## 1. Exports of Goods: Top Products

## Source details

**Canonical URL:** [1. Exports of Goods: Top Products](https://www.imf.org/-/media/files/publications/cr/2023/english/1colea2023002.pdf)

## Other formats

- [Markdown version](/-/media/files/publications/cr/2023/english/1colea2023002.pdf.md)
- [Structured JSON version](/-/media/files/publications/cr/2023/english/1colea2023002.pdf.json)

---

### Better Together: The Impact of Fiscal and Monetary Policies on Inflation — overview and main findings
- Context: Inflation rose above central bank targets globally and in Colombia after the pandemic due to supply-side shocks, unprecedented policy stimulus, and a sharp recovery in activity.
- Objective: Explore the role fiscal policy can play to support monetary policy in addressing inflationary pressures without unduly dampening growth using (i) empirical PVAR evidence and (ii) a calibrated DSGE model.
- Empirical sample and setup:
  - Panel VAR (PVAR) sample: 24 inflation targeting small-open economies, 2000-21 (quarterly), including Colombia.
  - Shock simulated in PVAR: 10 percent decline in commodity terms-of-trade (impulse responses reported as cumulative one-year responses).
- DSGE model calibration and features:
  - Calibrated to Colombian macroeconomic data.
  - Monetary policy follows a Taylor-type rule.
  - Fiscal policy follows a fiscal rule with a structural deficit target and debt anchor.
  - Welfare analysis uses Schmitt-Grohe and Uribe (2004) framework; TFP shock assumed AR(1) with persistence parameter equal to 0.9.
- Key findings:
  - PVAR evidence:
    - Terms-of-trade shocks have been accompanied by monetary policy tightening, particularly in countries with the largest deviation between medium-term inflation expectations and the central bank target.
    - In countries with better anchored inflation expectations, the monetary policy response was much less aggressive and fiscal policy played a more supportive role.
  - DSGE qualitative match and welfare implications:
    - Lower (larger) fiscal adjustment in the form of expenditure cuts is associated with lower (larger) overall demand adjustment and more (less) aggressive monetary policy tightening.
    - A faster reduction in inflation occurs when fiscal consolidation supports the monetary policy response.
    - Households’ welfare is higher if fiscal policy does not deviate from the fiscal rule; overburdening monetary policy increases consumption volatility and reduces lifetime expected utility.
    - Conclusion: Monetary and fiscal policy moving in tandem is welfare-enhancing, especially in a low growth and high inflation setting.

### Identifying the role of fiscal policy — methodological points
- Limitation: PVAR is not structural, making causal attribution of fiscal policy effects on inflation and welfare/optimal policy mix difficult.
- Approach: Qualitative matching of PVAR impulse responses with a structural DSGE model and welfare analysis to address PVAR shortcomings.

### Identifying recent inflation drivers in Colombia (2001–2022 focus)
- Stylized narrative:
  - Inflation rose sharply starting in 2021Q2 due to global and domestic supply and demand factors: global value chain bottlenecks, transport logistics disruptions, Russia’s war in Ukraine, and demand recovery as COVID-19 lockdowns were lifted.
  - Key challenge: identify nature (demand vs supply) and source (domestic vs external) of inflation drivers.
- Analytical approaches:
  - Macro approach:
    - Bayesian VAR including external variables (food prices, oil prices, shipment costs proxied by the Baltic Dry Index (BDI)) and domestic variables (output gap, inflation or core inflation, nominal effective exchange rate) for 2000Q2–2022Q4.
    - Forecast error decomposition of inflation projections for 2022 based on different WEO vintages since 2017.
  - Micro approach:
    - Bartik shift-share design using global sectoral input-output linkages for over 1000 production sectors to decompose effects of lockdowns, maritime congestion, fiscal stimulus, monetary policy, etc.
- Empirical quantifications and sectoral findings:
  - Bayesian VAR results (2021Q4–2022Q4 average contributions to headline inflation):
    - External factors accounted for an average of around one-third of headline inflation.
    - Domestic factors (especially domestic demand and the exchange rate) account for an average of 36 percent.
  - For core inflation:
    - Domestic factors accounted for about 35 percent.
    - External factors accounted for less than 30 percent.
  - Additional idiosyncratic factors (weather events, 2021 social demonstrations, macro policy effects, indexation mechanisms) explain a significant residual share.
  - Forecast error decomposition for 2022 points to significant effects from surprises related to external inflation.
  - Micro (sectoral Bartik) results:
    - The strong post-pandemic recovery of demand worldwide was the biggest contributor to inflation in Colombia in 2021–2022.
    - Exchange rate depreciation became increasingly important since 2021H2.
    - The sectoral model significantly underpredicts inflation for 2021Q3–2022Q2, suggesting other domestic idiosyncratic factors and global commodity price shocks may have contributed to up to one-third of inflation during this period.

### Can supply-driven shocks be transmitted to core inflation? — second-round effects and persistence
- Second-round effects:
  - Supply-driven shocks in Colombia have often become broad-based and affected core inflation.
  - Persistent world food price increases and supply bottlenecks can warrant policy action because of transmission lags, effect on inflation expectations, and indexation mechanisms.
- VAR evidence on CPI components (monthly data 2000–2019):
  - Shocks to headline CPI have significant, positive subsequent effects on its components.
  - Responses of CPI to:
    - Food inflation shocks: persistence up to 2 months (monthly response).
    - Core inflation shocks: persistence up to 6 months.
    - Shocks to food and regulated items’ inflation can affect core inflation for up to 6 months and 4 months respectively.
  - Persistence in headline monthly inflation implies added persistence to annual inflation.
- Inflation persistence estimates:
  - Estimated hybrid Phillips Curve using rolling 10-year samples (2004–2022) shows inflation persistence increased over time: the coefficient on lagged inflation rose from 0.2 to 0.6 over two decades.
- Sources of indexation and persistence:
  - Minimum wage indexation:
    - A Constitutional Court ruling suggested the purchasing power of the minimum wage should be sustained, implying annual minimum wage adjustment is unlikely to be lower than headline inflation in the previous year.
    - Minimum wage adjustments can influence wages close to the minimum wage.
  - Regulated prices:
    - Rates for electricity, gas, and water/sewage are linked to past CPI or PPI.
  - Informal indexation practices also contribute to elevated inflation persistence despite inflation targeting.

### Policy implications highlighted in the analysis
- Fiscal policy can support monetary policy and improve welfare when:
  - Fiscal consolidation (expenditure cuts, including reductions in subsidies and tax expenditures) accompanies monetary tightening following a negative supply-side shock.
  - Fiscal policy adheres to a structural deficit target rule and debt anchor, avoiding ad hoc deviations that reduce households’ welfare.
- Monetary policy alone:
  - Overburdening monetary policy to tackle supply-driven inflation increases consumption volatility and can be welfare-reducing.
- Action warranted when:
  - Supply shocks are persistent and risk entrenching inflation expectations via indexation mechanisms and second-round effects.

*Source: IMF staff analysis prepared for Colombia mission papers (March 8, 2023).*

### Headline inflation can impact inflation expectations with knock on effects on core
- Objective and methodology:
  - Analyze how inflation surprises (inflation forecast errors) affect inflation expectations.
  - Time-t revision in inflation expectation i periods ahead: A_{t|t-1}^i = E_t π_{t+i} - E_{t-1} π_{t+i}
  - Inflation surprise: S_{t-1}^t = π_t - E_{t-1} π_t
  - Estimated relationship: A_{t|t-1}^i = α_i S_{t-1}^t
  - Interpretation: α_i gauges the impact of inflation surprises on revisions of inflation expectations; α_i close to one implies low anchoring.
  - Data and estimation:
    - Estimation period: 2000Q1-2022Q3.
    - Inflation expectation horizons: 6-month, 9-month, and 12-month ahead.
    - Expectations data source: central bank’s survey of economic expectations from entrepreneurs (quarterly).
- Empirical findings:
  - Inflation expectations are partially impacted by inflation surprises; impact is positive and statistically significant.
  - Transmission is not one-to-one (α_i < 1) and declines with the expectation horizon.
  - When distinguishing supply vs demand shocks using food price surprises, the impact on expectations is positive but significantly lower.
  - Robustness:
    - Using expectations from market analysts, the estimated coefficient for one-year ahead expectations was 0.29.
    - Regressions using 24-month ahead expectations (sample starts in 2015Q1):
      - Entrepreneurs: estimated coefficient 0.41.
      - Market analysts: estimated coefficient 0.13.
- Footnoted context:
  - Indexed items represent about 42 percent of the CPI basket (Morales and Jaramillo (1996)).
  - Current staff estimate that the fraction of the CPI basket indexed directly to last year inflation, PPI, and minimum wage would be around 35 percent.
- Interpretation and policy implications:
  - Persistent inflationary shocks can become broad-based, affect core inflation and expectations, and become entrenched.
  - Since end-2021, inflationary shocks in Colombia coincided with excess capacity and important domestic demand pressures due to lifting COVID mobility restrictions and faster-than-expected recovery.
  - Indexation to a higher level of inflation has contributed to increased inflationary pressures.
  - Policy recommendation: the central bank would need to react to preserve its hard-gained credibility to reduce inflationary pressures and mitigate the risk of expectations becoming unanchored.

*Source: IMF staff calculations.*

### Exchange rate flexibility will remain critical in supporting the growth in nontraditional exports
- Exchange-rate effects on nontraditional exports:
  - IMF (2017): external adjustments to large and persistent shifts in terms of trade reflect mainly increased flexibility of exchange rates.
  - A real depreciation leads to a small boost to exports and a stronger reduction in imports, with demand shifting toward locally produced goods.
  - Export performance responds more to changing relative prices for non-commodity products and exporters trading manufactured goods more heavily.
  - Local projection results:
    - A 20 percent depreciation of the real effective exchange rate increases the market share of nontraditional exports by 5 percent for a commodity exporter.
    - The same 20 percent depreciation lowers the market share of nontraditional exports for a non-commodity exporter by about 12 percent.
- Magnitude, timing, and frictions:
  - The magnitude is small: an estimated 5 percent increase in market share following a large real depreciation (consistent with IMF (2018)).
  - The effect takes long to materialize: about five years to fully materialize.
  - Trade costs blunt the export volume response (IMF, 2020).
  - Firm-level behavior in Colombia (Casas et al., 2017): manufacturing export firms tend to absorb about half of the price changes arising from exchange rate changes, and absorb even more in the short run than by adjusting production.
  - Sticky prices in a dominant currency can mute exports' response (Gopinath et al., 2020).
- Implications for the energy transition strategy:
  - New administration agenda centers on climate-related issues and aims to boost productivity and long-term growth by diversifying away from extractive industries and relying more on non-conventional renewable energy sources; five pillars listed in source.
  - Challenge: reducing high dependence on energy and mining exports, which represent around 6 percent of GDP.
  - Oil and coal comprise around half of Colombia’s export basket and attract most FDI into the mining industry.
  - Fuel-related revenues are an important source of fiscal revenues; dependence on the sector would increase following the 2022 tax reform (more than half of the reform’s total yield comes from taxes on the oil and coal sector).
  - Oil and coal are key for Colombia’s FX generating capacity and essential for safeguarding external and fiscal sustainability.
- No-replacement of oil/coal production scenario (illustrative):
  - Partial-equilibrium assumptions:
    - Oil production declines by about 90 percent by 2033.
    - Resulting impacts (partial-equilibrium):
      - GDP would go down by 1.3 percent (based on the sector’s value added in GDP).
      - The current account deficit ratio would hover at around 6 percent of GDP.
      - Fiscal revenues could fall by 2 percent of GDP.
  - General-equilibrium model-based simulations (Jakab, 2023, forthcoming):
    - GDP could drop by more than 2 percent in 2-3 years and only gradually recover.
    - The trade balance would deteriorate permanently by more than 1.5 percent of GDP.
- Policy considerations for a gradual energy transition:
  - Transition must be carefully calibrated and implemented gradually because:
    - Building policy fundamentals to boost export diversification, including services, will take time.
    - With oil consumption at about 0.460 million barrels per day (mmb/d), including gasoline imports of about 0.110 mmb/d, a fast reduction in domestic production could generate external gaps.

*Sources: OECD, BIS, IMF staff calculations.*

### 1. Exports of Goods: Top Products ____________________________________________________ 13

### 1. Exports of Goods: Top Products

### Better Together: The Impact of Fiscal and Monetary Policies on Inflation — overview and main findings
- Context: Inflation rose above central bank targets globally and in Colombia after the pandemic due to supply-side shocks, unprecedented policy stimulus, and a sharp recovery in activity.
- Objective: Explore the role fiscal policy can play to support monetary policy in addressing inflationary pressures without unduly dampening growth using (i) empirical PVAR evidence and (ii) a calibrated DSGE model.
- Empirical sample and setup:
  - Panel VAR (PVAR) sample: 24 inflation targeting small-open economies, 2000-21 (quarterly), including Colombia.
  - Shock simulated in PVAR: 10 percent decline in commodity terms-of-trade (impulse responses reported as cumulative one-year responses).
- DSGE model calibration and features:
  - Calibrated to Colombian macroeconomic data.
  - Monetary policy follows a Taylor-type rule.
  - Fiscal policy follows a fiscal rule with a structural deficit target and debt anchor.
  - Welfare analysis uses Schmitt-Grohe and Uribe (2004) framework; TFP shock assumed AR(1) with persistence parameter equal to 0.9.

Key findings from empirical and model analysis:
- PVAR evidence:
  - Terms-of-trade shocks have been accompanied by monetary policy tightening, particularly in countries with the largest deviation between medium-term inflation expectations and the central bank target.
  - In countries with better anchored inflation expectations, the monetary policy response was much less aggressive and fiscal policy played a more supportive role.
- DSGE qualitative match and welfare implications:
  - Lower (larger) fiscal adjustment in the form of expenditure cuts is associated with lower (larger) overall demand adjustment and more (less) aggressive monetary policy tightening.
  - A faster reduction in inflation occurs when fiscal consolidation supports the monetary policy response.
  - Households’ welfare is higher if fiscal policy does not deviate from the fiscal rule; overburdening monetary policy increases consumption volatility and reduces lifetime expected utility.
  - Conclusion: Monetary and fiscal policy moving in tandem is welfare-enhancing, especially in a low growth and high inflation setting.

### Identifying the role of fiscal policy — methodological points
- Limitations of PVAR: Not structural, hence difficult to disentangle causal role of fiscal policy in lowering inflation and to assess welfare/optimal policy mix.
- Approach to address shortcomings: Qualitative matching of PVAR impulse responses with a structural DSGE model and welfare analysis.

### Identifying recent inflation drivers in Colombia (2001–2022 focus)
- Stylized narrative:
  - Inflation rose sharply starting in 2021Q2 due to a combination of global and domestic supply and demand factors: global value chain bottlenecks, transport logistics disruptions, Russia’s war in Ukraine, and demand recovery as COVID-19 lockdowns were lifted.
  - Key policy challenge: identify nature (demand vs supply) and source (domestic vs external) of inflation drivers.

- Analytical approaches used:
  - Macro approach:
    - Bayesian VAR including external variables (food prices, oil prices, shipment costs proxied by the Baltic Dry Index (BDI)) and domestic variables (output gap, inflation or core inflation, nominal effective exchange rate) for 2000Q2–2022Q4.
    - Forecast error decomposition of inflation projections for 2022 based on different WEO vintages since 2017.
  - Micro approach:
    - Bartik shift-share design leveraging differential sectoral exposure using global sectoral input-output linkages for over 1000 production sectors (Chau et al., 2023) to decompose effects of lockdowns, maritime congestion, fiscal stimulus, monetary policy, etc.

Empirical quantifications and sectoral findings:
- Contribution estimates (2001Q4–2022Q4 and focused 2021Q4–2022Q4):
  - Bayesian VAR results (2021Q4–2022Q4 average contributions to headline inflation):
    - External factors accounted for an average of around one-third of headline inflation.
    - Domestic factors (especially domestic demand and the exchange rate) account for an average of 36 percent.
  - For core inflation:
    - Domestic factors accounted for about 35 percent.
    - External factors accounted for less than 30 percent.
  - Additional idiosyncratic factors (weather events, 2021 social demonstrations, macro policy effects, indexation mechanisms) explain a significant residual share.
- Forecast error decomposition analysis for 2022 points to significant effects from surprises related to external inflation.
- Micro (sectoral Bartik) results:
  - The strong post-pandemic recovery of demand worldwide was the biggest contributor to inflation in Colombia in 2021–2022.
  - Exchange rate depreciation became increasingly important since 2021H2.
  - The sectoral model significantly underpredicts inflation for 2021Q3–2022Q2, suggesting other domestic idiosyncratic factors and global commodity price shocks may have contributed to up to one-third of inflation during this period.

### Can supply-driven shocks be transmitted to core inflation? — evidence on second-round effects and persistence
- Second-round effects:
  - In Colombia, supply-driven shocks have often become broad-based and affected core inflation.
  - Policy relevance: Persistent world food price increases and supply bottlenecks can warrant policy action because of transmission lags, effect on inflation expectations, and indexation mechanisms.
- VAR evidence on CPI components (monthly data 2000–2019):
  - Shocks to headline CPI have significant, positive subsequent effects on its components.
  - Responses of CPI to:
    - Food inflation shocks: persistence up to 2 months (monthly response).
    - Core inflation shocks: persistence up to 6 months.
    - Shocks to food and regulated items’ inflation can affect core inflation for up to 6 months and 4 months respectively.
  - Persistence in headline monthly inflation implies added persistence to annual inflation.
- Inflation persistence estimates:
  - Estimated hybrid Phillips Curve using rolling 10-year samples (2004–2022) shows inflation persistence increased over time: the coefficient on lagged inflation rose from 0.2 to 0.6 over two decades.
- Sources of indexation and persistence in Colombia:
  - Minimum wage indexation:
    - A Constitutional Court ruling suggested the purchasing power of the minimum wage should be sustained, implying annual minimum wage adjustment is unlikely to be lower than headline inflation in the previous year.
    - Minimum wage adjustments can influence wages close to the minimum wage.
  - Regulated prices:
    - Rates for electricity, gas, and water/sewage are linked to past CPI or PPI.
  - Informal indexation practices also contribute to elevated inflation persistence despite inflation targeting.

### Policy implications highlighted in the analysis
- Fiscal policy can support monetary policy in bringing down inflation and improve welfare when:
  - Fiscal consolidation (expenditure cuts, including reductions in subsidies and tax expenditures) accompanies monetary tightening following a negative supply-side shock.
  - Fiscal policy adheres to a structural deficit target rule and debt anchor, avoiding ad hoc deviations that reduce households’ welfare.
- Monetary policy alone: Overburdening monetary policy to tackle supply-driven inflation increases consumption volatility and can be welfare-reducing.
- Action warranted when supply shocks are persistent and risk entrenching inflation expectations via indexation mechanisms and second-round effects.

*Source: IMF staff analysis prepared for Colombia mission papers (March 8, 2023).*

### 10.      Headline inflation can impact inflation expectations with knock on effects on core

### 10.      Headline inflation can impact inflation expectations with knock on effects on core

### Objective and methodology
- Analyze how inflation shocks impact inflation expectations by assessing how inflation surprises (inflation forecast errors) affect inflation expectations.
- Measure the extent to which inflation expectations are adjusted following an inflation “surprise.”
- Define the time-t revision in inflation expectation i periods ahead:
  - A_{t|t-1}^i = E_t π_{t+i} - E_{t-1} π_{t+i}
- Define the inflation surprise:
  - S_{t-1}^t = π_t - E_{t-1} π_t
- Estimate relationship:
  - A_{t|t-1}^i = α_i S_{t-1}^t
- Interpretation:
  - The coefficient α_i gauges the impact of inflation surprises on revisions of inflation expectations.
  - If α_i is close to one, surprises are fully transmitted to expectations (low anchoring).
- Data and estimation:
  - Estimation period: 2000Q1-2022Q3.
  - Inflation expectation horizons: 6-month, 9-month, and 12-month ahead.
  - Expectations data source: central bank’s survey of economic expectations from entrepreneurs (done quarterly).

### Empirical findings
- General result:
  - Inflation expectations are partially impacted by inflation surprises.
  - Impact is positive and statistically significant.
  - Transmission is not one-to-one (α_i < 1) and declines with the expectation horizon (partial and declining transmission).
- Distinguishing shocks:
  - When an inflation surprise is defined as the difference between food price inflation and the past expectation of current inflation (to distinguish demand vs supply shocks, or persistent vs short-lived shocks), results again show a positive impact on expectations adjustment, but of significantly lower magnitude.
  - Conclusion: shocks are not fully transmitted to inflation expectations.
- Robustness / alternative samples:
  - Using inflation expectations from market analysts, the estimated coefficient associated with one-year ahead inflation expectations was 0.29.
  - Regressions performed using 24-month ahead inflation expectations (sample starts in 2015Q1):
    - Entrepreneurs: estimated coefficient 0.41.
    - Market analysts: estimated coefficient 0.13.
- Footnoted context preserved in source:
  - Morales and Jaramillo (1996) mentioned indexed items represent about 42 percent of the CPI basket.
  - Current staff estimate that the fraction of the CPI basket indexed directly to last year inflation, PPI, and minimum wage would be around 35 percent.

### Interpretation and policy implications
- Macroeconomic interpretation:
  - Persistent inflationary shocks, even if originating from the supply side, can become broad-based, affect core inflation and inflation expectations, and could become entrenched in headline inflation.
  - Since end-2021, inflationary shocks in Colombia have coincided with a period of excess capacity, with important domestic demand pressures due to:
    - Lifting of COVID-related mobility restrictions.
    - Faster-than-expected recovery in economic activity.
  - Indexation to a higher level of inflation has also contributed to the increase in inflationary pressures.
- Policy recommendation:
  - To reduce inflationary pressures and mitigate the risk of expectations becoming unanchored, the central bank would need to react to preserve its hard-gained credibility.

*Source: IMF staff calculations.*

### 11.      Exchange rate flexibility will remain critical in supporting the growth in nontraditional

### 11.      Exchange rate flexibility will remain critical in supporting the growth in nontraditional exports

### Exchange-rate effects on nontraditional exports
- IMF (2017) finds that external adjustments to large and persistent shifts in terms of trade reflect mainly the increased flexibility of exchange rates.
- A real depreciation leads to a small boost to exports and a stronger reduction in imports, with demand shifting toward locally produced goods.
- Export performance, in terms of global shares, responds more significantly to changing relative prices for non-commodity products and for exporters that trade manufactured goods more heavily.
- Staff analysis using local projection (Jordá, 2005) results:
  - A 20 percent depreciation of the real effective exchange rate increases the market share of nontraditional exports by 5 percent for a commodity exporter.
  - The same 20 percent depreciation lowers the market share of nontraditional exports for a non-commodity exporter by about 12 percent.

### Magnitude, timing, and frictions
- The magnitude of the export-share effect of large real depreciations is small: an estimated 5 percent increase in market share following a large real depreciation (consistent with IMF (2018)).
- The effect tends to take long to materialize: about five years to fully materialize.
- Trade costs blunt the export volume response (IMF, 2020).
- Firm-level behavior in Colombia (Casas et al., 2017):
  - Manufacturing export firms tend to absorb about half of the price changes arising from exchange rate changes, and absorb even more in the short run than by adjusting production.
- Sticky prices in a dominant currency can mute exports' response (Gopinath et al., 2020).

### Implications for the energy transition strategy
- The new administration's agenda centers on climate-related issues and aims to boost productivity and long-term growth by diversifying away from extractive industries and relying more on non-conventional renewable energy sources. The agenda's five pillars are:
  1) reducing land use and deforestation and expanding the agricultural frontier;
  2) ensuring that territories are resilient to climate change and natural disasters;
  3) incentivizing the re-industrialization of the economy and promote more environmental-friendly economic activities and the use of renewable sources of energy;
  4) focusing on inclusive policy making to foster a more equitable and productive economy;
  5) developing a progressive energy transition strategy aimed at using surpluses from the oil and coal sectors to finance alternative industries.
- A key challenge is reducing high dependence on energy and mining exports, which represent around 6 percent of GDP.
- Oil and coal comprise around half of Colombia’s export basket and attract most FDI into the mining industry.
- Fuel-related revenues are an important source of fiscal revenues; dependence on the sector would increase following the 2022 tax reform (more than half of the reform’s total yield comes from taxes on the oil and coal sector).
- Oil and coal are key for Colombia’s FX generating capacity and essential for safeguarding external and fiscal sustainability.

### No-replacement of oil/coal production scenario (illustrative)
- Partial-equilibrium assumptions:
  - Oil production declines by about 90 percent by 2033.
  - Resulting impacts (partial-equilibrium):
    - GDP would go down by 1.3 percent (based on the sector’s value added in GDP).
    - The current account deficit ratio would hover at around 6 percent of GDP.
    - Fiscal revenues could fall by 2 percent of GDP.
- General-equilibrium model-based simulations (Jakab, 2023, forthcoming; general equilibrium New Keynesian two country model calibrated for Colombia):
  - GDP could drop by more than 2 percent in 2-3 years and only gradually recover.
  - The trade balance would deteriorate permanently by more than 1.5 percent of GDP.

### Policy considerations for a gradual energy transition
- The energy transition must be carefully calibrated and implemented gradually because:
  - The build-up and strengthening of policy fundamentals to boost export diversification, including services, will take time.
  - With oil consumption at about 0.460 million barrels per day (mmb/d), including gasoline imports of about 0.110 mmb/d, a fast reduction in domestic production could generate external gaps.

*Sources: OECD, BIS, IMF staff calculations.*

---


_Source: https://www.imf.org/-/media/files/publications/cr/2023/english/1colea2023002.pdf_