## 1andea2022002

## Source details

**Canonical URL:** [1andea2022002](https://www.imf.org/-/media/files/publications/cr/2022/english/1andea2022002.pdf)

## Other formats

- [Markdown version](/-/media/files/publications/cr/2022/english/1andea2022002.pdf.md)
- [Structured JSON version](/-/media/files/publications/cr/2022/english/1andea2022002.pdf.json)

---

### ESTIMATION OF ANDORRA'S POTENTIAL OUTPUT

- Definition and purpose
  - Potential output: maximum level of output achievable without creating inflationary pressures (Okun, 1962).
  - Uses: macroeconomic surveillance, fiscal stance assessment, returns on public investment, long-run anchor for forecasting.

- Data constraints and methodological approach
  - Data limitations: small sample, no quarterly GDP level series, no capital stock data; COVID-19 break complicates estimation.
  - Approaches used:
    - Univariate: Hodrick-Prescott (HP) filter on annual real GDP series for 2000–2021 plus projections for 2022–27 (from April 2022 World Economic Outlook). Limitation: purely statistical, end-of-sample problem.
    - Multivariate filter: based on Balgrave et al. (2015); conditions potential output on a Phillips curve (cyclical unemployment ↔ inflation) and Okun’s law (cyclical unemployment ↔ output gap). Uses annual real GDP level, CPI inflation, and unemployment.
    - Production function approach: not implemented due to lack of capital-stock data.
  - COVID-19 treatment in estimation:
    - Pre-crisis trend estimated using data to 2019 and pre-crisis projections (Andorran 2020 budget projections).
    - 2020:Q2 GDP negative shock attributed 50 percent to supply factors and 50 percent to demand; recovery in H2 2020 fully attributed to demand.
    - Only a small part of 2020 GDP decline is translated into lower potential output; potential growth assumed to converge gradually to pre-crisis estimate from 2021 onwards.

- Pre-COVID trends (all filters)
  - Three-phase pattern:
    - Phase 1 (2001–07): high potential growth averaging between 3.7 and 5.2 percent.
    - Phase 2 (2008–14): lower potential growth averaging between -2.0 and -0.3 percent.
    - Phase 3 (2015–19): positive but lower potential growth; output gap nearly closed by end-2019.
  - Recovery from 2015–19 largely supported by strong employment growth; contribution from physical capital and TFP was very modest.
  - Output gap took about 10 years from the trough of the GFC to close, reflecting significant scarring linked to supply-side factors, particularly labor.

- COVID-19 impact
  - Actual output contracted by 11.2 percent in 2020.
  - Potential output estimated to have declined nearly 1.5 percent in 2020.
  - Negative output gap in 2020 expected to close in 2024.
  - Pandemic-induced permanent income loss estimated at 3.2 percent (attributed to missed 2020/21 winter season and a sharp contraction in investment, partially offset by later construction boom and FDI inflows).
  - Comparative permanent income loss (gap relative to pre-crisis trend):
    - Andorra: -3.2%
    - France: -1.4%
    - Spain: -2.5%
    - Greece: -3.2%
    - Malta: -3.7%

- Medium-term potential growth
  - Prior to COVID-19, medium-term potential growth estimated at 1.5 percent.
  - Output gap expected to remain close to zero and inflation stabilized around 1.8 percent.
  - After recovery, potential growth projected to converge to 1.5 percent in the medium term, but the level of output remains permanently lower due to the pandemic.

- Drivers of low potential growth
  - Low investment:
    - Total investment in Andorra averages about 13 percent of GDP.
    - Comparators: high-income economies average 21.8 percent of GDP; euro area countries average 21.0 percent of GDP.
    - Public investment in Andorra is nearly 5 percent of GDP.
      - Comparators: high-income public investment 3.7 percent of GDP; euro area public investment 3.2 percent of GDP.
    - Low private investment is the main driver of the low total investment share.
  - Human capital:
    - 32 percent of the population older than 25 in Andorra has completed post-secondary non-tertiary education or higher.
    - Comparators: high-income economies average 36.6 percent; euro area countries average 37.2 percent.
    - Relationship between Andorran educational attainment and potential growth may be weakened by a large inflow of foreign seasonal workers representing a sizeable share of the labor force.

- Policy implications / recommendations to boost potential growth
  - Diversify the economy to reduce concentration in tourism and banking.
  - Boost investment:
    - Increase private investment by leveraging relatively strong public investment (nearly 5 percent of GDP).
    - Improve access to credit.
    - Reduce red tape and administrative rigidities to support private-sector activity.
  - Build human capital:
    - Encourage vocational training and tertiary education to raise educational attainment.
    - Ease immigration requirements and expand affordable housing supply to attract high-skilled workers.
  - Enhance resilience to climate-related risks:
    - Green transportation and energy.
    - Reduce dependency on imported energy.
  - Pursue international integration:
    - Current negotiations on an EU Association Agreement could unlock substantial benefits.

*Prepared by Ana Lariau (EUR); May 25, 2022 — International Monetary Fund*

---

### SUSTAINABILITY OF ANDORRA'S PENSION SYSTEM

- Executive summary and urgency
  - Reforming the Andorran pension system is a key priority to ensure sustainability and reduce contingent liabilities.
  - In the absence of reform, the social security system will accumulate deficits starting in 2024, rising to about 9 percent of GDP per year by 2040.
  - Parliament appointed a special commission to elaborate a reform plan before end-2022.
  - Measures will need to be comprehensive and adjust all three key parameters—the contribution rate, the conversion factor, and the retirement age.
  - Reform design should be guided by public policy choices on affordability, adequacy, and equity that go beyond the scope of sustainability analysis.

- Key features of the system
  - System type: Pillar I — mandatory pay-as-you-go by points defined contribution; compulsory for all workers including self-employed.
  - Current contribution rate: 12 percent of gross salary (8.5 percent paid by the employee, and 3.5 percent by the employer).
  - Statutory retirement age: 65 years after at least 15 years of contribution; early retirement option at 61 years with a minimum contribution period of 40 years.
  - Conversion coefficient history: 6.4 → 8.0 (2009) → 9.6 (2015).
  - Replacement rate: 50 percent.

- 2020 system statistics
  - Dependency ratio: about 3 (with an average of 47,439 members, 90 percent of the working-age population).
  - Members resident in Andorra: 97 percent.
  - Pensioners including widows: 15,840 (59 percent resided in the country; represented 83 percent of the Andorran elderly population).
  - Average annual contribution: €2,979.
  - Average annual pension: €7,009.

- Pillar 2 and investment strategy
  - Pillar 2: mandatory for government employees, voluntary for private sector; employer matches up to a total of 3 percent of salary.
  - FRJ investment strategy targets:
    - ~60 percent high credit quality fixed income (minimum 50 percent).
    - 30 percent equity (maximum 35 percent).
    - 10 percent other assets (maximum 15 percent).
    - Currency mix: about 90 percent euro, rest in other advanced economies' currencies.
  - Authorities reviewing investment strategy to upgrade risk management and increase geographic and asset-class diversification.

- Benchmarking and demographic outlook
  - Life expectancy: 84.5 years; life expectancy at retirement: about 19.5 years.
  - Old-age dependency: 20.7 percent (current), EU average 32.3 percent; expected to almost double by 2070.
  - Coverage rate: expected to decline from current low level of three members per pensioner.
  - Generosity rate (average pension / average contribution) in 2020: 2.35.
  - Contribution rate comparison: Andorra 12 percent of gross salaries vs EU average 21 percent.
  - Replacement rate: 50 percent (above EU average).
  - Andorra currently has the lowest pension expenditure in the region but is expected to have the largest increase in the next two decades.

- Methodology and scenarios
  - Diagnostic approach uses demographic and system parameters to project pension expenditure and deficits.
  - Inputs: actuarial reports, EC's 2021 Ageing Report, OECD, Eurostat, UN population projections, IMF, Andorran authorities, IMF staff calculations.
  - No Policy Change Scenario assumptions:
    - Long-term GDP growth 1.5 percent, inflation 1.7 percent, real return on fund investment 1 percent, population projections in line with France and Spain.
    - Projection: annual deficits starting in 2024; reserves depleted by 2039 without government support.
    - Deficits rise to about 9 percent of GDP per year by 2040.

- Alternative reform scenarios modeled
  - i) One-time increase in the contribution rate from 12 to 16 percent of gross salary in 2023.
  - ii) Gradual increase in the conversion factor from 9.6 to 17 in 25 years.
  - iii) Gradual increase in the retirement age from 65 to 67 years in 8 years.
  - Combined reform: implementation of all three measures.

- Scenario results (summary)
  - Contribution rate increase (12 → 16 percent):
    - Delays deficits to 2027; reserves decline from 62.7 percent of GDP in 2022 to 17 percent of GDP in 2040.
  - Conversion factor increase (9.6 → 17 over 25 years):
    - Deficits still start in 2024 (gradual implementation); estimated deficit by 2040 is 4.2 percent of GDP; reserves decline to 27.9 percent of GDP by end of projection period.
  - Retirement age increase (65 → 67 over 8 years):
    - Positive impact but insufficient alone because of high old-age dependency ratio and life expectancy.
  - Combined reform (all three implemented):
    - Projected increase in expenditure: 3.3 percent of GDP (present discounted value of 28.3 percent of GDP) vs 7.8 percent of GDP under no policy change (present discounted value of 63 percent of GDP).
    - System accumulates small deficits starting in 2031, reaching up to 3 percent of GDP by 2040, with government support starting in 2037.
    - Reserves decline from 62.3 percent of GDP currently to 59.9 percent of GDP at the end of the projection period.
    - Conclusion: only joint implementation of the three reforms substantially improves sustainability.

- Quantitative decomposition (2022–2040 present discounted values, % of GDP)
  - Conversion factor: -31.5
  - Contribution rate: 3.8
  - Retirement age: -7.0
  - Net change (All reforms): 28.3

- Policy implications and considerations
  - Reforms need to be comprehensive, combining increases in the contribution rate, conversion factor, and retirement age.
  - Standalone measures produce improvements but are insufficient; the joint package yields the largest and most durable improvement in sustainability.
  - Reform design should weigh trade-offs and policy goals beyond sustainability: pension adequacy, affordability, intergenerational equity, and distributional considerations.
  - Role of second-pillar pensions:
    - Expanding the second pillar could help achieve pension adequacy, but should not supplant measures needed for first-pillar financial sustainability.

*Prepared by Michelle Tejada (EUR).*

---

### BANKING SECTOR: SIZE, BUSINESS MODEL, INTERNATIONALIZATION, AND RISKS

- Size and economic importance
  - Financial sector (banks and insurance companies) accounts for 14 percent of GDP and employs approximately 5 percent of the workforce.
  - Consolidated bank assets amount to about 600 percent of GDP.
  - Assets under management (AUM)—largely off-balance sheet—are nearly 23 times the GDP.
  - A banking sector of this size, comprising only three banking groups, implies that all banks are systemic in nature.

- Ownership and business focus
  - Three banking groups mainly owned by Andorran families.
  - Business model: domestic commercial banking combined with private banking both in Andorra and internationally.
  - Private banking and wealth management specialization: about 80 percent of banks’ income corresponds to non-interest income sources (largely commissions).

- Performance, efficiency, and buffers
  - Profitability remained relatively stable in recent years despite large exogenous shocks.
  - Private banking model associated with higher cost-to-income ratios than commercial banking; cost efficiency has been weak.
  - Ongoing consolidation should help reduce the cost-to-income ratio, but supervisor should ensure it does not significantly hinder competition.
  - Continued AUM growth would increase efficiency by covering fixed costs and bringing additional commissions.
  - Capital and liquidity ratios have been stable and well-above the EU weighted average in recent years.
  - LCR, under baseline assumptions, results in an LCR of 217 percent.

- Internationalization and external risks
  - Internationalization geographically diversified business and significantly increased AUMs over the past 10 years.
  - By end-2021, about 65 percent of AUMs were managed in 10 countries outside Andorra.
  - AUMs more than doubled since 2009 and by end-2021 they were nearly 23 times Andorra’s GDP, of which nearly 80 percent are off-balance sheet.
  - Internationalization exposes banks to foreign-market risks and vulnerabilities, including foreign exchange fluctuations, local competition, regulatory environment, economic and political instability, and operational diseconomies associated with monitoring from a distance.
  - Empirical evidence cited: Berger et al (2016) find a positive relationship between degree of internationalization and bank risk for U.S. banks.

- Key risks and stress findings
  - Reliance on foreign deposit funding exposes the system to liquidity risks.
  - In an extremely adverse scenario (outflow rates up to four times baseline), the system LCR would fall below the 100 percent requirement and there would be a liquidity gap of 4.8 percent of GDP.
    - Adverse scenario LCR = 94 percent.
    - Baseline scenario LCR = 217 percent.
  - Banks that display sharp declines in the LCR tend to rely more on funding from retail deposits (particularly less stable ones), unsecured funding from nonfinancial corporates and sovereigns not covered by deposit insurance, or on undrawn but committed credit facilities to retail and other nonfinancial customers.
  - Private banking activities and operations through subsidiaries abroad are vulnerable to money laundering and terrorism financing (ML/TF) risks; the 2020 National Risk Assessment (UIFAND) qualifies the vulnerability to ML/TF of private banking and of subsidiaries abroad to be medium-high, mitigated to medium by relatively good control systems.
  - Operating in a small and low-diversified economy raises the risk of large exposures and related-party lending, leading to concentration of credit risks.
  - Loan concentration and private banking focus may lower banks’ incentives to expand domestic credit; observed pattern: credit to the private non-financial sector tends to be lower than in other advanced economies.

- Policy recommendations and supervisory actions
  - Enhance liquidity supervision, including:
    - Further reinforce supervision of liquidity risk by enhancing monitoring and stress testing of net liquidity outflows.
    - Undertake analysis of liquidity risks for longer horizons (the LCR considers only a 30-day horizon).
    - Include onsite liquidity inspections in supervisory planning.
    - Accelerate development and review of liquidity contingency plans by banks.
    - Strengthen the AFA more broadly, including by more solid funding and staffing.
  - AML/CFT actions:
    - Continue improving effectiveness of the AML/CFT framework to ensure financial stability.
    - Persevere in aligning the AML/CFT framework to FATF standards, including in the area of virtual assets.
    - Continue refining the national risk assessment and act upon findings of the 2020 assessment.
    - Analyze data on cross-border transactions and exchange information with counterparts in relevant originating countries.
    - Focus AML/CFT supervisory resources increasingly on higher risk areas and effectively impose sanctions for breaches.
  - Concentration and exposure management:
    - Reduce concentration of credit risks by limiting related party lending and large exposures.
    - Implementation of EU-derived regulation limiting large exposures to less than 25 percent of the bank’s capital is critical.
    - AFA reinforced reporting requirements on large exposures; full implementation will be important as banks increase financing of the economy.
  - Absence of lender of last resort (LOLR) and reserves:
    - No LOLR is in place.
    - Recent buildup of international reserves—up to the reference level estimated by IMF staff of 8.7 percent of GDP—may help the government cope with materialization of financial-sector-related contingent liabilities.
  - Maintain vigilance given reliance on foreign funding and systemic nature of the three banking groups.

*Source: 1andea2022002*

### References ______________________________________________________________________________ 9

### ESTIMATION OF ANDORRA'S POTENTIAL OUTPUT

### Introduction
- Potential output is defined as the maximum level of output achievable without creating inflationary pressures (Okun, 1962).
- Estimates of potential output guide macroeconomic surveillance, fiscal stance assessment, returns on public investment, and act as the long-run anchor for forecasting.
- Andorra faces data limitations (small sample, no quarterly GDP level series, no capital stock data) and a COVID-19 break that complicate estimation.
- Multiple estimation approaches are considered to exploit available information and mitigate data constraints.

### Data and Methodology
- Univariate approach:
  - Hodrick-Prescott (HP) filter applied to annual real GDP series for 2000–2021 plus projections for 2022–27 (from April 2022 World Economic Outlook).
  - Limitations: purely statistical, end-of-sample problem.
- Multivariate filter:
  - Based on Balgrave et al. (2015), conditions potential output on Phillips curve (cyclical unemployment ↔ inflation) and Okun’s law (cyclical unemployment ↔ output gap).
  - Uses annual data on real GDP level, CPI inflation, and unemployment.
- Production function approach:
  - Not implemented due to lack of capital-stock data.
- COVID-19 treatment:
  - Pre-crisis trend estimated using data to 2019 and pre-crisis projections (Andorran 2020 budget projections).
  - 2020:Q2 GDP negative shock attributed 50 percent to supply factors and 50 percent to demand; recovery in H2 2020 fully attributed to demand.
  - Only a small part of 2020 GDP decline is translated into lower potential output; potential growth assumed to converge gradually to pre-crisis estimate from 2021 onwards.

### Results — Pre-COVID Trends
- All filters depict a three-phase pattern:
  - Phase 1 (2001–07): high potential growth averaging between 3.7 and 5.2 percent.
  - Phase 2 (2008–14): lower potential growth averaging between -2.0 and -0.3 percent.
  - Phase 3 (2015–19): positive but lower potential growth; output gap nearly closed by end-2019.
- Recovery from 2015–19 was largely supported by strong employment growth; contribution from physical capital and TFP was very modest.
- The output gap took about 10 years from the trough of the GFC to close, reflecting significant scarring linked to supply-side factors, particularly labor.

### Results — Impact of the COVID-19 Pandemic
- Actual output contracted by 11.2 percent in 2020.
- Potential output is estimated to have declined nearly 1.5 percent in 2020.
- The larger decline of actual relative to potential GDP created a negative output gap in 2020, expected to close in 2024.
- The pandemic induced a permanent income loss estimated at 3.2 percent (attributed to missed 2020/21 winter season and a sharp contraction in investment, partially offset by later construction boom and FDI inflows).
- Comparative permanent income loss (gap relative to pre-crisis trend):
  - Andorra: -3.2%
  - France: -1.4%
  - Spain: -2.5%
  - Greece: -3.2%
  - Malta: -3.7%

### Results — Potential Growth in the Medium Term
- Prior to COVID-19, medium-term potential growth was estimated at 1.5 percent.
- Output gap expected to remain close to zero and inflation stabilized around 1.8 percent.
- After recovery, potential growth is projected to converge to 1.5 percent in the medium term, but the level of output remains permanently lower due to the pandemic.

### Drivers of Low Potential Growth
- Low investment:
  - Total investment in Andorra averages about 13 percent of GDP.
  - Comparators: high-income economies average 21.8 percent of GDP; euro area countries average 21.0 percent of GDP.
  - Public investment in Andorra is nearly 5 percent of GDP.
    - Comparators: high-income public investment 3.7 percent of GDP; euro area public investment 3.2 percent of GDP.
  - Low private investment is the main driver of the low total investment share.
- Human capital:
  - 32 percent of the population older than 25 in Andorra has completed post-secondary non-tertiary education or higher.
  - Comparators: high-income economies average 36.6 percent; euro area countries average 37.2 percent.
  - Note: the relationship between Andorran educational attainment and potential growth may be weakened by a large inflow of foreign seasonal workers representing a sizeable share of the labor force.

### Policy Implications / Recommendations to Boost Potential Growth
- Diversify the economy to reduce concentration in tourism and banking.
- Boost investment:
  - Increase private investment by leveraging relatively strong public investment (nearly 5 percent of GDP).
  - Improve access to credit.
  - Reduce red tape and administrative rigidities to support private-sector activity.
- Build human capital:
  - Encourage vocational training and tertiary education to raise educational attainment.
  - Ease immigration requirements and expand affordable housing supply to attract high-skilled workers.
- Enhance resilience to climate-related risks:
  - Green transportation and energy.
  - Reduce dependency on imported energy.
- Pursue international integration:
  - Current negotiations on an EU Association Agreement could unlock substantial benefits.

*Prepared by Ana Lariau (EUR); May 25, 2022 — International Monetary Fund*

### References

### SUSTAINABILITY OF ANDORRA'S PENSION SYSTEM

### Executive summary
- Reforming the Andorran pension system is a key priority to ensure its sustainability and reduce contingent liabilities.
- In the absence of reform, the Andorran social security system will accumulate deficits starting in 2024, rising to about 9 percent of GDP per year by 2040.
- The Andorran Parliament appointed a special commission to elaborate a reform plan before end-2022.
- Measures will need to be comprehensive and adjust all three key parameters—the contribution rate, the conversion factor, and the retirement age.
- Reform design should be guided by public policy choices on affordability, adequacy, and equity that go beyond the scope of sustainability analysis.

### A. Introduction — key findings
- The social security fund is projected to have annual deficits starting in 2024 and to deplete its reserve fund by 2039 in the absence of government financial support.
- Drivers: aging population and a structural imbalance between low contribution rates and a relatively high replacement rate.
- The paper uses scenario analysis to identify reform options that ensure long-term sustainability.

### B. Overview of the pension system — structural features and past reforms
- System type: Pillar I — mandatory pay-as-you-go by points defined contribution; compulsory for all workers including self-employed.
- Current contribution rate: 12 percent of gross salary (8.5 percent paid by the employee, and 3.5 percent by the employer).
- Statutory retirement age: 65 years after at least 15 years of contribution; early retirement option at 61 years with a minimum contribution period of 40 years.
- Conversion coefficient history: 6.4 → 8.0 (2009) → 9.6 (2015).
- Replacement rate: 50 percent.
- 2020 system statistics:
  - Dependency ratio: about 3 (with an average of 47,439 members, 90 percent of the working-age population).
  - Members resident in Andorra: 97 percent.
  - Pensioners including widows: 15,840 (59 percent resided in the country; represented 83 percent of the Andorran elderly population).
  - Average annual contribution: €2,979.
  - Average annual pension: €7,009.
- Pillar 2: mandatory for government employees, voluntary for private sector; employer matches up to a total of 3 percent of salary.
- Investment strategy of FRJ (pension fund):
  - ~60 percent high credit quality fixed income (minimum 50 percent).
  - 30 percent equity (maximum 35 percent).
  - 10 percent other assets (maximum 15 percent).
  - Currency mix: about 90 percent euro, rest in other advanced economies' currencies.
- Authorities are reviewing investment strategy to upgrade risk management and increase geographic and asset-class diversification.

### C. Benchmarking and cross-country comparison — salient metrics
- Life expectancy: 84.5 years; life expectancy at retirement: about 19.5 years.
- Old-age dependency: 20.7 percent (current), EU average 32.3 percent; expected to almost double by 2070.
- Coverage rate: expected to decline from current low level of three members per pensioner.
- Generosity rate (average pension / average contribution) in 2020: 2.35.
- Contribution rate comparison: Andorra 12 percent of gross salaries vs EU average 21 percent.
- Replacement rate: 50 percent (above EU average).
- Andorra currently has the lowest pension expenditure in the region but is expected to have the largest increase in the next two decades.

### D. Methodology — diagnostic approach and data sources
- Uses a simplified diagnostic to estimate pension expenditure and deficits, analyzing:
  - Share of elderly in the population, retirement age, benefit indexation, and pension system structure.
- Compares measures of long-term sustainability and contingent liabilities under:
  - No policy change scenario.
  - Alternative scenarios with changes to key pension parameters.
- Inputs and data sources: background materials on the pension system and past reforms; actuarial reports and underlying data; EC's 2021 Ageing Report; OECD; Eurostat; UN population projections; IMF; Andorran authorities; IMF staff calculations.

### E. Scenario analysis — projections and reform scenarios
- No Policy Change Scenario:
  - Assumptions: long-term GDP growth 1.5 percent, inflation 1.7 percent, real return on fund investment 1 percent, population projections in line with France and Spain.
  - Projection: annual deficits starting in 2024; reserves depleted by 2039 without government support.
  - Deficits rise to about 9 percent of GDP per year by 2040.
  - Under the law, deficits are covered by the Fons de Reserva de Jubilació up to the amount of its previous year gains on investments and by the central government for the rest.
  - Pension income set to decline due to shrinking working-age population; pension expenses set to increase driven by population aging.

- Alternative reform scenarios modeled:
  i) One-time increase in the contribution rate from 12 to 16 percent of gross salary in 2023.
  ii) Gradual increase in the conversion factor from 9.6 to 17 in 25 years.
  iii) Gradual increase in the retirement age from 65 to 67 years in 8 years.

- Scenario results (summary):
  - Contribution rate increase (12 → 16 percent): delays deficits to 2027; reserves decline from 62.7 percent of GDP in 2022 to 17 percent of GDP in 2040.
  - Conversion factor increase (9.6 → 17 over 25 years): deficits still start in 2024 (gradual implementation), estimated deficit by 2040 is 4.2 percent of GDP; reserves decline to 27.9 percent of GDP by end of projection period.
  - Retirement age increase (65 → 67 over 8 years): positive impact but insufficient alone because of high old-age dependency ratio and life expectancy.
  - Combined reform (all three implemented):
    - Projected increase in expenditure: 3.3 percent of GDP (present discounted value of 28.3 percent of GDP) vs 7.8 percent of GDP under no policy change (present discounted value of 63 percent of GDP).
    - System accumulates small deficits starting in 2031, reaching up to 3 percent of GDP by 2040, with government support starting in 2037.
    - Reserves decline from 62.3 percent of GDP currently to 59.9 percent of GDP at the end of the projection period.
    - Conclusion: only joint implementation of the three reforms substantially improves sustainability.

- Quantitative decomposition and outcomes reported:
  - Projected spending increase 2022-2030: 3.3 percentage points of GDP; present discounted value 13.5 percent of GDP.
  - Projected spending increase 2022-2040: 7.8 percentage points of GDP; present discounted value 63.0 percent of GDP.
  - Impact of reforms (2022-2040 present discounted values, % of GDP):
    - Conversion factor: -31.5
    - Contribution rate: 3.8
    - Retirement age: -7.0
    - Net change (All reforms): 28.3

### F. Policy implications — recommendations and considerations
- Reforms need to be comprehensive, combining:
  - Increases in the contribution rate.
  - Increases in the conversion factor.
  - Increases in the retirement age.
- Standalone measures produce improvements but are insufficient; the joint package yields the largest and most durable improvement in sustainability.
- Reform design should weigh trade-offs and policy goals beyond sustainability: pension adequacy, affordability, intergenerational equity, and other distributional considerations.

*Prepared by Michelle Tejada (EUR).*

### 17.      A comprehensive reform package is essential to reduce contingent liabilities and

### 17.      A comprehensive reform package is essential to reduce contingent liabilities and 

### Pension system: reform scenarios, projections, and trade-offs
- Context and urgency
  - The system has been adequately managed, but the country’s aging population combined with substantial structural imbalances is exacerbating spending pressures.
  - Parliament’s special commission is working on a reform plan to be approved by end-2022.

- Scenario analysis and recommended combination of measures
  - All three individual scenarios improve the financial health of pensions compared to the no-policy change scenario.
  - Only a combination of the three alternative scenarios would significantly ensure the financial soundness of the Andorran pension system:
    - Increase the contribution rate from 12 to 16 percent of gross salary.
    - Increase the conversion factor from 9.6 to 17.
    - Increase the retirement age from 65 to 67.

- Projected fiscal outcomes under the combined reform
  - The system accumulates small deficits starting in 2031.
  - Deficits reach up to 3 percent of GDP by 2040.
  - Government support would start in 2037.

- Role of second-pillar pensions
  - Expanding the “second pillar” pension scheme could help achieve pension adequacy.
  - Expansion of the second pillar should not supplant the measures needed to achieve financial sustainability of the “first pillar” scheme.

- Equity and policy trade-offs
  - Financial sustainability considerations need to keep in mind equity issues and other important principles and trade-offs that are public policy choices beyond the scope of this analysis.
  - A public pension system has two basic objectives: to protect participants against income poverty and to limit the decline in consumption after retirement.
  - Public policy choices must be made to achieve the right balance between actuarial fairness, intra and intergenerational equity, incentive compatibility, affordability, and adequacy of pensions.

### Banking sector: size, business model, and structural vulnerabilities
- Size and economic importance
  - The financial sector (banks and insurance companies) accounts for 14 percent of GDP and employs approximately 5 percent of the workforce.
  - Consolidated bank assets amount to about 600 percent of GDP.
  - Assets under management (AUM)—largely off-balance sheet—are nearly 23 times the GDP.
  - A banking sector of this size, comprising only three banking groups, implies that all banks are systemic in nature.

- Ownership and business focus
  - The three banking groups in the system are mainly owned by Andorran families.
  - The banking groups combine domestic commercial banking with private banking both in Andorra and internationally.
  - Banks are specialized in private banking and wealth management; about 80 percent of the banks’ income corresponds to non-interest income sources (largely commissions).

- Performance and efficiency
  - Profitability of Andorran banks remained relatively stable in recent years despite large exogenous shocks.
  - The private banking model is associated with higher cost-to-income ratios than those in commercial banking; cost efficiency in Andorran banks has been weak.
  - Ongoing consolidation should help reduce the cost-to-income ratio, but the supervisor should ensure it does not significantly hinder competition.
  - Continued AUM growth would increase efficiency by covering fixed costs and bringing additional commissions.

- Internationalization and geographic footprint
  - The internationalization strategy of the past 10 years geographically diversified business and led to a significant increase of AUMs.
  - By end-2021, about 65 percent of AUMs were managed in 10 countries outside Andorra.
  - AUMs more than doubled since 2009 and by end-2021 they were nearly 23 times Andorra’s GDP, of which nearly 80 percent are off-balance sheet.

- Capital and liquidity buffers
  - Despite shocks (transparency process, BPA crisis, COVID-19), Andorran banks maintained solid solvency and liquidity ratios.
  - Capital and liquidity ratios have been stable and well-above the EU weighted average in recent years.
  - The LCR, under baseline assumptions, results in an LCR of 217 percent.

- Key risks and stress findings
  - Reliance on foreign deposit funding exposes the system to liquidity risks.
  - In an extremely adverse scenario (outflow rates up to four times baseline), the system LCR would fall below the 100 percent requirement and there would be a liquidity gap of 4.8 percent of GDP.
    - Adverse scenario LCR = 94 percent.
    - Baseline scenario LCR = 217 percent.
  - Banks that display sharp declines in the LCR tend to rely more on funding from retail deposits (particularly less stable ones), unsecured funding from nonfinancial corporates and sovereigns not covered by deposit insurance, or on undrawn but committed credit facilities to retail and other nonfinancial customers.
  - Private banking activities and operations through subsidiaries abroad are vulnerable to money laundering and terrorism financing (ML/TF) risks; the 2020 National Risk Assessment (UIFAND) qualifies the vulnerability to ML/TF of private banking and of subsidiaries abroad to be medium-high, mitigated to medium by relatively good control systems.
  - Operating in a small and low-diversified economy raises the risk of large exposures and related-party lending, leading to concentration of credit risks.

- Recommended supervisory and policy actions
  - Enhance liquidity supervision.
  - Continue improving the understanding of AML/CFT risks.
  - Reduce the concentration of credit risks by limiting related party lending and large exposures.
  - Maintain vigilance given the reliance on foreign funding and the systemic nature of the three banking groups.

*Italic: Source: 1andea2022002 - 17.      A comprehensive reform package is essential to reduce contingent liabilities and*

### 13.      The internationalization strategy exposes the system to external risks which may limit

### 13.      The internationalization strategy exposes the system to external risks which may limit

### Internationalization and external risks
- Andorran banks expanded international presence to support profitability and geographical diversification and to limit exposure to volatility of the local economy.
- Internationalization exposes banks to foreign-market risks and vulnerabilities, including:
  - foreign exchange fluctuations
  - local competition
  - local culture
  - regulatory environment
  - economic and political instability
  - market imperfections and asymmetric information
  - operational diseconomies associated with monitoring from a distance
- Empirical evidence: Berger et al (2016) find a positive relationship between the degree of internationalization and bank risk for U.S. banks.

### Large banking sector in a small and low-diversified economy
- Total consolidated assets of the Andorran banking entities are 600 percent of GDP.
- Assets under management (AUMs)—largely off-balance sheet—amount to nearly 23 percent of GDP.
- Any exogenous shocks to Andorran banks would directly affect Andorra’s economy given their systemic nature; public finances would be impacted via the materialization of contingent liabilities in an extremely adverse scenario.
- Operating in a low-diversified economy (LDE) entails an inherent risk of loan concentration (large exposures to a few clients or sectors), increasing sensitivity of loan losses to economic downturns.
- Loan concentration and a focus on private banking may lower banks’ incentives to expand domestic credit:
  - To compensate for concentration risks, banks require a higher return and may select only the best lending opportunities, generating less lending in equilibrium.
  - Narain et al (2003) show domestic credit tends to be lower in LDEs, although higher income LDEs may mitigate risks and be less constrained by lack of diversification.
- Observed pattern: Credit to the private non-financial sector tends to be lower than in other advanced economies (chart referenced: Total credit to the private non-financial sector in advanced economies (Percent of GDP, 2020)).

### Policy implications — Liquidity risks
- Recommendations for the AFA:
  - Further reinforce supervision of liquidity risk by enhancing monitoring and stress testing of net liquidity outflows.
  - Undertake analysis of liquidity risks for longer horizons (the LCR considers only a 30-day horizon).
  - Include onsite liquidity inspections in supervisory planning.
  - Accelerate development and review of liquidity contingency plans by banks.
  - Strengthen the AFA more broadly, including by more solid funding and staffing.

### Policy implications — AML/CFT risks
- Continue improving effectiveness of the AML/CFT framework to ensure financial stability.
- Actions advised:
  - Persevere in aligning the AML/CFT framework to FATF standards, including in the area of virtual assets.
  - Continue refining the national risk assessment and act upon findings of the 2020 assessment to mitigate risks and make supervision more effective.
  - Carefully analyze data on cross-border transactions and exchange information with counterparts in relevant originating countries to improve understanding of ML/TF risks.
  - Focus AML/CFT supervisory resources increasingly on higher risk areas.
  - Effectively impose sanctions for breach of regulatory requirements.

### Policy implications — Absence of lender of last resort (LOLR) and reserves
- No LOLR is in place.
- Recent buildup of international reserves—up to the reference level estimated by IMF staff of 8.7 percent of GDP—may help the government cope with materialization of financial-sector-related contingent liabilities.

### Policy implications — Loan concentration and regulatory measures
- Andorra has transposed into its legal framework the EU regulation limiting large exposures to a given client to be less than 25 percent of the bank’s capital.
- The AFA reinforced reporting requirements on large exposures.
- Implementation of this regulation will be critical to keep risks at bay, especially as banks increase their financing of the economy.
- Planned diversification of the economy should help mitigate concentration risks and create incentives to expand domestic credit, supporting medium-term growth.

*Source: IMF staff compilation from the provided chapter content.*

---


_Source: https://www.imf.org/-/media/files/publications/cr/2022/english/1andea2022002.pdf_