit Insurance of Individual Savings (2011) and Credit Histories (2009) were adopted and were largely aligned with good international practice. Table 8 outlines the specific steps that have been taken and those remaining to be addressed.

Table 3.

Tajikistan: Liquidity Stress Test Results

article image
Sources: NBT; and IMF staff estimates.
Table 4.

Tajikistan: Large Depositor Outflow Test

article image
Sources: NBT; and IMF staff estimates.
Table 5.

Tajikistan: Selected Economic Indicators, 2011–20

(Quota: SDR 87 million)

(Population: 8.3 million; 2014)

(Per capita GDP: US$1121; 2014)

(Poverty rate: 42 percent; 2011)

(Main exports: aluminum, cotton, 2013)

article image
Sources: Data provided by the Tajikistan authorities, and Fund staff estimates.

Private investment and savings are estimates. Investment includes changes in stocks.

Including statistical discrepancy, except in 2013 and 2014 where statistical discrepancy is treated as below the line domestic financing item.

Figures differ from those reported earlier due to structural revision to monetary and financial sector statistics based on recent IMF TA.

Slowdown in 2012 is due to bad loans write-off at Agroinvestbank.

Receipts from aluminium exports under the tolling arrangements are booked as services exports.

Adjusting for unrecorded oil imports in 2012-13.

Excluding imports related projects financed with loans from China.

Table 6.

Tajikistan: Financial Soundness Indicators, 2010–14 1/

(In percent, unless specified)

article image
Sources: National Bank of Tajikistan.

Nonperforming loans include 4 loan classifications: substandard (1-30 days past due), doubtful (30-60), problem (60-180), and bad (180 days +).

Nonperforming loans include 3 loan classifications: doubtful (30-60), problem (60-180), and bad (180 days +).

Table 7.

Tajikistan: Risk Assessment Matrix (January 2015) 1/

article image
article image

The Risk Assessment Matrix (RAM) shows events that could materially alter the baseline path (the scenario most likely to materialize in the view of IMF staff). The relative likelihood of risks listed is the staff’s subjective assessment of the risks surrounding the baseline (“low” is meant to indicate a probability below 10 percent, “medium” a probability between 10 and 30 percent, and “high” a probability between 30 percent and 50 percent). The RAM reflects staff views on the source of risks and overall level of concern as of the time of discussions with the authorities. Non-mutually exclusive risks may interact and materialize jointly.

Table 8.

Tajikistan: Status of 2007 FSAP Main Recommendations

article image
article image
article image
article image

Financial Sector Stability

A. Key Macro-Financial Risks and Vulnerabilities

17. Dollarization in the financial sector has been increasing and remains a challenge for FX risk and credit risk management. Prior to the global financial crisis, it was 82 percent and 76 percent, for bank loans and deposits, respectively, before tapering off some. From early 2011, both indicators of dollarization have again been on the rise, reflecting a lack of confidence in the somoni and the business environment. The overwhelming majority of private enterprises that borrow in foreign exchange earn income in somoni and thus are not hedged against depreciation. In the past, external shocks and substantial exchange rate depreciations have imposed significant costs on the banking sector. High dollarization also complicates monetary policy management and crisis management and resolution.

18. Rapid credit growth in a worsening macroeconomic environment is likely to make the system more vulnerable. A credit boom in 2014, with private credit expanding to 31½ percent, was accompanied by a slowing real economy, exchange rate pressures, and loss of foreign exchange reserves. The placement of government deposits with commercial banks supported liquidity in the system and fueled already high credit growth. Weak underwriting standards, a legacy of directed lending, and regulatory forbearance have encouraged undue risk taking. In addition, credit quality is impacted by weaknesses in formal credit assessment methodologies and insufficient credit information. Moreover, credit growth has exceeded deposit growth, pushing the loan-to-deposit (LTD) ratio above 100 percent. FX loans have been expanding at a faster rate than general credit, exposing banks to indirect credit risk by customers, who maintain unhedged short foreign exchange exposures. (Figure 3).4 An early sign of mounting vulnerabilities is the NPLs, which increased significantly to 25 percent at end-December 2014 compared to 16 percent a year ago.

Figure 3.
Figure 3.

Tajikistan: Private Credit and Money Aggregates

Citation: IMF Staff Country Reports 2016, 041; 10.5089/9781498303019.002.A001

Sources: NBT; IFS; and IMFstaff estimates.

19. Weaknesses among public sector borrowers represent another macro-financial risk. The weaker macroeconomic environment means that state-owned companies will likely have greater difficulties servicing their loans, with consequences for banks.5 At end-2014, the share of government and SOEs in total bank credit was about 11 percent; 74 percent of these loans were overdue for 30+ days, the worst ratio among all economic sectors. Reform of the SOE sector will be a critical precondition for improving asset quality and securing financial stability.

20. Government-directed lending and poor governance at banks have increased credit and concentration risks. Exposure to the construction sector of one large bank increased five-fold in 2014. Another bank, which accounts for 50 percent of government on-lending, has an NPL ratio that is more than triple the industry average. Insider and connected lending is prevalent and interference in lending decisions is not uncommon. This practice is often motivated by efforts to support flagship public enterprises facing difficulties or by food and fuel security considerations.6 Unless curbed, directed lending will continue to take a toll on the financial sector and productivity (see Box 1 on AIB, the main vehicle for directed lending in recent years).

21. Loan concentration exacerbates the banks’ vulnerabilities. Loans are concentrated in the commercial sector (46 percent of the total NPLs), to government entities (30 percent of total NPLs), and to SMEs (14 percent of total NPLs) (Figure 4). The impact on banks is exacerbated by the fact that these sectors are where NPLs are the highest.

Figure 4.
Figure 4.

Tajikistan: Bank Loans and NPLs Concentration by Economic Sector

Citation: IMF Staff Country Reports 2016, 041; 10.5089/9781498303019.002.A001

Sources:NBT, IFS;and IMF staff estimates.

22. Interest rates remain high, reflecting the poor business environment and elevated credit risk (Figure 5). High interest margins/spreads point to inefficiencies in intermediation and impede economic growth by discouraging productive investments. Lending-deposit spreads have been increasing since mid-2011 and remain elevated: 10 percent in somoni and 13¾ percent in FX on average during 2014.

Figure 5.
Figure 5.

Tajikistan: Interest Rates and Spreads

Citation: IMF Staff Country Reports 2016, 041; 10.5089/9781498303019.002.A001

Sources: NBT; World Bank and IMF staff estimates.

B. Banking Sector Vulnerabilities

Overview and scenarios

23. Quantitative tests were conducted to assess banks’ resilience to credit and liquidity shocks. These included bottom-up macroeconomic stress tests conducted by the banks themselves, single factor sensitivity analysis, and top-down solvency stress tests on nine banks using supervisory data through end-September 2014.7 The results should be interpreted with care due to data limitations and quality concerns.8

24. The banking system’s resilience was assessed against three scenarios: baseline, adverse moderate, and adverse severe, which correspond broadly to the main risks identified in the Risk Assessment Matrix (RAM, Table 7).9 The baseline scenario is based on IMF staff projections as of January 2015 (Appendix I).10 Shocks were calibrated based on historical data (especially the outcome of the 2009 crisis) and satellite models. For the credit risk model, the growth in NPLs is modeled as a function of real GDP growth, inflation, remittances, interest rates, and exchange rates. Five-year macroeconomic projections (Table 9 and Figure 6) were then quantified with real GDP as the leading dependent variable for each of these scenarios. Quarterly bank-specific data for nine banks were used in the regressions.

Table 9.

Tajikistan: Macro Projections for Stress Test Scenarios

(In percent)

article image
Source: World Economic Outlook, IMF; and IMF staff estimates.1/ These scenarios are based on projections made in January 2015. Therefore these numbers (in particular the projections for 2014) need to be interpreted in that context.
Figure 6.
Figure 6.

Tajikistan: Macro Projections for Stress Test Scenarios

Citation: IMF Staff Country Reports 2016, 041; 10.5089/9781498303019.002.A001

Source: IMF, World Economic Outlook; and IMF staff estimates.

Solvency risk

25. Macro credit risk stress tests suggest that the banking sector is vulnerable even under the baseline scenario. Most banks are under-provisioned, as specific provisions only account for 42¾ percent of NPLs. When full provisioning was assumed (Figure 7),11 the system-wide CAR falls to 16¾ percent in 2015, with three institutions (including two large banks) representing more than a third of the system assets falling below the regulatory threshold.

Figure 7.
Figure 7.

Tajikistan: Macro Scenarios Solvency Stress Test Results

(Regulatory capital in percent of risk weighted assets)

Citation: IMF Staff Country Reports 2016, 041; 10.5089/9781498303019.002.A001

Source: IMF and National Bank of Tajikistan staff estimates (top-down stress tests).1/ The top down stress tests covers nine banks and AIB is excluded from the large institutions. Adjusted baseline takes into account full provisioning requirements.

26. Significant strain appears in the system under the moderate and severe alternate scenarios. In the adverse moderate scenarios, loan quality deteriorates with the NPL ratio for the nine banks rising to 23¼ percent in 2015. The deterioration intensifies under the adverse severe scenario: the system-wide NPL increases to 30 percent, bringing the share of banks with the CAR below the regulatory minimum to nearly two-thirds of total assets.12

27. Credit risk is the most critical vulnerability. The results of the sensitivity analysis (see Tables 10A and 10B) are consistent with those of the macro stress test. Increases in NPLs have a significant impact on CAR, which turns negative. The system-wide CAR falls by 14 percentage points to 3 percent while the potential recapitalization amount in percent of GDP increases to 2½ percent.

Table 10A.

Tajikistan: Sensitivity Stress Test Results on Credit Risks

(In percent unless indicated otherwise)

article image
Sources: National Bank of Tajikistan: and IMF staff calculations.1/ Additional NPLs are provisioned at 33 percent regardless of their category.2/ NPLs are downgraded by one category.
Table 10B.

Tajikistan: Sensitivity Stress Test Results on Credit Risks with Regulatory Provisioning

(In percent unless indicated otherwise)

article image
Sources: National Bank of Tajikistan: and IMF staff calculations.1/ Additional NPLs are provisioned at 33 percent regardless of their category.2/ NPLs are downgraded by one category.

28. The largest banks are more vulnerable to concentration risk. The tests show that in the event of a default by the largest borrower and full provisioning rates, the CAR of the largest banks will fall from 11¼ percent to 9 percent. For the smaller banks, the CAR will decline by 3 percentage points (to 36 percent). If the top five borrowers default, the largest banks’ CAR will drop to 4 percent, while the CAR for smaller banks will decline to 34 percent.13

29. Interest rate risk appears to be contained, but banks face considerable exchange rate risk. Interest rate risk is not significant for banks due to only small maturity mismatches. Even with an interest rate shift of up to 800 basis points, the impact on CAR is marginal, falling by only 2 percentage points, remaining above the regulatory threshold (Tables 11A and 11B). Banks are less resilient to a severe currency shock: a (direct impact of) depreciation of the somoni by 50 percent lowers the banking system’s CAR to 19 percent. Five banks, including three large banks, fail to meet the minimum capital requirement: the number of banks in breach of the regulatory minimum increases to 6 and the system-wide CAR falls to 13 percent.

Table 11A.

Tajikistan: Sensitivity Stress Test Results on Market Risks

(In percent unless indicated otherwise)

article image
Sources: National Bank of Tajikistan and IMF staff calculations.