Republic of Tajikistan
Financial System Stability Assessment, including Reports on Observance of Standards and Codes on the following topics, Banking Supervision, and Monetary and Financial Policy Transparency

The Republic of Tajikistan’s Financial System Stability Assessment and reports on the Observance of Standards and Codes are examined. The Tajik financial system is small despite recent rapid growth. Overall, banks remain well capitalized and liquid, but the brisk expansion of their loan portfolios is rapidly eroding capital buffers in a context where governance and supervision of the financial system raise concerns. The governance, regulatory, and supervisory frameworks for the financial sector should be strengthened. The key areas for the banking sector include licensing, remedial actions, and disclosure of significant shareholders and beneficial ownership.


The Republic of Tajikistan’s Financial System Stability Assessment and reports on the Observance of Standards and Codes are examined. The Tajik financial system is small despite recent rapid growth. Overall, banks remain well capitalized and liquid, but the brisk expansion of their loan portfolios is rapidly eroding capital buffers in a context where governance and supervision of the financial system raise concerns. The governance, regulatory, and supervisory frameworks for the financial sector should be strengthened. The key areas for the banking sector include licensing, remedial actions, and disclosure of significant shareholders and beneficial ownership.

I. Risks to the Financial System

A. Macroeconomic Background and Risks

1. Tajikistan is a landlocked, low-income country with a population of about 7 million and a per capita GDP of less than $600 (Table 1). The poverty rate, while declining, is high—estimated by the World Bank at 57 percent in 2004. The country suffered greatly from the breakup of the Soviet Union and the civil war that followed in 1992. Peace was restored in 1997, followed by macroeconomic and political stabilization. Macroeconomic performance has since been fairly robust, albeit with a few lapses.

Table 1.

Tajikistan: Selected Macroeconomic Indicators, 2005–08

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

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

2006 overall balance excludes the MDRI debt relief that is reflected in grants. PIP expenditure includes investment financed by loans from China.

Actuals are based on official exchange rates.

Starting from 2005, the export and import figures reflect the transition to tolling arrangement for aluminum exports. Therefore, the export and import figures are lower than earlier years.

Gross reserves are net of the pledged deposits of the NBT.

2. Real GDP has expanded at an average annual rate of about 8½ percent during the last five years. In this process, the sectoral composition of GDP has gradually shifted toward services and away from industry and agriculture, although the latter still accounted for about 20 percent of GDP in 2007 and almost 70 percent of employment, primarily in the cotton sector.

3. Inflation, after falling to single digits in 2004–05, has accelerated in the past two years, reaching 20 percent at end-2007. This was mostly due to a surge in international wheat prices and energy, but also reflected rapid monetary expansion from quasi-fiscal operations by the central bank associated with cotton sector financing. Fiscal performance on the other hand has been solid, the government having run small surpluses (excluding the externally financed public investment program) for the last five years.

4. The current account deficit averaged about 3 percent of GDP during 2002–06, but tripled in 2007. In the last few years, worker remittances—about 38 percent of GDP in 2007—have become the main source of foreign exchange overtaking receipts from cotton and aluminum exports. Most remittances originate in Russia, thus negative downturns in the Russian economy could substantially affect these inflows.3 Reflecting the impact of pledges and guarantees for cotton financing, the NBT’s international reserves fell to very low levels at end-2007 (about ½ month of imports or less than one fifth of dollar-denominated deposits). The national currency is de-facto pegged to the U.S. dollar.

5. The public and publicly guaranteed external debt stood at a modest 33 percent of GDP in 2007 with a debt-service-to-goods-export ratio of less than 13 percent, but both ratios are set to increase starting this year.4 The debt ratio, which exceeded GDP in the late 1990s, has been brought down to the current level by a combination of debt restructuring—including through the Multilateral Debt Relief Initiative—and rapid economic growth.

6. The authorities have requested a SMP for the second half of 2008 to address emerging macroeconomic risks. Economic growth could slow down sharply—first quarter annual GDP growth is estimated at 3.2 percent compared to 5.3 percent in Q1 2007. Relatively high inflation, which was running at around 20 percent in March 2008, could become entrenched with the risk that last year’s rapid growth in monetary aggregates and rising food prices could push it even higher. Finally, the low level of international reserves limits the options for dealing with adverse shocks to the balance of payments under Tajikistan’s conventional peg.

B. Risk Exposures of Tajik Banks and the NBT’s Balance Sheet

7. The rapid growth of Tajik banks has brought new risk exposures to the system. Albeit from a low base, bank deposits have grown very rapidly since 2004—at accelerating annual rates that topped 110 percent in 2007.5 Net bank loans have also grown very rapidly, broadly matching the growth of deposits in 2007 (Figure 1).

Figure 1.
Figure 1.

Key Characteristics of the Tajik Financial System

Citation: IMF Staff Country Reports 2008, 371; 10.5089/9781451837155.002.A001

Sources: International Financial Statistics; National Bank of Tajikistan; and Fund staff estimates.1/ Lending rates minus deposit rates, national currency, excluding demand deposits.

8. In the absence of strong credit risk management capacities, the rapid growth in bank lending is a reason for concern. While this credit growth supports much needed financial deepening, an excessively rapid increase could lead to a build-up of unrecognized credit risk. In particular, the surge in credit to the agriculture sector in 2007 should be carefully watched given its poor loan repayment record (21 percent of the loans to the sector were nonperforming at end-2007). More broadly, this expansion is taking place in the context of serious weaknesses in credit risk management practices, and in the governance, regulatory, and supervisory frameworks (see ROSC for BCP).

9. Dollarization is rising, thereby increasing the vulnerability of banks. Loan dollarization increased from 57 percent at end-2004 to 68 percent at end-2007. The main driving factor appears to be the evolving currency composition of banks’ funding, which is increasingly skewed toward the dollar. It is a source of concern that there seems to be a lack of awareness by borrowers and banks of the exchange rate–induced credit risk—i.e., the risk of default by unhedged borrowers in the event of a sizable exchange rate depreciation—although the availability of foreign exchange from remittances could mitigate this risk.

10. The rapid accumulation of nonresident deposits without clear knowledge of their sources may expose the banking system to reputational and ML risks. Between 2004 and 2007, deposits grew by more than 600 percent with a third of the growth accounted for by nonresidents’ foreign currency demand deposits; during 2007 those deposits displayed a very erratic pattern likely driven by changes to reserve requirement regulations. The bulk of those deposits reportedly accrue to a few accounts at the largest bank in the system. The authorities should investigate the sources of this funding to ensure that no illegal activity is involved.6 At this point in time these funds do not appear to pose a liquidity risk as they are largely deposited in correspondent accounts abroad.

11. In the last few years, the NBT has been an active, indirect, and nontransparent participant in the financing of the cotton sector with dire consequences for its balance sheet and the discharge of its mandates. The NBT provided guarantees and pledges to foreign lenders and somoni-denominated loans to KI, an asset management company.7 These operations have put the NBT’s balance sheet in serious jeopardy. There is a risk that the pledges and guarantees (equivalent to about 85 percent of the NBT’s gross foreign assets at end-2007) could be called in the coming months. Most of the direct loans (accounting for the bulk of the NBT’s rapid NDA expansion since 2004) are nonperforming. As supervisor, the NBT allowed the continued operation and growth of KI, even though there were serious doubts about its solvency (see ¶34); reportedly the license of KI was suspended last year.

II. Strengths and Vulnerabilities: Institutions

12. The Tajik financial system is relatively small, and dominated by banks and KI. The monetization of the economy—broad money to GDP—remains relatively low at about 21 percent of GDP in 2007, reflecting the lingering impact of the civil war and hyperinflation. Financial sector assets, in turn, have grown very rapidly as percent of GDP, more than doubling in the last three years to about 44 percent (Table 2). Banks and KI held 63 and 34 percent respectively of the financial system’s assets in 2007; other intermediaries had a negligible participation. Bank holdings are heavily concentrated and securities markets are virtually nonexistent.8

Table 2.

Financial System Structure, 2004–07

(In percent, unless otherwise indicated)

article image
Sources: National authorities; and IMF/World Bank staff estimates.

1/ Excludes the Tajik Development Bank.

Figures prior to 2007 exclude assets other than loans

13. Foreign borrowing for cotton financing is a significant share of financing to the economy. In the past, the cotton financing system was based on loans extended by various international cotton trading concerns to a network of local intermediaries known as “investors,” who in turn provided farms with equipment and seed in exchange for produce. This funding was channeled through AIB until its resolution in 2003. Elements of this financing system continued to be used until recently, with KI playing the role of AIB, but with the funding provided by foreign banks (rather than international trading concerns) backed by the NBT’s pledges and guarantees.

A. Banks

14. Despite the rapid increase in intermediation, the banking system is not yet in a position to serve the needs of the economy. Only about 45 percent of bank assets were loans (net of NPLs) at end-2007, and while deposit growth has been rapid, access to banks by the less-serviced population remains constrained.9 In this vein, expanding access to financial services outside the main urban areas should be encouraged by, inter alia, replacing current licensing requirements for the opening of bank branches and agencies with less taxing NBT guidelines. Further, while the handling of workers’ remittances is a profitable source of fee income, deficiencies in other services and products prevent banks from turning these funds into a stable source of financing.

15. Loans to individuals and small- and medium-sized enterprises (SMEs), mainly for trading activities, have grown most rapidly, but commercial and industrial loans still account for more than half the outstanding volume (Table 3). Loans to individuals and SMEs had a five-fold rise between end-2005 and end-2007 outpacing the growth of other sectors. Average loan maturity is about six months (Figure 2), and average loan sizes in the portfolio of the top three banks range from $10,000 to $60,000 (25 to 150 times GDP per capita).

Table 3.

Bank Credit by Sector, 2004–07

article image
Sources: National authorities; and IMF staff estimates.

Includes loans to individual entrepreneurs for commercial purposes.

16. Interest rate spreads in Tajikistan are the highest in the region and could contribute to adverse selection of borrowers (see Figure 1). Average ex-ante bank spreads have stayed above 20 percent both in somoni and foreign currency in the last few years—compared to an overall median worldwide of about 6 percent in 2003.10 The high spreads are driven by three main factors: (i) limited credit risk assessment and mitigation mechanisms, (ii) weak competition, and (iii) deficiencies in the tools and infrastructure for liquidity management. These spreads make loans unaffordable for large sections of the economy because bank lending is directed to more risky, but higher yielding, projects.

Figure 2.
Figure 2.

Loans by Maturity, 2007

Citation: IMF Staff Country Reports 2008, 371; 10.5089/9781451837155.002.A001

17. While banks’ financing to agriculture is increasing, anecdotal evidence suggests that access by potential rural clients remains limited.11 The record of bank lending to agricultural SMEs has been particularly disappointing. This reflects products and procedures that are not adapted to the needs of farmers, serious problems in the use of agricultural land as collateral,12 and maturity mismatches—shorter loan maturities than the length of most production cycles in agriculture.

18. Banks appear profitable, adequately capitalized, and highly liquid but in the context of a weak supervisory framework. Against a mandated minimum capital adequacy ratio (CAR) of 12 percent, the average for the industry was 21.3 percent at end-2007 (Table 4). 13 Notwithstanding additions to their capital base, reported ratios have fallen sharply (some 17 percentage points since 2005) reflecting the very rapid growth of the loan portfolio; they still appear adequate but just barely (see stress tests below). With intermediation margins at very high levels, returns on assets of Tajik banks are very high in international comparison (see Figure 1). However, competitive pressures in the coming years are likely to increase as a result of foreign bank entry, already seen with a bank from Kazakhstan starting operations earlier this year.

Table 4.

Financial Soundness Indicators, 2004–07

(In percent, unless otherwise indicated)

article image
Sources: National authorities; and IMF staff estimates.

Nonperforming loans include substandard (1–30 days), doubtful (30–60 days), dangerous (60–180 days), and hopeless (180 days +).

Maximum single borrower limit is defined as 25 percent of capital.

19. Asset classification practices and provisioning rules may give an upward bias to reported profits and capital. The reported NPLs ratio has been on a rapid downward trend in the last few years falling from 21 percent in 2004 to just below 5 percent in 2007. This reflects the expansion of the loan portfolio, but also a surprising fall in the stock of NPLs in 2007. Loan classification and provisioning rules were recently tightened,14 but loopholes remain, which raises questions about the quality of assets and bank capital. Loans can be restructured once with no effect on classification.15 Reclassification of loans into less-performing categories has occurred frequently as a result of onsite inspections. Provisions are relatively low (at about 40 percent of NPLs), reflecting the high share of substandard loans in NPLs, provisioning requirements calculated net of collateral, and weaknesses in collateral valuation.

20. The FSAP team conducted stress tests to evaluate the potential vulnerabilities of the banking system. The tests assessed the impact of a deterioration in the quality of credit portfolio, fluctuations in exchange and interest rates, and liquidity withdrawals. The reliability of the tests depends on the underlying quality of banks’ financial statements and there are reasons for concerns on this regard.

21. The stress tests suggest that the banking system is broadly resilient to a range of extreme but plausible shocks, owing to banks’ still adequate capitalization and high profits. The tests revealed limited, but rising, strain for most banks under most scenarios. Tajik banks are most vulnerable to credit risk, including indirect credit risk related to exchange rate fluctuations. In particular, the largest two banks appear vulnerable to a significant simultaneous deterioration in the financial health of their major borrowers.

22. The risk profile of Tajik banks has deteriorated considerably in the last year, largely owing to the reduction in capital relative to assets. A review of the stress test results based on December 2006, June 2007, and December 2007 data highlights the overall increase in risk-sensitivity of the banks. Declining CARs, while more in line with the capital buffers expected from a dynamic and competitive banking system, makes banks’ risk profile more pronounced. Details on the methodology and results of the stress tests are in the appendix.

B. Nonbank Financial Institutions

23. Nonbank financial institutions (NBFIs) comprise microfinance organizations (MFOs), insurance firms, leasing companies, and KI. There were 67 MFOs, 14 insurance firms, and 5 leasing companies as of June 2007. KI is covered in the next section.

Microfinance organizations

24. The MFO sector is growing fast and already provides services to a sizeable number of borrowers. Lending by MFOs grew by more than 200 percent in 2006 and a further 120 percent in 2007. This growth has been possible mainly due to access to foreign microfinance credit lines. However the sector is still very small, with a $50 million aggregate loan portfolio at end-2007, roughly equivalent to 10 percent of the commercial banks’ portfolio. MFOs’ client base is significant with more than 50,000 borrowers in 2007.16

25. There are three types of MFOs licensed under the 2004 law on microfinance—microcredit funds (38 institutions), microcredit commercial organizations (23) and microcredit deposit-taking institutions (6). Each type is subject to different regulatory requirements (e.g., minimum capital and types of services they are allowed to offer) and supervision processes, and can access a broad range of financing sources, including deposits and commercial borrowing.

26. MFOs, which hitherto have focused on financing commercial activities, are increasingly targeting rural clients. After commercial banks, MFOs are the second largest source of financing for trade, providing 12 percent of all financing to these activities. Given commercial banks’ competition in urban areas, MFOs are looking to find unmet niches in rural areas. Rural credit from MFOs has significant constraints—loans are small, short term, and subject to rigid group lending procedures—although they are apparently adequate for the financing of small-scale agricultural activities, like agro processing.


27. Despite rapid growth in recent years, the insurance sector is underdeveloped, highly concentrated, and the majority of companies are undercapitalized. Total gross premiums were equivalent to 0.3 percent of GDP in 2006, one of the lowest in the region (Table 5). Fourteen insurance companies were operating in Tajikistan in 2006, including two state-owned that have a monopoly on compulsory classes of insurance and all government business. Two companies dominate the market, one private and one state-owned, collecting 60 and 25 percent, respectively, of the premiums in 2006. To level the playing field and foster competition, the monopoly on compulsory insurance should be abolished.

Table 5.

Insurance Penetration and Density in Selected Markets, 2006

article image
Sources: Sigma and SISS.

28. Several factors hold back the development of the sector, including limited managerial experience. The low insurance penetration reflects, inter alia, widespread poverty, a very skewed income distribution (e.g., growth of life insurance is dependent on the size of the middle class), and an out-of-date legal framework. Further, there is little understanding of the role reinsurance can play in reducing risk and managing loss exposures. Although the Tajik insurance sector has very limited capacity, some of the local companies retain all the business that they write, while others are only reinsuring single large risks on an optional basis.


29. Leasing is an emerging source of finance, but the lack of long-term sources of funding is a major obstacle to its development. The volume of leasing remains negligible (0.1 percent of GDP), with only 5 leasing companies in operation in mid-2007. According to operators in the sector, reluctance by financial institutions to invest in mid- and long-term projects hinders the growth of the sector. Leasing could be especially important for Tajik businesses without an adequate credit history, because it enables the financing of equipment based on collateral.

30. The legal framework is satisfactory but there is room for improvement. Leasing is governed by the civil code—with specific norms that supersede general norms for rent relationships—and the 2003 law on leasing, and regulated by the NBT. The legal framework could be further strengthened by including in legislation provisions that allow for the releasing of equipment.

C. Cotton Debt

31. The resolution of the large outstanding stock of distressed cotton debt features high on the authorities’ agenda. Over the years, with some variations in the scheme (see ¶s 11 and 13), the financial sector has provided significant financing to the cotton sector. The loan repayment record has been abysmal, driven largely by falling world cotton prices since the mid-1990s, weather-related bad harvests in 1999 and 2001, and weak record-keeping on both loans and collateral. The result has been a steady build-up of bad debts.17

32. KI has been a key player in the financing of the cotton sector in the last few years, straying away from its intended role. Upon inception KI had no capital but received a private capital injection of $300,000 soon thereafter.18 By end-2004, impaired assets with a book value of somoni 650 million (about $210 million equivalent) had been transferred to KI, along with the corresponding liabilities. Thereafter, rather than shrinking as could be expected from a gradual disposition of assets and paying of debts, KI’s balance sheet started to grow rapidly to about $550 million by end-2007. This growth was spurred by the government using the NBT’s balance sheet to support the financing of KI (see ¶11). KI’s record in asset management has been lackluster, having reportedly recovered only about 10 percent of assets.

33. Although KI is likely deeply insolvent, it does not appear to pose a threat to the stability of the rest of the Tajik financial system other than the NBT. Based on the information available to staff, almost all of KI’s assets at end-2007 were loans, of which 70 percent were restructured. Provisions for bad loans were nil, and reported capital was a meager $1.5 million. In all likelihood the loans are still booked at highly unrealistic values, and thus KI is likely technically insolvent. The NBT and foreign creditors hold 99 percent of KI’s liabilities.

34. With assistance from donors, the authorities have developed a “road map” that aims to improve the profitability of the cotton sector and resolve the cotton debt.19 The road map foresees the development of a resolution strategy by the end of 2009.

35. An effective cotton debt resolution strategy must be transparent and encompass all parties involved. These include the NBT and the government, foreign lenders, KI, local investors, and farmers. This would allow for a transparent recognition of all disputed claims and fair negotiation of how the debt burden is to be allocated. The strategy should consider the impact of various debt resolution options on the financing of the sector and include implementation and enforcement mechanisms to prevent the emergence of a similar debt situation in the future. Finally, it should also address the situation of KI, with a first step being the immediate halt of its activities other than the management of its distressed assets. Moreover, a comprehensive due diligence audit of KI should be launched and performed by an internationally recognized audit company.20 If necessary KI should be resolved, but in the context of a comprehensive plan for the cotton sector.

36. Going forward, access to finance for current cotton farmers will require a comprehensive reform of the cotton sector—as envisaged in the authorities’ road map. This would require a halt to practices by the central and local authorities that restrict farmer’s freedom to select crops, and the de-linkage between farmers and cotton investors (in terms of input supplies and processing and sale of cotton).21 It would also require strengthening the legal framework on land use and secured lending (see below).

37. The cotton sector financing scheme put in place by the government for 2008, while greatly improving transparency, could undermine the liquidity and solvency of banks. The government budget has allocated somoni 130 million for the financing of the cotton sector in 2008 to be channeled to cotton farmers through banks. These funds are provided at subsidized interest rates—apparently bank’s spreads are significantly lower than current market averages. The allocation is equivalent to about 30 percent of the banking system’s capital at end-2007. In case banks fail to repay the funds (for example, because farmers fail to pay back loans), the government is contractually authorized to seize the funds banks hold in their correspondent accounts at the NBT. Given the very poor track record of loan repayment by cotton farmers, the scheme puts banks’ liquidity—with potentially serious repercussions for the operation of the payment system—and solvency at significant risk. The government should introduce, at the earliest opportunity, changes to the scheme that preserve banks’ incentive for thorough due diligence in the management of the funds while shifting the credit risk back to the government’s budget.

III. Strengths and Vulnerabilities: Financial Policy Framework

A. The National Bank of Tajikistan’s Issues

38. The NBT is facing significant challenges that need to be decisively addressed. Its significant involvement in commercial lending activities, the acceleration of inflation, and the very rapid credit growth and erosion of bank capital adequacy have added urgency to iis task. Failure to address these challenges will continue to impinge on the NBT effective discharge of its monetary authority and financial sector supervisory responsibilities.

39. The NBT autonomy and governance require considerable strengthening. The NBT should be capitalized to ensure its financial independence. Its chairman and his deputies should be appointed for fixed terms—that do not coincide with the electoral cycle—and the reasons for their dismissal made explicit in the law.22 The Board should be fully staffed—the three nonexecutive members provided for in the law have never been appointed.23 The parameters governing the Board’s remuneration policy should be publicly disclosed. Clear provisions covering conflicts of interest should be introduced in the law, and require the NBT staff (and all government officials) to disclose personal holdings in financial institutions and recluse themselves from official decisions affecting those institutions.

40. The NBT is deeply insolvent and will need to be recapitalized to enhance its credibility as a monetary authority. The most recent audit puts the NBT’s negative equity at $18 million, but this estimate is likely to grow several times over as a result of the recently disclosed exposure to KI. The NBT’s financial position drastically limits its ability to effectively use monetary policy instruments to mop up excess liquidity, with its credibility hinging on the government’s willingness to underwrite its current and future losses.24

41. NBT’s recent decision to stop its commercial lending activities is welcomed and they should not be resumed in the future. This should be a precondition for its recapitalization. The NBT has grappled with conflicting policy objectives. While by law its main objective is to maintain the purchasing power of the national currency, the NBT has found itself under pressure to support cotton sector development (Box 2). The NBT funds have been allocated in a nontransparent manner, with politically well-connected institutions (including KI) receiving the bulk of these funds.

B. Regulatory and Supervisory Framework


42. Much work has been done in recent years to enhance banking supervision and regulation. The two key laws governing the financial sector—the NBT law and the law on banks and banking—provide the NBT with adequate authority in most areas. Considerable efforts have also been made to strengthen the NBT’s supervisory capacity, particularly in the area of on-site examination, the introduction of a robust off-site reporting regime, and its capacity to analyze the data that it receives.

43. However, important shortcomings remain. Laws and regulations require strengthening in several key areas: licensing, remedial actions, AML/CFT, and the NBT autonomy and governance. Of at least equal concern is the ability of the NBT bank supervisors to effectively address compliance, safety, or soundness problems given the perceived or real position of power of many bank owners. Until the existing environment changes, supervisory effectiveness will likely remain restricted. The degree of compliance with the BCP is low.

44. The licensing regime and the remedial action approach need enhancements:

  • On licensing: the law does not provide for the fit and proper assessment of shareholders (including beneficial owners), board members, and senior management, nor for the review of the proposed corporate governance vis-à-vis best practice. These weaknesses were evident in a recent decision. A license was granted notwithstanding that the business plan was considered weak, there was no information on the shareholders, and the source of capital had not been ascertained.

  • On remedial action: laws and process require substantial improvement. This is the case even though the NBT has annulled 20 bank licenses during the period 1996–2007. The NBT is not legally required to take action should it consider that the safety and soundness of an institution is compromised. There is no prompt corrective action framework in law, regulation, or practice. A predictable and transparent remedial action regime will strengthen the supervisory framework.

NBT Lending to the Private Sector

The NBT retains on its balance sheet commercial loans extended to the private sector. These include loans extended to state-owned enterprises by the NBT on behalf of the government in the 1990s, an old stock of loss loans to the cotton sector that were guaranteed by the NBT, and more recent loans extended by the NBT to the private sector during 2006–07. These loans are reflected in the NBT balance sheet as follows:

  • A Somoni 154 million government bond issued in 2001 to cover for a portfolio of loss loans to state-owned enterprises. In 2006, the NBT provisioned Somoni 83 million against that amount to reflect the fact that the bonds carry below-market interest rates.

  • A Somoni 24 million portfolio of loss loans to agriculture extended by AIB and guaranteed by the NBT at the time of AIB’s restructuring at end-2003. The portfolio was assumed by the NBT at end-2006 upon expiration of the guarantee.

  • The equivalent of $25 million in loans to commercial banks for agricultural development financed through concessional credit lines from the China Development Bank (CDB). The funds received so far were priced at 2.5 to 5 percent by the CDB and on-lent to selected commercial banks by the NBT at subsidized interest rates (4, 6, or 14 percent). Banks are free to lend the money on commercial terms.

  • A Somoni 14 million loan to KI for food imports, funded by the NBT’s own funds. The loan was used to finance imports of potatoes to alleviate shortages and the attendant increase in prices. The loan had a term of three months, subsequently rolled over for another six months, and an interest rate of 10 percent.

  • Guarantees and pledges to foreign creditors of KI for $321 million and a Somoni 874 million loan. These operations were intended to help mobilize and provide direct funding for the financing of the cotton sector.

45. Recent legal amendments have further weakened the law in those areas and are contrary to best international practice. The government is now empowered to designate the entity that licenses and de-licenses banks. The NBT is no longer entitled to curtail a bank’s activities as part of its corrective action powers. Finally, the list of allowable bank activities no longer includes taking deposits, undoubtedly a parliamentary mistake. Notwithstanding the government’s proclamation giving the licensing and de-licensing authority to the NBT, the changes to the law have created the perception that the NBT’s independence has been undermined. It is strongly recommended that the authorities make the necessary amendments to the legal framework for bank licensing to align it with international good practice.

46. There are a number of other areas that require the attention of supervisors. These include corporate governance, ethical standards, and internal controls and audit; insider lending; the application of International Financial Reporting Standards (IFRS) accounting for all banks; legal protection for the NBT and its staff for supervisory actions taken in good faith; and consolidated supervision. The NBT is well aware of the need for improvements in these areas, and has already taken steps to draft regulatory and legal amendments, for instance on consolidated supervision.

47. The NBT should consider strengthening prudential regulation to induce banks to better manage rising risks. Among possible measures: (i) it could apply higher capital charges to foreign currency-denominated loans to unhedged borrowers to deal with rising dollarization, (ii) it should strictly enforce prudential rules (e.g., on maximum exposure limits), and increase “haircuts” in the value of loan collateral to deal with rapid credit growth and falling capital adequacy. The NBT should also consider targeted inspections of those banks that have reported a substantial reduction in the stock of NPLs to verify that it is warranted.

48. A review of the NBT’s transparency practices in banking supervision found them to be broadly satisfactory. But a few areas need further enhancement, including the timeliness and extent of dissemination of information on financial policies and the health of the financial system, and standards of conduct for the NBT staff.


49. An adequate legal, regulatory, and supervisory framework is lacking, which undermines the sustainable growth of the insurance sector. The 1994 insurance law is far from international standards in a number of areas, including the protection of policyholders—from abuses and insolvency risks of insurers—and supervisors. Ongoing legislative efforts should continue, offering a timely opportunity to align the legal framework with good international practice in this area.

50. In particular, the market supervisor needs substantial strengthening to be effective. The State Insurance Supervisory Service (SISS), a department of the ministry of finance, oversees the market. The knowledge, technical skills, and experience of SISS technical staff (currently five) need considerable enhancement, inter alia, in the areas of licensing, prudential requirements, and on-going supervision. Staffing decisions are constrained by a salary scale that is well below that of bank supervisors. In considering options to strengthen the SISS, the authorities should consider alternative institutional structures, including a possible transfer to the NBT.

51. The absence of standardized accounting rules across companies hinders proper evaluation of the condition and performance of insurers. Accurate accounting provides authorities with the practical means to perform proper supervision, while at the same time being a resourceful instrument for the management of companies.

Deposit insurance

52. The Tajik deposit insurance system needs improvement, with current draft legislation addressing some of the shortcomings. The system established in 2003 has significant shortcomings: balances are equivalent to less than one percent of insured deposits, there are no contingency financing arrangements, participation is voluntary, there is no cap on insurance coverage (70 percent of eligible deposits are covered independent of their size), and only large shareholders and related parties are excluded from coverage.25 Draft legislation makes membership mandatory, and caps coverage to somoni 2,500 per depositor/per bank, while improving the governance of the system.

53. Additional improvements to the draft legislation should be considered. These include: (i) excluding bank shareholder and related parties from coverage; (ii) establishing a minimum time frame to reimburse deposits; (iii) excluding commercial bank representatives from the fund’s supervisory council to avoid conflict of interest; 26 (iv) restricting the fund’s investment policy to highly rated (domestic and foreign) government securities; and (v) authorizing emergency borrowing from the government but not from commercial banks in the event that the fund is insufficient to cover insured deposits of a failed bank.

C. Financial Infrastructure

Legal framework

54. The legal framework governing Tajikistan’s financial sector has undergone significant reform over the last decade. Although individually the laws and regulations are reasonably comprehensive and broadly consistent with international best practices, they are often conflicting, open to varying interpretations and not widely disseminated. Implementation and enforcement has generally been weak.

55. Additional reforms are still needed to foster the development of businesses and the financial system. They include:

  • The full implementation of the legal entities registration law. The law adopted the “single window” principle and provided for the simplification of procedures. Registration for small businesses has become easier (it takes less than a week for individual entrepreneurs to register), but it remains complex and time-consuming for larger ones.

  • An overhaul of the 2003 Bankruptcy Law, which is deficient in almost all key areas. Shortcomings include the lack of reorganization provisions and vague criteria for the commencement of insolvency proceedings. Cases of insolvency are uncommon and expertise in this area is almost nonexistent. A new model law on insolvency for the region is currently under development.

  • Substantial improvement in the administration of justice on contract enforcement and property rights matters. Court proceedings are slow, unpredictable, and at times biased, with opinions and decisions often secret and not revealed to the parties. In part, this reflects the poor working conditions of judges, including pay and infrastructure. Thus banks resort to the courts only when absolutely necessary. However efforts are underway to improve this situation, including by establishing a judicial council and developing a time-bound action plan for the overall reform of the judicial sector.

Collateral and creditor information systems

56. Problems with secured lending and the lack of a creditor information system are additional barriers to the efficient development of the financial system. Both problems increase the risk to lenders and affect the cost and availability of lending. The authorities are well aware of the problems and efforts are underway to address them.

57. The laws governing secured lending are often overlapping and in conflict with each other. For example, the land code prohibits the mortgaging of immovable property while the civil code and the mortgage law provide for this possibility.27 This lack of coherence and redundancy creates confusion as to the effectiveness of the collateral. The authorities are planning a reform of the legal framework governing the use of land.

58. The framework for registration and property rights enforcement is subject to a number of weaknesses:

  • There is no uniform system for title registration and related transactions (e.g., title transfer and mortgage registration), with multiple bodies involved depending on the type of property. Work is underway developing a new law on the registration of rights on real estate and creating a unified registry for movable property.

  • The information contained in the property registries is not publicly available, which, coupled with the lack of appropriate legal mechanisms, fails to protect good faith property purchasers during the title registration process.

  • Judicial foreclosure carries significant risks for mortgage lenders. They include uncertainty regarding procedures for the eviction of residents and repossession of foreclosed properties, and lengthy court proceedings.28 Out-of-court settlement is possible in the case of default, but it is based on the voluntary cooperation of the borrower to ensure that the sale of the property is performed as efficiently and quickly as possible.

59. The establishment of a creditor information system would support better credit risk assessment by banks. To this effect, the authorities are drafting legislation that would allow for the establishment of a private credit bureau.29 But this will take time, and therefore as a transitional arrangement the NBT should consider establishing a credit registry.

Corporate governance

60. The authorities have taken major steps to improve governance in the corporate and banking sectors but substantial weaknesses remain. Notably among the reforms were the enactments earlier this year of a new Law on Joint Stock Companies, the establishment of a central share registry for all open joint stock companies and the privatization of medium and large companies. In particular, the new law incorporates much improved provisions on related-party transactions, major asset sales, meetings of shareholders and election of members of boards of directors—with clear fiduciary duties for them and other senior corporate officials. Nevertheless, the standard of corporate governance remains weak.30

61. To further improve corporate governance, the government should:

  • Increase the transparency of ownership and control by publishing the names of significant owners of banks and other financial institutions (immediately for direct owners and over time for beneficial owners); placing on-line and making available to the public the names of company founders and company statutes filed with the business registry; and establishing a central registry of land use certificates with access made available through the internet; and

  • Subject by law all credit institutions to annual independent audits of their financial statements. This will extend the current requirement applicable to all open joint stock companies—including all private banks—to public banks (currently externally audited once every two years) and to KI (a closed joint stock company not subject to external audits).31 Tajikistan also needs to license sufficient professional audit firms capable of auditing financial institutions.

D. Systemic Liquidity Management

62. Systemic liquidity management and the monetary policy framework are rudimentary. The NBT’s main monetary policy instruments are unremunerated reserve requirements and outright sales of the NBT certificates of deposit; its monetary policy interest rate, the refinancing rate, is not tied to any actual transaction; and there are no standing facilities. The NBT’s auction volumes are small and the state-owned bank purchases almost every issue at negative real interest rates suggesting a lack of competitive market pricing. There is no secondary market for the NBT certificates, nor a market-determined interest rate or yield curve.

63. The limited stock of tradable government securities constrains monetary policy implementation and banks’ liquidity management. In the absence of domestic financing requirements, the ministry of finance has not issued securities since 2001. The NBT certificates were first issued in 2003 but the NBT’s insolvency discourages larger placements and market-based pricing. The resulting paucity of freely transferable collateral drastically limits the use of repos and holds back the development of an interbank money market. These constraints, the lack of standing facilities at the NBT, and payment system’s deficiencies, heighten banks’ incentives to hold ample idle liquidity.

64. The issuance of government bonds would be beneficial for several reasons. First, these securities could be used to mop up excess liquidity via open market operations, while the cost of implementing monetary policy would be born directly and transparently by the ministry of finance.32 Second, the bonds could serve as collateral in the banking system. Third, a reference yield curve would gradually be established, facilitating the pricing of credit risk in the private sector and helping to develop the securities markets.

65. A useful first step could be the securitization of the nonmarketable government bond on the NBT’s books. The securitized debt could then replace the NBT’s certificates as its main short-term discretionary market instrument. Gradually, bonds with longer maturities could be sold to the market, while auctioning procedures would need to be strengthened.33

66. In the meantime, the NBT should ensure that reserve requirements remain an effective instrument of monetary policy. The NBT changed reserve requirement regulations on several occasions during the last year and a half. It excluded nonresident deposits in January 2007 but reversed course later in the year as advised by the FSAP mission.34 The rate applied to nonresident deposits has at times also differed from that applied to resident deposits. Going forward, it will be important that the treatment of both types of deposits remain unified, compliance with the requirement is strictly enforced, and the NBT gradually develops other tools to implement monetary policy, including setting up a standing facility.

67. The NBT’s transparency practices in monetary policy have improved significantly but need further enhancement. The NBT’s law lacks key provisions, among others, for a monetary policy making body, for the central bank’s engagement in commercial lending, and the specific grounds for dismissal of the chairman and his deputies, term limits on their appointments, and qualifications that they should possess. Furthermore, there is no guidance in law as to what constitutes conflict of interest, how it is to be monitored, and what sanctions apply in case of violation. On other matters, the general framework regulating the NBT’s activities is broadly satisfactory, although recent legislative changes on the powers to license banks constitute a step backwards. The NBT strives to inform relevant parties and the general public about its activities in the monetary policy area, although providing fuller and more future oriented analysis would constitute an improvement. Similarly, formalizing a process seeking public comments on changes in policies would be helpful.

E. Payment Systems

68. Tajikistan is largely a cash-based economy. In 2002, the NBT launched a RTGS system which is handling increasing volumes of electronic payment orders. The use of other payment instruments, like debit cards, is still very limited largely because only a small portion of the country’s population holds a bank account.

69. The NBT-operated RTGS system is the only major interbank payments system in the country, and it is therefore considered systemically important. It has contributed to a better integration between the central bank and the banks, and to an overall improvement of payment and settlement arrangements in the country. The system only processes Somoni-denominated orders, with the value settled growing by about 30 percent a year since launching.

70. Tajik banks continue to settle a significant portion of domestic interbank payment transactions through bilateral correspondent accounts, thereby increasing systemic risk. Like banks in most other countries, Tajik banks use their correspondent accounts with foreign banks to settle foreign currency-denominated payments. But Tajik banks also frequently settle somoni-denominated payments using domestic bilateral accounts with one another, thus increasing systemic risk from settlement failures.35 Bilateral correspondent arrangements are not regulated from a payment system perspective, and the only control is the obligation for banks to report statistics to the NBT on their bilateral settlements with other banks.

71. There is a need to ensure business continuity for the RTGS system. A key measure in this regard is the need for a back-up secondary server(s) capable of resuming operations quickly without loss of transaction data, with the primary and secondary processing equipment located in different sites. In the medium term, a more comprehensive business continuity plan should be developed, addressing not only technical issues but also the availability of staff and information procedures.

72. The legal framework lacks many basic concepts that underpin the operation, regulation and oversight of a modern payments system. While the NBT law provides the central bank with clear, albeit somewhat general powers to operate and regulate payment systems, it lacks key concepts which are also not found in the regulations. Missing concepts include the timing of settlement finality, the system’s legal protection from insolvency effects, and the timely seizure of collateral pledges. At a minimum, these and other relevant concepts (e.g., access rules, electronic signatures) should be addressed through the NBT regulations or other equivalent forms of secondary legislation.

73. The NBT should play a leading role in promoting the development of electronic retail payment instruments and systems. Depending on the circumstances and the degree of cooperation among banks this role could range from acting as facilitator, catalyst for change, or even as a direct service provider.

74. In the medium term, the NBT also needs to develop and fully implement its payment system oversight function. At present, the NBT performs some basic payment system oversight functions such as the analysis of data and the publication of statistics for the RTGS and card payment systems. However, it has not established specific objectives in this field, or the requirements that the various payment systems must meet and ways to enforce such requirements. For the NBT to perform this function effectively, it will need to establish appropriate organizational arrangements and staffing.

75. Banks could help promote the deposit of remittances in the banking system and the NBT could foster competition. Remittances are paid almost exclusively in cash. However new products, such as payment cards, can effectively reduce the costs of cash handling and make it possible for remittance beneficiaries to have a deposit account. For this purpose, it is indispensable to develop awareness campaigns, highly concrete education mechanisms, and specific incentives to encourage remittance recipients to use these products. The NBT could foster competition among money transfer operators by increasing market transparency (e.g., publishing comparative data on the various pricing and service alternatives offered by banks and other paying agents).

Annex—Observance of Financial Sector Standards and Codes—Summary Assessments

This annex contains summary assessments of two international standards and codes relevant for the financial sector. The assessments have helped to identify the extent to which the supervisory and regulatory frameworks are adequate to address the potential risks in the financial system.

The following detailed assessments of financial sector standards were undertaken:

  • The BCP, by Ms. Laura Ard (World Bank, formerly U.S. Office of the Comptroller of the Currency), and Mr. Peter Phelan (IMF consultant, formerly Bank of England).

  • Transparency in Monetary and Financial Policies, by Mr. Mariusz Sumlinski (IMF staff).

The BCP assessment was carried out during the February 2007 FSAP mission to Tajikistan, and it was benchmarked against BCP I using both the “essential” and “additional” criteria. The CTMFP assessment was carried out during the August 2007 FSAP mission and was based on the IMF Code of Good Practices on Transparency in Monetary and Financial Policies (MFP Transparency Code). All assessments were based on the laws, regulations, policies, and practices in place at the time the assessments were made. In addition, they take note of changes in legislation that were introduced after the mission, in particular with regard to bank licensing. Those changes had an important bearing on the assessments of BCP, in particular CP1, which was downgraded.

The assessments were based on several sources including:

  • Review of relevant legislation, regulations, policies, and other documentations including information made available on the NBT’s website and in its Annual Reports and other publications, and responses to a pre-FSAP questionnaire response;

  • Self assessments by supervisory authorities and the NBT’s current advisors, USAID-Bearing Point;

  • Detailed interviews with NBT’s officials, including supervisors, and other government officials;

  • Meetings with financial sector firms.

A. Basel Core Principles for Effective Banking Supervision


76. Tajikistan’s banking supervision function is a part of the NBT responsibilities and therefore, the NBT was the primary counterpart for the review.

77. Although the 1999 Basel Core Principles for Effective Banking Supervision were used in the review and rating process, in the future the revised BCP will be the goalpost standard for banking supervision. Given the changes to the BCP, particularly the increased emphasis on sound governance and risk management, the NBT would be well served to address the issues cited herein while at the same time simultaneously measuring itself against the revised, more elevated standards.

78. The assessment team reviewed available bank legislation, regulation, and policies; however, a comprehensive list of relevant regulation was somewhat difficult to obtain as was the full description of the various forms and enforceability of regulation and the NBT issuances. Furthermore, the enforceability of regulations and issuances has not been concretely tested through court cases or formal appeals process, and therefore, has not been clearly observed. As a result, the ability of the assessors to evaluate all aspects of the legal framework was hindered somewhat.

Institutional and macroeconomic setting and market structure—overview

79. The central bank regulates and supervises most financial institutions. The NBT oversees all deposit-taking institutions as well as all the nondeposit-taking microcredit organizations. The banking system has undergone substantial consolidation in the last few years against the backdrop of tighter prudential regulations.

Preconditions for effective banking supervision
Macroeconomic policies

80. Relatively prudent macroeconomic policies have supported strong growth and a broadly stable macroeconomic environment in recent years. Real GDP has expanded at an average annual rate of about 8½ percent during the last five years, with the sectoral composition of GDP gradually shifting toward services. Inflation fell from double digits to 6–7 percent in 2004–05, but accelerated to about 20 percent by end-2007. Fiscal performance has been solid, with the government having run small surpluses (excluding the externally financed public investment program) for the last five years.

Public infrastructure

81. The legal framework governing Tajikistan’s financial sector is relatively young and untested. Six laws govern the sector (i) the Law on the NBT (LNBT) (1996, amended 2001–02); (ii) the Law on Banks and Banking (LBB) (1998); 36 (iii) the Law on Guarantee of the Deposits of Physical Persons (2003); (iv) the Law on Securities and Stock Exchange (1992); (v) the Law on Insurance (1994); and (vi) the Law on Microfinance Organizations (2004). There is no specific law for nonbank financial institutions; the latter are governed by the NBT Regulation No. 118 “On Credit Companies.” There is virtually no private ownership of real property, and land reform is a major component of needed future reform of the legal and public systems. International accounting standards were introduced in 1999.

82. The legal framework of the financial system has not been tested in the courts. Challenges to the NBT’s oversight and regulation are seldom, although the level of influence the financial sector has over prevailing regulation and NBT decisions is not clear. The local audit industry is shallow, and firms responsible for auditing banks are domiciled in neighboring countries.

Market discipline

83. The mechanisms for market discipline are nascent. Banks are required to disclose financial statements which are audited annually by firms licensed by the NBT. No interim results are disclosed. By law the NBT sets the accounting standards, and the central banks and the industry adopted IFRS in 2006. However, international auditing standards are not required.37 The incentive structure for hiring and removal of managers and Board members is not concretely established, and the tools for the exercise of market discipline (e.g., deposit mobility and general public education about bank soundness) are not evident.

Systemic protection/resolution

84. Bank resolution procedures remain ad hoc. In the last 10 years the NBT has withdrawn 21 bank licenses and initiated 1 major bank restructuring. The latter, in 2003–04, involved a “good bank–bad bank” split of the then largest bank in the system, Agroinvestbank, which had suffered serious problems with its cotton sector exposures. Impaired claims amounting to some ¾ of its total assets (7½ percent of GDP) were transferred to a new asset management company, KI; a sizable portion of the remaining claims were underwritten by the government/NBT; and the rump bank was privatized. The new asset-management entity, however, has made little subsequent progress in resolving this debt.

85. A limited-scope deposit insurance scheme is in place, but suffers a number of design flaws. With the fund’s balances equivalent to less than 1 percent of estimated insured deposits, the absence of contingency financing arrangements is a key weakness of the scheme. Other problems include the voluntary nature of the scheme, no cap on insurance coverage, and insufficient exclusion of bank shareholders and related parties from coverage. A number of revisions to this law are under consideration.

Main findings
Objectives, autonomy, powers, and resources (CP 1)

86. The powers of the NBT to license and supervise banks had been established in law, but these provisions have recently been repealed, and the only legal basis for the NBT supervision is a presidential regulation, which is contrary to best international practice and raises the important question of the level and effectiveness of enforceability. The other provisions of the LNBT and the LBB apparently remain in force, however, and these provide a reasonable basis for supervision, although the LNBT, while addressing the stability of the banking system, does not specifically deal with safety and soundness concerns. The NBT is officially independent, although there is some evidence of interference by the Tajik politico-economic establishment. The NBT lacks a capital base to ensure its financial independence and its chairman is appointed for an indefinite term and can only be dismissed for ‘abuse of power.’ Staff professionalism is growing, resources are adequate, and staff turnover is low. There is no protection for either the NBT or its staff from being sued, nor are costs reimbursed should a lawsuit arise. There are no MOUs with other supervisors.

Licensing and structure (CPs 2–5)

87. Recent legal changes have weakened the effectiveness of the law on the use of the term “bank” and made it less clear. Assuming that the NBT is still operating a licensing regime based de facto on the LNBT and the LBB, there are still legal shortcomings in the system, with no requirement to check on shareholders, directors and senior management, nor on the operational structure of the proposed bank. Practice is also seriously deficient with decisions made for reasons that are not clear and are not adequately minuted. Probity checks are not carried out on new shareholders, although holdings of more than 20 percent can be rejected for financial reasons, and there is no system of prior approval for investment holdings of banks, just a limit based on a percentage of capital.

Prudential regulations and requirements (CPs 6–15)

88. There are no capital charges for market risks and no consolidated capital requirements, although many of the other capital adequacy calculations are consistent with Basel requirements. The NBT has taken action against banks in breach of capital requirements, but there is no formalized planning to deal with capital problems before they become critical.

89. There is currently no regulation that deals with credit policies, although this area is covered in detail during on-site examinations. There is an adequate, past-due based, regulation for provisioning, and large exposure limits are defined in terms of international norms, although there is evidence of frequent breaches of the rules. Connected lending rules give no powers to the NBT to require decision-making at a senior level without the connected person being present. There is no country risk, nor market risk regime, although in both cases such risks are minimal at present. Some supervision of liquidity and foreign exchange risk takes place, albeit more quantitative through verification of calculations and minimal, if any, review of the nature and effectiveness of the attendant risk systems. Other risks are monitored by an ‘internal control unit’ set up in banks at the behest of the NBT. The approach is more compliance than risk-focused, however. More needs to be done to develop management information systems in banks, as well as an effective internal audit function. There is no law or regulations setting forth customer diligence requirements and reporting of suspicious transactions for the prevention of money laundering or terrorist financing activities, nor has a FIU been established. A draft law is pending before Parliament.

Methods of on-going supervision (CPs 16–20)

90. A comprehensive set of returns has been developed by the NBT with the help of international consultants, and these consultants have also assisted with training for on-site examiners. There is an annual on-site examination of each bank, with the prime focus on credit risk (the principal risk in Tajik banks), but some work on internal control matters as well as validation of off-site returns. There are regular contacts with senior management, ad hoc meetings with the chairman, but no regular meetings with the full board. There is no consolidated supervisory regime at present, although this will be needed soon as the financial system develops, and the NBT has been drafting a regulation to cover this.

Information requirements (CP 21)

91. Banks have to have their accounts audited in accordance with IFRS and published annually. The audit profession is not well developed in Tajikistan and many of the audits are conducted by staff based in adjacent countries. The effectiveness with which IFRS is applied is in doubt, given in part the cited issues surrounding timeliness and accuracy of asset classification by banks. International audit standards are not formally applied. Dialogue between the NBT and the audit profession is limited, although the NBT does license auditors (in the absence of any other body capable of doing the work).

Formal powers of supervisors (CP 22)

92. The recent changes to the laws mentioned above call in question the foundation on which the NBT can take remedial action against banks. Notwithstanding these changes, the NBT’s powers in this area needed substantial strengthening. Although fines, staff changes, and business restrictions can be imposed, a bank cannot be put into liquidation without a prior period of temporary administration. There is no system of a structured remedial action approach triggered by specific events or any requirement for action to be taken on a timely basis. Although the NBT has taken action against banks in recent years, it is not always on a consistent and proportionate basis.

Cross-border banking (CPs 23–25)

93. No Tajik banks are established outside the country, and only one foreign bank operates in Tajikistan, with a specialized role serving two embassies. Nevertheless, since the Tajik authorities would like to see more foreign banks in the country, the NBT should develop the necessary regime to deal with this, including the ability to assess the effectiveness of a home country’s supervision.

Table 6.

Summary Compliance with the Basel Core Principles

article image
article image
Recommended action plan and authorities’ response
Table 7.

Recommended Actions to Improve Compliance with the Basel Core Principles

article image
article image
article image
Authorities’ response to the assessment

The authorities were in broad agreement with the assessment.

B. Transparency in Monetary and Financial Policies

Summary, key findings, and recommendations

94. The NBT is the central bank of the country and is responsible for formulation and implementation of monetary policy in Tajikistan, as well as for banking supervision and regulation.

Institutional and market structure—overview
Monetary policies

95. Systemic liquidity management and the monetary policy framework in Tajikistan are rudimentary. The NBT’s main monetary instruments are unremunerated reserve requirements and sales of the NBT certificates of deposit; there are no standing facilities. Auction volumes are small, and there is no secondary market for the NBT certificates or market-determined interest rates or yield curve. The NBT is insolvent. An interbank money market is virtually nonexistent—some of the banks transact among each other on an ad hoc basis. Hence, the banks maintain excess reserves also as insurance as the lender-of-last-resort arrangements are not formalized. A small foreign exchange market serving bilateral exchanges is dominated by the NBT.

Financial policies

96. The NBT regulates and supervises most financial institutions. The NBT oversees all deposit-taking institutions. The banking system has undergone substantial consolidation in the last few years.

97. Banking supervision and regulation are governed by the NBT law and the law on banks and banking, which provide the NBT with authority in these areas. However, important shortcomings remain—areas of licensing, remedial actions, and NBT autonomy and governance require strengthening. Also, the effectiveness of the supervisory effort remains restricted as some bank owners rely on (the perceived or real) position of power to resist the recommendations of supervisors.

Main findings
Monetary policies

98. The NBT’s transparency practices in monetary policy have improved significantly but need further enhancement. The general framework regulating the NBT’s activities is broadly satisfactory and the NBT strives to inform relevant parties and the general public about its activities in the monetary policy area. However, there are a number of areas were action is required to improve transparency. Specifically, there are no provisions in the law for the monetary policy making body, as well as the NBT’s engagement in commercial lending and the rest of the economy. The grounds for dismissal of the chairman and his deputies, term limits on their appointments, and the qualifications that the appointees should possess are not specified in the law. Another notable omission is the lack of legislative guidance as to what constitutes conflict of interest, how it is monitored, and what sanction to apply in case of violation. Also, the recent amendments relating to licensing of banks constitute a step backwards. As regards the communication with the public at large, providing fuller and more future oriented analysis would constitute an improvement, and formalizing a process seeking public comments on changes in policies would be helpful.

Table 8.

Summary Observance of the Transparency Code (Monetary Policy)

article image
article image
Financial policies

99. The authorities have put much effort into improving the transparency of financial policies. A review of the NBT’s transparency practices in banking regulation and supervision found them to be broadly satisfactory with few exceptions. These include (i) the clarity of roles, responsibilities, and objectives of financial agencies; and (ii) standards of conduct for the NBT staff. The key findings show that independence and institutional authority of the NBT need increased support especially as the recent legislative changes seem to have undermined them. The integrity and accountability of the NBT needs strengthening to overcome the lack of effective legislation defining conflict of interest, its applicability, and legal responsibilities, and the protection of employees. Observance in two other areas—open process for formulating and reporting financial policies and public availability of information on financial policies—is fairly satisfactory.

Table 9.

Summary Observance of the Transparency Code (Financial Policies)

article image
article image
article image
Recommended action plan and authorities’ response
Monetary policies
Recommended action plan
Table 10.

Recommended Action Plan to Improve Observance of the Transparency Code (Monetary Policy)

article image
article image
Authorities’ response to the assessment

100. The authorities were in broad agreement with the assessment. They provided detailed comments on specific items, all of which were incorporated. The authorities also pointed out that a number of revisions to the central bank law and banking law have been drafted that would enhance observance of the transparency code.

Financial policies
Recommended action plan
Table 11.

Recommended Action Plan to Improve Observance of the Transparency Code (Financial Policies)

article image
Authorities’ response to the assessment

101. The authorities were in broad agreement with the assessment.

Appendix I. Stress Testing

102. The stress tests of the Tajik banking system were conducted using detailed balance sheet data for eight commercial banks (accounting for 95 percent of total banking sector assets). They were initially based on December 2006 data, and subsequently updated with June and December 2007 data.38 The purpose of the tests was to examine the potential effects of a set of specified changes in risk factors, corresponding to exceptional but plausible events, on banks’ financial conditions. Hypothetical scenarios based in part on discussions with the authorities and on the FSAP analysis were used to calibrate the shocks. The stress tests gauged the vulnerabilities of banks to a deterioration in the quality of their credit portfolios, fluctuations in the exchange and interest rates, and a liquidity shock in the form of bank runs.

103. The stress tests results suggest that overall the banking system is broadly resilient to a large range of extreme but plausible shocks, but capital buffers are being eroded rapidly (Table 12). Tajik banks remain most vulnerable to credit risk, including indirect credit risk related to exchange rate fluctuations. The risks are considerably amplified by uncertainties regarding banks’ asset quality, especially in the current context of rapid credit growth and weak credit assessment and monitoring tools.

Table 12.

Summary Stress Tests Results

article image
Source: National Bank of Tajikistan and IMF/World Bank staff estimates.

Number of banks with a CAR that falls below the 12 percent prudential requirement.

Worst/Best corresponds to the highest/lowest CAR of individual banks under each scenario.

Credit risk

104. Three separate credit risk shocks were performed: (i) a two-category downward shift of 20 percent of all loans (i.e., a “migration” shock with 20 percent of “normal” loans moving to the “doubtful” loan category, 20 percent of “sub-standard” loans moving to the “dangerous” loan category, etc.); (ii) a deterioration in the quality of the exposures of the largest borrower (re-classification to the “dangerous” loan category); and (iii) a deterioration in the quality of 50 percent of all large exposures (re-classification to the “dangerous” loan category). While the first shock focuses on a hypothetical worsening of the overall credit quality of banks’ lending portfolios, the last two shocks focus on the concentration of banks’ lending activities to a few large customers and the impact of their potential financial difficulties on banks’ capital.

105. The credit risk tests took account of a number of characteristics of the Tajik banking sector that could potentially increase bank’s sensitivity to shocks substantially. In particular, weak accounting and auditing standards, relatively lax implementation of asset classification and provisioning guidelines, limited regulation on connected lending, and lack of reliable data on borrowers’ creditworthiness and on the value of their collateral raise the potential for a build-up of unrecognized credit risk in banks’ balance sheets. To account for these uncertainties, the tests were conducted under two assumptions: a 50 percent collateral coverage assumption (in line with the Tajik collateral coverage requirement for real estate), and a zero collateral coverage assumption (to account for the possibility that the collateral is unrealizable or unmarketable).39 The results discussed below focus on the more conservative assumption.

106. Based on data at end-2007, the results indicate that banks’ still have an adequate (but just barely) level of capitalization. In response to a 2-category downward loan “migration”, and allowing for no collateral coverage, two banks breach the 12 percent minimum CAR requirement (but by only 2 percentage points in the worst case) and the system-wide CAR falls from about 19½ percent to a still comfortable 15 percent. However, these two banks are the largest in the system holding close to 70 percent of the assets.

107. Some major banks are vulnerable to a deterioration in the credit quality of their largest exposure. When the exposures of each bank’s largest borrower is reclassified to the “dangerous” loan category (61 to 180 days past due), three banks would see their CAR fall below minimum requirement. In the worst case, the shortfall is significant (6.5 percentage points) but that bank holds only a very small fraction of the system’s assets; in the other two cases the shortfall ranges between 1 and 2 percentage points. Nevertheless, there is a clear deterioration of banking sector soundness (70 percent of total banking sector assets would be covered by a CAR below 12 percent), even though system-wide CAR is still above 14½ percent.

108. When all large exposures of the banks are downgraded simultaneously, there is a large impact on bank and system-wide soundness. When 50 percent of all large exposures are reclassified to the “dangerous” loan category, the percent of the banking sector assets covered by a CAR below 12 percent also rises to 70 percent, as under the previous case. A major difference however, is the very severe deterioration experienced by the largest bank in the system (its after-shock CAR is only 5.5 percent). Furthermore, the after-shock system-wide CAR falls below minimum requirement (by about one percentage point). The current degree of interconnectedness of large exposures is difficult to assess. Therefore, the potentially sizable impact of a simultaneous deterioration of large exposures on the banking system calls for a close monitoring of the nature of banks’ large exposures, especially in the absence of a clear regulation on connected lending.

Foreign exchange risk

109. Overall, most banks were found to be sufficiently well capitalized to withstand the impact of a substantial appreciation of the currency. On average, banks have a long net open position in foreign exchange (i.e., foreign currency denominated assets exceed foreign currency denominated deposits) so that an appreciation of the somoni relative to U.S. dollars reduces banks’ CARs. The overall impact of an exchange rate fluctuation remains contained by the small size of banks’ net open positions (relative to capital). For example, a 30 percent appreciation of the somoni lowers the system-wide capital adequacy level to an still adequate 15½ percent. However, the largest bank in the system sees its CAR fall below the 12 percent minimum requirement.

110. The high degree of loan dollarization can be an important source of financial vulnerability in the case of a substantial depreciation in somoni. While the direct impact of a somoni depreciation raises banks’ CARs, the indirect credit risk associated with such an exchange rate fluctuation can reverse this effect. When the somoni depreciates relative to foreign currency, unhedged foreign currency borrowers may find it difficult to repay foreign denominated loans. For example, a 30 percent depreciation of the somoni, combined with a 30 percent deterioration in the quality of foreign currency denominated loans (and assuming a zero collateral coverage), leads one bank to violate the 12 percent capital adequacy ratio requirement by 1½ percentage points. Overall, however, the banking system appears robust to such a shock, with a system-wide capital adequacy level remaining slightly below 18 percent. The robustness of these tests depends largely on the sensitivity of banks’ loan portfolios to an exchange rate depreciation, a parameter that is difficult to assess given the limited information on borrower’s access to foreign currency denominated income, such as remittances. In the context of such uncertainties, it is important that banks price this exchange rate induced credit risk appropriately, possibly with additional provisions on foreign currency denominated loans.

Interest rate risk

111. Tajik banks appear able to absorb a sizable interest rate shock, largely because of their unusually high net interest rate margins and small maturity mismatches. A uniform 400 basis points upward shift in the deposit interest rate across all maturities (which represent a 16 percent decrease in net lending margins) leads to a five percentage point reduction (to about 14½ percent) in the system-wide capital adequacy ratio. Two banks, holding about 10 percent of the system’s assets, are particularly vulnerable to such a shock; one because of the relatively longer-term structure of its loan portfolio and the other one because of its heavy reliance on nonresident deposits that are deposited abroad. While considerable, the results do not necessarily suggest an acute vulnerability to interest rate risk, as the applied shock is particularly strong and the methodology to quantify its impact necessarily very basic. Furthermore, if the interest rate rises over time, rather than as an instantaneous shock, banks would be able to modify the maturity structure of their balance sheet and so reduce their exposure to this risk.

Liquidity risk

112. The results suggest that banks are able to withstand a sizable liquidity shock with minimal outside funding. In the case of a somoni deposit run (of 10 percent of deposits each day), two banks would not be able to survive a daily deposit withdrawal for more than three days without securing additional liquidity and, at the end of the fifth day, all banks would be technically considered illiquid (Table 13). The overall liquidity shortfall needed to pay off all deposit withdrawals, however, remains very small (less than ½ percent of pre-shock total banking sector assets), suggesting that banks are able to overcome a relatively dramatic deposit drain (equivalent to a 40 percent cumulative deposit loss) with minimal outside funding. A similar picture emerges from a drain on foreign currency denominated deposits, except that the overall liquidity shortfall is slightly higher (about 2 percent of pre-shock total assets). Nonetheless, banks’ capacity to absorb a deposit run depends largely on the assumption that they can sell their liquid assets when needed, which may not be possible if several banks are hit by the same shock simultaneously.

Table 13.

Summary Results of the Liquidity Stress Tests 1/

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

Deposit run based on a daily withdrawal of 10 percent of demand deposits and term deposits for five consecutive days.

In percent of pre-shock stock of assets.

Dynamic developments

113. The risk profile of Tajik banks has markedly sharpened in the last twelve months due to their more active role in financial intermediation. A comparison of the stress test results based on December 2007 data with the ones based on June 2007 and December 2006 data highlight the overall increase in risk sensitivity of Tajik banks, largely due to a rise in their risk-weighted asset base. Since December 2006, banks’ total assets rose by about 120 percent, most of which is attributed to a growth in lending activities. Accordingly, banks’ capital adequacy ratio on a system-wide basis fell by 7½ percentage points to about 19½ percent, a level that is becoming more comparable to the capital buffers expected from a dynamic and competitive banking system. Nonetheless, if banks continue to expand their balance sheets at a fast pace, it becomes even more important for banks to price appropriately the risks contained in their balance sheets, including the potential build-up of unrecognized credit risk associated with weak credit assessment and management tools.


Remittances are typically relatively stable and counter-cyclical (flows tend to increase in times of hardship in the recipient country), but are highly dependent on economic performance in the host country.


The debt to GDP ratio figure excludes the NBT’s pledges.


Last year’s growth rate reflects the impact of a reclassification of bank liabilities associated with a shift between non-resident (foreign liability) and resident (deposit liability) accounts.


Tajikistan has not yet passed a law or regulations setting forth customer due diligence requirements and reporting of suspicious transactions, nor has it established a FIU. The authorities are strongly encouraged to pass AML/CFT legislation as soon as possible.


KI is a nonbank intermediary involved in cotton sector financing. It was established in 2003 as an asset management company to deal with the bad loans portfolio in the “good bank–bad bank” resolution of the formerly-state owned Agroinvestbank (AIB).


The three systemically important banks hold about 80 the banking system’s assets. Two of these banks are believed to have strong political connections; the fourth largest bank is state-owned.


In 2006, there were about 220 bank branches, translating into about one branch for every 30,000 people, which is low by international standards.


Gaston Gelos, 2006, “Banking Spreads in Latin America,” IMF Working Paper WP/06/44.


See EBRD and FAO, “Tajikistan: Expanding Finance in Rural Areas,” August 2006. This study estimates the universe of potential clients in rural areas at around 700,000 producers on household plots, 27,000 Dekhan farms, and an unknown number of rural nonfarming enterprises.


Recently, the government introduced changes to the legislation on the use of land certificates that could help overcome some of the problems, but all land ownership is still with the state.


CAR is calculated according to the Basel I methodology.


Regulation 139 of July 2005 requires that loans be classified into four past-due categories: substandard (1–30 days), doubtful (30–60 days), dangerous (60–180 days), and hopeless (180 days +). A 2 percent general reserve is required on the performing portfolio, and provisions on the other classes are 5 percent, 30 percent, 75 percent, and 100 percent, respectively. Provisions are calculated net of realizable collateral (with predetermined haircuts).


Loans must be classified as at least substandard if restructured more than once. Restructured loans were 10 percent of the total at end-2007.


Access to microfinance in Tajikistan remains limited compared to other countries. About 1 percent of the active population had access to microcredit in 2005, compared to 2.5 percent in Central and Eastern Europe.


The cotton debt problem has been closely intertwined with the land reforms of the late 1990s, which parceled out the land but not all of the debts of the former collective farms. In the early phases of land reform, some individuals managed to establish farms without carrying a share of the liabilities of the former collective farm. A 2003 Presidential Decree ruled that all debts of reorganized farms should be allocated according to land size.


It is believed that its current owners include politically well-connected individuals, including cotton investors.


Resolution of the Government of the Republic of Tajikistan No. 111 of March 5, 2007 and Box 4 of “Republic of Tajikistan–Staff Monitored Program” report (IMF Country Report No. 08/197).


Many elements of a due diligence audit of KI (e.g., the identity of owners, borrowers and creditors, aspects of credit allocation decisions, loan disbursement and repayment records) are included in the audit of the NBT.


Cotton farming is often not viable but is imposed by the authorities as part of land use agreements. Moreover, the NBT’s system of export authorization prevents farmers in arrears from exporting cotton or other produce.


Currently, the chairman can only be dismissed for abuse of power.


The fist deputy chairman position has also been vacant for some time.


The NBT’s limited ability to mop-up liquidity may also reflect its reluctance to accept market clearing prices.


Shareholders are excluded only if their share is above 20 percent of bank equity (which is easily eluded); there is no related party concept in law.


The fund could consider creating an advisory board which could count commercial bankers as members.


Pursuant to the constitution, all land is owned by the state. The right to use land (and structures on the land) can be mortgaged, but not the land itself.


The court may delay, at the request of the mortgagor, the sale of the foreclosed property through public auction for up to one year. Further appeals by the borrower may delay the trial process for an additional year.


It is reasonable that financial institutions, as the main beneficiary of the services, bear the costs of establishing and maintaining the bureau.


According to the 2006 enterprise restructuring and governance indicator of the European Bank for Reconstruction and Development, Tajikistan was rated at “2−” out of a possible “4+” with only one out of about 30 European and Central Asian countries ranked lower. See also Box 5 in the Staff Report for the 2006 Article IV Consultations.


Tajikistan has begun adopting IFRS standards, and all financial institutions as well as other corporations are obliged to implement the IFRS—although the degree of enforcement reportedly varies across institutions.


The government is ultimately responsible for recapitalizing the central bank, hence currently these costs are already born by the ministry, although only indirectly.


For details, see IMF, 2006, “Strengthening the Monetary Policy Framework and Liquidity Management” (technical assistance report; 05/31/2006).


A large share of nonresident deposits are reportedly deposited abroad (about 75 percent as of June 2007), but this could be easily reversed. Moreover, the exclusion creates an incentive for banks to book deposits as nonresident deposits (this may partly explain the exceptional increase in nonresident deposits in the first half of 2007 and its subsequent sharp decline after the exclusion was reversed).


Banks hold deposits with one another, or a group of smaller banks with bigger banks. This implies fragmentation of liquidity and potential settlement risks as the settlement agent (in this case a commercial bank) is not a risk-free agent even for the domestic currency. The NBT should discuss with the banks involved the risk implications of such practices.


See also NBT Instruction No. 132 “On order of regulating banks activity” (October 2004), and Government Decision No. 206 “On fixing of volume (quotas) of participation of foreign capital in banking system of RT” (May 2003).


The NBT has recently begun involving the banks’ external auditors in high level discussions through tripartite meetings to discuss key issues and focus.


Omitted institutions included a foreign bank branch and a newly established small bank.


Provisioning requirements are computed net of collateral, with a predetermined haircut depending on the nature of the collateral. In practice, the collateral coverage of loans does not seem to follow a regular pattern and can significantly vary across banks and loans, as banks can either increase their collateral coverage or reclassify them to a less-performing category, when they become less-performing. Furthermore, loans can be restructured once without penalty, and upon a subsequent restructure, as at least substandard.

Republic of Tajikistan: Financial System Stability Assessment, including Reports on Observance of Standards and Codes on the following topics, Banking Supervision, and Monetary and Financial Policy Transparency
Author: International Monetary Fund