## BOSNIA AND HERZEGOVINA Selected Issues (content unit: _cr07269)

## Source details

**Canonical URL:** [BOSNIA AND HERZEGOVINA Selected Issues (content unit: _cr07269)](https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/scr/2007/_cr07269.pdf)

## Other formats

- [Markdown version](/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/scr/2007/_cr07269.pdf.md)
- [Structured JSON version](/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/scr/2007/_cr07269.pdf.json)

---

### Overview — Medium-term outlook and risks
- Staff macroeconomic projections: robust growth in 2006 expected to continue; medium-term baseline outlook described as stable.
- Fiscal and external positions: projected to be in a comfortable zone in the medium term despite recent spending pressures.
- Persistent structural weaknesses that could erode growth:
  - Weak policy coordination; absence of a comprehensive fiscal strategy; structural reform drift.
- Key cross-cutting risks:
  - External sustainability: persistently high current account deficits would raise net foreign liabilities (NFL) and increase vulnerability to external shocks.
  - Fiscal sustainability: growing primary deficits and increasing non-concessional borrowing could lead to explosive debt dynamics without appropriate rules and policy frameworks.
  - Labor market rigidities: high unemployment despite growth; supply-side problems include wage policies, heavy taxation of labor, and limited portability of pension and health insurance.
  - Limited internal economic integration: convergence between Entities but weak integration and mobility of capital and labor.
  - Financial stability: rapid credit growth; policy measures to curb credit growth ineffective; foreign-owned banks appear to run a higher risk of insolvency.

*Approved by the European Department; June 29, 2007.*

### External balance sheet and required current account adjustment
- Current position and context:
  - NFL at end-2006: about 50 percent of GDP (NFL reached about 50 percent of GDP at end-2006); staff estimate: 48.6 percent of GDP.
  - Gross external liabilities at end-2006: about 90 percent of GDP.
  - Gross official reserves comfortable; external public debt low and largely concessional; FDI increasing.
  - Staff baseline (no policy change): current account deficit remains high, implying a steady rise in NFL.
- Methodology:
  - CGER external sustainability approach (debt-dynamics accounting framework covering debt and equity).
  - Required current account adjustment estimated around 8 percentage points of GDP.
- Scenarios and timing:
  - Adjustment horizon: 10 years.
    - Early adjustment: starts in 2007 (2007–2017).
    - Delayed adjustment: starts at end-2012 (2012–2022).
- Main quantitative results:
  - Early adjustment: NFL stabilizes at around 56 percent of GDP by 2017.
  - Delayed adjustment: NFL reaches about 90 percent of GDP before stabilizing.
  - FDI inflows raise private sector foreign liabilities and future income outflows (profit repatriation) and are included in required adjustment estimates.
- Policy implications:
  - Early adjustment important and feasible if recent export trends persist.
  - Postponing reforms leads to much higher NFL and greater vulnerability.
  - Required adjustment (around 8 percentage points of GDP) implies substantial and sustained policy efforts over the adjustment horizon.

### External balance sheet — end‑2006 composition and 2007–12 projection
- End-2006 headline: gross external liabilities about 90 percent of GDP.
- End-2006 composition (percent of GDP):
  - Assets: Monetary authority 30.6; Commercial banks 13.1; Government 23.4; Portfolio 18.4; FDI 27.9; Total assets 43.7.
  - Liabilities: Commercial banks 22.6; Government (of which London Club) 1.4; Total liabilities 92.4.
- Comparative figures (selected):
  - NFL end-2006 (BiH staff estimate): 48.6 percent of GDP (average excl. BiH: 48.2).
- Medium-term projection (2007–12, baseline unchanged policies):
  - GDP growth: averaging 5.2 percent.
  - Current account deficit expected to remain around 13 percent of GDP.
  - Significant FDI inflows over next two years, abating thereafter.
  - Public debt-to-GDP: after an initial jump, projected to decline.
- Projected external balance sheet at end-2012 (percent of GDP):
  - Assets: Monetary authority 28.7; Commercial banks 13.6; Government 15.8; Portfolio 18.0; FDI 44.2; Total assets 42.3.
  - Liabilities: Commercial banks 25.6; Government (of which London Club 0.2; of which non concessional 3.0); Total liabilities 103.6.
- Projected NFL at end-2012: about 61 percent of GDP.

### Methodology and key assumptions (explicit)
- Framework: CGER (Lane and Milesi-Ferretti, 2006) decomposing debt and equity; debt-dynamics accounting to derive stabilizing current account.
- Core assumptions:
  - All debt instruments carry the Euro Libor as base interest rate.
  - External liabilities carry a sovereign risk premium of 200 basis points.
  - Inflation set at 2 percent in Bosnia & Herzegovina and in Europe.
  - Expected real returns on private equity = projected real GDP growth + 100 basis points.
  - Transfers and non-financial income assumed stable.
  - Assumed real GDP growth: 5.2 percent.
  - LIBOR Euro (WEO average over 2006-2011): 3.8.
  - FDI assumed at 3.3 percent of GDP during transition (set at end-of-medium-term projected level).
  - Net public debt accumulation set at 1.5 percent of GDP during transition.
  - Depreciation rate of capital: 5 percent.
  - Net external borrowing by commercial banks linked to GDP; if financing sources insufficient, international reserves decline.

### Main results — Early adjustment (2007–2017)
- Stabilized NFL: 56 percent of GDP by end-2017.
- Current account deficit needed to stabilize NFL in 2017: around 5.5 percent of GDP.
- Projected current account deficit in 2007: 13.5 percent of GDP.
- Implied required adjustment magnitude: about 8 percentage points of GDP.
- Projected external balance sheet at end-2017 (percent of GDP):
  - Assets: Monetary authority 46.4; Commercial banks 13.6; Government 23.3; Portfolio 18.0; FDI 49.3; Total assets 60.0.
  - Liabilities: Commercial banks 25.6; Total liabilities 116.1.

### Main results — Delayed adjustment (2012–2022)
- NFL would build to about 90 percent of GDP by end-2022.
- Net income payments increase to about 5.7 percent of GDP (compared with 4 percent of GDP at end-2012).
- Current account deficit required to stabilize NFL at this higher level: around 5.7 percent of GDP.
- Average projected current account 2007–12: 13.5 percent of GDP.
- Required adjustment magnitude: about 8 percent of GDP.
- Projected external balance sheet at end-2022 (percent of GDP):
  - Assets: Monetary authority 22.8; Commercial banks 13.6; Government 30.0; Portfolio 18.0; FDI 53.2; Total assets 36.4.
  - Liabilities: Commercial banks 25.6; Total liabilities 126.8.
- Vulnerabilities:
  - NFL ≈ 90 percent of GDP increases sensitivity to interest rates, liquidity conditions, and sentiment.
  - Government external debt projected about 40 percent of GDP.
  - Central bank reserves much lower, weakening capacity to sustain currency board.

### Conclusions and policy implications on external adjustment
- Core conclusion: stabilize NFL only by lowering the current account deficit.
- Estimated adjustment: about 8 percent of GDP reduction in the current account deficit required.
- Timing matters:
  - Early adjustment stabilizes NFL below 60 percent of GDP by 2017.
  - Delayed adjustment yields NFL near 90 percent of GDP, undermining resilience and currency board viability.
- Implementation considerations:
  - Adjustment must be gradual.
  - Persisting recent export trends would make early adjustment feasible, conditional on structural reforms to boost competitiveness and FDI attractiveness.
  - Ambitious infrastructure investments (e.g., corridor Vc road projects) would generate sizeable additional foreign liabilities and require an even larger current account adjustment.

### Fiscal sustainability and borrowing rules (summary)
- Past findings: growing primary deficits + higher share of non-concessional borrowing may lead to explosive debt dynamics long term.
- New debt laws: introduce borrowing rules limiting future debt service as percentage of last year’s revenue.
- Stochastic assessment (Chapter III) finds:
  - New borrowing rules would help maintain debt sustainability in the long term.
  - Implementing rules without a comprehensive policy framework and long planning horizon may cause inefficient expenditure compression and undermine credibility of borrowing limits.

### Extended-horizon DSA — key premise and purpose
- Extended-horizon DSA reveals longer-term vulnerabilities not shown in standard medium-term DSA.
- Over 5 years, public debt remains stable even with relatively large primary deficits due to high concessionality.
- Over 15–20 years, growing primary deficits and rising non-concessional borrowing can lead to explosive debt dynamics.
- Purpose: analyze impact of newly-introduced borrowing rules and compare implementations with 3-year and 5-year planning horizons.

### Borrowing rules — exact provisions (Box 1)
- Federation:
  - Debt service limit: 18 percent of previous year’s consolidated revenue.
  - Within this limit: maximum value for cantons: 5 percent of their previous year’s revenue.
  - Municipal limit: 3 percent of revenues for first two years after law, 5 percent for next two years, and 10 thereafter.
  - All limits cover debt guarantees.
- RS:
  - Debt service limit: 18 percent of previous year’s consolidated revenue.
  - Municipal limit: 18 percent.
- State:
  - State debt service cannot exceed 18 percent of previous year’s revenue for a period of three years from law effectiveness, subject to annual review.
  - State guarantees cannot exceed 30 percent of State revenue.
- Institutional framework: State Ministry of Finance prepares a five-year State Debt Management Strategy; Debt Committee composition specified.

### Fiscal pressures and financing assumptions (Box 2 — exact items)
- Revenues:
  - Gradual decline due to trade liberalization, slowdown in import and consumption growth, worsening demographics, and declining grants.
- Expenditures:
  - Increase with unification of Military and Police wages and worsening demographics affecting social security benefits.
- Domestic claims:
  - Add 18 percentage points of GDP to debt level in 2007; additional 10 and 5 percentage points in 2008 and 2009 respectively.
  - Average grant element: 25 percent.
- External and concessional financing:
  - IDA disbursements assumed to fall from ½ percentage point of GDP in 2007 to zero after 2011.
  - Commercial borrowing: 5-year maturity and interest rate LIBOR at 5 percent + 200 basis points.
- Domestic commercial borrowing: five-year maturity and interest rate equal to euro LIBOR simulated + 200 basis points.
- Privatization receipts:
  - 7½ percent of GDP in 2007 in the RS; 1½ percent in the Federation.
  - No further receipts assumed after 2007.

### Fiscal adjustment assumptions when limits bind
- Past adjustment (2003–2006) equivalent to 4½ percent of GDP: capital expenditures cut by 4 percent of GDP and wage bill cut by 1½ percent of GDP.
- When limits bind, assumed rules:
  - Future wage increases restricted to inflation.
  - Capital expenditures unchanged in nominal terms.
  - Pension increases indexed to inflation.
  - Transfers to households unchanged in nominal terms.
- Caveat: such adjustments follow past patterns and may not be optimal; cuts in investment can harm growth.

### Stochastic DSA results — main findings (exact reported outcomes)
- New borrowing rules prevent adverse debt dynamics under both 3-year and 5-year planning horizons.
- Debt initially increases but declines from 2009 in both Entities.
- Faster debt reduction in RS due to stronger 2007 fiscal position and substantial privatization receipts.
- Three-year horizon:
  - Initially relatively high deficits that recede slowly; high probability of exceeding borrowing limits.
  - Probability of exceeding limits increases to one in the Federation and to ¾ in the RS after 2015.
  - Despite breaches, assumed adjustments still sufficient for debt sustainability.
- Five-year horizon:
  - Initial fiscal relaxation less pronounced and reverses faster.
  - Fiscal position improves to close-to-balance by 2015 and beyond.
  - Probability of exceeding limits: slowly approaches ¾ in the Federation in 2019 then declines; negligible in the RS throughout.
- Policy implications:
  - Borrowing rules useful but full implementation difficult given capacity and political constraints.
  - Three-year horizon too short to fully assess future debt service impact.
  - MTEF not yet effective in guiding annual budget; "last-minute" spending initiatives overrule MTEF targets.
  - Staff recommends a balanced budget as medium-term fiscal anchor to avoid large fiscal swings.

### Labor market, cohesion, and financial sector risks — high-level findings
- Labor market:
  - Unemployment estimated between 20 and 30 percent of labor force; April 2006 Labor Force Survey estimate: 31.1 percent.
  - Youth unemployment (15–24): 62.3 percent (Total 31.1; 25–49 29.5; 50–64 16.4; 65+ 2.1).
  - Labor participation (April 2006): Total 43.1 (working population 15+); 51.3 (15-64).
  - Employment rate (April 2006): Total 29.7 (15+); 35.0 (15-64).
  - Long-term unemployment: 75 percent of unemployed have been without a job for 24+ months; share out of work for more than a year 86 percent.
- Supply-side contributors:
  - Remittances: private transfers/remittances estimated at 30 percent of GDP.
  - Education: university unemployment 12.5 percent; elementary education unemployment 31 percent.
  - Secondary enrollment: 73 percent of 15–18 year olds enrolled.
  - Skill deficiencies: World Bank reports 40 percent of students lack basic skills.
- Structural rigidities:
  - High minimum wages and centralized wage-setting constrain flexibility.
    - Federation minimum wage: 55 percent of the average wage in 2005 General Collective Agreement (about 51 percent of the 2006 average wage).
    - RS minimum wage: estimated 40 percent of average wage in 2006.
  - Labor tax wedge (average wage):
    - Federation: 34–41 percent.
    - RS: 34 percent.
    - At wage = one third of average in RS: labor tax wedge reaches 42 percent of total wage cost (up from 34 percent for average wage).
  - Limited portability of pension and health insurance; multiple pension/social insurance systems.
  - Limited regional mobility: only about 2 percent of respondents moved for employment; 3.6 percent likely to move in next 12 months.
- Economic cohesion:
  - Convergence between Entities observed (RS catching up with Federation) but little evidence of increased integration or mobility.
  - Cantonal disparities widened 2001–2005 across most indicators; standard deviation of cantonal performance increased.
- Financial sector:
  - Real private sector credit growth averaged 22.8 percent between 2001 and 2006.
  - Credit-to-GDP ratio more than doubled during same period.
  - Household credit now accounts for 47 percent of total credit to private sector.
  - Long-term credits represent 76 percent of total lending.
  - Gross foreign liabilities reached 25 percent of GDP by end-2006.
- Banking sector vulnerabilities:
  - FSIs show good health: high capital adequacy, improving profitability, falling NPAs, high liquidity.
  - Weaknesses in asset classification and provisioning may mask credit risks.
  - Foreign-owned banks rely on long-term foreign funding and, controlling for other factors, appear to run higher insolvency risk.
  - Policy measures to curb credit growth ineffective; tightening maturity matching contributed to increased foreign liabilities.
  - Empirical results: large foreign banks keep capital-to-asset ratios low; distance-to-default regressions show negative interaction between foreign ownership and size.

### Credit expansion — key numeric features and financing
- Household credit share: 47 percent of total private sector credit.
- Housing-related credit: about 25 percent of household credit.
- Main cumulative sources of asset growth (percentage points of GDP):
  - Private sector deposits: 18.9
  - Net foreign liabilities: 8.1
  - Net assets at the CBBH: 9.1
  - Net central government deposits: 2.1
  - Noncentral government deposits: 2.0
  - Claims on noncentral government: 0.1
  - Other items, net: 2.9
- Bank-related inflows averaged 2.7 percent of GDP from 2001 to 2006.
- Short-term deposits: 56 percent of total deposits at end-Q3 2006; short-term lending: 24 percent of total lending.

### Financial soundness indicators (selected, 2001–06)
- Capital (Tier 1 capital to RWA): 22.4, 18.2, 15.9, 14.9, 13.7, 13.6 (2001–2006).
- Asset quality (Nonperforming loans to total loans): 17.9, 11.0, 8.4, 6.1, 5.3, 4.0 (2001–2006).
- Profitability (Return on equity): -4.8, -2.0, 3.4, 5.8, 6.2, 8.5 (2001–2006).
- Liquidity (Liquid assets to total assets): 44.6, 33.8, 35.1, 35.7, 36.1, 35.9 (2001–2006).
- Foreign currency exposure (selected): Foreign currency and indexed loans to total loans: 64.9, 65.4, 68.8, 71.0 (selected years shown).

### Empirical methodology and key econometric findings
- Estimation method: system GMM (Arellano and Bover (1995); Blundell and Bond (1998)); Windmeijer correction applied to two-step robust standard errors.
- Credit growth regressions (number of observations: 226) — main outcomes:
  - Sounder banks tend to have higher credit growth; coefficients small and statistically insignificant.
  - Policy dummies (reserve requirement increases, maturity matching tightening) statistically insignificant for credit growth.
  - Real interest rate (t-1) positively and significantly associated with credit growth in reported regressions.
- Bank soundness (distance-to-default) regressions:
  - Distance-to-default shows strong persistence: Distance-to-default (t-1) highly significant.
  - NPA ratio (t-1) negative impact on distance-to-default in some specifications (e.g., NPA ratio (t-1) = -111.350 *).
  - Ownership effects: Foreign ownership * size interaction negative and significant (e.g., -0.454 **); EU parent * size negative and significant (e.g., -0.978 * to -0.953 *).
- Banks’ foreign liabilities regressions (observations: 289):
  - Elasticity of foreign liabilities to credit growth ≈ 0.3–0.7 (if credit grows 1 percent both last quarter and current, foreign liabilities rise 0.3–0.7 percent).
  - Policy tightening of maturity matching contributed to increased foreign liabilities (policy dummy coefficients positive and significant, e.g., 0.402 * and 0.461 *).
  - Macroeconomic drivers significant: interest rate (t-1) positive and significant (e.g., 0.775 ***, 0.717 ***); inflation (t-1) positive and significant (e.g., 7.998 **).
  - Macroeconomic implication: if credit growth slows to zero in two quarters, foreign liabilities would fall by around 7–16 percent; translating to a decline of €864 million in international reserves, or reserves from 5.1 months to 4.3 months of imports (end-2006 figures) under a one-to-one assumption.

### Prudential regulation — international comparisons and recommendations (Appendix III)
- Foreign exchange risk:
  - FSAP recommended comprehensive FX and market risk guidelines.
  - Supervisors should require capital for foreign exchange exposure.
  - Example international FX risk capital charges: Croatia 10 percent; Poland 8 percent; Romania 8 percent; Lebanon 12 percent (Table entries).
- Liquidity and maturity matching:
  - Bosnia’s maturity matching requirements described as strict by international practice.
  - Minimum liquidity ratio in Bosnia: average 10-day minimum liquidity up to at least 10 percent of short-term fund sources by book value.
  - International comparators: Croatia (35 percent for foreign currency), Latvia (30 percent for <30 days), Romania (minimum liquidity ratio measured cumulatively), Slovenia (100 percent for foreign currency up to 6 months), Chile (100 percent for up to 30 days assets).
- Policy recommendations:
  - Introduce comprehensive FX and market risk guidelines.
  - Require capital charges for FX exposure to deter weak banks from taking on new FX risk and to allocate capital between credit and market risks.
  - Review maturity matching and liquidity requirements to ensure consistency with country-specific conditions; consider that Bosnia’s framework is strict relative to many peers.

*Source: IMF staff chapter _cr07269 (provided content).*

### 2007. The views expressed in this document are those of the staff team and do not necessarily reflect

### BOSNIA AND HERZEGOVINA Selected Issues

### Overview — Medium-term outlook and risks
- Staff macroeconomic projections show Bosnia & Herzegovina’s medium-term baseline outlook is stable (see accompanying staff report); robust growth in 2006 is expected to continue in the coming years.
- Fiscal and external positions are projected to be in a comfortable zone in the medium term, even with recent spending pressures.
- Persistent structural weaknesses could erode growth momentum over the longer term: weak policy coordination, absence of a comprehensive fiscal strategy, and structural reform drift.
- Key cross-cutting risks identified:
  - External sustainability: maintaining persistently high current account deficits would raise net foreign liabilities (NFL) and increase vulnerability to external shocks.
  - Fiscal sustainability: growing primary deficits and increasing non-concessional borrowing could lead to explosive debt dynamics without appropriate rules and policy frameworks.
  - Labor market rigidities: high unemployment despite growth, with supply-side problems including wage policies, heavy taxation of labor, and limited portability of pension and health insurance.
  - Limited internal economic integration: convergence between Entities has occurred, but integration and mobility of capital and labor remain weak.
  - Financial stability: rapid credit growth consistent with regional trends; policy measures to curb credit growth have not been effective; foreign-owned banks appear to run a higher risk of insolvency.

### External balance sheet and required current account adjustment
- Context and current position:
  - At about 50 percent of GDP, Bosnia & Herzegovina’s net foreign liabilities today are close to the average of Eastern European countries (NFL reached about 50 percent of GDP at end-2006).
  - Gross official reserves are at a comfortable level, external public debt is low and largely on concessional terms, and foreign direct investment (FDI) has recently been increasing.
  - In the staff baseline (no policy change), the current account deficit remains high and is projected to stay at a level that would imply a steady rise in NFL over the medium term.
- Methodology:
  - The paper uses the external sustainability approach of the Consultative Group on Exchange Rate Issues (CGER) methodology for exchange rate assessment, a debt-dynamics accounting framework that takes into account all forms of foreign liabilities (both debt and equity).
  - The required current account adjustment estimated is around 8 percentage points of GDP.
- Scenarios and timing:
  - Adjustment is assumed to take place over a 10-year horizon. Two scenarios are considered:
    - Early adjustment: the adjustment starts in 2007.
    - Delayed adjustment: the adjustment starts only in the next decade.
- Main quantitative results:
  - Early adjustment scenario: NFL would stabilize at around 56 percent of GDP by 2017.
  - Delayed adjustment scenario: NFL would reach about 90 percent of GDP before it stabilizes.
  - FDI inflows, while not creating debt, raise private sector foreign liabilities and generate increasing income outflows in the form of profit repatriation; these future outflows are explicitly taken into account in estimating the required adjustment.
- Policy implications:
  - An early start to the required current account adjustment is important and feasible if recent export trends persist.
  - Postponing reforms and delaying adjustment would lead to much higher NFL and greater vulnerability to external shocks.
  - The size of the required adjustment (around 8 percentage points of GDP) implies substantial and sustained policy efforts over the adjustment horizon.

### Fiscal sustainability and borrowing rules (summary of appraisal)
- Previous staff analysis showed that growing primary deficits combined with an increasing share of non-concessional borrowing may lead to explosive debt dynamics over the longer term.
- Debt laws recently adopted by the State and Entities introduce borrowing rules limiting future debt service as a percentage of last year’s revenue.
- Stochastic assessment in Chapter III finds:
  - The new borrowing rules would help maintain debt sustainability in the long term.
  - Implementing borrowing rules without a comprehensive policy framework and a long planning horizon may result in inefficient expenditure compression and could undermine the credibility of the borrowing limits.

### Labor market, cohesion, and financial sector risks (high-level findings)
- Labor market:
  - Unemployment remains stubbornly high despite robust growth; data limitations complicate firm conclusions.
  - Several supply-side and structural factors likely contribute: wage policies, heavy taxation of labor, limited portability of pension and health insurance, and labor mobility constraints.
- Economic cohesion:
  - Strong evidence of economic convergence between the Entities, with the Republika Srpska (RS) catching up with the Federation.
  - Little evidence of increased economic integration between the two Entities or within the Federation; eliminating remaining barriers to capital and labor mobility is necessary.
- Banking and credit growth:
  - Financial deepening in Bosnia & Herzegovina has been in line with regional trends.
  - Policy measures to curb credit growth have not been effective, highlighting the importance of prudential supervision.
  - Empirical results indicate foreign-owned banks appear to run a higher risk of insolvency, underlining the need for cooperation with home-country supervisors.

*Approved by the European Department; June 29, 2007.*

### 10.      This Chapter provides an estimate of the current account adjustment required to

### _cr07269 - 10.      This Chapter provides an estimate of the current account adjustment required to

### Overview
- Purpose: Estimate the current account adjustment required to stabilize net foreign liabilities (NFL) using the CGER external sustainability/debt-dynamics accounting framework that accounts for all forms of foreign liabilities (debt and equity).
- Adjustment drivers: outstanding stock of accumulated liabilities, output growth, and expected returns on external assets and liabilities.
- Three analytical steps:
  - Estimate composition of the external balance sheet at end-2006.
  - Project the external balance sheet at end-2012 in line with staff medium-term baseline.
  - Use debt dynamics accounting to estimate the current account adjustment required to stabilize NFL.
- Scenarios considered:
  - Early adjustment: adjustment starts immediately and spans a 10-year horizon (2007–2017).
  - Delayed adjustment: adjustment starts at end-2012 and spans a 10-year horizon (2012–2022), allowing continued build-up of liabilities before stabilization.

### The external balance sheet of Bosnia & Herzegovina (end-2006, comparisons, and medium-term projection)
- End-2006 headline: gross external liabilities about 90 percent of GDP (Table 1).
- End-2006 composition (in percent of GDP, Table 1):
  - Assets: Monetary authority 30.6; Commercial banks 13.1; Government 23.4; Portfolio 18.4; FDI 27.9; Total assets 43.7.
  - Liabilities: Commercial banks 22.6; Government (of which London Club) 1.4; Total liabilities 92.4.
- Observations:
  - Debt assets about 44 percent of GDP, two-thirds held by the monetary authority; average in the region is 3 percentage points higher.
  - Equity liabilities estimated at 46 percent of GDP, 4 percentage points higher than the region’s average.
  - NFL at end-2006: 48.6 percent of GDP (compared with average 48.2 for other Eastern European countries).
- Comparative composition of net external position (Table 2, percent of GDP, selected entries):
  - Estonia: Debt Assets 59.6; Equity Assets 16.3; Debt Liabilities 70.7; Equity Liabilities 104.9; NFL 99.7.
  - Hungary: Debt Assets 35.4; Equity Assets 6.7; Debt Liabilities 64.9; Equity Liabilities 74.0; NFL 96.8.
  - BiH (end-2006, staff estimates): Debt Assets 43.7; Equity Assets 0.0; Debt Liabilities 46.0; Equity Liabilities 46.3; NFL 48.6.
  - Average (excl. BiH): Debt Assets 47.3; Equity Assets 3.9; Debt Liabilities 56.8; Equity Liabilities 42.6; NFL 48.2.
- Medium-term projection (2007–12, baseline unchanged policies):
  - GDP growth: averaging 5.2 percent.
  - Current account deficit expected to remain around 13 percent of GDP.
  - Significant FDI inflows expected over the next two years, abating thereafter.
  - Public debt-to-GDP ratio: after an initial jump, projected to decline over the period.
- Projected external balance sheet at end-2012 (Table 3, in percent of GDP):
  - Assets: Monetary authority 28.7; Commercial banks 13.6; Government 15.8; Portfolio 18.0; FDI 44.2; Total assets 42.3.
  - Liabilities: Monetary authority (not repeated); Commercial banks 25.6; Government (of which London Club 0.2; of which non concessional 3.0); Total liabilities 103.6.
- Projected NFL at end-2012: about 61 percent of GDP (driven by FDI flows, increased banks’ borrowing, and declining central bank foreign assets).

### Methodology and key assumptions
- Analytical framework: CGER approach (Lane and Milesi-Ferretti, 2006) linking net external position, current account, trade balance, and rates of return on the external portfolio; decomposes assets and liabilities into debt and equity components and uses debt-dynamics accounting to derive stabilizing current account.
- Core equations and logic:
  - Change in net foreign asset position B: B_t − B_{t-1} = CA_t + KG_t + E_t (with definitions in text).
  - Expressed relative to GDP and assuming zero capital gains leads to equations relating current account needed to stabilize b to nominal GDP growth, inflation, and transfers/errors.
  - Expanded to decompose assets A and liabilities L and their nominal and real returns; further decomposed into debt (D) and equity (EQ) components to yield equation (5) for stabilization calculation.
- Main parameter and modeling assumptions (explicit):
  - All debt instruments carry the Euro Libor as the base interest rate.
  - External liabilities assumed to carry a sovereign risk premium of 200 basis points.
  - Inflation set at 2 percent, both in Bosnia & Herzegovina and in Europe.
  - Expected real returns on private equity assumed equal to projected real GDP growth plus a constant spread of 100 basis points over the growth rate.
  - Transfers and non-financial income assumed to remain stable.
  - Assumed real GDP growth: 5.2 percent (Table 4).
  - LIBOR Euro (WEO average over 2006-2011): 3.8 (from Table 4 presentation).
  - FDI assumed at 3.3 percent of GDP during transition (set at end-of-medium-term projected level).
  - Net public debt accumulation set at 1.5 percent of GDP during transition.
  - Net external borrowing by commercial banks linked to GDP; if financing sources insufficient, international reserves decline.
  - Depreciation rate of capital assumed at 5 percent.
- Scenario timing:
  - Early adjustment: transitional path beginning in 2007, stabilizing NFL in 2017 (10-year span).
  - Delayed adjustment: transitional path beginning in 2012, stabilizing NFL at end-2022 (10-year span); stabilizing current account recalculated recursively as liabilities build up.

### Main results (quantitative outcomes and trajectories)
- Early adjustment scenario (stabilize by end-2017):
  - Stabilized NFL: 56 percent of GDP by end-2017 (Table 5).
  - Current account deficit needed to stabilize NFL in 2017: around 5.5 percent of GDP.
  - Projected current account deficit in 2007: 13.5 percent of GDP.
  - Implied required adjustment magnitude: about 8 percentage points of GDP.
  - Projected external balance sheet at end-2017 (Table 5, percent of GDP):
    - Assets: Monetary authority 46.4; Commercial banks 13.6; Government 23.3; Portfolio 18.0; FDI 49.3; Total assets 60.0.
    - Liabilities: Commercial banks 25.6; Total liabilities 116.1.
- Delayed adjustment scenario (stabilize by end-2022):
  - NFL would build up to about 90 percent of GDP by end-2022 (Table 6).
  - Net income payments increase to about 5.7 percent of GDP (compared with 4 percent of GDP at end-2012).
  - Current account deficit required to stabilize NFL at this higher level: around 5.7 percent of GDP.
  - Average projected current account 2007–12: 13.5 percent of GDP.
  - Required adjustment magnitude again about 8 percent of GDP.
  - Projected external balance sheet at end-2022 (Table 6, percent of GDP):
    - Assets: Monetary authority 22.8; Commercial banks 13.6; Government 30.0; Portfolio 18.0; FDI 53.2; Total assets 36.4.
    - Liabilities: Commercial banks 25.6; Total liabilities 126.8.
  - Vulnerabilities of delayed adjustment:
    - NFL as high as 90 percent of GDP increases vulnerability to shocks in interest rates, liquidity conditions, and market sentiment.
    - Government external debt projected about 40 percent of GDP.
    - Central bank reserves would be much lower than at onset, weakening capacity to sustain currency board.

### Conclusions and policy implications
- Medium-term baseline: external position remains sound with large inflows supporting investment and growth, but NFL continue to grow.
- Core conclusion: The only way to stabilize net external liabilities is to lower the current account deficit.
- Estimated adjustment magnitude: about 8 percent of GDP reduction in the current account deficit required to stabilize NFL.
- Timing matters:
  - Early adjustment (starting now) would stabilize NFL below 60 percent of GDP by 2017 — a relatively low level by international standards.
  - Delayed adjustment results in very high NFL (about 90 percent of GDP), leaving the country vulnerable to external shocks and undermining the viability of the currency board.
- Implementation considerations:
  - Adjustment magnitude must be achieved gradually.
  - Persisting recent export trends would make early adjustment feasible, conditional on further structural reforms to increase competitiveness and attractiveness to foreign investors.
  - Ambitious infrastructure investments (e.g., corridor Vc road projects) would generate sizeable additional foreign liabilities (debt or equity) and would require an even larger current account adjustment.

*Source: IMF staff chapter _cr07269 (provided content).*

### 43.      An extended-horizon debt sustainability analysis (DSA) (see for example Chapter

### _cr07269 - 43.      An extended-horizon debt sustainability analysis (DSA) (see for example Chapter

### Extended-horizon DSA: key premise
- An extended-horizon debt sustainability analysis (DSA) shows longer-term vulnerabilities not revealed in a standard medium-term debt sustainability analysis.
- Over a 5-year horizon, public debt would remain stable even with relatively large primary deficits, due to high concessionality of the outstanding debt.
- Growing primary deficits and the increasing share of non-concessional borrowing would lead to explosive debt dynamics over the longer term (15-20 years).

### Purpose of the paper
- Analyze the impact of newly-introduced borrowing rules on longer-term debt dynamics.
- Compare two possible implementations of the rules given lack of a precise implementation framework:
  - a three-year planning horizon (consistent with the current MTEF)
  - a five-year planning horizon (stipulated in the State Debt Law)
- Assume scope for adjustments to comply with limits is limited—aligned with past adjustment efforts—to slowing down certain categories of expenditures.
- The analysis does not specify the exact mechanisms for how adjustments are undertaken when limits are expected to be breached.

### Methodology (stochastic framework)
- Use a stochastic method to evaluate risks to debt dynamics from macroeconomic uncertainty, simulating future paths for macroeconomic and fiscal variables and summarizing risks with frequency distributions of debt-to-GDP ratio paths (‘fan charts’).
- Steps in stochastic projections:
  1. For a given year, simulate macroeconomic variables using a simplified model of the Bosnian economy with stochastic shocks.
  2. Construct ‘unadjusted’ fiscal projections assuming revenues and expenditures grow in line with GDP or respective tax bases, accounting for known fiscal pressures and financing assumptions.
  3. Extend projection either 3 or 5 years ahead using a non-stochastic version of the model to check if borrowing rules’ limit is exceeded.
  4. If the limit is exceeded, reduce the current year’s budget by slowing expenditures while still accommodating specified fiscal pressures.
- Projections are constructed for every year between 2008 and 2020 and the exercise is repeated to build a random sample of possible outcomes; each outcome yields a debt path and frequency distributions of debt-to-GDP.

### Borrowing rules (Box 1) — exact provisions
- Federation:
  - Debt service limit in any future year: 18 percent of the previous year’s consolidated revenue.
  - Within this limit, maximum value for cantons: 5 percent of their previous year’s revenue.
  - Municipal limit initially: 3 percent of revenues for the first two years after the law is passed, increases to 5 percent for the two following years, and to 10 thereafter.
  - All limits cover debt guarantees.
- RS:
  - Debt service limit: 18 percent of the previous year’s consolidated revenue.
  - Municipal limit: 18 percent.
- State:
  - State debt service cannot exceed 18 percent of the previous year’s revenue for a period of three years from the day the law becomes effective, subject to annual review.
  - State guarantees cannot exceed 30 percent of State revenue.
- Institutional framework highlights:
  - State Ministry of Finance prepares a five-year State Debt Management Strategy in cooperation with Debt Committee.
  - Debt Committee composition: two representatives of the Council of Ministers (including a Minister), a representative of the Central Bank, two representatives of the Federation (including the Minister of Finance), two representatives of the RS (including the Minister of Finance), and the Director of Revenue Administration of the Brcko District.
  - Federation Debt Committee: Federal and cantonal Ministers of Finance.
  - RS municipalities: must request agreement from the Ministry of Finance to borrow in first three years after law adoption; after transition, agreements required for foreign currency borrowing or if projected debt service exceeds 10 percent.

### Fiscal pressures and financing assumptions (Box 2) — exact items
- Revenues:
  - Gradual decline due to trade liberalization and slowdown in import and consumption growth, worsening demographics, and declining grants.
- Expenditures:
  - Increase with the unification of Military and Police wages and worsening demographics affecting social security benefits.
- Domestic claims:
  - Add 18 percentage points of GDP to the debt level in 2007, and an additional 10 and 5 percentage points in 2008 and 2009 respectively.
  - The average grant element is 25 percent.
- External and concessional financing:
  - IDA disbursements assumed to fall from ½ percentage point of GDP in 2007 to zero after 2011.
  - Grants decline gradually.
  - Commercial borrowing projected maturity: five years and interest rate at LIBOR at 5 percent + 200 basis points.
- Domestic borrowing on commercial terms:
  - Five year maturity and interest rate equal to euro LIBOR simulated from the model + 200 basis points.
- Privatization receipts:
  - 7½ percent of GDP in 2007 in the RS; 1½ percent in the Federation.
  - No further receipts assumed after 2007.

### Fiscal adjustment assumptions when limits bind
- Past adjustment (2003–2006) equivalent to 4½ percent of GDP; reductions concentrated in capital expenditures (4 percent of GDP) and wage bill (1½ percent of GDP).
- When borrowing limit binding, assume:
  - Future wage increases restricted to the rate of inflation.
  - Capital expenditures unchanged in nominal terms.
  - Pension increases indexed to inflation.
  - Transfers to households unchanged in nominal terms.
- Note: These assumed adjustments follow past patterns and may not be optimal; cuts in investment spending may harm growth.

### Representation of uncertainty and outcomes
- Debt dynamics summarized by ‘fan charts’ with deciles; black zone represents a 20 percent confidence interval; progressively lighter grey zones show 40, 60 and 80 percent intervals.
- If the upper bound of the 80 percent interval is above the initial debt ratio for some years, it may be interpreted as an over 20 percent probability that the debt-to-GDP will increase above the initial level in those years.
- Probabilities of exceeding borrowing limits are derived as frequency of exceedance across simulated sample.

### Results — main findings (exact reported outcomes)
- The new borrowing rules would prevent adverse debt dynamics under both three-year and five-year planning horizons.
- Debt would initially increase but return firmly onto a declining path from 2009 in both Entities.
- Debt reduction faster in the RS where:
  - The initial 2007 fiscal position is stronger.
  - Substantial privatization receipts reduce borrowing needs.
- Three-year planning horizon:
  - Initially produces relatively high deficits which recede slowly, leading to a high probability of exceeding borrowing rules’ limits.
  - Governments do not fully anticipate impact of new borrowing on future debt service and increase expenditures.
  - Consolidated budget position improves slowly and approaches close-to-balance only at end of simulation period.
  - Probability of exceeding limits increases to one in the Federation and to ¾ in the RS after 2015.
  - Despite breaches, assumed adjustment effort still sufficient to ensure debt sustainability.
- Five-year planning horizon:
  - Also generates initial fiscal relaxation but less pronounced and reverses faster to balance.
  - Extended planning horizon limits initial expenditure increases and allows stronger fiscal position.
  - Fiscal position improves from a small deficit at beginning to close-to-balance in 2015 and beyond.
  - Probability of exceeding limits:
    - Slowly approaches ¾ in the Federation in 2019 and then starts declining.
    - Negligible throughout simulation period in the RS.
  - Lower initial deficits and faster adjustment reflect better anticipation of consequences of increased expenditures on future debt service costs.

### Conclusions and policy implications (exact recommendations and caveats)
- Borrowing rules are a useful tool to ensure sustainability, but full implementation may be difficult due to capacity and political constraints.
- The simplified rules can stop adverse debt dynamics, but weak planning may hinder implementation in Bosnia & Herzegovina.
- The three-year horizon (consistent with current MTEF) is too short to fully assess the impact of new borrowing on future debt service costs.
- MTEF does not yet guide the annual budget process effectively; "last-minute" spending initiatives often overrule MTEF targets, indicating weak ownership by policymakers.
- In absence of a longer-term expenditure plan, ad-hoc adjustments to meet rules will likely conflict with country needs and spending priorities.
- A well-planned adjustment aligned with World Bank Public Expenditure Review recommendations would be more efficient and reduce risks of unsustainable debt dynamics through deeper and more durable adjustments.
- Staff recommendation: a balanced budget as a medium-term fiscal anchor avoids large fiscal swings generated by simplified rules; maintaining a balanced-budget position would help avoid politically difficult subsequent adjustments.
- Methodological caveat: stochastic methodology better assesses risks than deterministic approach, but results should be interpreted with caution due to uncertainty in model structure, estimated parameters, and future policy assumptions; these factors are not reflected in the reported ‘fan charts,’ so actual uncertainty is likely larger.

### Annex — projection model (selected exact model specifications and projections)
- GDP growth:
  - Cobb-Douglas with labor and capital coefficients 0.7 and 0.3 respectively.
  - Investment rate fixed at 20 percent of GDP.
  - Capital depreciation at 5 percent annually.
  - Initial capital stock (in 2000) at 150 percent of GDP.
  - Labor grows with projected working-age population; average productivity growth 1.4 percent per year (average for 2000-05).
- Net exports:
  - Metal exports expand till 2012 then flat.
  - Imports of petroleum products elasticity to real KM oil prices: -0.2.
- External environment:
  - VAR with one lag estimated on 1980-2006 sample; IDA disbursement and grant assumptions noted in Box 2.
- Exchange rate and prices:
  - Real exchange rate vis-à-vis the euro is constant; CPI and import prices follow euro-area level.
- Shocks:
  - Stochastic shocks to external variables and domestic productivity; shocks assumed normally distributed with covariance matrix from residuals.
- Model projections (exact phrasing):
  - GDP growth gradually decelerates from nearly 6 percent in 2008 to around 3½ percent in 2040.
  - Net export improves from projected 2007 level of -31 percent of GDP to -22 percent of GDP in 2040.
  - Reduction in VAT base generates a declining revenue ratio.

*Source: _cr07269.*

### References

### _cr07269 - References

### References
- Celasun, O., X. Debrun and J. Ostry, 2006, "Primary Surplus Behavior and Risks to Fiscal Sustainability in Emerging Market Countries: A 'Fan-Chart' Approach," IMF Working Paper No. 06/67 (Washington: International Monetary Fund).
- Garcia, M. and R. Rigobon, 2004, "A Risk Management Approach to Emerging Market's Sovereign Debt Sustainability with an Application to Brazilian Data," NBER Working Paper No. W10336 (Cambridge: National Bureau of Economic Research).
- Samake, I. and E. Tanner, 2007, "Probabilistic Sustainability of Public Debt: A Vector Autoregression Approach for Brazil, Mexico, and Turkey," IMF Working Paper No. 06/295, (Washington: International Monetary Fund).

### IV. UNEMPLOYMENT AND LABOR MARKET — overview
- A flexible, well-functioning labor market is essential. High and persistent unemployment in Bosnia & Herzegovina—estimated between 20 and 30 percent of labor force—is a source of concern because it represents a waste of economic resources and suggests underlying rigidities that may complicate adjustment.
- Limited and poor quality statistics, large informal sector, and differing definitions complicate definitive assessment.

### A. Introduction
- Chapter scope: overview of unemployment facts; international comparison with other transition economies; exploration of alternative explanations (demand, supply, structural); policy considerations.
- Prepared by Milan Cuc.

### B. Basic facts — unemployment, employment, and labor participation
- Unemployment has remained high despite cumulative GDP increase of almost 30 percent between 2000–05.
- Estimates of unemployment rate vary substantially, ranging from 22.3 percent to 33.5 percent in 2004; April 2006 Labor Force Survey estimate: 31.1 percent.
- Informal economy estimates between 20 and 40 percent.
- Assumed unemployment rate for discussion: in the 20–30 percent range.
- Between 2001 and 2004 unemployment rate increased from 16 percent to the 22–24 percent indicated by sources using a standardized methodology.
- Increase in unemployment occurred despite robust GDP growth and rising labor participation and employment rates — employment growth was outstripped by increases in labor force.

Key tables (selected numeric entries preserved):
- Labor Force Survey, April 2006: unemployment rate 31.1 (working population 15+); 31.8 (working population 15-64).
- LSMS, 2001–04: "self-reported" unemployment rate up to 38.8, 37.0, 32.4, 33.5 (years shown); ILO definition World Bank, December 2005: 15.9, 21.4, ..., 22.0 (selected years).
- Participation rate (Labor Force Survey, April 2006): 43.1 (working population 15+); 51.3 (working population 15-64).
- Employment rate (Labor Force Survey, April 2006): 29.7 (working population 15+); 35.0 (working population 15-64).

### C. International comparisons
- Bosnia & Herzegovina mirrors experience of other transition economies: initial contraction in output/employment, later growth with lagged employment recovery.
- Growth 2000–05: 5.0 percent (BiH) versus 4.9 percent for CEE.
- In 2005, GDP was still below its 1989 level.
- Labor market outcomes close to other Balkan countries; outcomes lag CEE countries.

### The characteristics of unemployment
- Main features: high youth unemployment, high long-term unemployment, persistent regional differences.
- Youth (15–24) unemployment: 62.3 percent (table shows Total 31.1; 15–24 62.3; 25–49 29.5; 50–64 16.4; 65+ 2.1).
- Participation rates by age group: Total 43.4; 15–24 33.4; 25–49 66; 50–64 36.6; 65+ 5.6.
- Employment rates by age group: Total 29.7; 15–24 12.6; 25–49 46.5; 50–64 30.6; 65+ 35.0.
- Education and unemployment (2006): unemployment rate for university education 12.5 percent; for elementary education 31 percent.
- Secondary enrollment: only 73 percent of 15-18 year olds enrolled.
- Long-term unemployment: 75 percent of the unemployed have been without a job for two years or more; share out of work for more than a year is 86 percent (compare 40–60 percent for transition economies).
- Unemployment by duration (Labor Force Survey, 2006): months 0: 0.7 percent; 0 months means currently unemployed have found a job. 0–5 months 5.6 percent; 6–11 months 7.8 percent; 12–23 months 10.8 percent; 24+ months 75.1 percent.
- Regional persistence: cantonal unemployment rates in 2005 closely mirror rates in 2003 and 2001 — limited mobility across cantons.

### C. Explaining Unemployment — framework
- Three alternative explanations considered: deficient demand, supply-side problems, structural unemployment.

Deficient demand (key findings)
- Bosnia & Herzegovina has exhibited strong sustained growth: 5.0 percent in 2000–05 (4.9 percent for CEE).
- HP filter estimates suggest an output gap is unlikely to exist.
- Widening trade deficit: almost doubled from $2.8 to $5.0 billion between 2000 and 2005.
- Wage growth 2000–06: close to 8 percent in nominal terms and 6 percent in real terms annually.
- Conclusion: deficient demand unlikely to be the primary cause; expansionary macroeconomic policies could be ineffective and harm external competitiveness.

Supply-side factors (key points)
- Two supply-side channels:
  - High reservation wages: unemployed may refuse private-sector jobs because alternative incomes (remittances, work abroad prospects) or preference for public-sector jobs raise reservation wage above market wages.
  - Skill mismatch: workers may lack skills demanded by private sector.
- About one quarter of Bosnia & Herzegovina’s population estimated to live abroad.
- Private transfers/remittances estimated at 30 percent of GDP.
- Public sector size and wages:
  - Public sector large and average public wages particularly high.
  - Public administration wages (percent of average national wage): Bosnia and Herzegovina: 132; Federation: 148; Republika Srpska: 123; comparisons: Bulgaria 106; Croatia 113; Czech Republic 130; Hungary 172; Romania 128; Slovak Republic 118; Slovenia (data shown).
  - Education sector: Bosnia and Herzegovina 102; Federation 88; Republika Srpska ...; comparisons include 97, 91, 112, 116, 82, 123.
  - Health and social services: Bosnia and Herzegovina 114; Federation 102; Republika Srpska ...; comparisons include 117, 94, 89, 93, 79, 112.
- Government wage bill (percent of GDP) shows Bosnia & Herzegovina as an apparent outlier among 25 emerging markets (time series 1990–2006 referenced).

High reservation wage (detail)
- Reservation wage depends on alternative income (unemployment benefits, remittances), valuation of leisure, and other factors.
- Lack of job vacancy data prevents direct measurement of workers forgoing available employment.
- Large emigrant population and remittances (30 percent of GDP) may discourage work effort and raise reservation wages.
- Public employment and high public wages can raise reservation wages and reduce attractiveness of private-sector employment; literature example: Demekas and Kontolemis (1999) for Greece.

Skill deficiencies
- Low quality of education: World Bank reports 40 percent of students lack basic skills and knowledge.
- Many vocational students receive inadequate general education; low secondary and tertiary enrollment rates are a serious problem.
- Example: 2003/04 survey: at least 40 percent of eight-grade students failed to obtain a satisfactory score on the mathematics assessment (World Bank (2006)).
- Youth (15–19) lack work experience, limiting job possibilities.

Demographics
- Relatively young population may cause "cohort crowding" where larger youth cohorts face reduced job opportunities in presence of wage rigidities and imperfect substitutability between workers of different ages.

*Source: _cr07269 - References*

### 81.      In addition to demand or supply deficiencies, unemployment may also be caused

### _cr07269 - 81.      In addition to demand or supply deficiencies, unemployment may also be caused

### Structural rigidities and unemployment: overview
- Structural transformation and sectoral shifts require commensurate shifts in employment across sectors and regions; reallocation takes time and can generate persistent unemployment.
- Indicators of serious labor‑market shortcomings in Bosnia & Herzegovina:
  - Existence of a widespread informal sector.
  - Reported unemployment rates are twice the level of its level estimated from household surveys.
  - Informal employment functions as a coping mechanism for employers and job seekers unable to find formal jobs.
- Informal economy signals institutional barriers preventing normal functioning of product and labor markets.
- Possible structural contributors to unemployment: low wage flexibility; limited labor mobility within the country.

### Low wage flexibility: key factors
- With inflexible wages, unemployment adjusts slowly; wages remain relatively high instead of falling to restore employment.
- Contributing factors to wage rigidity identified:
  - High minimum wages.
  - Centralized wage‑setting process.
  - Large labor tax wedge.

#### High minimum wage
- In the Federation:
  - Minimum wages are set at relatively high levels and are binding for all regions within the entity.
  - The 2005 General Collective Agreement in the Federation stipulated a minimum wage of 55 percent of the average wage (about 51 percent of the 2006 average wage).
  - Some Branch Collective Agreements set minimum wages even higher (telecommunication and postal service sector: 100–110 percent of the average wage in the Federation).
  - High minimum wages constrain regional wage responses to local shocks and disproportionately affect youth unemployment.
- In the RS:
  - Minimum wage estimated at 40 percent of average wage in 2006, which does not appear binding and is in line with the 30–40 percent range observed in other CEE countries.

#### Centralized wage‑setting process
- General collective agreements specify wages in detailed and rigid ways, reflecting job complexity and education level.
- These stipulations are binding mainly for state‑owned and mass‑privatized sectors; generally not respected in the new private sector.

#### Labor tax wedge
- Labor tax wedge defined as the difference between the cost to the firm of employing a worker and the worker’s take‑home pay; includes labor income and payroll taxes and is expressed as a ratio of these payments divided by total labor cost.
- Calculated at the average wage:
  - 34–41 percent in the Federation.
  - 34 percent in RS.
- High labor tax wedge discourages firms from hiring and workers from seeking employment; distortion may be strong in an open economy with labor migration.
- In the RS, a legal provision sets the basis for social contributions at not less than 50 percent of average net salary in the RS in the preceding month:
  - Example: at a wage equal to one third of the average, the labor tax wedge reaches 42 percent of total wage cost (up from 34 percent for the average wage).
  - This discourages formal employment of low‑wage workers.

### Limited regional mobility
- Limited spatial mobility weakens the equalization of regional unemployment differences.
- Factors contributing to low mobility:
  - Shortage of adequate housing.
  - Regional differences in physical infrastructure endowments.
  - Apparent reluctance of households to leave their communities.
  - Multitude of pension and social insurance systems at entity and cantonal levels and limited portability of these plans.
- 2004 household survey findings:
  - Only about 2 percent of respondents indicated employment was the reason for their move to their current residence.
  - Only 3.6 percent indicated it was very likely or quite likely that they would move in the upcoming 12 months.

### Policy implications and recommendations
- Diagnostic caution:
  - Analysis does not allow firm conclusions on the nature of unemployment or a single best policy prescription; avoid quick fixes.
  - Stimulating aggregate demand is likely ineffective currently and could raise wages, undermine external competitiveness, and exacerbate macroeconomic imbalances.
- Preferred emphasis:
  - Strengthen supply capacity by improving the business environment and attracting private investment to raise sustainable growth and absorb excess labor.
  - Recognize that some unemployment stemming from structural and supply‑side factors will decline only over time; begin corrective steps early.

- Education and training:
  - Education and training policies are key to reducing skill mismatches.
  - Early education reform is essential to improve education quality and reduce waste in education spending by streamlining and refocusing expenditure on priority areas.

- Public wage and employment policy:
  - Public sector is a sizable employer; public sector wages and nonmonetary benefits may be excessive relative to productivity.
  - Public sector should promote wage moderation through public wage and benefits policy.

- Privatization:
  - Accelerating privatization could reduce distorting influence of public wage and employment policies, harden budget constraints for privatized firms, cut transfers to cover losses, limit excessive wage settlements, and reduce the size of the public sector.

- Labor tax wedge reforms:
  - Explore cutting social insurance contribution rates to encourage formal sector employment growth and reduce incentives for tax evasion.
  - To avoid financing shortfalls in social funds, supplement revenues with transfers from the budget (for example, using part of the VAT proceeds).

- Pension and social insurance portability:
  - Prioritize harmonizing provisions of pension and social insurance systems and increasing portability to promote labor market flexibility and reduce disincentives to regional mobility.

*Source: https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/scr/2007/_cr07269.pdf*

### 103.     The apparent economic convergence between the Entities stands in contrast to

### _cr07269 - 103.     The apparent economic convergence between the Entities stands in contrast to

### Convergence within the Entities and Cantons
- Available data suggest there has been very little convergence among the cantons.
- Economic disparities among the cantons appear to have grown across a range of indicators.
- The gap between the best and the worst performing canton increased between 2001 and 2005 across almost all indicators.
- There was a generalized increase in the differences between economic performance of the cantons over this period, evidenced by an increase in the standard deviation (or variance) of cantonal economic performance relative to the average cantonal performance.
- Exception: differences in the purchasing power of an average wage have narrowed slightly.  
  - This is measured as the cantonal net wage relative to the average price of the basket of necessary goods in the same canton.
  - This narrowing is perhaps in part explained by efforts to equalize public sector pay across lower levels of government in the Federation.

### Evidence of Economic Integration
- Core premise: economic integration depends on the mobility of factors of production within an economic space.
- Available evidence suggests the Entities are not economically integrated. The noted economic convergence between Entities reflects the relative strengthening of the RS economy alongside minimal interaction between the Entities.

Subtheme — Labor mobility
- There are no direct measures of labor mobility available in Bosnia & Herzegovina; inferences are drawn from unemployment data.
- Evidence of increasing differences in registered unemployment rates between the Entities points to low labor mobility.
  - Registered unemployment rates appear to be much lower in the Republika Srpska than in the Federation.
  - Trends in registered unemployment in the Federation and the Republika Srpska have been moving in opposite directions.
  - While some divergence can be expected because of differences in the pace and nature of structural reforms, these differences would be expected to narrow over time as workers find employment in the Entity with less excess labor supply. There appears little evidence, to date, that this has occurred in Bosnia & Herzegovina.
- Evidence of persistent differences in unemployment rates across cantons is also consistent with low labor mobility.
  - Cantons with relatively higher unemployment in 2003 tend to be the same cantons with higher unemployment in 2005.
  - This pattern also appears to hold over the longer period from 2001 to 2005.
- Note: The 2006 Labor Force Survey also shows large differences in unemployment rates between the Entities: youth unemployment was 52.4 percent in the RS and 66.7 percent in the Federation; and male youth unemployment was 48.7 percent in the RS and 65.2 percent in the Federation.

Subtheme — Capital mobility
- There has been a steady increase in the number of companies with headquarters in the Federation or the RS that have established branch offices in the other Entity.
- Two possible interpretations:
  - Capital mobility is increasing; or
  - Cross-border branch activity reflects contortions to overcome legal and institutional barriers when conducting inter-Entity commerce.
- Available data do not allow distinguishing between these explanations; anecdotal evidence suggests the latter should not be discounted.
  - Example anecdote: “without a physical presence in both Entities it is more difficult for a company to conduct inter-Entity business than it is to undertake trade with a foreign country.”

### Institutional and Legislative Barriers to Integration
- Progress achieved:
  - Single currency accepted throughout the country, facilitating inter-Entity transactions.
  - Harmonized customs rules and duties applied at all points of entry.
  - Country-wide VAT has replaced an unharmonized domestic sales tax, facilitating price transparency.
  - Modern and harmonized bankruptcy laws have been adopted.
  - Bank deposit holders are insured by a single and country-wide deposit insurance agency.
  - Establishment of a country-wide banking system has aided efficient investment of domestic savings regardless of geographic origin.
- Significant remaining barriers:
  - Little portability of pension rights and health care insurance.
  - Capital mobility affected by varying and non-transparent property rights.
  - Costs and procedures of doing business differ across the Entities because of varying degrees of red tape.
  - No single banking supervisor despite deeper financial intermediation.
  - A host of industry-level legislation and labor regulations remain Entity specific.

### Conclusions on Internal Cohesion and Policy Needs
- The RS appears to be catching-up in economic terms with the Federation; this may reflect implementation of delayed reforms in the RS.
- The RS is reforming faster and has fostered competitive advantages over the Federation that may lead it to pull ahead in coming years.
- The apparent lack of economic integration between the Entities underscores the prospect of divergence.
- Promoting internal economic cohesion will require:
  - Accelerating structural reforms in the Federation.
  - Eliminating remaining barriers to capital and labor mobility within and between the Entities.

### Credit Growth, Bank Soundness, and Banks’ Foreign Liabilities — Key Findings
- Bosnia & Herzegovina experienced rapid credit growth in recent years.
  - Real private sector credit growth averaged 22.8 percent between 2001 and 2006.
  - Credit to GDP ratio more than doubled during the same period.
- Foreign-owned banks are the main sources of credit growth and rely on long-term foreign funding to finance the expansion.
- The 2006 Financial Sector Assessment Program (FSAP) identified weaknesses in asset classification and provisioning that may mask underlying credit risks.
- Reliance on potentially volatile foreign funding may create risks for both financial stability and the macroeconomy, especially given the currency board arrangement.
- Empirical results summarized:
  - There is no evidence that stronger banks expand credit faster.
  - The policy measures of the Central Bank of Bosnia & Herzegovina (CBBH) have had no significant impact on credit growth.
  - Credit growth improves bank soundness, possibly through boosting profits.
  - Controlling for other factors, large foreign banks appear to run a higher risk of insolvency because foreign bank subsidiaries keep capital-to-asset ratios low to raise return on equity; insurance is provided by parent banks that typically can be expected to provide additional capital when needed.
  - Credit expansion, robust economic activity and strong profitability are the main determinants of banks’ foreign borrowing.
  - The tightening of the maturity matching requirements has also contributed to the increase in banks’ foreign liabilities.
- Policy implications emphasized:
  - Since policy measures to curb credit growth have not been effective, it is even more important to strengthen prudential supervision to minimize risks to financial stability.
  - Supervisors should ensure large foreign subsidiaries build up adequate reserves reflecting underlying risk through tightening provisioning requirements or requesting additional capital buffers.
  - Cooperation with home-country supervisors is important to closely monitor local subsidiaries of foreign banks.
  - Relaxing the maturity matching requirements could help lessen reliance on foreign borrowing to some extent.

*Prepared by IMF staff (content unit: _cr07269 - 103).*

### 13.5 percent for the credit to enterprises. Starting from a very low base, credits to household

### _cr07269 - 13.5 percent for the credit to enterprises. Starting from a very low base, credits to household

### Key features of the credit expansion
- Household credit now accounts for 47 percent of total credit to private sector.
- Corporate lending picked up in 2003-04.
- Most household loans are long term; about 25 percent is housing-related credit.
- Anecdotal evidence suggests part of household credit expansion has been led by a housing boom (no aggregate housing price data available).
- Competition has led to a fall in lending interest rates.
- Long-term credits represent 76 percent of total lending.
- Short-term deposits accounted for 56 percent of total deposits at end-Q3 2006, while short-term lending accounted for only 24 percent of total lending.
- Share of foreign currency and indexed loans to total loans was high and rising (figures shown in the source).

### Financing of the credit expansion — aggregate flows and composition
- Private sector credit to GDP grew by 25 percentage points (aggregate asset increase).
- Net assets at the CBBH grew 9 percentage points.
- Main sources of asset growth (cumulative, percentage points of GDP):
  - Private sector deposits: 18.9
  - Net foreign liabilities: 8.1
  - Net assets held with the central bank: 9.1
  - Net central government deposits: 2.1
  - Noncentral government deposits: 2.0
  - Claims on noncentral government: 0.1
  - Other items, net: 2.9
- Bank-related inflows averaged 2.7 percent of GDP from 2001 to 2006.
- Gross foreign liabilities reached 25 percent of GDP by end-2006.
- Foreign funding typically comes in the form of euro-denominated loans or non-resident deposits; these are exchanged for KM at the CBBH because most lending is in KM indexed to the euro.
- Rising excess reserves at the CBBH reflect banks’ inability to intermediate short-term deposits.
- Bank-related capital inflows contributed significantly to increases in international reserves.

### Key risks identified
- FSIs indicate the banking system is in good health: improved profitability, high capital adequacy ratios, moderate and falling nonperforming loan ratios, and high liquidity.
- Rapid credit growth combined with weaknesses in asset classification and provisioning may mask underlying credit risks.
- Weaknesses in regulation of credit risk management could lead commercial banks to underestimate credit risks.
- A downgrade in asset quality could force parent banks to curb lending to Bosnian subsidiaries, jeopardizing financial and external stability.
- Indirect risks:
  - Most loan contracts contain adjustable interest rates linked to Euribor; many borrowers may not realize the implications.
  - Some banks have started offering Swiss franc-indexed loans in addition to euro-indexed loans, increasing exchange rate–induced credit risk.
  - Movements in exchange rates or interest rates could significantly affect borrowers’ capacity to pay.
- Reliance on potentially volatile foreign funding introduces risk to financial stability and the macroeconomy:
  - A fall in foreign funding could lead to a fall in international reserves; under the currency board arrangement this could have an immediate economic impact and potentially cause a sharp increase in non-performing loans.
  - FSAP stress tests (based on June 2005 data) suggest a slowdown in credit growth and capital flows, coupled with a shift from KM into euro assuming declining confidence, could negatively impact financial stability; however, the system appeared resilient at the time.
- Cross-border contagion risk:
  - Limited number of foreign parent banks in CEE and SEE raises possibility of contagion.
  - Parent–subsidiary relationships are asymmetric; Bosnia and Herzegovina’s portfolio is a small share of parent banks’ total portfolio, so small changes in parent bank lending policies can have large impacts on capital flows and lending in Bosnia and Herzegovina.

### Financial soundness indicators (selected, 2001–06)
- Capital:
  - Tier 1 capital to risk-weighted assets (RWA): 22.4, 18.2, 15.9, 14.9, 13.7, 13.6 (2001–2006)
  - Net capital to RWA: 25.1, 20.5, 20.3, 18.7, 17.8, 17.7 (2001–2006)
- Asset quality:
  - Nonperforming assets (NPAs) to total assets: 9.2, 6.9, 5.2, 3.8, 3.3, 2.5 (2001–2006)
  - Nonperforming loans to total loans: 17.9, 11.0, 8.4, 6.1, 5.3, 4.0 (2001–2006)
  - Provision to NPAs: 40.7, 42.5, 42.9, 44.6, 40.1, 39.6 (2001–2006)
- Profitability:
  - Return on assets: -1.1, -0.3, 0.4, 0.7, 0.7, 0.9 (2001–2006)
  - Return on equity: -4.8, -2.0, 3.4, 5.8, 6.2, 8.5 (2001–2006)
- Liquidity:
  - Liquid assets to total assets: 44.6, 33.8, 35.1, 35.7, 36.1, 35.9 (2001–2006)
  - Liquid assets to short-term financial liabilities: 65.1, 51.5, 58.7, 59.5, 61.9, 60.8 (2001–2006)
  - Short-term financial liabilities to total financial liabilities: 83.2, 77.6, 69.0, 68.7, 66.1, 67.1 (2001–2006)
- Foreign exchange risk:
  - Foreign currency and indexed loans to total loans: 64.9, 65.4, 68.8, 71.0 (selected years shown)
  - Foreign currency liabilities to total financial liabilities: 61.5, 64.0, 64.4, 62.8 (selected years shown)
  - Net open position: 42.2, 11.6, 8.3, 8.9 (selected years shown)

### Measures taken to curb credit growth and tighten regulation
- Foreign exchange exposure limits:
  - Banks required to phase in tighter limits on foreign exchange exposure from about 120 percent of capital in mid-2003 to 30 percent of capital by mid-2004.
  - Definition of foreign currency assets and liabilities amended to include KM assets and liabilities indexed to foreign currency.
- Maturity matching prudential regulations:
  - Banks required to comply by June 2004 (some allowed to delay compliance until September 2004).
  - Maturity matching requirements (Table 3):
    - Up to 90 days: At least 100 percent of fund sources must be placed in assets with the same maturity.
    - Up to 180 days: At least 95 percent of fund sources must be placed in assets with the same maturity.
    - Up to 365 days: At least 90 percent of fund sources must be placed in assets with the same maturity.
- Reserve requirements and monetary stance (Table 4):
  - June 2003: Base for calculation of required reserves expanded to include foreign currency deposits; required reserves ratio reduced from 10 to 5 percent; cash holding in banks’ vaults eliminated as eligible assets; all bank reserves remunerated at market rates (rate earned by the CBBH on its overnight deposits of foreign exchange reserves).
  - December 2003: Required reserves continued to be remunerated at market rates, but remuneration of excess reserves cut to 1 percent.
  - September 2004: Reserve requirements increased to 7.5 percent.
  - December 2004: Reserve requirements increased to 10 percent.
  - December 2005: Reserve requirements increased to 15 percent.
- Remuneration on excess reserves was cut to 1 percent to discourage excess reserves accumulation.

### Empirical investigation — data, methodology, and modeling
- Data:
  - Bank-level unpublished data from CBBH, FBiH Banking Agency, and RS Banking Agency.
  - Data include individual banks’ financial statements, ownership, detailed assets and liabilities classifications, and prudential reports.
  - Annual data available 2001–2005 (interest rates from 2002); quarterly data available Q2 2003–Q3 2006.
- Bank soundness measure:
  - Distance-to-default (DD) based on option pricing model (Black and Scholes, Merton); used in simplified balance-sheet form where market equity data are not liquid.
  - Formula summary: DD = (k + μ) / σ (where μ is average return on assets, k is equity capital as percent of assets, σ is standard deviation of returns on assets).
- Baseline econometric specifications:
  - Credit growth equation (quarterly panel):
    - CreditGri,t = αi + β1 CreditGri,t-1 + β2 DDi,t + β3 DDi,t-1 + β4 Macroi,t-1 + β5 BankSpecifici,t-1 + β6 Policyt + μi + εi,t
  - Bank soundness equation:
    - DDi,t = αi + β2 DDi,t-1 + β1 CreditGri,t-1 + β3 Macroi,t-1 + β4 BankSpecifici,t-1 + β5 Policyt + μi + εi,t
  - Foreign liabilities equation:
    - FLi,t = αi + β1 FLi,t-1 + β2 Crediti,t + β3 Crediti,t-1 + β4 Macroi,t-1 + β5 BankSpecifici,t-1 + β6 Policyt + μi + εi,t
  - Where FL is log banks’ foreign liabilities and Credit is log private sector credit; macro and bank-specific control sets included; policy captured by dummies; lagging controls to reduce endogeneity.

*Source: IMF staff analysis, extracted from the provided content unit.*

### 135.     The equations are estimated with the generalized method of moments (GMM)

### _cr07269 - 135.     The equations are estimated with the generalized method of moments (GMM)

### Methodology
- Estimation approach:
  - The equations are estimated with the generalized method of moments (GMM) system estimator (“system GMM”).
  - System GMM combines the original level equation and the differenced equation and uses lagged difference variables as instruments for level variables (Arellano and Bover (1995); Blundell and Bond (1998)).
  - System GMM was developed to improve efficiency over difference GMM (Arellano and Bond (1991)), which uses first differences and lagged levels as instruments.
- Advantages of system GMM:
  - Addresses endogenous or not strictly exogenous regressors, fixed effects, heteroskedasticity, and autocorrelation within individuals.
  - Captures the full effect of policy dummies that are equal to one only after implementation by including the level equation (unlike difference GMM which would estimate only the initial quarter effect).
- Panel unit-root testing:
  - Maddala and Wu (1999) tests for panel unit roots show that no series used in the empirical investigation has a unit root.

### Model diagnostics and estimation details
- Estimates reported:
  - Two-step robust estimates with standard errors corrected for finite sample bias (the Windmeijer correction) are reported.
  - All lags are used for each endogenous variable unless otherwise noted.
- Specification diagnostics:
  - All regressions show no sign of second-order serial correlation.
  - The Hansen test for overidentification is not rejected in reported regressions.
  - Adding the second lag of explanatory variables (notably for credit growth in equations (2) and (3)) did not add information and was excluded.
- Note on credit growth series:
  - Only year-on-year credit growth is used in some specifications because quarter-on-quarter credit growth does not pass the AR(1) test (i.e., no serial correlation is observed in the first difference).

### Results — Credit growth
- General findings:
  - Sounder banks seem to have higher credit growth, though coefficients are small and not statistically significant (Tables 5).
  - Distance-to-default and its first lag have positive but statistically insignificant coefficients on credit growth.
- Other macro/financial drivers:
  - Real interest rate (t-1) has a positive and significant impact on credit growth in the reported regressions.
  - Real retail sales growth (proxy for economic activity) has a negative, though insignificant, impact on credit growth.
  - Net interest margin and cost-to-income ratio coefficients are reported but generally not significant in the narrative summary.
- Policy dummies:
  - Policy measures (reserve requirement increases, tightening of maturity matching) do not have a statistically significant effect on credit growth.
  - Dummy variables for policy changes in Q3 2004 (tightening of maturity matching and increase in reserve requirements), Q4 2004, and Q4 2005 have positive but statistically insignificant coefficients.
- Diagnostic statistics (from Table 5, sample size and tests):
  - Number of observations: 226 (for the credit growth regressions reported).
  - P Value for Hansen test examples: 0.78, 0.765, 0.767, 0.769 (across columns reported).
  - P-Value for AR(1) test examples: 0.036, 0.039, 0.037, 0.037.
  - P-Value for AR(2) test examples: 0.173, 0.166, 0.136, 0.223.

### Results — Bank soundness (distance-to-default) 
- Credit growth effects:
  - Annual (year-on-year) credit growth appears to have no significant impact on bank soundness in some specifications, whereas quarter-on-quarter credit growth seems to boost bank soundness (Tables 6 and 7).
  - In quarter-on-quarter specifications, the coefficient on Credit Growth (t-1) is positive and statistically significant in several columns (e.g., 0.517 **, 0.606 **, 0.653 **, 0.638 **, 0.615 *** in Table 7).
- Persistence and core determinants:
  - Bank soundness shows positive persistence: Distance to default (t-1) has large positive and highly significant coefficients (e.g., 0.698 ***, 0.715 ***, 0.722 ***, etc., in Table 6; 0.673 ***, 0.671 ***, 0.711 ***, etc., in Table 7).
  - Higher NPA ratio (t-1) has a negative impact on distance-to-default; statistically significant in some specifications (e.g., NPA ratio (t-1) = -111.350 * in Table 6, column (1)).
  - Cost-to-income ratio (t-1) is associated with changes in bank vulnerability: in Table 6 it shows positive and significant coefficients (e.g., 2.770 **, 2.691 **, 2.919 **), while in Table 7 cost-to-income ratio coefficients are positive with some significance (e.g., 1.984 *).
  - Real interest rate (t-1) generally has a negative but statistically insignificant impact on bank soundness in several specifications.
- Ownership and size effects:
  - Large foreign banks appear to have higher risk of insolvency; interaction terms between foreign ownership and size, and between EU bank ownership and size, show negative and statistically significant coefficients (e.g., Foreign ownership * size (t-1) = -0.454 ** in Table 6; EU bank parent * size (t-1) = -0.978 * to -0.953 * in Table 6).
  - State ownership shows a positive and statistically significant coefficient in some specifications, reflecting higher capital ratios of state-owned banks (e.g., State ownership = 9.425 *; 9.535 in Table 6 columns).
- Policy effects:
  - The increase in the reserve requirement in December 2004 may have had a negative impact on bank soundness (Policy dummy: Q4 2004 = -1.782 * in Table 6, column (2)), while other policy changes show no significant effect.

### Results — Banks’ foreign liabilities
- Key elasticities and drivers:
  - Credit expansion, robust economic activity, and strong profitability all positively affect banks’ foreign borrowing (Table 8).
  - If credit grows by 1 percent both in the last quarter and in the current quarter, banks’ foreign liabilities will increase by between 0.3-0.7 percent (elasticity of foreign liabilities to credit ≈ 0.3-0.7).
  - Interest rate (t-1) positive and significant (e.g., 0.775 *** and 0.717 *** in Table 8), real retail sales growth (t-1) positive and significant at 10 percent (e.g., 1.761 *, 1.862 *), inflation (t-1) positive and significant (e.g., 7.998 **), and net interest margin (t-1) positive and significant at 10 percent (e.g., 9.939 *, 12.046 *).
- Policy and ownership:
  - The tightening of maturity matching requirements contributed to the increase in banks’ foreign liabilities; the policy dummy shows positive and significant coefficients (e.g., 0.402 * and 0.461 * in Table 8).
  - Controlling for other factors, ownership dummies (state, foreign, EU parent) do not have a statistically significant direct impact on the level of foreign liabilities in the reported specifications.
- Macroeconomic implication (empirical estimate linking credit, foreign liabilities, and reserves):
  - If credit growth slows to zero in two quarters, foreign liabilities would fall by around 7-16 percent.
  - Assuming a one-to-one fall in international reserves and taking the maximum value, this translates to a decline of €864 million in international reserves, or from 5.1 months to 4.3 months of imports (end-2006 figures).
- Diagnostic statistics (from Table 8, sample size and tests):
  - Number of observations: 289.
  - P Value for Hansen test examples: 0.963, 0.953.
  - P-Value for AR(1) test examples: 0.052, 0.047.
  - P-Value for AR(2) test examples: 0.376, 0.488.

### Regional comparison
- Contrast with other studies/regions:
  - Unlike findings for the new EU member states (NMS) and Bulgaria, Bosnia & Herzegovina shows credit growth having a positive and immediate impact on bank soundness (see Appendix Table comparisons).
  - Igan and Tamirisa (2006) find credit growth statistically insignificant for bank soundness in NMS; Herderschee and Ong (2006) find positive loan growth strategies raise solvency risk in Bulgaria.
  - Large foreign banks in Bosnia & Herzegovina seem to run higher solvency risk than domestic counterparts, contrary to NMS and Bulgaria where foreign ownership is associated with higher distance-to-default.

### Conclusions and policy implications
- Financial system assessment:
  - Financial deepening in Bosnia & Herzegovina has been supported by favorable economic conditions and consolidation/competition in the banking sector.
  - The banking system is well-capitalized and profitable, but inherent risks exist under rapid credit expansion.
- Effectiveness of macroprudential measures:
  - Policy measures to curb credit growth (notably reserve requirement increases) have not been effective at slowing credit growth; reserve requirement increases showed no statistically significant impact on credit growth.
  - Authorities therefore largely rely on prudential tools to contain financial-stability risks.
- Supervisory recommendations:
  - Supervisors should ensure large foreign subsidiaries build adequate reserves and capital buffers because such subsidiaries tend to keep minimum capital and rely on parent bank support.
  - Possible measures include tightening provisioning requirements, requesting additional capital buffers, and addressing weaknesses in asset classification and provisioning rules (as recommended by the FSAP).
  - Banking agencies could request additional capital requirements for market risks, such as foreign exchange risks.
  - Strengthened cooperation with home-country supervisors is important to monitor the health of parent banks that may be the source of capital injections.
- Funding and maturity-matching policy:
  - Empirical results show maturity matching tightening contributed to increased foreign liabilities.
  - Relaxing the maturity matching requirement — which is strict by international standards — could help lessen banks’ reliance on foreign funding and encourage domestic intermediation, though foreign borrowing is driven mainly by economic factors beyond the maturity rule.

### Data sources and variable definitions (selected)
- Bank-level data: CBBH, FBiH Banking Agency, RS Banking Agency.
- Macroeconomic data: Federation Statistics Office, RS Institute of Statistics.
- Variable definitions highlighted in the source:
  - Distance to default: Return on average assets plus equity as a percent of assets divided by the standard deviation of return on average assets over the period; balance sheet data used.
  - Private sector credit growth: Percentage change in total private sector credit in real term (deflated by CPI), year-on-year and quarter-on-quarter.
  - Banks’ foreign liabilities: Outstanding foreign liabilities in banks’ balance sheet in KM.
  - Real retail sales growth: Year-on-year percentage change in Federation retail sales index in real term (deflated by CPI).
  - Real interest rates: Commercial bank lending rates in KM, deflated by CPI inflation (data for lending rates in euro or euro-linked loans available only from January 2007).
  - Cost to income: Total operating expenses divided by total operating income.
  - Net interest margin: Net interest income divided by average assets.
  - Size: Logarithm of total assets.

*Source: _cr07269 - 135. The equations are estimated with the generalized method of moments (GMM) (IMF staff report excerpt).*

### APPENDIX III—PRUDENTIAL REGULATION: INTERNATIONAL EXAMPLES

### APPENDIX III—PRUDENTIAL REGULATION: INTERNATIONAL EXAMPLES

### Foreign exchange risk: findings and guidance
- The FSAP recommended that comprehensive guidelines on foreign exchange risk and other market risks should be introduced in Bosnia & Herzegovina.
- The foreign exchange exposure limit in Bosnia & Herzegovina is in line with practice in other countries with dollarized financial systems (Appendix Table 1).
- Supervisors should require capital for foreign exchange exposure.
  - Rationale: capital requirements make it more difficult for weakly capitalized banks to take on new risks and give banks flexibility to allocate capital between credit and market risk, including foreign exchange risk (Cayazzo and others,2006).

- Table 1. Foreign Exchange Risk Requirements: Selected Countries
  - Central and Eastern European countries
    - Croatia — Limits on Foreign Exchange Exposure: 20 percent; Foreign Exchange Risk Capital Charge: 10 percent
    - Latvia — Limits on Foreign Exchange Exposure: ...; Foreign Exchange Risk Capital Charge: 10 percent
    - Poland — Limits on Foreign Exchange Exposure: No.; Foreign Exchange Risk Capital Charge: 8 percent
    - Romania — Limits on Foreign Exchange Exposure: 20 percent; Foreign Exchange Risk Capital Charge: 8 percent
    - Slovenia — Limits on Foreign Exchange Exposure: ...; Foreign Exchange Risk Capital Charge: 8 percent
  - Other emerging markets
    - Argentina — Limits on Foreign Exchange Exposure: 30 percent; Foreign Exchange Risk Capital Charge: No.
    - Brazil — Limits on Foreign Exchange Exposure: 50 percent; Foreign Exchange Risk Capital Charge: 30 percent
    - Chile — Limits on Foreign Exchange Exposure: 30 percent; Foreign Exchange Risk Capital Charge: No.
    - Lebanon — Limits on Foreign Exchange Exposure: ...; Foreign Exchange Risk Capital Charge: 12 percent

- Source cited for Table 1: Cayazzo and others (2006).

### Liquidity and maturity matching: findings and international comparisons
- The maturity matching requirements in Bosnia and Herzegovina are strict by international practice.
- Bosnia’s banking agencies use:
  - a minimum liquidity ratio; and
  - maturity matching requirements to regulate banks’ liquidity.
- For the minimum liquidity ratio, banks in Bosnia have to maintain average 10-day minimum liquidity in cash funds up to at least 10 percent of short-term funds sources according to book value.
- Country practices vary; most countries with partially dollarized banking systems have laxer requirements.
- Romania appears to have a system as strict as Bosnia’s.
- Chile has a 100 percent requirement but it applies only to assets and liabilities of up to 30 days.
- In Slovenia coverage must be at least 100 percent, but demand deposits are given a weight smaller than 100 percent.

- Table 2. Liquidity Requirements: Selected Countries
  - Central and Eastern European countries
    - Croatia — Minimum Liquidity Ratio: 35 percent only for foreign currency; Limits on Maturity Gaps: No
    - Latvia — Minimum Liquidity Ratio: 30 percent for less than 30 days; Limits on Maturity Gaps: No
    - Poland — Minimum Liquidity Ratio: ...; Limits on Maturity Gaps: No
    - Romania — Minimum Liquidity Ratio: ...; Limits on Maturity Gaps: Yes. Minimum liquidity ratio is 1.0. It is measured for all time buckets on a cumulative basis.
    - Slovenia — Minimum Liquidity Ratio: 100 percent only for foreign currency of up to 6 months; Limits on Maturity Gaps: Yes. Liquidity coefficient (investments/liabilities) must be at least 1.0 separately for foreign and domestic currency for up to 30 days and up to 180 days. Household and corporate demand deposits are given a weight of 85 percent for up to 30 days, and 60 percent for up to 180 days.
  - Other emerging markets
    - Argentina — Minimum Liquidity Ratio: Rates range from 20 percent for current account deposits to 80 percent for demand deposits.; Limits on Maturity Gaps: The sum of the balances of the eligible items cannot be less than 60 percent of the total requirement for up to 30 days.
    - Brazil — Minimum Liquidity Ratio: 8-45 percent only for domestic currency; Limits on Maturity Gaps: No
    - Chile — Minimum Liquidity Ratio: ...; Limits on Maturity Gaps: Yes. Liabilities with a term of less than 30 days cannot exceed 1 time the assets of the same maturity. Liabilities with a term less than 90 days cannot exceed 2 times the assets of the same maturity. The rule holds separately for domestic and foreign currency and for total assets and liabilities. However, banks can consider an excess of foreign currency assets to cover domestic liabilities.
    - Lebanon — Minimum Liquidity Ratio: 10 percent; Limits on Maturity Gaps: No

- Sources cited for Table 2: Cayazzo and others (2006); and Hayward and Byskov (2005).

### Policy implications and recommendations
- Introduce comprehensive guidelines on foreign exchange risk and other market risks in Bosnia & Herzegovina, as recommended by the FSAP.
- Require capital charges for foreign exchange exposure to:
  - Deter undercapitalized banks from taking on new FX risks.
  - Allow banks to allocate capital between credit and market risks, including foreign exchange risk.
- Review maturity matching and liquidity requirements to ensure they are consistent with country-specific conditions, noting that Bosnia’s current maturity-matching framework is strict relative to many peers.

*Source: APPENDIX III—PRUDENTIAL REGULATION: INTERNATIONAL EXAMPLES (from the provided IMF content unit).*

---


_Source: https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/scr/2007/_cr07269.pdf_