## 1chlea2023004

## Source details

**Canonical URL:** [1chlea2023004](https://www.imf.org/-/media/files/publications/cr/2023/english/1chlea2023004.pdf)

## Other formats

- [Markdown version](/-/media/files/publications/cr/2023/english/1chlea2023004.pdf.md)
- [Structured JSON version](/-/media/files/publications/cr/2023/english/1chlea2023004.pdf.json)

---

### Executive summary and main projections
- Chile’s fiscal framework and sovereign wealth funds (SWFs) have supported macroeconomic stability and crisis response.
- Recent upgrades: adoption of a medium-term fiscal path, formalization of a prudent debt ceiling, and introduction of an escape clause.
- Macro tail risks identified: a significant global slowdown and tightening of global financial conditions; severe terms of trade shocks; major natural disasters; rising social discontents and political gridlocks.
- Quantitative assessment:
  - Keep debt well below the current prudent ceiling on central government gross debt at 45 percent of GDP.
  - Additional fiscal buffers may be required to manage tail risks.
- Illustrative guidance:
  - Maintain a liquid asset buffer in the Economic and Social Stabilization Fund (ESSF) equivalent to 5-7 percent of GDP over the medium term to help mitigate liquidity needs from tail risks.
- Policy trade-off:
  - Self-insurance via liquid assets has opportunity costs; optimal design requires an integrated asset-liability framework and social-choice considerations about risk tolerance.

### Key recommendations (timing indicated)
- Section II (SWF governance and investment)
  - Continue to maintain a high degree of flexibility on the in/outflows with the ESSF. (ST)
  - Continue to align the investment policies of SWF with its objectives and pursue its integration to the sovereign asset-liability management strategy. (ST)
  - Conduct periodic review of investment strategies, accounting for the government’s risk preference and investment horizon. (MT)
- Section III (fiscal risk management)
  - Continue to strengthen fiscal risk management such as developing an integrated sovereign asset and liability management framework. (ST)
- Section IV (rebuilding fiscal buffers)
  - Rebuild fiscal buffers over the medium term by gradual fiscal efforts; adjust the pace according to the risk tolerance of the society and the nature of shocks; avoid accumulating liquidity buffers through borrowings. (MT)
  - Apply an integrated asset-liability management to consider jointly the gross debt level and the size of prudent liquidity buffers of ESSF. (ST)
    - Maintain the current prudent ceiling on central government gross debt at 45 percent of GDP. Avoid setting dual anchors on government debt ceiling and a floor on stabilization fund in the Fiscal Responsibility Law. (ST)
    - Given the current debt level and the expected fiscal adjustments, a prudent range of liquidity buffers in the ESSF of 5-7 percent of GDP over the medium term can help respond to adverse shocks. (MT)
- Timing note: ‘Short-term’ indicates the recommendations can be completed by end-2023; ‘medium-term’ indicates implementation over the next 2-3 years.

### SWFs, purpose, balances, and operational design
- Fund roles and end-2022 balances:
  - Economic and Social Stabilization Fund (ESSF): purpose to finance fiscal deficits and stabilize primary expenditures; balance at end-2022: USD 7.5 billion (2.4 percent of GDP).
  - Pension Reserve Fund (PRF): purpose to support State guarantee for pension/disability solidarity benefits; balance at end-2022: USD 6.5 billion (2.1 percent of GDP).
  - Fund for Natural Disasters (FODEN): created in 2022; will receive contributions up to a total amount of USD 1 billion (0.3 percent of 2022 GDP).
- Governance and management:
  - Ministry of Finance has primary responsibility under the FRL; Minister decides investment implementation: Treasury, Central Bank of Chile (BCCh), or private asset managers.
  - Current practice: SWFs managed by central bank and private asset managers under an investment policy and framework established in the Fiscal Responsibility Law.
  - A Financial Committee acts as an external advisory board; ESSF and PRF governance and disclosures "match high standards".
- Contribution/withdrawal rules (selected):
  - PRF: minimum annual contribution of 0.2 percent of GDP of the previous year; maximum contribution of 0.5 percent of GDP in the previous year; overall cap set at 900 million Unidad de Fomento (approximately 39 billion USD); since 2022, withdrawals shall not exceed 0.1 percent of previous year GDP.
  - ESSF: inflows/outflows flexible; any positive budget balance after deducting PRF contributions and prior-year public debt payments advances is transferred to the ESSF; closely linked to the structural balance fiscal rules.
  - FODEN: can be used only when the government declares a state of emergency and costs represent at least 0.5 percent of total budget expenses of the current year.

### Recent fiscal operations and asset trends
- Pandemic-era withdrawals:
  - Withdrawals from ESSF in 2020-21: USD 10.2 billion.
  - Use of PRF in 2020-21: USD 3 billion.
  - Annual contributions to the PRF temporarily suspended during 2020-21.
- Replenishment in 2022:
  - Budget transfers to ESSF: USD 6 billion in 2022.
  - Budget transfers to PRF: USD 0.5 billion in 2022 (complies with 0.2 percent of GDP requirement).
- Observed trends:
  - Fund assets declined significantly during the pandemic; prior to the pandemic funds had experienced gradual diminution in years before 2020.

### Macro tail risks facing Chile and implications
- Key tail risks:
  - Significant global slowdown and tightening of global financial conditions.
  - Severe terms of trade shocks from commodities.
  - Major natural disasters and epidemics.
  - Other risks: high inflation, stagnated real wage growth, widening inequality, social unrests, political gridlocks, cyber-threats.
- Historical fiscal impacts and calibrations:
  - COVID-19 fiscal measures announced in 2020-2021: fiscal measures at 13 percent of GDP; debt rose by nearly 10 percentage points of GDP relative to pre-pandemic levels.
  - Major natural disasters can raise government expenditures by an average of 15 percent and lower revenue by 10 percent over five years (Melecky and Raddatz 2011); 2010 earthquake costs up to 14 percent of GDP; large earthquakes (damages over 2 percent of GDP) have occurred 6 times in Chile since 1960 and cost an average of 6 percent of GDP.
  - Mining revenue fell abruptly by over 2 percentage points of GDP during the global financial crisis.
- Management instruments:
  - Build fiscal buffers, develop risk-sharing instruments, quantify vulnerabilities, strengthen counter-cyclical responses.
  - Contingency financing mix: self-insurance, grants/borrowed resources, risk transfer (insurance, catastrophe bonds, multilateral instruments).
  - Example target: reserves could aim for 3 percent of total spending to manage fiscal risks from natural disasters (Cebotari and others 2009).

### Quantitative risk-based assessment and simulated anchors
- Current positions and anchors:
  - Current ESSF balance: 2½ percent of GDP.
  - Gross debt as of end-2022: 37 percent of GDP (35 percent net of ESSF assets).
  - Accounting for current ESSF balance, existing net debt anchor is around 42 percent of GDP.
  - Current prudent ceiling on central government gross debt: 45 percent of GDP.
- Models used:
  - Eyraud and others (2018) risk-based probabilistic approach (extended for asymmetric growth shocks).
  - Small-open economy DSGE model (DIGNAR / Melina and others 2016) with fiscal instruments and sovereign wealth fund; 1,000 shock scenarios based on a VAR.
- Key calibration choices and results:
  - Conservative debt limit used for calibration: 55 percent of GDP.
  - Under assumptions: elasticity of debt-to-GDP ratio to interest rates = 3.1; neutral interest rate assumed to be 6 percent (average between 2007 and 2022).
  - If policymakers accept a 5-percent probability that debt would exceed 55 percent of GDP, maintaining a 45 percent of GDP ceiling is appropriate.
  - Simulation result: high likelihood (90-95 percent probability) that debt would remain below the prudent ceiling over the medium term from current levels.

### Sensitivity and buffer sizing
- Sensitivity factors and illustrative impacts:
  - Asymmetric growth shock calibrated to cause a larger drop in output would require an additional safety buffer of about 5 percent of GDP.
  - A major natural disaster calibrated to a loss of 12 percent of GDP would require an additional safety buffer of about 7½ percent of GDP.
  - Stronger countercyclical fiscal responses tend to lower the debt anchor but increase the prudent buffer by about 2-3 percentage points of GDP.
  - Contingent liabilities of 0.8 percent of GDP per year (alternative) lead to -3pp change in debt anchor relative to baseline 0.3 percent of GDP.
- Overall indicative range:
  - Additional fiscal buffers of 3.5-7 percentage points of GDP could serve as fiscal buffers under plausible scenarios accounting for tail risks and recent volatility.
  - Lower bound 3.5 percent of GDP based on differences between baseline and “tail event on growth” scenarios (difference between 44.7 and 41.2 percent of GDP).
  - Upper bound 7 percent of GDP reflects empirical simulations in the IMF-FAD toolkit accounting for greater recent output volatility (sample 2000-2022 relative to 1990-2022).

### Illustrative DSGE-based simulations for liquidity needs and buffer guidance
- Simulation setup:
  - Cumulative gross financing needs for two years comprise rollover of maturing debt and counter-cyclical fiscal responses.
  - Debt rollover needs set at 0.5 percent of GDP each year.
  - In adverse shock: government refrains from borrowing in markets in first year; in second year issues debt to cover 40 percent of financing requirements.
- Simulated outcomes:
  - Simulated cumulative financing needs (two years) range from 5.1 to 11.6 percent of GDP.
  - Of that, 3.4 to 7.3 percent may require liquid assets or other financing.
  - Maintaining a liquid asset buffer in the ESSF of about 5-7 percent of GDP can help buffer against financing needs in a tail event for the first two years.
  - Baseline ESSF use example: 5.4 percent of GDP.
- Assumptions informing bounds:
  - 60 percent of simulated fiscal responses occur in year 1 and 40 percent in year 2.
  - Lower bound 3.4 percent corresponds to “less proactive countercyclical fiscal policy” scenario (2.9 percent year 1 + 0.5 percent year 2).
  - Upper bound 7.3 percent corresponds to “low risk tolerance” scenario (6.8 percent year 1 + 0.5 percent year 2).

### Policy implications, trade-offs, and operational design
- Maintain gross debt ceiling at 45 percent of GDP and keep debt well below that ceiling over the medium term.
- Preferred strategy to rebuild buffers:
  - Accumulate asset buffers through fiscal efforts to improve structural balances gradually rather than borrowing to save.
  - Avoid borrowing at high interest rates to accumulate assets in the stabilization fund.
- Composition trade-offs:
  - Liquid assets provide immediate liquidity during tail events but earn lower returns than government borrowing costs on average; accumulation implies an insurance premium.
  - Integrated sovereign asset-liability management (SALM) recommended to jointly consider gross debt level and size/composition of liquidity buffers.
  - Avoid statutory dual anchors (fixed debt ceiling and an explicit legal floor for stabilization fund) to prevent perverse incentives; use non-binding projections or indicative targets in MTFF documents instead.
- Operational recommendations:
  - Continue flexibility on ESSF in/outflows to facilitate fiscal rule implementation.
  - Align SWF investment policies with objectives and integrate with the Treasury’s SALM strategy; periodic reviews of investment strategies are appropriate.
  - Strengthen fiscal risk management and publish regular risk analyses and scenario analyses (Fiscal Risk Statements alongside budgets).

### Model results and illustrative tail-event impacts (DIGNAR DSGE highlights)
- Natural resource production shock (tail event: 4.2 percent decline in copper production growth):
  - GDP growth declines by more than 0.5 percentage point.
  - Primary deficits rise by 0.2 percentage of GDP.
  - Government debt rises by more than one percentage of GDP.
- Terms of trade (copper price) shock (tail event: 50 percent fall for a year):
  - GDP declines by 4 percent.
  - Fiscal deficits rise by 2.5 percentage points of GDP.
  - Government debt rises by 4.5 percentage points of GDP.
- Major natural disaster (TFP shock: 12-percent drop in TFP):
  - GDP growth drops sharply by more than 6 percentage points.
  - Primary deficits rise by 4.5 percentage points of GDP.
  - Government debt rises by over 14 percentage points of GDP.
- Surge in external risk premium (sovereign spreads rise by 250 basis points):
  - External debt limited at 14.6 percent of GDP; refinancing costs and interest payments rise; deficits widen; limited output effect in current specification.
- Probabilistic medium-term outcome:
  - With government debt at 37 percent of GDP as of end-2022, there is an 11 percent probability that debt would exceed the prudent level of 45 percent of GDP in the medium term based on 1,000-simulation VAR-based distribution.

*Source: IMF Fiscal Affairs Department mission report “Recent Fiscal Developments in Chile,” mission conducted March 17-23, 2023; IMF staff analysis and simulations contained in the chapter provided.*

### 1. Recent Fiscal Developments in Chile ___________________________________________________________ 9

### 1. Recent Fiscal Developments in Chile

### Executive summary: main findings and projections
- Chile’s fiscal framework and sovereign wealth funds (SWFs) have supported macroeconomic stability and crisis response.  
- Recent upgrades: adoption of a medium-term fiscal path, formalization of a prudent debt ceiling, and introduction of an escape clause.  
- Macro tail risks identified: a significant global slowdown and tightening of global financial conditions; severe terms of trade shocks; major natural disasters; rising social discontents and political gridlocks.  
- Quantitative assessment: keeping debt well below the current prudent ceiling on central government gross debt at 45 percent of GDP is appropriate; additional fiscal buffers may be required to manage tail risks.  
- Illustrative guidance: a liquid asset buffer in the Economic and Social Stabilization Fund (ESSF) equivalent to 5-7 percent of GDP over the medium term can help mitigate liquidity needs stemming from tail risks.  
- Policy trade-off: self-insurance via liquid assets has opportunity costs; optimal design requires an integrated asset-liability framework and social-choice considerations about risk tolerance.

### Key recommendations (timing indicated)
- Section II (SWF governance and investment)
  - Continue to maintain a high degree of flexibility on the in/outflows with the ESSF. (ST)
  - Continue to align the investment policies of SWF with its objectives and pursue its integration to the sovereign asset-liability management strategy. (ST)
  - Conduct periodic review of investment strategies, accounting for the government’s risk preference and investment horizon. (MT)
- Section III (fiscal risk management)
  - Continue to strengthen fiscal risk management such as developing an integrated sovereign asset and liability management framework. (ST)
- Section IV (rebuilding fiscal buffers)
  - Rebuild fiscal buffers over the medium term by gradual fiscal efforts; adjust the pace according to the risk tolerance of the society and the nature of shocks; avoid accumulating liquidity buffers through borrowings. (MT)
  - Apply an integrated asset-liability management to consider jointly the gross debt level and the size of prudent liquidity buffers of ESSF. (ST)
    - Maintain the current prudent ceiling on central government gross debt at 45 percent of GDP. Avoid setting dual anchors on government debt ceiling and a floor on stabilization fund in the Fiscal Responsibility Law. (ST)
    - Given the current debt level and the expected fiscal adjustments, a prudent range of liquidity buffers in the ESSF of 5-7 percent of GDP over the medium term can help respond to adverse shocks. (MT)
- Timing note: ‘Short-term’ indicates the recommendations can be completed by end-2023; ‘medium-term’ indicates implementation over the next 2-3 years.

### Introduction: context, recent actions, and risks
- Pandemic response and aftermath
  - During the pandemic the authorities adopted multi-pronged responses, including drawing down stabilization funds to protect people.  
  - Effective policy response aided recovery but built macroeconomic imbalances; authorities have tightened monetary and fiscal policies to preserve macro stability and rebuild buffers.
- Fiscal framework upgrades
  - Enhanced the structural balance rule, adopted a medium-term fiscal path, introduced a prudent debt ceiling, and disclosed sensitivity of fiscal projections to macro shocks.
  - Investment strategy of SWFs refined; further enhancement recommended for government balance sheet analysis, fiscal risks, and broadening the role of the Autonomous Fiscal Council (AFC).
- Need to rebuild buffers
  - The government used SWFs during the pandemic; moving forward there is a need to replenish funds to face future shocks, including macro tail risks (low probability, high impact).
  - The Flexible Credit Line (FCL) provides additional precautionary external buffers and insurance against tail risks.

### International experience in stabilization funds and Chile’s SWFs
- Purpose and structure of Chile’s SWFs (Box 1)
  - Chile maintains two types of funds: a stabilization fund to insulate the budget from volatile commodity prices, and a saving fund for longer-term accumulation.
  - Economic and Social Stabilization Fund (ESSF)
    - Purpose: finance fiscal deficits and stabilize primary expenditures arising from low growth and/or low copper prices; ESSF can be used for debt service and contributions to the PRF.
    - Balance at end-2022: USD 7.5 billion (2.4 percent of GDP).
  - Pension Reserve Fund (PRF)
    - Purpose: support the State guarantee for pension and disability solidarity benefits; complements funding of fiscal obligations from the Universal Guaranteed Pension and disability solidarity pillar.
    - Balance as of end-2022: USD 6.5 billion (2.1 percent of GDP).
  - Fund for Natural Disasters (FODEN)
    - Created in 2022 by modification of the 2006 Fiscal Responsibility Law.
    - Objective: finance expenses and fiscal initiatives related to natural disasters.
    - Will receive contributions from the sale of assets, cash surpluses from the Public Treasury, and from the ESSF, up to a total amount of USD 1 billion (0.3 percent of 2022 GDP).
- Cross-country lessons
  - Adequate buffers in stabilization funds facilitate government responses to shocks.
  - Rigid accumulation/withdrawal mechanisms can lead to sub-optimal risk and liquidity management; flexibility on in/outflows is important.
  - Investment policy for SWFs should align with objectives and integrate with sovereign asset-liability management; periodic reviews of investment strategies are appropriate.

### Recent fiscal operations and numbers (pandemic and near-term replenishment)
- Pandemic-era use of funds
  - Withdrawals from ESSF in 2020-21: USD 10.2 billion (to help finance sizable support to households and firms under the Economic Emergency Plan, decree 21.225/2020).  
  - Use of PRF in 2020-21: USD 3 billion (to support the solidarity pillar of pension for vulnerable households).  
  - Annual contributions to the PRF were temporarily suspended during 2020-21.
- Replenishment in 2022
  - Budget transfers to ESSF: USD 6 billion in 2022.
  - Budget transfers to PRF: USD 0.5 billion in 2022 (in compliance with the requirement of 0.2 percent of GDP under the Fiscal Responsibility Law).
- Observed asset trends
  - Fund assets declined significantly during the pandemic; prior to the pandemic Chile had experienced gradual diminution of fund resources in years before 2020.

### Fiscal strategy implications and policy design
- Role of SWF liquidity versus debt space
  - Maintaining liquid assets in the ESSF is optimal when tail risks could cause a sharp rise in borrowing costs or constrain market access; liquid assets provide time to recalibrate policy and implement fiscal responses.
  - Rebuilding buffers can occur by reducing structural deficits gradually or by saving stronger-than-expected revenues (including from natural resources).
  - The government should avoid borrowing at high interest rates to accumulate assets in the stabilization fund.
- Integrated approach
  - Apply integrated sovereign asset-liability management to jointly consider gross debt level and size/composition of liquidity buffers in ESSF.
  - Avoid dual statutory anchors that both fix a debt ceiling and mandate a floor for the stabilization fund in the Fiscal Responsibility Law.

### Quantitative assessment highlights and illustrative simulations
- Prudent debt ceiling
  - IMF analysis complements the structural balance rule by recommending debt remain well below the current prudent ceiling on central government gross debt at 45 percent of GDP.
- Illustrative liquidity buffer
  - Simulations indicate a liquid asset in the ESSF equivalent to 5-7 percent of GDP can help mitigate liquidity needs from tail risks.
- Social-choice and cost considerations
  - The size of prudent buffers depends on shock size/persistence, societal risk tolerance, and the degree of counter-cyclical fiscal response during crises.
  - Self-insurance through liquid buffers entails opportunity costs; balancing liquidity needs and fiscal costs is a policy choice.

*Source: IMF Fiscal Affairs Department mission report “Recent Fiscal Developments in Chile,” mission conducted March 17-23, 2023; IMF staff analysis and simulations contained in the chapter provided.*

### 7. Provisions for contributions to and withdrawals from the funds are well established

### 7. Provisions for contributions to and withdrawals from the funds are well established

### Contributions and withdrawals: PRF, ESSF, and FODEN
- PRF (Pensions Reserve Fund):
  - Receives a minimum annual contribution of 0.2 percent of GDP of the previous year, according to the FRL.
  - Additional contributions possible if the effective fiscal surplus is high, with a maximum contribution of 0.5 percent of GDP in the previous year.
  - Overall cap set at 900 million Unidad de Fomento (approximately 39 billion USD).
  - Funds are earmarked to the solidarity spending on pensions.
  - Since 2022, fund withdrawals shall not exceed 0.1 percent of previous year GDP.
- ESSF (Economic and Social Stabilization Fund):
  - Inflows/outflows are flexible and adhere to international best practices (similar to Norway).
  - Funds can be used to finance the budget deficit, to complement fiscal revenues to finance public expenditures, and to service public debt or contribute to the PRF.
  - Any positive balance in the budget, after deducting PRF contributions and payment of public debt and advances made the year before, will be transferred to the ESSF; the year-end balance is carried forward to next year.
  - The ESSF is closely linked to the structural balance fiscal rules.
  - Persistent deficits during 2015-19 contributed to a gradual decline in ESSF assets even before the pandemic.
- FODEN (disaster risk fund):
  - Resources can be used only when the government declares a state of emergency and the natural disaster costs represent at least 0.5 percent of total budget expenses of the current year.
  - Resources will be exclusively used to finance expenses derived from the event, and possible insurance premiums or hedging subscribed by Chile against natural disasters.

### Governance and management responsibilities
- Ministry of Finance has primary responsibility for managing the SWFs under the FRL; the Minister decides how funds are invested: directly by the Treasury, by the Central Bank of Chile (BCCh), or by private asset managers.
- Current practice: SWFs are managed by the central bank and private asset managers under an investment policy and framework established in the Fiscal Responsibility Law.
- A Financial Committee acts as an external advisory board to provide advice on the long-term investment policy of the SWFs.
- The ESSF and PRF governance and disclosures "match high standards," with regular publications of reports.
- Authorities recently undertook a review of the ESSF investment strategy.

### Asset allocation, liquidity, and recent performance
- Asset allocation differs between ESSF and PRF due to different objectives and horizons:
  - ESSF: emphasis on liquidity and low-risk highly liquid instruments for easy access at short notice; mainly invested in sovereign (or multilateral agencies) bonds.
  - PRF: longer investment horizon and functions as a saving fund; assets invested in equities with a conservative approach, including a majority of investments in highly-rated issuers.
- 2020-22: returns on the SWF in Chile were subject to large volatility and experienced negative returns as in SWFs of other countries.

### Fiscal rule linkages and operational design implications
- ESSF’s flexibility on use helps set operational limits on fiscal aggregates to promote fiscal discipline, rather than restricting means of financing (inflows/outflows).
- Structural balance rule:
  - Under the structural balance rule, cyclically-adjusted government revenues are considered to determine budget expenditures, detaching fiscal spending from cyclical fluctuations in copper prices and the economy.
  - The FRL requires the government to announce its targets for the structural budget balance for the next four years.
  - If the country registers a structural deficit, the ESSF funds can be withdrawn to cover budget expenditures provided the structural balance rule is complied.
- Rigid restrictions on contributions (e.g., fixed amount or percentage of revenues/output independent of economic conditions) could complicate SWF operations and reduce government flexibility in crises.
- Despite general compliance with the structural budget balance rule for most years over the last decade and a half, government debt continued to rise while SWF assets declined gradually, raising questions whether structural balance targets imparted sufficient constraint on fiscal policy.

### International experience and comparative insights (selected)
- Norway:
  - Government Pension Fund Global (GPFG) accumulated almost USD 1.3 trillion, equivalent to 256 percent of GDP as of end-2022.
  - The GPFG finances flexibly the non-oil budget deficit without rigid in/outflow rules; this feature is similar to Chile’s arrangement.
  - Norway’s structural balance rule requires the nonoil structural deficit to reflect the long-run expected return of the GPFG (at 3 percent since 2017).
  - GPFG asset allocation includes a high share of equity (around 70% of total portfolio), which has generated more than 50 percent of the current wealth of the Fund.
- Mexico:
  - Mexican Petroleum Fund (FMP) legislation requires an annual transfer equivalent to 4.7 percent of GDP to the Federal Treasury; the threshold of 4.7 percent of GDP has rarely been reached.
  - In 2022, the FMP received about USD 33 billion oil revenues (or 2.6 percent of GDP), which was transferred to the Federal Treasury and the other two funds.
  - Reserve fund balance was USD 1.2 billion (0.1 percent of GDP) as of end 2022.
  - Budget Revenues Stabilization Fund (FEIP) and Federal Entities Revenues Stabilization Fund (FEIEF) balances were USD 1.3 billion (0.1 percent of GDP) and USD 1.08 billion (0.1 percent of GDP) at end-2022, respectively.
  - During the pandemic, oil revenues and stabilization funds declined sharply; FEIP declined from USD 8.4 billion in 2019 (0.7 percent of GDP) to 476 million in 2020.
  - FMP assets are held in a Trust Fund managed by the Central Bank, mostly in foreign sovereign bills/bonds (81 percent), corporate bonds (10 percent), and inflation-linked instruments.
- Peru:
  - Fiscal Stabilization Fund (FEF) aims at ensuring fiscal balance or surplus in the medium-term; contributions come from budget surpluses and two prescribed transfers related to government asset sales and mining license proceeds.
  - Withdrawal mechanism: if current revenues decrease by more than 0.3 percent of GDP compared to the 3-year moving average, transfer to the budget is limited up to 40 percent of the FEF balance.
  - The FEF registered an extraordinary outflow during the pandemic (about USD 5.5 billion) and was almost depleted; balance was 0.6 percent of GDP in 2022.
  - FEF asset allocation: short-term highly-liquid assets (term deposits in USD with maturity of less than 2 months) constitute almost all (99 percent).
- Colombia:
  - Maintains several sovereign wealth funds with differing objectives; withdrawals from some funds were constrained, and one fund was liquidated in 2020.
  - Fund of Savings and Fiscal and Macroeconomic Stabilization: withdrawals cannot exceed 10 percent of the fund balance in the previous year.
  - Savings and Stabilization Fund (FAE) assets reached USD 3.6 billion (1.1 percent of GDP) in 2022; managed with investments in foreign sovereign and corporate bonds.
  - Fiscal rules include a structural primary balance target and a debt anchor aimed at maintaining a prudent debt level at 55 percent of GDP.
  - During the pandemic, the fiscal rule was suspended and sovereign funds were drawn down; one fund was liquidated.

### Key lessons and implications
- Flexible in/outflow mechanisms linked to structural fiscal rules (as in Chile and Norway) can preserve the stabilization role of SWFs and allow timely government responses to shocks.
- Rigid contribution or withdrawal rules independent of economic conditions can reduce the operational effectiveness of SWFs during crises.
- High pre-pandemic deficits and persistent fiscal weaknesses can erode SWF buffers over time even when fiscal rules exist.
- Investment policies and asset allocations matter: higher equity allocations (as in Norway’s GPFG) can substantially boost long-term fund wealth but imply higher volatility; stabilization funds with low balances tend to hold highly liquid short-term assets (e.g., Peru).
- Governance arrangements that assign clear ministerial authority, allow central bank/private manager involvement, and provide external advisory input support high disclosure and operational standards.

*Source: IMF staff compilation from chapter "7. Provisions for contributions to and withdrawals from the funds are well established" in 1chlea2023004.*

### 25. Lessons from international experience suggests that a strong commitment to fiscal

### 25. Lessons from international experience suggests that a strong commitment to fiscal discipline is needed for SWFs to work well.

### Governance and fiscal discipline for SWFs
- Countries running recurrent, large budget deficits often draw down their funds; insufficient fiscal discipline can eventually deplete the fund and lead governments to borrow to maintain a modest fund balance.
- Fiscal discipline should be imposed directly at the budget level with well-designed fiscal rules or medium-term fiscal frameworks (MTFFs); funds should help save for the future or stabilize spending during adverse shocks.
- Funds should not try to ‘discipline’ spending by forcibly removing resources from the budget; they should be well integrated into the budget process and foster transparency and public understanding without imposing rigidities on asset-liability management.

### Accumulation and withdrawal mechanisms for stabilization funds
- Rigid rules (fixed annual amounts or constant percentages of receipts) to transfer to/withdraw from stabilization funds often interfere with optimal risk and liquidity management.
- A rigid rule could force the government to borrow to save even when there is a budget deficit.
- Flexible mechanisms (e.g., transferring the amount exceeding the structural budget balance under the fiscal rule, as in Norway, Chile) allow automatic accumulation during good times and use of funds during difficult times.
- Flexibility should not be misused to finance regular spending within a fiscal year based on short-term fiscal cash requirements beyond macroeconomic stabilization.
- Alignment is important: ensure the in/outflow mechanism of stabilization funds aligns with fiscal rules and is integrated with the fiscal policy framework.
- Desired size of liquidity buffers can be estimated based on a risk-based approach considering the level of buffers adequate to protect spending and avoid abrupt cuts during adverse shocks.

### Flexibility and escape clauses to respond to crises
- Governments should build high flexibility in the use of stabilization funds alongside a well-defined escape clause in fiscal rules.
- The COVID-19 pandemic highlighted the need for adequate means and buffers to respond to extraordinary adversity.
- Governments should allow temporary suspension—ideally via activation of an escape clause—of fiscal rules during crises, with clear guidance and a credible correction mechanism to return to compliance.
- Flexible use of funds risks political abuse; risks can be limited by strengthening governance and accountability of in/outflow mechanisms.
- Some forward guidance on fund use helps avoid market perceptions that the state intervenes in market activity.

### Investment policy and asset-liability alignment
- Investment policy should correspond to fund objectives and align with the government’s asset-liability management:
  - For stabilization funds: assets need to be liquid and quickly available; focus on high credit quality and liquidity (foreign sovereign bonds, short-term repos and deposits).
  - For savings / intergenerational equity funds: longer horizon; consider broader asset classes (including equities or financial instruments linked to real estates) for higher expected returns and diversification.
- Investing in foreign assets is preferred to avoid amplifying volatility of natural resources and stress on the domestic financial system.
- Asset management benefits from autonomy; fund managers must have clear directives on investment objectives, eligible assets, strategic asset allocation, and exposure limits, disclosed in the fund’s investment policy and aligned with the Treasury’s broader asset-liability management strategy.

### D. Recommendations (SWF and ESSF)
- Continue to maintain flexibility on the in/outflows with the ESSF, which can facilitate the implementation of fiscal rules.
- Continue to align the investment policies of Chile’s SWF with its objectives and pursue its integration to the sovereign asset-liability management strategy. As the SWF gradually accumulates assets, periodic review of investment strategies, accounting for the government risk preference and investment horizon, will be appropriate.

*Source: IMF staff chapter content.*

### III. MACRO TAIL RISKS FACING CHILE

#### Key tail risks and macro-fiscal implications
- Tail events have low probability but high impact and can trigger large fiscal needs and abrupt disruption in access to or cost of financing.
- Chile has faced significant adverse shocks: global financial crisis in 2008-10; volatile commodity cycles (surge between 2004 and 2013 and reversal since 2014); COVID-19 pandemic; Latin American crisis in early 1980s; Asian Financial Crisis in 1998-99; 2010 earthquake; political turmoil in late 2019.
- Chile’s government debt has remained below 40 percent of GDP, with strong credit ratings and lower spreads than other emerging markets even conditional on debt levels.
- Tail risks can trigger an abrupt rise in fiscal needs for more than one year and/or a temporary surge in spreads and sharp exchange depreciation for about 2-10 months (sometimes even more prolonged), constraining rollover of external debt.

- Specific tail risks discussed:
  - A significant global slowdown and tightening of global financial conditions: could lead to synchronized sharp global slowdown; sharp swings in real interest rates and risk premia; insolvencies; higher fiscal needs; rising debt ratio and borrowing costs.
  - Severe terms of trade shocks from commodities: mining revenue fell abruptly by over 2 percentage points of GDP during the global financial crisis; recurrent commodity price volatility could persist.
  - Major natural disasters: can raise government expenditures by an average of 15 percent and lower revenue by 10 percent over five years (Melecky and Raddatz 2011); large earthquakes in Chile have costs reaching as much as 14 percent of GDP (2010 earthquake); large earthquakes (damages over 2 percent of GDP) have occurred 6 times in Chile since 1960 and cost an average of 6 percent of GDP.
  - Epidemics: COVID-19 prompted sizeable fiscal measures (at 13 percent of GDP) announced in 2020-2021; debt rose by nearly 10 percentage points of GDP relative to pre-pandemic levels; additional fiscal support of 1.3 percent of GDP was part of its Inclusive Recovery Plan in 2022.
  - Other tail risks: high inflation, stagnated real wage growth, widening inequality, social unrests, political gridlocks, cyber-threats (example: Costa Rica in 2022).

### Ways to manage tail risks
- A holistic framework is needed, combining fiscal, monetary, financial policies and emergency measures.
- Fiscal tools to enhance resilience include building fiscal buffers, risk-sharing instruments, quantifying vulnerabilities, and strengthening counter-cyclical responses.

- Building fiscal buffers:
  - Enables swift, forceful response without jeopardizing credibility and sustainability.
  - Can involve keeping prudent debt well below its limit and/or accumulating liquid financial assets.
  - A risk-based approach helps estimate buffer size and optimal instrument mix.

- Enhancing risk-sharing mechanisms:
  - Contingency financing should rely on self-insurance (contingency reserves/stabilization funds), grants/borrowed resources, and risk transfer (insurance or state-contingent debt).
  - Examples: regional insurance (Caribbean Catastrophic Risk Insurance Facility), catastrophe bonds (Barbados, Mexico), African Risk Capacity, IMF Catastrophe Containment and Relief Trust, World Bank Cat DDO.
  - International experience suggests reserves could aim for 3 percent of total spending to manage fiscal risks from natural disasters (Cebotari and others 2009).

- Quantifying vulnerabilities:
  - Regularly integrate probability assessments of frequency/severity of tail risks, potential fiscal costs and financing strategies into fiscal frameworks (budget design, stress tests, public investment planning, debt/asset management).
  - Use IMF Fiscal Risks Toolkits and publish scenario analyses.

- Strengthening counter-cyclical response:
  - Upgrade social protection systems to be more scalable and targeted.
  - Adopt semi-automatic stabilizers that expand in adverse situations in a predetermined way (IMF 2022b).
  - Flexible fiscal rules with well-defined escape clauses allow quick deployment of fiscal support while preserving credibility (Caselli and others 2022).

### Chile-specific assessments and policy implications
- Chile needs to rebuild fiscal buffers after large draw-downs during the pandemic; stabilization fund balances declined and need replenishment in coming years.
- Chile maintains multiple fiscal buffers: stabilization funds, contingency funds, catastrophic bonds, and a precautionary buffer through its IMF FCL program.
- Rebuilding fiscal buffers alongside proposed escape clauses will enhance government flexibility to respond to shocks.

- Analytical and institutional recommendations:
  - Strengthen and publish risk analyses (including scenario analysis) more regularly; a summary reporting of fiscal risks alongside comprehensive debt sustainability analysis—including stochastic projections and scenario analysis—can help quantify a range of fiscal risks.
  - An increasing number of countries publish annual Fiscal Risk Statements alongside budgets (examples cited: Brazil, United Kingdom, New Zealand, Indonesia).
  - Develop an integrated sovereign asset and liability management framework (SALM) to incorporate long-term fiscal risks (demographics, climate change) into fiscal frameworks. SALM can include:
    - types of assets and liabilities considered;
    - assessment of public sector balance sheet exposures and mismatches;
    - government risk tolerance and ability to transfer risks;
    - regular publications and periodic evaluations of implementation.

### C. Recommendations (macro tail risk management)
- Continue to improve fiscal risk management, including developing an integrated sovereign asset-liability management framework.

*Source: IMF staff chapter content.*

### 35. A risk-based approach sets a medium-term fiscal anchor to ensure macro stability

### 35. A risk-based approach sets a medium-term fiscal anchor to ensure macro stability

### Overview and purpose of a risk-based fiscal anchor
- The fiscal anchor is expressed as a prudent level of government debt to output to provide space (buffers) to respond to shocks with a reasonably low risk that debt would exceed a level posing debt distress or large economic costs.
- A risk-based approach simulates shocks to determine buffer size that can prevent an abrupt cut in spending over a chosen horizon within a confidence interval.
- The size of the buffer can be based on what the government considers a minimum degree of confidence that it is adequate to meet adverse shocks.

### Determinants of the size and composition of fiscal buffers
- Size depends on:
  - Size and persistence of shocks.
  - Impact of shocks on the economy.
  - Dependence on natural resources.
  - Risk tolerance in society.
  - Exposure to exchange rate risks (for example, a higher share of foreign currency denominated debt would require a larger buffer).
  - Strength of private-sector resilience (stronger private-sector resilience may require less fiscal buffers).
- Composition considerations:
  - Financing options during adverse shocks: borrow, draw down liquid assets, activate risk-transfer instruments (e.g., regional insurance or catastrophe bonds).
  - The government’s access to liquidity at low cost during tail events determines composition: limited access implies a need for a higher liquidity buffer.
  - The government balance sheet (assets and liabilities) and interactions of fiscal policy in the overall policy mix must be considered.

### Chile’s current framework and options
- Chile has adopted a prudent gross debt ceiling that complements the structural balance rule to maintain debt sustainability.
- Current prudent ceiling of central government gross debt: 45 percent of GDP (established to ensure a prudent safety margin for typical macroeconomic shocks according to 2023 Public Finance Reports, MOF).
- Consideration to accumulate liquid assets given natural resource revenues (e.g., lithium) to reinforce resilience against tail risks by building liquid assets such as in the ESSF.
- Current ESSF balance: 2½ percent of GDP.
- Gross debt was 37 percent of GDP as of end-2022 (or 35 percent net of ESSF assets).
- Accounting for current ESSF balance, existing net debt anchor is around 42 percent of GDP.

### Quantitative models used to estimate prudent anchor and safety buffer
- Two related quantitative, risk-based probabilistic models calibrated to Chile:
  1. Eyraud and others (2018) risk-based approach:
     - Determines a prudent debt level such that debt will stay below an exogenous limit with high likelihood, accounting for shocks affecting debt dynamics.
     - Safety buffer = difference between the limit and the estimated debt anchor.
     - Accounts for a fiscal reaction (in reduced form) from output changes when shocks hit.
     - Stochastic simulations based on a multivariate normal distribution of shocks produce distributions of debt trajectories and probabilities of exceeding limits.
     - Extended using Akanbi, Gbohoui, and Lam (2023) to allow for an asymmetric growth shock for tail events.
  2. Small-open economy DSGE model (Melina and others 2016):
     - Contains fiscal instruments and a sovereign wealth fund, plus a production sector for natural resources.
     - Models interactions between output, consumption, and countercyclical fiscal responses.
     - Government conducts counter-cyclical fiscal policy subject to fiscal rule limits and an escape clause; financing can come from debt or drawing down sovereign fund assets.
     - Calibrated to Chile and simulated with 1,000 shock scenarios where shocks’ magnitude and persistence are estimated using a VAR.

- Historical data used to estimate magnitude and persistence of adverse shocks: 1990 to 2022 (with sensitivity to alternative samples, e.g., 2000-22).

### Key quantitative findings and calibration choices
- Capacity to borrow is uncertain; sustainable debt limit varies with global interest rates and elasticity between interest rates and debt.
- Calibration based on Mian, Straub, and Sufi (2022) suggests a debt limit wide range: 50 percent of GDP if interest rates are high to as much as 120 percent of GDP if interest rates are low and less sensitive to debt.
- For calibrating the prudent debt anchor, a conservative debt limit of 55 percent of GDP is used.
- Maintaining a prudent gross debt ceiling at 45 percent of GDP remains appropriate for typical macroeconomic shocks:
  - If policymakers accept a 5-percent probability that debt would exceed 55 percent of GDP, maintaining a 45 percent of GDP ceiling is appropriate over the medium term.
  - Simulation results indicate a high likelihood (90-95 percent probability) that debt would remain below the prudent ceiling over the medium term from current levels.
- Under the prudent scenario assumptions:
  - Elasticity of debt-to-GDP ratio to interest rates is assumed to be 3.1.
  - Neutral interest rate is assumed to be 6 percent (average level between 2007 and 2022).
  - Note: debt limit would reduce to 50 percent of GDP if neutral rates are at 7 percent.

### Simulated anchors, buffers, and current position
- Table 3 (excerpted values):
  - Debt anchor: 45
  - Safety buffer: 9
  - Debt limit: 54 (baseline) and 60 (alternative scenario?) [Table shows Debt limit 54 60 50 with associated changes; preserved as in source]
  - Risk tolerance (in percent): baseline 5 percent; alternative 10 percent leads to +4pp change in debt anchor.
  - Maximum sustainable primary balance: baseline 2; alternative 1 leads to -8pp change in debt anchor.
  - Contingent liabilities: baseline 0.3 percent of GDP per year; alternative 0.8 percent of GDP per year leads to -3pp change.
  - Sample horizon: baseline 1992-2022; alternative 2000-22 leads to -7pp change.
- Accounting for ESSF at 2½ percent of GDP implies existing net debt anchor around 42 percent of GDP.
- With gross debt at 37 percent of GDP (35 percent net of ESSF) as of end-2022, simulations indicate a 90-95 percent probability debt would remain below the prudent ceiling over the medium term.

### Sensitivity analysis: factors affecting prudent debt level and buffer size
- Primary factors and illustrative impacts:
  - Size and persistence of shocks, particularly tail-risk events: larger/more persistent shocks require larger buffers. An asymmetric growth shock calibrated to cause a larger drop in output would require an additional safety buffer of about 5 percent of GDP.
  - Tail-events (example): a major natural disaster calibrated to a loss of 12 percent of GDP would require an additional safety buffer of about 7½ percent of GDP.
  - Countercyclicality of fiscal responses: stronger countercyclical responses (higher fiscal stimulus during recessions) tend to lower the debt anchor but increase the prudent buffer by about 2-3 percentage points of GDP.
  - Risk tolerance: lower risk appetite (lower tolerance for exceeding debt limit) raises the prudent buffer.
  - Fiscal risks from contingent liabilities (e.g., recapitalization of failing banks or SOEs) increase required buffers.
  - Other available buffers (e.g., liquid assets) can allow higher government debt without harming sustainability; anchor can be interpreted in net debt terms.
- Overall indicative range from judgment and scenarios:
  - Additional fiscal buffers of 3.5-7 percentage points of GDP could serve as fiscal buffers under plausible scenarios accounting for tail risks and recent volatility.
  - The lower bound 3.5 percent of GDP is based on differences between baseline and “tail event on growth” scenarios (difference between 44.7 and 41.2 percent of GDP).
  - The upper bound 7 percent of GDP reflects empirical simulations in the IMF-FAD toolkit accounting for greater recent output volatility (sample 2000-2022 relative to 1990-2022).

### Policy implications and composition trade-offs
- Prudence is warranted: debt too close to the prudent ceiling increases risks of breaching fiscal rules and impairs counter-cyclical policy capacity in adverse shocks.
- Composition choice between debt reduction and liquid assets:
  - Accumulating liquid assets provides a liquidity buffer to avoid rolling over maturing debt at high interest rates during tail events and frees up resources for stabilization.
  - A stabilization fund serves as a liquidity buffer during tail events with extreme financial conditions, providing time to recalibrate and implement fiscal responses.
  - The risk-adjusted return on financial assets is likely lower than the cost of borrowing on average; net loss from accumulating assets rather than repaying debt is an insurance premium against costly rollover in tail events.

*Source: IMF staff estimates.*

### 48. Maintaining certain liquidity in assets could be optimal in face of tail risks. A

### 48. Maintaining certain liquidity in assets could be optimal in face of tail risks. A

### Tail risk characteristics and simulation setup
- Tail events for Chile often:
  - Call for substantial fiscal needs for more than one year, with most occurring in the first year.
  - Feature sovereign spreads surging and exchange rate depreciating sharply in the first year before returning gradually to pre-shock levels.
- Simulation design:
  - Simulate cumulative gross financing needs for two years, comprising (i) rollover of maturing debt and (ii) needs for counter-cyclical fiscal responses.
  - Deficit paths for the first two years are simulated using the DSGE model for the 5-percent shock level over 2 years.
  - Debt rollover needs are set at 0.5 percent of GDP each year (approximately the size of foreign currency-denominated debt service).
  - In the adverse shock, the government refrains from borrowing in markets during the first year; in the second year it is assumed to issue debt to cover 40 percent of financing requirements.

### Illustrative simulation results and recommended buffer size
- Simulated cumulative financing needs (two years) across plausible scenarios:
  - Range from 5.1 to 11.6 percent of GDP for two years.
  - Of that, 3.4 to 7.3 percent may require liquid assets or other financing.
- Stabilization fund role and size:
  - Maintaining a liquid asset buffer in the Economic and Social Stabilization Fund (ESSF) of about 5-7 percent of GDP can help buffer against financing needs in a tail event for the first two years.
  - The use of ESSF assets at 5.4 percent of GDP refers to the baseline scenario.
- Key modeling assumptions relevant to buffer sizing:
  - 60 percent of simulated fiscal responses occur in the first year and 40 percent in the second year.
  - The 40 percent assumption for second-year market financing is based on historical patterns (e.g., large fiscal needs concentrated in year one; ESSF drawdowns tend to be sizable in year one and modest in year two).
  - The lower bound of required liquid assets (3.4 percent of GDP) corresponds to a “less proactive countercyclical fiscal policy” scenario (2.9 percent year 1 + 0.5 percent year 2).
  - The upper bound (7.3 percent of GDP) corresponds to a “low risk tolerance” scenario (6.8 percent year 1 + 0.5 percent year 2).

### Determinants of buffer size and trade-offs
- Size of liquid asset buffers depends on:
  - Duration of spikes in sovereign spreads.
  - How strong fiscal policy responds to shocks.
  - Policymakers’ risk tolerance.
  - Availability of other buffers (domestic borrowing, contingent credit lines, catastrophe bonds).
- Trade-offs and costs:
  - Large liquid assets in the ESSF are likely not optimal because they typically earn lower returns on average than government bond yields and create political pressure to spend in normal times.
  - Self-insurance via liquidity accumulation entails opportunity costs and social choices.
  - Governments can often still borrow domestically even when international borrowing is costly because sovereign bonds can be perceived as safe assets during crises.
  - Accumulated assets may improve credit ratings and reduce debt servicing costs, lowering the insurance premium of building buffers.

### Asset-liability management and fiscal anchor choices
- Integrated asset-liability management:
  - Designing composition of fiscal buffers calls for a holistic balance-sheet approach that assesses assets and liabilities together, accounting for portfolio diversification, currency composition, maturity, and interaction with monetary policy.
  - Managing assets and liabilities in isolation is often suboptimal.
  - The Autonomous Fiscal Council can broaden oversight to provide independent assessment.
- Net debt versus gross debt anchors:
  - Net debt ceiling could better capture solvency by encompassing both liabilities and assets, but valuing government liquid assets is difficult and uncertain.
  - Inclusion feasibility:
    - ESSF assets (mainly treasury bills or foreign sovereign bonds) could be included.
    - PRF assets are legally not freely or fully usable at times of stress and thus may be excluded; since 2022 the PRF use cannot exceed 0.1 percent of previous year GDP.
  - Given the recent establishment of the debt rule (2022) and the relatively small size of ESSF (2.4 percent of GDP at end-2022), keeping a prudent gross debt ceiling is appropriate.
  - A dual anchor (gross debt ceiling and explicit asset floor) would complicate rules and could induce suboptimal borrowing to meet an asset floor; non-binding projections or indicative targets in MTFF documents are an alternative to an explicit floor.

### How to build buffers and scenario comparisons
- General principles:
  - Fiscal efforts to reduce structural deficits are important to build buffers sustainably.
  - Relying solely on saving windfalls from volatile natural resource prices is not sufficient.
- Three illustrative scenarios (effects on debt and ESSF balances):
  - Baseline:
    - Gradual improvement in structural primary balance to a small structural surplus by 2026 (primary balance follows April 2023 WEO projections).
    - Fiscal savings help accumulate ESSF liquidity buffers to about 6 percent of GDP over the medium term while stabilizing government debt well below the prudent ceiling.
  - “Borrow to save”:
    - Assets are quickly accumulated in the ESSF (gives impression of strong buffer).
    - If funded by debt, structural balance remains unchanged, net debt does not reduce much, and gross debt can approach 45 percent of GDP, risking breach of the ceiling if macro conditions deteriorate.
    - This strategy imposes fiscal costs and is generally undesirable unless policymakers have very high risk aversion and returns on liquidity assets exceed borrowing costs.
  - “Aggressive tightening”:
    - Achieves rapid asset accumulation in 5 years while keeping debt low.
    - Requires a sharp fiscal tightening of 0.75 percentage points of GDP in a year, which may be socially infeasible due to abrupt public service cuts.
- Sensitivity to interest-growth differential:
  - A higher interest-growth differential implies higher debt in the interim and a slightly lower pace of asset accumulation.

*Source: IMF staff estimates.*

### 57. On balance across scenarios, accumulating asset buffers through fiscal efforts to

### 1chlea2023004 - 57. On balance across scenarios, accumulating asset buffers through fiscal efforts to

### Rebuilding buffers: scenarios and interpretation
- Accumulating asset buffers through fiscal efforts to improve structural balances is the preferred option across scenarios.
- Building buffers can be gradual over the next few years to:
  - keep debt below the prudent ceiling;
  - allow time to build consensus and undertake sustained gradual fiscal adjustments;
  - accumulate fiscal surpluses to effective fiscal policy management.
- If there are transitory favorable revenues from natural resources, the government can save those in liquid assets to build buffers sooner.
- Figures referenced:
  - Figure 22. Illustrative Scenarios for Rebuilding Buffers (Percent of GDP).
  - Figure 23. Sensitivity Analysis for Rebuilding Buffers (Percent of GDP) — the alternative scenario assumes interest-growth differential is 1.1 percentage points higher than the baseline.

### Key numerical anchors and buffer guidance
- Maintain the existing prudent debt ceiling level on gross central government debt at 45 percent of GDP and keep the current debt level well below the ceiling over the medium term.
- Prudent range of liquidity buffers recommended: 5-7 percent of GDP to help respond to tail events.
- Avoid borrowing government debt at high interest rates to accumulate assets in the stabilization fund.
- Do not set explicit dual anchors on government debt ceiling and a floor on stabilization fund given communication challenges.

### Fiscal management and institutional recommendations
- Apply integrated asset-liability management to jointly consider gross debt levels and the size of prudent liquidity buffers of ESSF under the government balance sheet, accounting for other available instruments.
- Rebuild buffers gradually by reducing structural deficits over the medium term.
- Adjust size and pace of accumulation according to societal risk tolerance and shock size.
- Avoid high-cost debt-financed accumulation of stabilization assets.

### Methodology for estimating prudent anchor (IMF Calibration Toolkit summary)
- Macroeconomic shock generation:
  - 1,000 sequences of macroeconomic shocks are drawn for the six-year projection horizon from a multivariate normal distribution.
  - Variance-covariance matrix estimated using historical data for Chile spanning 1992-2022.
  - Macroeconomic variables modeled include real interest rates, real GDP growth, and the exchange rate.
- Fiscal reaction function:
  - A reduced-form panel regression for 26 emerging economies (including Chile) estimates how the primary fiscal balance responds to lagged primary balance, contemporaneous output gap (or terms of trade for commodity exporters), lagged gross public debt-to-GDP, and country fixed effects.
  - The maximum primary surplus after a shock is capped at 2 percent of GDP for Chile.
  - Fiscal policy shocks φit ~ N(0, σ2φi), with σ2φi calibrated to country-specific variance of residuals.
- Debt trajectories and debt anchor:
  - For each shock sequence, projected debt-to-GDP paths computed using the conventional debt accumulation equation.
  - The debt anchor is defined as the starting level of debt for which simulated debt trajectories over a six-year period would exceed a given debt limit with a probability given by the risk tolerance.
- Extension for natural disaster risks (Akanbi and others 2023):
  - Introduces asymmetric distribution with thicker left tail in growth shocks.
  - User-specified growth impact following a disaster shock assumed to be 12 percent of GDP for Chile.
  - Growth shock Yt = Yt* + I{Dt=1} Zt where Dt ~ Bernoulli(p) and Zt ~ Pareto, parameters calibrated to match average growth effects and skewness.
  - Tool can incorporate mitigation mechanisms (e.g., natural disaster funds, catastrophe bonds) via a user-defined parameter (percent of GDP) that disburses buffers in some years after the tail event, lowering required safety buffers.
- Caveats:
  - High uncertainty around defining an exogenous ‘debt limit’.
  - Validity depends on quality of generated shocks.
  - Other tail risks may not be accounted for.
  - Relationships estimated using past data may not hold in the future; model goodness-of-fit can be limited.

*Source: IMF staff estimates and IMF chapter content.*

### Annex III. A Dynamic Stochastic General Equilibrium Model

### Annex III. A Dynamic Stochastic General Equilibrium Model with Public Debt and Natural Resources

### Model features
- Model name and type:
  - Debt, Investment, Growth, and Natural Resources (the DIGNAR model): a small-open economy DSGE model with three production sectors and an expanded public sector including multiple debt instruments and a sovereign wealth fund.
- Households:
  - Two household types: intertemporal optimizing households and hands-to-mouth households.
  - Both earn labor income, consume domestic and imported goods, pay taxes, and receive government transfers.
  - Hands-to-mouth households are financially constrained and cannot save or borrow.
- Production sectors:
  - Three sectors: (i) a non-tradable goods sector; (ii) a (non-resource) tradable goods sector; and (iii) a natural resource sector.
  - Non-tradable and tradable sectors use Cobb-Douglas technology combining capital and labor.
  - Natural resource production is exogenous and calibrated to sector size; in simulations, all natural resource output is assumed exported (model allows domestic consumption split but results do not change materially).
- Government sector and fiscal instruments:
  - Government receipts: tax revenues (consumption tax and labor income tax), resource revenues and user fees on infrastructure services, foreign aid, interest earnings, transfers from the sovereign wealth fund, and proceeds from bond sales.
  - Government expenditures: government consumption, public investment, transfers to households, debt service payments, and transfers to the sovereign wealth fund.
  - Governments face annual budget constraints and intertemporal debt sustainability constraints.
  - Countercyclical fiscal policy is modeled with fiscal rule limits and an escape clause; deficit financing can be through debt issuance, higher taxes, or drawing down sovereign wealth fund assets.
  - Baseline fiscal response function:
    - Gct / Yt = Gc/Y + ρg (Yt / Yt−1 − 1) where ρg < 0, and government spending as percent of GDP Gct / Yt increases during economic recessions.
  - Sovereign risk premium specification (exogenous link to debt-to-GDP):
    - Rt = Rf + vdc exp[ηdc (dt / yt − d_y)]
    - Interest on external debt Rt equals risk-free rate Rf plus risk premium that depends on deviation of total external debt to GDP dt / yt from steady state d_y, steady state risk premium vdc, and elasticity ηdc.
- Model limitations noted by authors:
  - Abstracts from monetary and other nominal rigidities.
  - Does not include amplification channels such as endogenous sovereign spreads, working capital constraints, or financial accelerator mechanisms; results may therefore be a lower bound of fiscal impacts for tail events.

### Model calibration
- Calibration targets and data:
  - Calibrated to align with key macroeconomic features of Chile: GDP growth, private investment, size of the mining sector, imports and exports as percent of GDP.
  - Calibrated to match government tax revenues, total expenditures, and investment (measured as total government expenditures).
  - Government debt structure matches data on relative shares of external and domestic debt and financing.
  - Calibration captures relative size of government and private sector debt and key parameters of fiscal reaction functions (see source Annex Table A3.1).
  - Calibration based on initial steady state using publicly available data and the IMF WEO database.

### Impact of adverse economic shocks (tail-event scenarios and quantitative results)
- Shock types considered:
  - (i) Production shock in natural resources (copper).
  - (ii) Terms of trade shock (copper price shock).
  - (iii) External risk premium shock.
  - (iv) Major natural disaster shock (modeled as TFP shock).
- Shock: decline in natural resource production
  - Assumed tail event: temporary decline in copper production growth by 4.2 percent from the steady state (corresponds to the big decline in 2008).
  - Impacts:
    - GDP growth declines by more than 0.5 percentage point.
    - Copper and related sectors contract by over 4.2 percent.
    - Exports of copper decline, worsening current account deficits and real exchange rate.
    - Natural resources revenue is about 1 percent of GDP in the data.
    - Primary deficits rise by 0.2 percentage of GDP.
    - Government debt rises by more than one percentage of GDP.
- Shock: terms of trade (copper price) shock
  - Assumed tail event: copper prices fall by 50 percent for a year (similar magnitude to 2008-09).
  - Impacts:
    - Severe decline in GDP by 4 percent due to contraction in the resource sector.
    - Current account deficits widen.
    - Fiscal deficits rise by 2.5 percentage points of GDP.
    - Government debt rises by 4.5 percentage points of GDP (reflecting deficit financing and denominator effects from GDP decline).
- Shock: major natural disaster (TFP shock)
  - Assumed tail event: 12-percent drop in TFP (calibrated to levels identified for Chile’s 2010 earthquake, which led to a loss of 12-15 percent of GDP).
  - Impacts:
    - Production output and GDP growth drop sharply by more than 6 percentage points.
    - Primary deficits rise by 4.5 percentage points of GDP due to countercyclical policies (humanitarian relief and reconstruction).
    - Government debt rises by over 14 percentage points of GDP.
    - TFP shocks are longer-lasting than resource price shocks due to persistent activity effects.
- Shock: surge in external risk premium
  - Assumed tail event: sovereign spreads on external debt rise by 250 basis points.
  - Context: external debt is limited at 14.6 percent of GDP.
  - Impacts:
    - Refinancing costs and interest payments on newly issued debt rise, and real exchange rate depreciates.
    - Overall deficits widen but limited output growth impact in this model specification.
    - Authors note that including amplification channels (working capital constraints, financial accelerator) would likely increase adverse growth effects.

### Caveats and model limitations (authors' assessment)
- The model cannot fully capture all institutional features of Chile’s economy and may understate nonlinear effects and proactive policy responses during tail events.
- Missing amplification mechanisms: debt overhang, financial sector channels, endogenous sovereign spreads; model results may be a lower bound of fiscal impact for tail-risk events.
- Some tail events are complex and harder to model (e.g., pandemics, large-scale social unrest) because they involve simultaneous demand and supply shocks, changes in household preferences, distributional shifts, or political economy channels; in the model, such events might be proxied by a negative TFP shock.

### Model simulations and probabilistic medium-term outcomes
- Stochastic shock generation:
  - VAR estimation: a vector autoregression (one lag) estimated on annual data 2000-2021 with four variables: GDP growth, copper prices, copper production, and sovereign spreads.
  - Data sources: World Bank (GDP US dollar in PPP terms), IMF WEO database, Chilean Copper Commission, and FRED database.
  - Estimated VAR yields volatility, persistence, and covariance of shocks (Annex Table A.3.2).
- Simulation design:
  - 1,000 simulated scenarios of shocks drawn from VAR estimates.
  - In each scenario, shocks apply for the initial 10 periods and the economy returns to steady state according to rules in Melina et al. (2016).
  - Simulations produce a distribution of debt-to-GDP ratios over the medium term (Annex Figure A.3.5).
- Key probabilistic result:
  - With current government debt level at 37 percent of GDP (as of end 2022), there is an 11 percent probability that debt would exceed the prudent level of 45 percent of GDP in the medium term based on the 1,000-simulation VAR-based distribution.
- Sensitivity analysis:
  - A regime-switching approach (estimated on historical data) is used to capture higher volatility during tail events; the high volatility state correlates with historical tail events and is used in sensitivity analysis (Annex Figure A.3.6).

*Source: IMF staff estimates and simulations.*

---


_Source: https://www.imf.org/-/media/files/publications/cr/2023/english/1chlea2023004.pdf_