## 1tjkea2021005

## Source details

**Canonical URL:** [1tjkea2021005](https://www.imf.org/-/media/files/publications/cr/2021/english/1tjkea2021005.pdf)

## Other formats

- [Markdown version](/-/media/files/publications/cr/2021/english/1tjkea2021005.pdf.md)
- [Structured JSON version](/-/media/files/publications/cr/2021/english/1tjkea2021005.pdf.json)

---

### Public spending and sources of financing
- Public spending in Tajikistan increased by some 8 percentage points of GDP on average over the last 15 years.
- More than half of this increase was used for capital spending; capital spending in 2016 explained close to half of the spending envelope.
- Public spending was overwhelmingly financed by domestic tax revenue, which increased from 16 to 21 percent of GDP between 2002–06 and 2012–16.
- Cross-country context:
  - Similar patterns in other non-oil importing CCA countries (Armenia, Georgia, and Kyrgyz Republic) but those increases were more oriented toward current spending and showed a greater increase in domestic tax revenues.
  - CCA oil exporters showed a more modest increase in public spending and a decline in domestic tax revenue on average.

### Tax revenue performance and potential
- Performance and composition
  - Tajikistan’s tax revenue to GDP was second only to Georgia among CCA oil importing countries.
  - Tajikistan boosted its tax revenue at twice the rate of the median low income country over the period compared.
  - Tax revenue has made up on average about 88 percent of total domestic revenue.
  - Indirect taxes have made up on average about 76 percent of tax revenue.
  - Domestic taxes on goods and services (including VAT) explained about half of the tax collection in 2016 and 77 percent of the increase in the tax to GDP ratio over the last 15 years.
  - VAT rate: 18 percent on domestically produced and imported goods and services.
  - Direct taxes on income make up about a third of tax revenue and are identified as not yet realizing their full potential.
- Development since 1998
  - Domestic revenue doubled to 25 percent of GDP since 1998, with tax revenue the key driver.
- Estimating tax potential (Fenochietto and Pessino methodology; Mundlak random effects model with explanatory variables LDG, LDG2, TR, AVA, GINI, GOV, PE, OIL)
  - Tajikistan’s estimated tax potential ranges from a low of 5.9 percentage points of GDP (restricted sample of developing and emerging market economies) to a high of 8.7 percentage points of GDP (full sample estimation).
  - Cross-country positioning: lower tax potential than richer CCA countries, much higher than Azerbaijan and the Kyrgyz Republic, broadly in line with low income peers.
  - Dynamic note: as GDP per capita rises, the applicable tax frontier typically moves up, increasing potential tax capacity over time.
- Policy implications and recommended revenue measures
  - Scope to raise additional revenue implied by tax potential estimates between 5.9 and 8.7 percentage points of GDP.
  - Recommendations to raise revenues while improving fairness, efficiency, simplicity, and productivity:
    - Limit exemptions that jeopardize revenue and good governance and are hard to reverse; eliminate VAT exemptions to broaden the base and enhance revenue.
    - Improve tax compliance using information from financial institutions and mobile banking.
    - Make greater efforts to ensure tax compliance from high income individuals and companies.
    - Modernize large taxpayer management: redefine large taxpayers to rely largely on turnover; centralize management under the Large Taxpayers Inspectorate (LTI); increase and upskill LTI audit staff.
    - Strengthen land and immovable property tax, subject to tax administration capacity limitations.
- Key statistics (preserved exactly as in source)
  - Public spending increase: some 8 percentage points of GDP (last 15 years).
  - Domestic tax revenue increase: from 16 to 21 percent of GDP between 2002–06 and 2012–16.
  - Tax revenue share of total domestic revenue: on average about 88 percent.
  - Indirect taxes share of tax revenue: on average about 76 percent.
  - VAT rate: 18 percent.
  - Domestic revenue level since 1998: doubled to 25 percent of GDP.
  - Estimated tax potential for Tajikistan: between 5.9 and 8.7 percentage points of GDP.

### Exemptions by tax categories (percent of GDP) — selected table values
- 2013: VAT Domestic 8.3; External 4.0; Excises 0.1; Customs duties and fees 1.4; Total 9.8; Nominal GDP (millions of Somoni) 40,520.
- 2014: VAT Domestic 7.1; External 4.4; Excises 0.0; Customs duties and fees 1.4; Total 8.6; Nominal GDP (millions of Somoni) 45,610.
- 2015: VAT Domestic 8.1; External 5.1; Excises 0.0; Customs duties and fees 1.5; Total 9.7; Nominal GDP (millions of Somoni) 48,400.
- 2016: VAT Domestic 9.2; External 5.4; Excises 0.3; Customs duties and fees 1.6; Total 11.0; Nominal GDP (millions of Somoni) 54,470.

### Banking sector: development, access, and depth
- Financial development defined via Access, Depth, Efficiency.
- Access indicators (exact figures)
  - 2014 account ownership: 11 percent of Tajiks older than 15 years had an account at a financial institution (versus 18 percent in the Kyrgyz Republic and 41 percent in Uzbekistan).
  - Branches per 100,000 adults (2013): 7 in Tajikistan (versus 8 in the Kyrgyz Republic and 38 in Uzbekistan).
  - ATMs per 100,000 adults (2013): 10 in Tajikistan (versus 21 in the Kyrgyz Republic and 8 in Uzbekistan).
  - Borrowers per 1,000 adults (2013): 34 in Tajikistan (versus 42 in the Kyrgyz Republic and 47 in Uzbekistan).
- Recent improvements (exact figures)
  - In 2016, percent of Tajiks (15 years and older) with a bank account rose to 55 percent.
  - Bank payment cards: from 36,872 in 2007 to 1,620,334 in 2016.
  - Remote banking accounts increased from 1,570,400 in 2015 to 1,776,988 in January 2017.
  - Internet banking accounts: from 37,110 to 59,748.
  - Mobile device banking accounts: from 37,215 to 48,742.
  - Project launched October 2016 to introduce electronic and digital financial services; law “On Payment Services and Payment Systems” adopted to develop legal and institutional framework.
- Depth of intermediation
  - Deposits to GDP ratio: around 10 to 12 percent.
  - Cross-country comparison for credit to private sector is more favorable to Tajikistan (no numeric value provided in source).

### Banking sector efficiency, asset quality, and solvency — AQR findings and exact ratios
- Diagnostic findings: very low efficiency, lack of profitability, weak revenue generation, high cost structure, high credit risk and impairment, high lending-deposit spread.
- Profitability and capital (selected exact values)
  - Profitability remained negative or close to zero since 2014.
  - Capital adequacy ratio: 27.1% (2010), 24.4% (2011), 25.9% (2012), 22.1% (2013), 14.7% (2014), 11.1% (2015), 15.1% (2016).
  - Regulatory Tier 1 capital to risk-weighted assets: 23.7% (2010), 21.2% (2011), 23.3% (2012), 20.2% (2013), 12.0% (2014), 8.3% (2015), 26.4% (2016).
  - ROA: 1.2% (2010), 0.2% (2011), 0.8% (2012), 1.5% (2013), -3.7% (2014), 0.4% (2015), -3.2% (2016).
  - ROE: 6.3% (2010), 0.9% (2011), 3.9% (2012), 7.8% (2013), -24.2% (2014), 3.7% (2015), -27.2% (2016).
  - Interest margin to gross income: 50.4% (2010), 46.4% (2011), 33.2% (2012), 40.4% (2013), 32.0% (2014), 41.9% (2015), 44.5% (2016).
  - Noninterest expenses to gross income: 62.3% (2010), 86.3% (2011), 75.8% (2012), 65.9% (2013), 66.5% (2014), 64.1% (2015), 171.9% (2016).
  - Nonperforming loans net of provisions to regulatory capital: 5.3% (2010), 3.8% (2011), 5.7% (2012), 27.3% (2013), 51.3% (2014), 92.4% (2015), 131.4% (2016).
  - Nonperforming loans to total gross loans: 7.4% (2010), 7.2% (2011), 9.5% (2012), 15.9% (2013), 25.0% (2014), 29.9% (2015), 54.0% (2016).
- AQR qualitative findings
  - More than half of banks’ portfolios are now in arrears.
  - Even if ‘expected losses’ fully covered by provisions, ‘unexpected losses’ would absorb around 1.3 times the total banks’ capital.
  - Large amounts of repossessed properties and foreclosed assets expose banks to valuation risk.
  - Large exposures relative to regulatory capital: 165.7 percent as of end-2016 (against 58.5 percent as of end-2012); overwhelming majority of large exposures is non-performing.

### Bank restructuring, recapitalization plan, and assessment
- Recapitalization plan (adopted December 11, 2016)
  - Initial aim: recapitalization of four ailing banks for a total amount of TJS 3.85 billion (USD 488 million at that date) in five-year government local currency bonds at 2 percent interest.
  - Amended plan: recapitalization of two systemic banks (TSB and AIB) for a total amount of TJS 3.32 billion (i.e. 6.1 percent of 2016 GDP); two smaller banks, TPB and Fonombank, will be liquidated.
  - Ownership after recapitalization: AIB is 87 percent owned by the state and TSB 85 percent owned.
  - Banks sold these government bonds to the NBT, while retaining the 2 percent per annum charge on the amount of their “recapitalization”.
  - Existing chairmen/CEOs retained; those removed during temporary administration have been reinstated.
- Assessment and shortcomings
  - Recapitalization was implemented with 5-year government bonds instead of cash.
  - Consequences:
    - Full collateralization converts the State’s stake into a secured claim that ranks ahead of unsecured claims; the core function of capital (to absorb losses first in liquidation) is not met.
    - Short-term maturity raises doubt about permanence: bonds must be redeemed in regular instalments (5 percent in 2017, 20 percent in 2018, and 25 percent per annum thereafter), reducing capital over time.
    - Fixed remuneration of 2 percent per annum drags down banks’ profitability.
  - Even if recognized as regulatory capital, amounts fall significantly short of prudential requirements, especially on large exposures and related parties.
  - Continued supervisory forbearance is required for banks to operate as going concerns, fostering moral hazard; entities covered make up for half of banking system deposits.
  - Shielding shareholders and reinstating managers increases moral hazard: shareholders exempted from bearing first losses; managers lack credible sanctions.
  - NBT imposed lending restriction: prohibited the two banks from extending new credits in excess of TJS 300,000, which may deprive households and corporates of financing and depress the economy.

### Policy recommendations for financial-sector reform and stability
- Reduce state intervention and promote competition
  - State should refrain from using banks to pursue developmental objectives or support loss-making SOEs.
  - State-owned banks should originate, price and monitor credit on an arms-length basis like private banks.
  - Banks need to cut operating costs (staff and/or branch network) significantly to generate earnings to cover impairment and rebuild capital internally.
  - Promote development of capital markets as alternative sources of finance to bank loans.
- Promote financial transparency and borrower information
  - Improve disclosure of quantitative and qualitative information on exposures, risk management, and profit and loss formation.
  - Strengthen IFRS-compliant disclosures (IFRS adoption in 2015 improved disclosure quality but many banks still lack timely IFRS disclosure).
  - Develop central credit registers and/or credit bureaus; complement bureau data with independent assessments of clients’ capacity to meet commitments based on cash flows.
  - NBT submitted draft amendments to law on Credit History Bureau to make provision of complete, accurate, and timely data compulsory.
- Upgrade institutional environment: regulation and supervision
  - NBT should drive an ambitious policy for banking-sector development.
  - Regulatory priorities:
    - Ensure capital requirements calculated and complied on a consolidated basis; include all financial subsidiaries and controlled entities.
    - Tailor capital requirements to banks’ risk profiles, including operational risk coverage.
    - Update corporate governance regulation (dated 2011) consistent with Basel Committee revisions; improve identification of ultimate beneficial owners.
  - Supervisory priorities:
    - Shift from rules-compliance to risk-based, forward-looking supervision emphasizing governance, internal control, and risk management.
    - Complement quantitative monitoring with qualitative assessments; enhance off-site monitoring (stress testing) and on-site examination staffing and expertise.
    - Obtain technical assistance (TA) for stress testing.
- Strengthen financial safety net: resolution framework, deposit insurance, ELA, insolvency reform
  - Enable NBT to be designated as resolution authority with transfer and bridge bank powers; align legislation with FSB’s Key Attributes of Effective Resolution Regimes.
  - Empower NBT with non-suspendable and non-reversible resolution decisions (except in cases of criminal or egregious misconduct).
  - Prevent NBT from acquiring shares or guaranteeing liabilities of troubled banks; allow government funding for systemic resolution only when prerequisites met: (i) necessary to preserve financial stability, (ii) alternative private sources exhausted, (iii) losses allocated to shareholders.
  - Deposit Insurance Fund (DIF):
    - DIF established in 2003; improvements in 2011; first operational test with liquidation of two small banks in early 2017.
    - If DIF proves paybox capacity, consider broadening mandate to contribute to resolution funding subject to least cost test and limits tied to insured deposit reimbursement.
  - Emergency Liquidity Assistance (ELA):
    - Establish formal ELA policy clarifying eligibility, preconditions, collateral, terms, and disclosure; disentangle monetary policy operations from ELA provision.
  - Insolvency and bankruptcy modernization:
    - Modernize bankruptcy law to enable prompt workout and restructuring of NPLs, allow out-of-court settlements, specialized courts, and arbitration options to maximize recoveries.
  - NPL workout and asset-resolution options:
    - NBT should set ambitious NPL workout objectives for boards and senior managers with time-bound plans, operational targets, early intervention triggers, and supervisory monitoring.
    - System-wide asset spin-off strategies:
      1. AMC approach: transfer legacy NPLs and non-earning assets to an Asset Management Company; advantages (clean balance sheets, centralize recoveries) and risks (contingent liabilities, governance, taxpayer exposure).
      2. Bridge bank / purchase-and-assume: transfer good assets and some liabilities to another bank or bridge bank; advantages (maintain value of performing assets, minimize customer disruption) and risks (perception of unfair competition, political interference).
    - Both approaches require fair valuation, rehabilitation strategies, compliance with prudential standards, and adherence to “least cost test” and NCWO benchmarking.

### State-Owned Enterprises (SOEs): diagnostics, fiscal risks, and reforms
- SOE sector scale and structure
  - Enterprises with state ownership of > 50%: 908 (Owned by central Govt. 716; Owned by local Govt. 192).
  - Non-financial sector 898; Financial sector 10.
  - By type: SUEs 797 (Owned by central Govt. 608; Owned by local Govt. 189), open JSC 104 (Owned by central Govt. 101; Owned by local Govt. 3), closed JSC 6 (Owned by central Govt. 6), LLC 1 (Owned by central Govt. 1).
- Performance and fiscal exposure (exact figures)
  - Gross income of the 24 largest SOEs declined to TJS 5.3 billion (9.7 percent of GDP) in 2016 from TJS 6.4 billion (13.2 percent of GDP) in 2015.
  - The five largest SOEs (Amonat Bank, Barki Tajik, Tajik Rail Road, Tajik Air Company, TALCO) contributed over 75 percent of gross income.
  - Expenditures declined by 16.5 percent in 2016 compared to the previous year.
  - Net losses decreased from TJS 2.6 billion (5.4 percent of GDP) in 2015 to TJS 2.2 billion (4.0 percent of GDP) in 2016.
  - About 99 percent of these losses can be attributed to Barki Tajik and TALCO.
  - Current accrued liabilities of the 24 largest SOEs were TJS 3.6 billion (6.6 percent of GDP) at end-2016 after increasing substantially by over 30 percent annually between 2014-16.
  - Outstanding sub-loans, including interests and penalties, were TJS 13.3 billion (24.4 percent of GDP) at the end of 2016.
  - Barki Tajik accounts for approximately 95 percent of total on-lent funds (sub-loans).
  - Direct lending from the budget to SOEs: TJS 155.9 million in 2014 and TJS 147.1 million in 2015; about 90 percent of these credits were given to Barki Tajik.
  - Over the period 2013-2015 the government guaranteed TJS 44.6 million loans on average.
  - Contingent liabilities estimated to be about 7-8 percent of GDP in 2016 (considering accrued liabilities of the 24 largest SOEs and the guarantee of deposits in Amonat bank).
  - Dividends paid by state enterprises increased from TJS 13.7 million in 2014 to TJS 17.8 million in 2015 (almost 30 percent increase).
- Quasi-fiscal activities and banking linkages
  - Main quasi-fiscal activities: below-cost energy pricing, noncommercial services, soft budget constraints, barter/offset arrangements, subsidized lending, bailouts.
  - Quasi-fiscal activities generate liabilities that can accumulate and undermine public debt sustainability; rising SOE fiscal costs translate into NPLs in the banking system.
- Institutional developments and monitoring capacity
  - SOE Monitoring Department (SOEMD) established in Ministry of Finance in 2008; monitoring expanded to 24 SOEs by 2013.
  - Government publishes statements of fiscal risks as part of budget documentation since 2013.
  - Dividends policy introduced in 2013 and expanded in 2016.
  - Strategy for managing fiscal risks developed and approved in 2016.
  - SOEMD capacity constraint: two units with nine employees limiting detailed monitoring.
- SOE reform recommendations
  - Phase out and end quasi-fiscal activities; fully fund and disclose any government-directed quasi-fiscal activity.
  - Put transparent framework in place to monitor, assess and disclose fiscal risks from contingent liabilities; consider budgetary reserve fund though constrained by revenue shortfalls.
  - Require all large SOEs to publish audited financial statements annually; require medium and small SOEs to publish summarized financial data annually.
  - Assess SOE viability and earmark nonviable SOEs for restructuring, privatization and winding-up.
  - Expand Statement of Fiscal Risks in the budget to include SOEs’ accumulated arrears to domestic banks.
  - Strengthen capacity of SOEMD in the Ministry of Finance.

*Source: IMF staff estimates and national authorities as presented in “1. Public Spending and Sources of Financing.”*

### 1. Public Spending and Sources of Financing ___________________________________________  3

### 1. Public Spending and Sources of Financing

### Context
- Public spending in Tajikistan increased by some 8 percentage points of GDP on average over the last 15 years.
- More than half of this increase was used for capital spending; capital spending in 2016 explained close to half of the spending envelope.
- Public spending was overwhelmingly financed by domestic tax revenue, which increased from 16 to 21 percent of GDP between 2002–06 and 2012–16.
- Comparative context: similar increases in public spending and domestic tax revenues occurred in other non-oil importing CCA countries (Armenia, Georgia, and Kyrgyz Republic), but on average those countries’ increases were more oriented toward current spending and showed a greater increase in domestic tax revenues. CCA oil exporters showed a more modest increase in public spending and a decline in domestic tax revenue on average.

### Tax Revenue Performance
- Tajikistan’s tax revenue to GDP was second only to Georgia among CCA oil importing countries.
- Tajikistan boosted its tax revenue at twice the rate of the median low income country over the period compared.
- Tax revenue has shown significant reliance on indirect taxes:
  - Tax revenue has made up on average about 88 percent of total domestic revenue.
  - Indirect taxes have made up on average about 76 percent of tax revenue.
  - Domestic taxes on goods and services (including VAT) explained about half of the tax collection in 2016 and 77 percent of the increase in the tax to GDP ratio over the last 15 years.
- The value-added tax is levied at 18 percent on domestically produced and imported goods and services.
- Direct taxes on income make up about a third of tax revenue and are identified as not yet realizing their full potential.

### Development in Domestic Revenue, 1998–2016
- Domestic revenue doubled to 25 percent of GDP since 1998, with tax revenue the key driver of this increase.
- Composition dynamics (1998–2016):
  - Total domestic revenue increased to 25 percent of GDP.
  - Tax revenue accounted for the bulk of the increase; non-tax revenue remained a smaller component.
  - Within tax revenue, direct taxes and indirect taxes evolved as noted above, with indirect taxes becoming dominant.

### Estimating Tax Potential
- The study applies the Fenochietto and Pessino methodology, using a Mundlak random effects model with explanatory variables:
  - LDG: log of real GDP per capita (PPP constant 2005)
  - LDG2: square of LDG
  - TR: trade openness (imports plus exports as percent of GDP)
  - AVA: value added of agriculture in percent of GDP
  - GINI: GINI coefficient
  - GOV: dummy for use of central government revenue in place of general government revenue
  - PE: total public expenditure on education in percent of GDP
  - OIL: dummy for revenue-dependent oil-exporting countries
- Regression results (full sample and restricted sample) are broadly consistent with Fenochietto and Pessino (2010, 2013).
- Tax potential is computed as the difference between estimated maximum tax revenue and the 2013–16 average actual tax revenue. The estimations produce a time-invariant tax effort exp(−u_i).
- Tajikistan’s estimated tax potential:
  - Range from a low of 5.9 percentage points of GDP (restricted sample of developing and emerging market economies) to a high of 8.7 percentage points of GDP (full sample estimation).
- Cross-country positioning:
  - Tajikistan’s tax potential is lower than richer CCA countries (suggesting higher tax effort there), much higher than Azerbaijan and the Kyrgyz Republic, and broadly in line with low income peers.
- Dynamic note: as GDP per capita rises, the applicable tax frontier typically moves up, increasing potential tax capacity over time.

### Challenges and Prospects
- The empirical results indicate room to improve Tajikistan’s tax system; tax potential estimates between 5.9 and 8.7 percentage points of GDP imply scope to raise additional revenue.
- Given recent declines in tax revenue as a share of GDP and pressing fiscal consolidation needs, mobilizing additional revenue is important to maintain macroeconomic stability and sustainable development efforts.
- Recommended measures to raise revenues while improving fairness, efficiency, simplicity, and productivity of the tax system:
  - Limit exemptions that jeopardize revenue and good governance and are hard to reverse.
    - The VAT base has been narrowed by exemptions and special preferences; eliminate VAT exemptions to broaden the base and enhance revenue.
  - Improve tax compliance to reduce distortions and raise revenues, including:
    - Better mobilizing information from the increasing number of transactions via financial institutions and mobile banking to improve compliance.
    - Making greater efforts to ensure tax compliance from high income individuals and companies, as they account for a large share of taxable income.
    - Modernizing large taxpayer management, including changing the definition of large taxpayers to rely largely on turnover; centralize management of large taxpayers and bring many large taxpayers currently managed outside the Large Taxpayers Inspectorate (LTI) under it.
    - Increasing the number and improving skills of LTI audit staff.
    - Strengthening land and immovable property tax, subject to tax administration capacity limitations.

### Key Statistics and Figures (preserved exactly as in source)
- Public spending increase: some 8 percentage points of GDP (last 15 years).
- Domestic tax revenue increase: from 16 to 21 percent of GDP between 2002–06 and 2012–16.
- Tax revenue share of total domestic revenue: on average about 88 percent.
- Indirect taxes share of tax revenue: on average about 76 percent.
- VAT rate: 18 percent.
- Domestic revenue level since 1998: doubled to 25 percent of GDP.
- Estimated tax potential for Tajikistan: between 5.9 and 8.7 percentage points of GDP.

### Exemptions by Tax Categories (percent of GDP)
- VAT Domestic / External / Total and other items by year:
  - 2013: VAT Domestic 8.3; External 4.0; (implied VAT Total 8.3 4.0 4.2 as presented); Excises 0.1; Customs duties and fees 1.4; Total 9.8; Nominal GDP (millions of Somoni) 40,520.
  - 2014: VAT Domestic 7.1; External 4.4; (implied VAT Total 7.1 4.4 2.8 as presented); Excises 0.0; Customs duties and fees 1.4; Total 8.6; Nominal GDP (millions of Somoni) 45,610.
  - 2015: VAT Domestic 8.1; External 5.1; (implied VAT Total 8.1 5.1 3.1 as presented); Excises 0.0; Customs duties and fees 1.5; Total 9.7; Nominal GDP (millions of Somoni) 48,400.
  - 2016: VAT Domestic 9.2; External 5.4; (implied VAT Total 9.2 5.4 3.7 as presented); Excises 0.3; Customs duties and fees 1.6; Total 11.0; Nominal GDP (millions of Somoni) 54,470.
- (Table header presented as "Table 3. Tajikistan: Exemptions by Tax Categories (percent of GDP)"; source: Authorities Data.)

*Source: IMF staff estimates and national authorities as presented in “1. Public Spending and Sources of Financing.”*

### 11.      These measures would ensure a fairer and more efficient tax collection system and

### 11.      These measures would ensure a fairer and more efficient tax collection system and

### Tax collection reforms and considerations
- These measures would ensure a fairer and more efficient tax collection system and support income growth and job creation.
- Concerns noted:
  - Some individual reforms may be regressive.
  - Benefits in terms of improved governance and the progressivity of the additional expenditure being financed should be considered when assessing the measures.
  - Constraints arising from limited tax administration capacity in a low-income country context will need to be addressed.

### The challenges of the Tajik banking sector — overview
- Despite recent progress in access to banking services, the banking system is still characterized by:
  - Low level of financial inclusion.
  - High financial stability risks.
- The paper’s case:
  - Design a higher mix of financial development and stability.
  - Require a comprehensive, ambitious and outcome-oriented strategy to:
    - Enable proper market discipline.
    - Upgrade the institutional environment.
    - Strengthen individual players through a clean-up of legacy problem assets.

### Financial development — definition and indicators
- Financial development defined as combination of:
  - Access (ability of individuals to access financial services).
  - Depth (size and liquidity of markets).
  - Efficiency (ability of institutions to provide financial services at low cost and with sustainable revenues).
- Indicators used to measure financial development include:
  - Access: accounts per thousand adults, number of branches, percentage of firms with lines of credit.
  - Depth/size: deposits to GDP, private sector credit to GDP.
  - Efficiency/structure: assets’ market share of the 3 largest banks, cost to income ratio, net interest margin, non-interest income/total income, overhead costs/total assets, return on assets, return on equity, credit to government and SOEs to GDP, lending-deposit spread.
- Due to limited data availability, analysis focused on a subset of these indicators.

### Access to financial services — findings and exact figures
- 2014 account ownership:
  - In 2014, 11 percent of Tajiks older than 15 years were reported to have an account at a financial institution (versus 18 percent in the Kyrgyz Republic and 41 percent in Uzbekistan).
- Branches and ATMs (2013):
  - Commercial bank branches per 100,000 adults: 7 in Tajikistan (versus 8 in the Kyrgyz Republic and 38 in Uzbekistan).
  - Automated teller machines per 100,000 adults: 10 in Tajikistan (versus 21 in the Kyrgyz Republic and 8 in Uzbekistan).
- Borrowers (2013):
  - Number of borrowers from commercial banks per 1,000 adults: reached 34 in Tajikistan (versus 42 in the Kyrgyz Republic and 47 in Uzbekistan).
- Recent improvements:
  - In 2016, the percent of Tajiks (15 years and older) with a bank account rose to 55 percent.
  - Growth in bank payment cards: from 36,872 in 2007 to 1,620,334 in 2016.
  - Development of leasing and insurance companies, pawnshops, and credit unions noted.

### Remote banking services expansion — exact figures and legal framework
- Accounts with remote banking services increased:
  - From 1,570,400 in 2015 to 1,776,988 in January 2017.
  - Internet banking accounts: from 37,110 to 59,748.
  - Mobile device banking accounts: from 37,215 to 48,742.
- Project launched October 2016 to introduce electronic and digital financial services (including mobile money products).
- Law adopted: “On Payment Services and Payment Systems” to develop legal and institutional framework for payment services and supervision; establishes requirements for interaction of credit organizations and mobile operators.

### Depth of banking sector — deposits and credit
- Deposits to GDP ratio:
  - Intermediation seems very limited, with a ratio of deposits to GDP stuck around 10 to 12 percent, which lags behind neighboring countries.
- Credit to private sector:
  - Cross-country comparison more favorable to Tajikistan in respect of credit to the private sector (no numeric exact value given in source text for this ratio).

### Efficiency and structure — key diagnostic findings and table references
- Indicators point to very low efficiency:
  - Lack of profitability (large negative returns on assets or equity).
  - Weak revenue generation (low net interest margin).
  - High cost structure (high cost-to-income, high overheads costs to total assets).
  - High cost of credit risk (significant impairment costs on non-performing loans).
  - Resulting high lending-deposit spread.
- Since 2015, profitability has deteriorated further compared to peers.

### Financial stability — asset quality, solvency, and resilience (exact figures)
- Longstanding weaknesses: poor asset quality, weak solvency, low efficiency and profitability.
- Impact on profitability and capital:
  - Profitability remained negative or close to zero since 2014.
  - Capital adequacy ratio decreased from a peak 25.9 percent in December 2012 to 7.1 percent in June 2016, before state recapitalization in December 2016 raised it to 15.1 percent.
- Funding base:
  - Essentially deposit-based; lack of access to long-term resources led some banks to breach regulatory thresholds or resort to emergency refinancing with the central bank.

### Asset quality review (AQR) findings — exact ratios and implications
- AQRs highlighted credit risk and deterioration:
  - More than half of the banks’ portfolios are now in arrears.
  - Even if ‘expected losses’ fully covered by provisions, ‘unexpected losses’ would absorb around 1.3 times the total banks’ capital.
  - Large amounts of repossessed properties and foreclosed assets expose banks to valuation risk.
- Table of indicators (selected exact values from source):
  - Regulatory capital to risk-weighted assets: 27.1% (2010), 24.4% (2011), 25.9% (2012), 22.1% (2013), 14.7% (2014), 11.1% (2015), 15.1% (2016).
  - Regulatory Tier 1 capital to risk-weighted assets: 23.7% (2010), 21.2% (2011), 23.3% (2012), 20.2% (2013), 12.0% (2014), 8.3% (2015), 26.4% (2016).
  - Return on assets (ROA): 1.2% (2010), 0.2% (2011), 0.8% (2012), 1.5% (2013), -3.7% (2014), 0.4% (2015), -3.2% (2016).
  - Return on equity (ROE): 6.3% (2010), 0.9% (2011), 3.9% (2012), 7.8% (2013), -24.2% (2014), 3.7% (2015), -27.2% (2016).
  - Interest margin to gross income: 50.4% (2010), 46.4% (2011), 33.2% (2012), 40.4% (2013), 32.0% (2014), 41.9% (2015), 44.5% (2016).
  - Noninterest expenses to gross income: 62.3% (2010), 86.3% (2011), 75.8% (2012), 65.9% (2013), 66.5% (2014), 64.1% (2015), 171.9% (2016).
  - Nonperforming loans net of provisions to regulatory capital: 5.3% (2010), 3.8% (2011), 5.7% (2012), 27.3% (2013), 51.3% (2014), 92.4% (2015), 131.4% (2016).
  - Nonperforming loans to total gross loans: 7.4% (2010), 7.2% (2011), 9.5% (2012), 15.9% (2013), 25.0% (2014), 29.9% (2015), 54.0% (2016).

### Peculiar characteristics compounding credit risk
- High contribution of related party-lending to NPLs and foreclosed assets; auditors often unable to obtain comprehensive information.
- Concentration of credit risk:
  - Frequency of breaches of large exposures regulation.
  - Large exposures relative to regulatory capital: 165.7 percent as of end-2016, against 58.5 percent as of end-2012.
  - Overwhelming majority of large exposures is non-performing.
- Collateral and foreclosure inefficiencies:
  - Inefficiency of registrations by lenders or foreclosure processes, especially in judicial proceedings.
- Indirect credit risk:
  - Materiality of unhedged borrowers with foreign currency indebtedness who are exposed to an exchange rate shock.
- These factors complicate identification, monitoring and management of credit risk and require guidance from the NBT.

### Cross-country perspective and measurement differences
- Key performance indicators for solvency and asset quality were favorable until around mid-2014, then deteriorated sharply.
- Note on NPL definition:
  - NPLs reported by Tajikistan include exposures more than 30 days past due, which is more stringent than the FSI criterion limited to exposures more than 90 days past due.
  - Regulatory definition also includes exposures to debtors assessed as unlikely to pay their credit obligations, regardless of past due days; in practice, Tajik banks usually detect such unlikelihood to pay after loans are overdue.

### Bank restructuring and government response — exact figures and outcomes
- AQRs provided insight into banks’ financial and prudential situation, though viability was not addressed in depth.
- Recapitalization plan adopted December 11, 2016:
  - Aimed at recapitalization of four ailing banks for a total amount of TJS 3.85 billion (USD 488 million at that date) in five-year government local currency bonds at 2 percent interest.
  - Plan amended to confirm recapitalization of two systemic banks, TSB and AIB, for a total amount of TJS 3.32 billion (i.e. 6.1 percent of 2016 GDP); two smaller banks, TPB and Fonombank, will be liquidated.
  - Resulting ownership: AIB is 87 percent owned by the state and TSB 85 percent owned.
  - Banks sold these government bonds to the NBT, while retaining the 2 percent per annum charge on the amount of their “recapitalization”.
  - Existing chairmen/CEOs have been retained; those removed during temporary administration have been reinstated.
- Policy rationale and merits of the plan:
  - Broadened scope of resolution measures to four banks, enabling a comprehensive approach.
  - Included liquidation of two small banks deemed obviously insolvent and unviable, avoiding costly rehabilitation at public expense and limiting moral hazard.

### Assessment of the recapitalization plan and outstanding shortcomings
- Concerns that the plan is not fully consistent with best international practices in bank resolution:
  - Does not tackle the source of banking distress (inadequate bank governance).
  - Does not necessarily solve the banking distress itself and may foster adverse developments.
- Source of distress — inadequate bank governance:
  - AQRs show TSB and AIB saddled with huge legacy of NPLs inherited from inadequate governance manifesting as:
    - Inappropriate loan approval, monitoring, and collection processes.
    - Widespread lending to related parties.
    - Material shortcomings in accounting and prudential reporting.
  - Reinstatement of top managers unlikely to allow prompt overhaul of governance and risk management because of:
    - Their own track record in managerial competence and risk strategy.
    - Their links with the largest non-performing borrowers.
- Inadequate recapitalization (issue raised in text; further specifics beyond this heading not provided in excerpt).

*International Monetary Fund — Republic of Tajikistan*

### 16.      The plan falls short of providing capital that would be eligible for regulatory purposes.

### 16.      The plan falls short of providing capital that would be eligible for regulatory purposes.

### Recapitalization design and immediate implications
- Recapitalization was implemented with 5-year government bonds instead of cash.
- Consequences identified:
  - "The full collateralization of these government bonds converts the State’s stake into a secured claim, such that this claim ranks ahead of unsecured claims and, accordingly, the core function of capital (i.e., to absorb losses first in liquidation) is not met."
  - "The short-term maturity of the government bonds raises a doubt about the permanence of the scheme – also a core feature of regulatory capital. Until their 5-year maturity, the bonds must be redeemed by the banks in regular instalments 8, and the bank’s capital is reduced accordingly."
    - Footnote 8: "5 percent in 2017, 20 percent in 2018, and 25 percent per annum thereafter."
  - "The fixed remuneration of the “capital” provided (2 percent per annum) drags down the banks’ profitability, heightening the risk of (re) failure, even more since these bonds account for a substantial share of assets; the viability of these banks is thus critically threatened."

### Regulatory-capital adequacy and gaps
- Even if the mentioned amounts were recognized as regulatory capital, they "fall significantly short of allowing the banks to meet prudential requirements."
- Comparison basis: "gaps estimated on the basis of call reports for each type of regulation, without taking account of FX-related requirements9 and capped by the total amount of liabilities."
  - Footnote 9: "The compliance with such FX-related capital requirements is assumed to be achieved by closing the excessive FX positions, rather than through a capital increase."
- "Even after taking into account the recapitalization of December 2016, the capital shortages to meet the regulatory requirements are very significant, especially those on large exposures and related parties."
  - Note: "Ac Actually, the amount of “recapitalization” has been allocated primarily for the purpose of improving the capital adequacy ratio, which is not necessarily the most stringent prudential requirement."

### Potential adverse aftermath and moral hazard
- Continued supervisory forbearance is required for banks to operate as going concerns, fostering moral hazard.
  - Forbearance is being applied to entities which together make up for half of the banking system’s deposits.
- Shielding shareholders and reinstating managers increases moral hazard:
  - Shareholders are "exempted from bearing the first losses despite government intervention" and thus "preserve their shares—at most on a diluted basis—and will benefit from any upside."
  - Lack of credible sanctions undermines incentives for boards and senior managers to meet "duty of care" and "duty of loyalty."
- Lending restrictions could have macroeconomic and bank-quality consequences:
  - "The NBT has prohibited these two banks from extending new credits in excess of TJS 300,000."
  - "Restrictions on lending ... will deprive households and corporates from new financing from the two largest systemic banks and contribute to further depressing the economy," potentially increasing provisioning needs and exacerbating borrower difficulties.

### Market-enabling policies — reducing state intervention and promoting competition
- Problem statement: "The mix of low financial development and high financial fragility ... is an impediment to economic growth."
- Reduce government-directed lending and strengthen bank governance:
  - "The state should refrain from using banks to pursue developmental objectives or support loss-making state-owned enterprises."
  - "State-owned banks should be allowed to originate, price and monitor credit on an arms-length basis, in the same way as privately-owned banks."
- Promote competition and efficiency in the financial system:
  - As of end-2016, the system composition: 18 banks, 38 micro-credit deposit organizations, 14 micro-credit organizations, and 34 micro-credit funds.
  - Banks account for 80.6 percent of the total (net) loan portfolio.
  - "Commercial banks dominate the financial system ... report (net) loans for less than 15 percent of GDP."
  - Four insolvent banks have a market share of 46 percent and have long weighed on the system’s capacity to properly select and price credits.
  - Recommendation: banks need to "cut their operating costs (in terms of staff and/or branch network) significantly" to generate earnings to cover impairment and rebuild capital internally.
  - Longer-term: "competition should also be enhanced by promoting the development of capital markets as alternative sources of finance to bank loans."

### Promoting financial transparency and borrower information
- Bank-side transparency:
  - Improve disclosure of quantitative and qualitative information on exposures (credit, market, operational, liquidity, and interest rate risks), risk management and mitigation techniques, and profit and loss formation.
  - IFRS adoption in 2015 improved disclosure quality but "persistent absence or lag of IFRS-compliant disclosures by most banks, including the largest ones."
- Borrower-side transparency:
  - Develop central credit registers and/or credit bureaus to reduce reliance on collateral-based lending.
  - Current credit bureau data should be complemented by independent assessments of clients’ capacity to meet commitments based on cash flows.
  - Ongoing authority projects to improve comprehensiveness and reliability of credit-bureau information are "worth commending."
  - Footnote 14: "The AQRs have confirmed that the information on borrowers’ cash-flows or business plans was frequently lacking in the banks’ credit files."
  - Footnote 15: "The NBT has submitted to the government draft amendments and additions to the Republic of Tajikistan Law on the Credit History Bureau, to make the provision of such (complete, accurate, and timely) data compulsory."

### Upgrading the institutional environment — regulation and supervision
- Role of the NBT:
  - The NBT "should become the driving force of an ambitious policy dedicated to the development of the banking sector."
  - Tasks: ensure regulatory environment, governance, and supervisory capacity to assure prudent and sustainable financial sector growth.
- Regulatory priorities:
  - Increase the appropriateness of capital requirements: ensure calculation and compliance on a consolidated basis, include all financial subsidiaries and controlled entities.
  - Tailor capital requirements to banks’ risk profiles, including coverage of operational risk.
  - Strengthen internal governance: update corporate governance regulation (dated 2011) consistent with Basel Committee revisions.
  - Improve transparency of ownership structures: better identification of ultimate beneficial owners to monitor related-party lending and governance.
- Supervisory priorities:
  - Shift supervision from rules-compliance to risk-based, forward-looking supervision emphasizing governance, internal control, and risk management.
  - Complement quantitative monitoring with qualitative assessment of banks’ internal processes.
  - Enhance supervisory capacity for off-site monitoring (stress testing) and on-site examination (increase staffing in quantity and expertise).
  - Suggestion: obtain technical assistance (TA) from the World Bank for stress testing.

### Financial safety net — resolution framework and deposit insurance (start of section)
- Resolution framework needs strengthening to give the NBT adequate resolution powers to minimize taxpayer cost without disrupting financial stability:
  - Objective: enable NBT to be designated as resolution authority with transfer and bridge bank powers.
  - Align Tajik legislation with the FSB’s Key Attributes of Effective Resolution Regimes.
- Current status and legislative needs:
  - NBT cannot require banks to prepare and test recovery plans; no recovery planning framework exists.
  - NBT has limited intervention powers that "can be countered by the courts, i.e. either suspended or reversed."
  - Drafted amendments would strengthen resolution powers (transfer of assets and deposits to a bridge bank, transfer to asset management companies, transfer of shares, write-offs of shareholders, bail-in of some unsecured and uninsured creditors).
  - To align with international standards, proposals should provide for "the non-suspension and non-reversal of resolution decisions of the NBT18," while preserving the right to monetary compensation.
    - Footnote 18: "Except in cases where the NBT acted in a criminal manner or was otherwise in breach of applicable norms."
- Resolution funding and safeguards against moral hazard:
  - Law should prevent the NBT from acquiring shares or guaranteeing liabilities of a troubled bank to protect the central bank’s balance sheet and avoid conflicts of interest.
  - Law should allow government funding of resolution transactions for systemic cases subject to prerequisites:
    - (i) the provision of such public funds is necessary to preserve financial stability,
    - (ii) alternative private sources of funding have been exhausted,
    - (iii) the losses have been allocated to shareholders.

*REPUBLIC OF TAJIKISTAN INTERNATIONAL MONETARY FUND*

### 37.      Parallel initiatives could usefully strengthen the deposit guarantee mechanism, to

### 1tjkea2021005 - 37.      Parallel initiatives could usefully strengthen the deposit guarantee mechanism, to

### Deposit Insurance Fund (DIF) — current status and potential mandate extension
- Tajik Deposit Insurance Fund (DIF) established in 2003; significant improvements in 2011; first operational test occurred with the liquidation of two small banks in early 2017.
- The execution of the 2017 test — in both amount and promptness of payouts — is crucial to support the DIF’s credibility in depositor protection and confidence in the banking system.
- If the current payout experience confirms the DIF’s capacity to perform the ‘paybox’ mandate assigned by law, there would be strong arguments for broadening the mandate to enable contribution to resolution funding.
- Under an extended mandate:
  - DIF could finance a restructuring transaction that involves a transfer of insured deposits provided the least cost test is met (cost of restructuring transaction is less than cost of reimbursing insured deposits in liquidation, net of expected recoveries).
  - If a bank failure is systemic and public funds are authorized by the Ministry of Finance, the DIF could make payments toward the cost of a restructuring transaction that involves a transfer of insured deposits even if the least cost test is not met, provided DIF resources used do not exceed the amount that would be necessary for reimbursement of insured deposits in liquidation, net of expected recoveries.

### Emergency Liquidity Assistance (ELA) framework
- A legally-grounded lender-of-last-resort framework is necessary alongside strong supervision, timely resolution, and effective deposit protection; current NBT law already entrusts the central bank with such a function.
- Given increasing dependency of the banking system on central bank refinancing, particularly for public banks, and potential adverse impact on monetary policy autonomy, a proper ELA framework needs to be established.
- Key legal and operational features recommended:
  - Better disentangle monetary policy operations from ELA provision; ELA should preserve financial stability by handling banks’ idiosyncratic refinancing difficulties.
  - Law should explicitly state ELA can be provided at the NBT’s sole discretion only under very limited circumstances and toward banks that meet predetermined preconditions.
  - NBT should set up a formalized, comprehensive ELA policy to operationalize detailed eligibility criteria, processes, terms and conditions, and disclosure requirements applicable to ELA.
  - NBT should require banks to be prepositioned so that collateral can be accessed quickly and reliably to upgrade operational preparedness.

### Insolvency framework and bankruptcy law modernization
- Modernization of the bankruptcy law and solvency regime is needed to improve efficiency of the financial system, given the track record in banks’ NPL recovery processes.
- Principles and reforms proposed:
  - Enable prompt workout and/or restructuring of NPLs supported by legal framework protecting rights of lenders and borrowers, and allowing out-of-court settlements.
  - Give banks more flexibility to compromise with troubled borrowers (especially the most creditworthy) before initiating collection lawsuits.
  - Develop arbitration with borrowers as a last-ditch option to optimize trade-offs between amount, timing, and cost of recoveries to maximize present value of assets.
  - Maintain banks’ right to go to court when workout fails; courts should be specialized to better take commercial/banking considerations into account, speed up proceedings, and reduce enforcement costs.
  - Revision of legal framework should aim to reduce bias toward borrowers’ rights, speed up court proceedings, and reduce cost of enforcing judicial decisions.
  - Carry out bankruptcy law revision within a broader legislative overhaul to improve enforceability of contractual and property rights, particularly regarding collateral realization and foreclosure.

### Addressing the legacy of high NPLs — supervisory and restructuring measures
- NBT should set ambitious NPL workout objectives for banks’ Board members and senior managers, with individual, time-bound plans:
  - Assign operational targets to implement NPL workout processes.
  - Trigger early intervention measures or resolution actions if targets are not met.
  - Hold boards and senior managers responsible for implementation, using specific and verifiable measures and milestones to improve identification, measurement, management and write-off of NPLs or foreclosed assets.
  - Off-site supervision division should monitor follow-up of plans; on-site examinations targeted on problem banks or those presenting delays.

### System-wide schemes to facilitate asset spin-off — two alternative splitting strategies
- Two conceivable strategies depend on banks’ sustainability and implied taxpayer cost.

1) AMC approach (transfer all non-performing and non-earning assets to an Asset Management Company):
- Structure:
  - Transfer legacy NPLs and non-earning assets to a winding down non-banking entity (AMC); banks remain with deposits and are required to be viable after asset sales.
  - AMC would need significant long-term funding, but only little capital.
- Advantages:
  - Cleans up banks’ balance sheets.
  - Breaks links between banks and their debtors (especially related parties), potentially improving collectability.
  - Allows banks to focus on core business (lending).
  - Safeguards banks’ brands.
  - Serves as vehicle for getting NPLs out of troubled banks based on uniform valuation criteria, centralizing collateral ownership, providing more leverage over debtors.
  - Could allow some payoff as impaired assets could be sold over time when market conditions are more favorable.
- Disadvantages:
  - Leaves banks with contingent liabilities likely to arise from specific asset transfers to the AMC or from non-transferred business lines.
  - Loses banks’ informational and operational advantages (knowledge of borrowers; ability to provide additional financing facilitating restructuring).
  - May become bloated and depend increasingly on government funds or government guaranteed bonds, potentially becoming a black hole for taxpayer money.
  - May lessen banks’ incentives to improve recovery processes and risk management.
  - Without sound governance, AMC creation may lead to deterioration of payment discipline, asset values, moral hazard, and corruption.
  - Difficult to insulate AMC against external pressure from insiders and/or related parties.

2) Bridge bank / purchase-and-assume approach (transfer good assets and some liabilities to another bank or bridge bank):
- Structure:
  - Transfer good assets and some liabilities (including, but not necessarily limited to, insured deposits) to another bank; revoke license of transferor bank and liquidate remaining entity.
  - Envisaged on a closed basis: license of transferor revoked to leave contingent liabilities in liquidation.
  - Where no private bank is available to assume assets, a temporary bridge bank may be established, temporarily owned wholly or partially by the Government with aim to privatize later.
- Advantages:
  - Maintains value of performing assets by transferring them immediately to a healthy bank; facilitates sale to buyers not interested in entire failed bank.
  - Immunizes bridge bank (and subsequent buyer) against contingent liabilities left in liquidation.
  - Minimizes disruption to customers, particularly insured depositors and possibly other depositors, by providing access to bridge bank.
- Disadvantages:
  - May feed perception of unfair competition if bridge bank is state-owned and subject to special/different prudential requirements.
  - May result in excessive time in state hands and risk of political interference.
  - May raise cost overruns relative to original estimate if good assets deteriorate after transfer.
  - May lead to reduction of deposit base and potential loss in franchise value.

- Common prerequisites and safeguards for both approaches:
  - Attention to immediate asset-quality improvement and upgrading of risk management instruments.
  - Rehabilitated banks must strictly comply with all prudential standards and financial discipline; require drafting of rehabilitation strategy and action plan with monitoring for next few years.
  - Availability of a fair, prudent and realistic economic valuation of banks’ assets and liabilities (whether transferred or not) is fundamental.
  - Benchmarking economic value in resolution against gone-concern value in liquidation provides basis for ‘No Creditors Worse Off’ (NCWO) test, whereby “creditors should have a right to compensation where they do not receive at a minimum what they would have received in a liquidation of the firm under the applicable insolvency regime”.

- Cost-effectiveness assessment:
  - Should follow best international practices, in particular the Key Attributes of Effective Resolution Regimes for Financial Institutions.
  - The “least cost test” (minimization of cost to taxpayers) is a key criterion.
  - Main cost components typically include capital injection in transferee structure (AMC or bridge bank) and cost of funding transferred assets over expected recovery process.

### The way forward — strategic priorities
- Low levels of financial development and financial stability in Tajikistan weigh significantly on macroeconomic outlook; underdeveloped banking sector fails to finance growth and high vulnerability stifles depositor confidence and savings mobilization.
- Strong case for a comprehensive, ambitious and outcome-oriented strategy driven top-down by a few authorities (such as the Ministry of Finance and the Central Bank) within a clearly articulated and publicly communicated policy framework.
- Desirable pillars of the strategy:
  - (i) enable proper market discipline;
  - (ii) upgrade the institutional environment;
  - (iii) strengthen individual players through clean-up of legacy problem assets.

### State-Owned Enterprises (SOEs) — summary diagnostics (introduction)
- The paper assesses performance of the 24 largest State-Owned Enterprises (SOEs) in Tajikistan and fiscal risks they pose.
- Key findings and data points:
  - Although SOEs contribute to growth and development, poor financial and operational performance imposes serious fiscal costs and risks.
  - Revenue shortfalls and rising fiscal risks limit fiscal space; government should fast-track reforms to improve SOE governance to improve service delivery and minimize fiscal risks.
  - Privatization since 1991: more than twelve thousand enterprises privatized by 2015, but privatization slowed as remaining SOEs lack privatization potential.
  - During 2013-2015, SOEs accounted for an average of 14 percent of total economic activity.
  - Preliminary information indicates SOEs account for a significant share of NPLs in the banking sector.
  - The analysis uses data for the 24 largest SOEs monitored by the SOE Monitoring Department (SOEMD).

### SOE sector structure (key numbers)
- Enterprises with state ownership of > 50%: 908 (Owned by central Govt. 716; Owned by local Govt. 192).
- Grouped by sector:
  - Non-financial sector 898
  - Financial sector 10
- Grouped by type:
  - SUEs 797 (Owned by central Govt. 608; Owned by local Govt. 189)
  - open JSC 104 (Owned by central Govt. 101; Owned by local Govt. 3)
  - closed JSC 6 (Owned by central Govt. 6)
  - LLC 1 (Owned by central Govt. 1)

*Source: 1tjkea2021005 - 37.      Parallel initiatives could usefully strengthen the deposit guarantee mechanism, to*

### 7.      In recent years, the profitability of some of the large SOEs has deteriorated. In 2016,

### 7.      In recent years, the profitability of some of the large SOEs has deteriorated. In 2016,

### Summary of recent performance and fiscal exposure
- Gross income of the 24 largest SOEs declined to TJS 5.3 billion (9.7 percent of GDP) in 2016 from TJS 6.4 billion (13.2 percent of GDP) in 2015.
- The five largest SOEs—Amonat Bank, Barki Tajik (electricity company), Tajik Rail Road, Tajik Air Company, and Tajikistan Aluminum Company (TALCO)—contributed over 75 percent of the gross income.
- Expenditures declined by 16.5 percent in 2016 compared to the previous year.
- Net losses decreased from TJS 2.6 billion (5.4 percent of GDP) in 2015 to TJS 2.2 billion (4.0 percent of GDP) in 2016.
- About 99 percent of these losses can be attributed to the two largest SOEs: Barki Tajik and TALCO.
- Continued losses by Barki Tajik and TALCO, if not contained, can threaten fiscal and debt sustainability.

### Accrued liabilities, contingent liabilities and on-lending
- Current accrued liabilities of the 24 largest SOEs increased substantially by over 30 percent annually between 2014-16.
- At end 2016, current accrued liabilities of the 24 largest SOEs were TJS 3.6 billion (6.6 percent of GDP).
- These liabilities include: borrowing directly from the government’s budget; borrowing from domestic banks and other non-bank financial institutions; payments to suppliers for goods received; unpaid taxes and unpaid wages.
- Outstanding sub-loans, including interests and penalties, were TJS 13.3 billion (24.4 percent of GDP) at the end of 2016.
- Barki Tajik accounts for approximately 95 percent of total on-lent funds (sub-loans).
- Direct lending from the budget to SOEs was TJS 155.9 million in 2014 and TJS 147.1 million in 2015; about 90 percent of these credits were given to Barki Tajik with the remainder to Tajik Trans Gas. Preliminary information indicates that a majority of these loans are still outstanding.
- Over the period 2013-2015 the government guaranteed TJS 44.6 million loans on average.
- Considering current accrued liabilities of the 24 largest SOEs and the guarantee of deposits in Amonat bank, contingent liabilities are estimated to be about 7-8 percent of GDP in 2016.

### Dividends and revenue contribution
- Dividends paid by state enterprises increased by almost 30 percent in 2015 – from TJS 13.7 million in 2014 to TJS 17.8 million in 2015.
- Not all profitable enterprises transfer dividends to the budget due to poor compliance with the policy; therefore, revenues from dividends are lower than what the government should receive.

### Quasi-fiscal activities and banking-sector linkages
- Main quasi-fiscal activities include: the use of below cost prices for energy, utilities, and other services; provision of noncommercial services; soft budget constraints; barter and offset arrangements; subsidized lending; and bailouts.
- Quasi-fiscal activities generate liabilities that can accumulate over time, undermining public debt sustainability.
- Rising SOE fiscal costs translate into nonperforming loans (NPLs) in the banking system, creating additional fiscal and financial sector risks.
- Contingent liabilities realizations, likely to coincide with banking crises, are comparable to recent cross-country estimates (study cited finds about 6 percent of GDP on average).

### Institutional developments and monitoring capacity
- Key government actions to strengthen oversight:
  - Establishing and operationalizing an SOEMD in the Ministry of Finance in 2008; monitoring scope expanded from 10 SOEs to 16 and then to 24 SOEs by 2013.
  - Preparing and publishing statements of fiscal risks as part of the budget documentation since 2013. The statement includes: explicit direct liabilities from on-lending arrangements; sub-loans from the state budget; direct loans to SOEs from state budget; contingent liabilities from guaranteed loans; the materialization of risks; legal disputes against SOEs; quasi-fiscal activities by the SOEs; statistics on direct subsidies; statistics on dividends transferred by the SOEs to the state budget; and general recommendations to mitigate and manage risks.
  - Introducing a dividends policy in 2013 and expanding it in 2016 to cover additional large SOEs that are majority-owned by the government.
  - Developing and approving a strategy for managing fiscal risks in 2016.
- Capacity constraint: The SOEMD currently includes two units consisting of nine employees, limiting capacity to conduct detailed monitoring of SOE financial performance.

### Conclusions and policy recommendations
- SOEs play a pivotal role in Tajikistan’s economy but are a consistent fiscal drain due to deteriorating financial positions, lack of transparency, and limited capacity to manage and monitor fiscal risks.
- Recommended reforms and actions:
  - Phase out and end quasi-fiscal activities of SOEs. SOEs that engage in quasi-fiscal activities directed by the government should be compensated; quasi-fiscal activities should be fully funded in the budget and disclosed in financial reports prepared by the government and SOEs.
  - Put a transparent framework in place to monitor, assess and disclose fiscal risks from contingent liabilities. The government indicated intention to establish a budgetary reserve fund as a buffer against contingent liabilities, but given revenue shortfalls there may not be enough budgetary resources to establish this fund.
  - Require all large SOEs to publish audited financial statements on an annual basis; require medium and small SOEs to publish summarized financial data annually.
  - Assess SOE viability and earmark nonviable SOEs for restructuring, privatization and winding-up.
  - Expand the Statement of Fiscal Risks included in the budget to include information on SOEs’ accumulated arrears to domestic banks to monitor banking-sector viability.
  - Strengthen the capacity of the SOE Monitoring Department (SOEMD) in the Ministry of Finance.

*Source: IMF staff calculations and analysis in the chapter on Fiscal Risks of SOEs from the provided IMF document.*

---


_Source: https://www.imf.org/-/media/files/publications/cr/2021/english/1tjkea2021005.pdf_