## _wp11207

## Source details

**Canonical URL:** [_wp11207](https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2011/_wp11207.pdf)

## Other formats

- [Markdown version](/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2011/_wp11207.pdf.md)
- [Structured JSON version](/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2011/_wp11207.pdf.json)

---

### I. Introduction and scope
- Purpose:
  - Provide a range of estimates for the long-run external current account balance for eight oil-producing SSA countries using three IMF methodologies and common simplifying assumptions.
- Countries analyzed:
  - Angola, Cameroon, Chad, Côte d’Ivoire, Equatorial Guinea, Gabon, Nigeria, and the Republic of Congo.
- Key motivations:
  - The horizon for oil production in most countries is relatively short and oil depletion requires management of foreign assets.
  - Vulnerability to external shocks (volatile world oil prices, weak non-oil sectors) makes maintaining a foreign asset cushion important.
  - Fiscal sustainability is not a perfect substitute for external sustainability because governments are often a relatively small player in the oil sector and private-sector saving and investment behavior can diverge from public fiscal policy.
- Core assumption:
  - No change in the exchange rate; the current exchange rate level is assumed to be at the optimum.

### II. Methodologies employed
- Three complementary approaches applied consistently across the eight countries:
  - Consumption smoothing (Thomas, Kim, and Aslam, 2008): forward-looking optimizing representative agent consumes only return on oil and non-oil wealth; optimal non-oil current account financed by “permanent” revenue from total wealth.
  - Precautionary saving (Bems and de Carvalho Filho, 2009a): forward-looking model adding oil-price uncertainty and precautionary accumulation.
  - Macroeconomic balance (CGER / Lee and others, 2008): backward-looking regression-based derivation of a current account “norm”.
- Modeling strategy:
  - Apply the three models using shared simplifying assumptions (see Box 1 in the source) and perform robustness checks via parameter variations and alternative regression specifications.

### III. Key underlying assumptions and data sample (Box 1)
- World oil prices:
  - WEO nominal oil price projection: US$80 in 2010, rising to US$87 in 2015, and remaining constant in real terms thereafter.
  - Precautionary saving approach: oil price follows a stochastic (multiplicative error) process.
- Oil reserves and production:
  - Gas excluded due to missing and inconsistent data across SSA countries.
  - Domestic consumption assumed at 4 percent of total production.
- Macroeconomic data:
  - Nearly all macroeconomic data sourced from the Fall 2010 WEO.
- Calibration and sample:
  - Consumption-smoothing and precautionary-saving models calibrated on data from 2005; in-sample results examined for 2005–09.
  - Macroeconomic balance regression estimated using data covering 1993–2009 with non-overlapping four-year averages.
  - Macroeconomic balance panel includes 13 non-SSA oil-producing countries plus the eight SSA oil countries for some specifications.
- Theoretical and parameter choices:
  - CRRA utility with curvature parameter θ (risk aversion).
  - Real return on investment set as 2½ percent.
  - Discount rate used in numerical implementation: 4 percent (baseline).
  - Population growth projections: UN population growth projections for 2010–50.
  - Deterministic baseline for consumption smoothing; stochastic element for precautionary saving.
- Notable numerical assumptions called out:
  - Log of the relative price of oil follows a multiplicative error process with mean 1, ρ < 0.9, σε = 0, and initial value p2005 = 53.5.
  - Non-oil sector grows at the same pace as the labor force.
  - Real return on investment: 2½ percent; discount rate for precautionary implementation: 4 percent; CRRA curvature = 2.

### IV. Background facts and descriptive statistics
- Proven oil reserves (sample range):
  - 0.3 billion barrels in Cameroon and Côte d’Ivoire to about 33 billion barrels in Nigeria.
  - Nigeria accounts for about 73 percent of SSA’s total oil reserves (within the paper’s sample).
- Annual oil production in 2010 (sample extremes):
  - 18 million barrels in Côte d’Ivoire to 908 million barrels in Angola.
- Production shares in 2010:
  - Angola about 44 percent of subregion production; Nigeria about 36 percent.
- Oil-dependency ratio (share of oil sector in GDP, 2010, percent):
  - Angola 45.3
  - Cameroon 6.1
  - Chad 36.9
  - Republic of Congo 69.4
  - Côte d’Ivoire 1.8
  - Equatorial Guinea 57.4
  - Gabon 46.7
  - Nigeria 29.1
- Observed behavior:
  - Non-oil GDP growth rose during 2005–08 oil-price boom and declined during 2009–10 for highly oil-dependent countries.
  - Correlation between oil and non-oil sector growth about 0.5 over 2005–10 in highly oil-dependent countries, reinforcing precautionary-saving rationale.
- Recent external balance movements:
  - Average deterioration: overall balance deteriorated by about 8 percentage points of GDP on average in 2009–10 compared to 2006–08.
  - Non-oil current account improved: non-oil deficit reduced by an average of about 26 percentage points between 2006–08 and 2009–10.
- Selected exact values (Non-oil Current Account as percent of non-oil GDP; Overall Current Account as percent of total GDP) for 2001–05, 2006–08, 2009–10, projected 2011–15:
  - Angola: Non-oil Current Account -145.4, -121.2, -96.2, -74.8; Overall Current Account -0.4, 16.2, -1.9, 0.7
  - Cameroon: Non-oil -11.6, -12.2, -10.8, -11.3; Overall -3.4, 0.4, -3.2, -1.8
  - Chad: Non-oil -70.5, -100.3, -99.6, -53.9; Overall -38.0, -11.1, -32.1, -9.7
  - Republic of Congo: Non-oil -142.5, -201.3, -164.2, -138.9; Overall -1.4, -2.0, -1.9, 5.4
  - Côte d’Ivoire: Non-oil 1.3, -2.9, 0.5, -8.6; Overall 2.0, 1.3, 6.8, -0.2
  - Equatorial Guinea: Non-oil -511.5, -305.8, -170.9, -107.7; Overall -20.2, 6.8, -9.3, -7.4
  - Gabon: Non-oil -53.1, -55.0, -53.5, -49.3; Overall 12.3, 18.9, 15.4, 10.8
  - Nigeria: Non-oil -50.7, -18.1, -11.9, -19.0; Overall -0.4, 20.3, 13.2, 11.1
  - Average: Non-oil -123.0, -102.1, -75.8, -57.9; Overall -6.2, 6.4, -1.6, 1.1

### V. Present value of oil wealth and annual return (authors' estimates)
- Present Value of Oil Wealth (bn USD) and Annual Return (percent of 2006 GDP):
  - Angola 80 18.0
  - Cameroon 30 1.5
  - Chad 55 2.1
  - Congo 128 15.1
  - Côte d'Ivoire 27 1.1
  - Equatorial Guinea 114 6.9
  - Gabon 148 18.1
  - Nigeria 2,365 15.3
- Aggregate statement using baseline assumptions:
  - The return on oil wealth for the eight countries amounts to US$3 trillion with an average of 8½ percent of GDP in 2006.
  - Country extremes: highest about 18 percent in Gabon and lowest is 1.1 percent in Côte d’Ivoire.

### VI. Consumption-smoothing and precautionary-saving quantitative results (2006, percent of GDP)
- Consumption-smoothing (optimal external current account, percent of GDP, 2006):
  - Angola 54.4
  - Cameroon 12.7
  - Chad 48.6
  - Congo 57.5
  - Côte d'Ivoire 6.8
  - Equatorial Guinea 76.0
  - Gabon 37.4
  - Nigeria 26.0
- Precautionary-saving component (percent of GDP, 2006):
  - Angola 1.0
  - Cameroon 0.0
  - Chad 0.0
  - Congo 0.7
  - Côte d'Ivoire 0.0
  - Equatorial Guinea 0.9
  - Gabon 0.5
  - Nigeria 0.6
- Notable observations:
  - Countries with low oil-dependency (Cameroon, Côte d’Ivoire) have projected precautionary saving levels very low—less than 0.03 percent of GDP in 2006 (text note).
  - In examples such as Equatorial Guinea precautionary saving peaks in 2011 at about 1½ percent of GDP and then declines.

### VII. Cross-method comparison for 2006 (optimal current account balances, percent of GDP)
- Table 6 comparison (Consumption Smoothing; Precautionary Saving; Macro Balance; Actual):
  - Angola: 54.4; 1.0; 9.8; 25.2
  - Cameroon: 12.7; 0.0; 0.5; 1.6
  - Chad: 48.6; 0.0; -30.7; -9.0
  - Congo: 57.5; 0.7; 5.9; 1.5
  - Côte d'Ivoire: 6.8; 0.0; 0.4; 2.8
  - Equatorial Guinea: 76.0; 0.9; 3.7; 7.1
  - Gabon: 37.4; 0.5; 13.7; 15.8
  - Nigeria: 26.0; 0.6; 0.1; 26.5
- Interpretation:
  - Consumption-smoothing yields the largest optimal surpluses.
  - Macroeconomic balance estimates are most similar to actual historical outcomes.
  - Precautionary-saving components are positive but small relative to consumption-smoothing surpluses.

### VIII. Macroeconomic balance empirical findings (selected regression results and interpretations)
- Robust determinants (expected signs summarized in source):
  - Non-oil fiscal balance (positive).
  - Demographics: old-age dependency ratio and population growth (expected negative).
  - NFA (ambiguous).
  - Lagged current account (persistence; positive).
  - Oil current account balance (positive).
  - Economic growth: growth of real per capita GDP (negative); ratio of per capita PPP income to US level (positive).
- Fixed effects estimates (column 3 of Table 7a, as reported):
  - A 1 percentage point increase in old-age dependency worsens the current account balance by about 1 percent of GDP.
  - A 1 percentage point increase in population growth worsens the current account balance by about 1 percent of GDP.
  - A 1 percentage point increase in the oil balance improves the current account balance by about 0.4 percent of GDP.
  - A 1 percentage point increase in relative income improves the current account balance by about 0.4 percent of GDP.
  - Positive coefficient on lagged current account implies persistence.
- Non-oil current account (fixed effects, columns 5 and 6):
  - Unexpected negative and significant coefficient on the oil balance: the non-oil current account deteriorates as the oil balance improves.
  - Possible mechanisms: higher oil production stimulates correlated non-oil GDP, raises imports of investment goods, and increases oil-related imports; net effect worsens non-oil trade balance.
  - Negative lagged non-oil current account suggests lack of persistence in the non-oil current account.
  - Positive sign on NFA indicates higher NFA generates more investment income, improving the non-oil current account.

### IX. Robustness, sensitivity, and scenario implications
- Consumption-smoothing sensitivity:
  - Sensitive to initial conditions and model parameters (rate of return, oil prices, non-oil cash flow).
  - Under extreme assumptions (rate of return above 4 percent, doubling of oil prices, or substantial non-oil cash flow), estimated current account could approach 2006 observed levels.
  - Baseline parameterization: risk averseness (σ)=2 and discount rate (β)=4 percent.
- Precautionary-saving sensitivity:
  - Higher risk aversion raises precautionary saving; consumption-smoothing component unchanged.
  - Raising the discount rate from 4 percent to 10 percent triples the precautionary-saving component in most countries while the consumption-smoothing component declined by about 34 percent (average). Net effect: lower overall optimal current account as consumption-smoothing decline dominates.
  - A 1 percentage point increase in labor supply growth would halve the precautionary-saving component in some countries and increase the consumption-smoothing component by about 17 percent (average); net effect: overall optimal current account rises because consumption-smoothing dominates.
- Sensitivity to initial values:
  - Increasing NFA by 10 percentage points of non-oil GDP changes outcomes (direction and magnitude reported in robustness tables in source).
  - Increasing share of non-oil GDP reduces both precautionary saving and consumption smoothing, substantially lowering sustainable current account versus baseline.
  - A higher initial oil price reduces the sustainable current account.
- Overall robustness conclusion:
  - Outcomes are sensitive to initial conditions (particularly the share of the non-oil sector) and less sensitive to some model parameters; across variants, the recommendation to accumulate additional precautionary savings is robust.

### X. Policy-relevant interpretations and recommendations
- External sustainability requires explicit attention in oil-producing SSA countries, particularly those with fixed exchange rate regimes.
- Countries should maintain a foreign asset cushion and use external-balance benchmarks to guide policy adjustments against oil depletion and price volatility.
- Precautionary saving is especially relevant given weak non-oil sectors and their correlation with oil-sector fluctuations.
- Fiscal sustainability analyses should not substitute for external sustainability assessments; private-sector behavior and partial government participation in oil sectors can create divergence between fiscal and external outcomes.
- Benchmarks should be forward looking and allow for:
  - External shocks and sensitivity to model parameters and initial conditions.
  - Potential front-loaded public investment in infrastructure to develop the non-oil sector, raise return on investment, spur non-oil growth, and broaden the export base beyond oil.
- Policy imperative:
  - Shift from backward-looking behavior (consumption tied to current oil income) to forward-looking saving and investment strategies given short extraction horizons; accumulate precautionary savings and consider front-loaded investment where appropriate.

### XI. Conclusions
- Baseline finding:
  - Macro balance (backward-looking) estimates align more closely with historical outcomes; forward-looking consumption-smoothing and precautionary-saving approaches imply larger adjustments and higher optimal surpluses for many countries.
- Uncertainty:
  - Long-term external sustainability estimates are model-dependent and sensitive to initial conditions and key parameters; large margins of uncertainty exist.
- Urgency:
  - Without new discoveries and given relatively short extraction horizons for most sample countries, forward-looking policy adjustments and precautionary saving accumulation are urgent.
- Potential mitigation:
  - Front-loaded public investment to develop non-oil sectors could materially alter sustainability outlook by raising returns and changing core model parameters.

*Source: _wp11207 - References................................................................................................31*

### References................................................................................................31

### _wp11207 - References................................................................................................31

### I. Introduction
- Purpose: Provide a range of estimates for the long-run external current account balance for eight oil-producing SSA countries using three IMF methodologies and common simplifying assumptions.
- Countries analyzed: Angola, Cameroon, Chad, Côte d’Ivoire, Equatorial Guinea, Gabon, Nigeria, and the Republic of Congo.
- Key motivations:
  - The horizon for oil production in most countries is relatively short and oil depletion requires management of foreign assets.
  - Vulnerability to external shocks (volatile world oil prices, weak non-oil sectors) makes maintaining a foreign asset cushion important.
  - Fiscal sustainability is not a perfect substitute for external sustainability because governments are often a relatively small player in the oil sector and private-sector saving and investment behavior can diverge from public fiscal policy.
- Assumption: No change in the exchange rate; the current exchange rate level is assumed to be at the optimum (justified by the prevalence of fixed exchange rate regimes and low trade elasticity to exchange rate movements in oil-producing SSA countries).
- Organization: Review of literature (Section II), country background (Section III), modeling approaches and results (Section IV and V), conclusions and policy implications (Section VI).

### II. Review of the empirical literature
- Standard IMF methodologies referenced: Lee and others (2008) / CGER framework covering (i) external sustainability (NFA stabilization), (ii) equilibrium real exchange rate, and (iii) macroeconomic balance (current account “norm”).
- Prior work on oil producers:
  - Bems and de Carvalho Filho (2009b): non-oil fiscal balance, benchmark NFA via inter-temporal optimization; included Nigeria and Cameroon.
  - Morsy (2009): dynamic panel estimation, added proven oil reserves and maturity of oil production; argued newer oil producers could have larger current account deficits.
  - Thomas, Kim, and Aslam (2008): consumption smoothing (stabilize real value of total wealth; consume return on oil and non-oil wealth with correction for consumption tilting).
  - Bems and de Carvalho Filho (2009a): added precautionary saving for oil-price uncertainty; found precautionary motive can account for sizable external saving.
- Low-income country literature: Christiansen and others (2009) find aid, financial liberalization, capital account liberalization, institutions, shocks, and demographics significantly explain external balance for 134 low-income countries.
- Specific studies for CEMAC / low-income oil producers:
  - Deléchat and Kireyev (2008): permanent-income framework; showed CEMAC current account deficits are excessive.
  - Akitoby and others (2011): including public investment feedback reduces sustainable external balance estimates; showed actual/projected IMF staff current account deficits exceed model “optimal” levels for some countries.

### III. Background
A. Oil resources and dependence
- Proven oil reserves range reported in sample: 0.3 billion barrels in Cameroon and Côte d’Ivoire to about 33 billion barrels in Nigeria.
- Nigeria accounts for about 73 percent of SSA’s total oil reserves (within the paper’s sample).
- Annual oil production in 2010: low of about 18 million barrels in Côte d’Ivoire to a high of 908 million barrels in Angola.
- Production shares in 2010: Angola about 44 percent of subregion production; Nigeria about 36 percent.
- Projection: Oil production expected to peak in most countries over the next few years; except for Gabon and Nigeria, oil reserves could be exhausted (without further exploration and development) within the next two decades (Figure 1).
- Oil-dependency ratio (share of oil sector in GDP) in 2010:
  - Angola 45.3
  - Cameroon 6.1
  - Chad 36.9
  - Republic of Congo 69.4
  - Côte d’Ivoire 1.8
  - Equatorial Guinea 57.4
  - Gabon 46.7
  - Nigeria 29.1
  (Table 1: percent)
- Observed behavior: Highly oil-dependent countries (Angola, Congo, Equatorial Guinea, Gabon) have failed to develop non-oil sectors; non-oil GDP growth rose during 2005–08 oil-price boom and declined during 2009–10. Correlation between oil and non-oil sector growth is about 0.5 over 2005–10 in those countries, reinforcing importance of precautionary savings.

B. Recent developments in the external balance
- World oil price movements (nominal and real) directly affected current accounts (Figure 2).
- During 2006–08 oil-price boom, overall current accounts moved into surplus in most countries; shifted into deficit after price declines in 2009–10.
- Average deterioration: overall balance deteriorated by about 8 percentage points of GDP on average in 2009–10 compared to 2006–08 (attributable to volatile oil exports).
- Non-oil current account: actually improved during the same period; non-oil deficit reduced by an average of about 26 percentage points between 2006–08 and 2009–10.
- IMF World Economic Outlook (staff projections): over the medium term 2011–15 both overall and non-oil current account balances expected to improve in most countries; overall balance expected to turn into a surplus on average; non-oil deficit expected to decline by about 18 percentage points of GDP in 2011–15 compared with 2009–10, partly driven by fiscal consolidation.

- Selected exact values from Table 3 (Non-oil Current Account as percent of non-oil GDP; Overall Current Account as percent of total GDP), shown for periods 2001–05, 2006–08, 2009–10, and projected 2011–15:
  - Angola: Non-oil Current Account -145.4, -121.2, -96.2, -74.8; Overall Current Account -0.4, 16.2, -1.9, 0.7
  - Cameroon: Non-oil -11.6, -12.2, -10.8, -11.3; Overall -3.4, 0.4, -3.2, -1.8
  - Chad: Non-oil -70.5, -100.3, -99.6, -53.9; Overall -38.0, -11.1, -32.1, -9.7
  - Republic of Congo: Non-oil -142.5, -201.3, -164.2, -138.9; Overall -1.4, -2.0, -1.9, 5.4
  - Côte d’Ivoire: Non-oil 1.3, -2.9, 0.5, -8.6; Overall 2.0, 1.3, 6.8, -0.2
  - Equatorial Guinea: Non-oil -511.5, -305.8, -170.9, -107.7; Overall -20.2, 6.8, -9.3, -7.4
  - Gabon: Non-oil -53.1, -55.0, -53.5, -49.3; Overall 12.3, 18.9, 15.4, 10.8
  - Nigeria: Non-oil -50.7, -18.1, -11.9, -19.0; Overall -0.4, 20.3, 13.2, 11.1
  - Average: Non-oil -123.0, -102.1, -75.8, -57.9; Overall -6.2, 6.4, -1.6, 1.1

### IV. Modeling external sustainability: methodology overview
- Objective: Estimate optimal or “normal” current account balances using three approaches:
  1. Consumption smoothing (Thomas, Kim, and Aslam, 2008): forward-looking optimizing representative agent consumes only return on oil and non-oil wealth; optimal non-oil current account financed by “permanent” revenue from total wealth to maintain real value of total wealth stock.
  2. Precautionary saving: adds precautionary motives when households face oil-price uncertainty; forward-looking and similar to deterministic case but incorporates uncertainty and precautionary accumulation.
  3. Macroeconomic balance: backward-looking derivation of a current account “norm” from macro fundamentals; parsimonious and based on simple assumptions (CGER / Lee and others, 2008).
- Key point: First two approaches are intertemporal and forward-looking; third is backward-looking and regression-based.
- Models are applied consistently across the eight countries with similar simplifying assumptions (see Box 1 in the source for detailed assumptions and data sample).
- Consumption-smoothing formulation: uses the maximization problem and intertemporal budget constraint as per Thomas, Kim, and Aslam (2008) (equations cited in the source).

### V. Research strategy, robustness, and comparative perspective
- Strategy: Apply three models to the eight SSA oil producers using the same lens to highlight similarities and differences; assess robustness by varying parameters and assumptions.
- Comparison: Results are compared with high- and middle-income oil-producing countries to highlight unique characteristics of SSA oil producers.
- Robustness checks indicated in the source include:
  - Varying parameters and assumptions in consumption-smoothing and precautionary-saving models (Tables 8 and 9: robustness tests for consumption smoothing and precautionary saving approaches).
  - Alternative macroeconomic-balance panel regressions and determinants analysis (Tables 7a and 7b).
  - Different methodologies for setting optimal current account balance for 2006 (Table 6).

### VI. Key implications highlighted in the source text
- External sustainability requires explicit attention in oil-producing SSA countries, particularly those with fixed exchange rate regimes.
- Maintaining a foreign asset cushion and benchmarking external balances can guide policymakers on magnitude of policy adjustments needed to provide adequate cover against oil depletion and volatile oil prices.
- Precautionary saving motives are especially relevant because non-oil sectors are weak and correlated with oil-sector fluctuations, amplifying the need for external buffers.
- Fiscal sustainability analyses should not be used as a stand-in for external sustainability assessments because private-sector behavior and partial government participation in oil sectors can create divergence between fiscal and external outcomes.

*Source: _wp11207 - References................................................................................................31*

### Box 1. Key Underlying Assumptions and Data Sample

### Box 1. Key Underlying Assumptions and Data Sample

### World oil prices
- Data source: IMF’s Fall 2010 edition of the World Economic Outlook (WEO); based on the average of West Texas, Intermediate Brent, and Dubai Fateh crude oil.
- Macroeconomic balance approach: does not rely directly on oil prices but uses variables that depend on WEO-projected oil price evolution.
- Consumption-smoothing assumption: WEO’s nominal oil price projection of US$80 in 2010, rising to US$87 in 2015, and remaining constant in real terms thereafter.
- Precautionary saving approach: oil price follows a stochastic (multiplicative error) process, introducing uncertainty into the model.

### Oil reserves and production
- Data based on information used by IMF staff in their WEO projections.
- Gas is excluded because of lack of missing and consistent data across SSA countries.
- Acknowledged limitation: exclusion of gas may be restrictive for countries where gas is important (e.g., Equatorial Guinea and Nigeria).
- Reference to empirical sensitivity: York and Zhan (2009) show results can change significantly if gas reserves are included, extending production life and increasing wealth accumulation.

### Oil trade balance
- Data from Fall 2010 WEO used through 2015.
- For 2016 and beyond: oil income = oil production × oil price, corrected for domestic consumption.
- Assumption: domestic consumption is 4 percent of total production based on observable data.

### Macroeconomic data
- Nearly all macroeconomic data sourced from the Fall 2010 WEO except where specific assumptions are used.

### Data sample and calibration
- Consumption-smoothing and precautionary-saving models calibrated on data from 2005 to minimize distortions from the run-up in oil prices in the second half of the decade.
- Allows examination of in-sample results for the period 2005–09.
- Macroeconomic balance approach and current versus permanent income assessment:
  - Regression estimated using data covering 1993–2009 with not-overlapping four-year average.
- For all models: results compared with outturns from 2005–09 and IMF staff projections extending over 2010–15.

### Sample countries
- Consumption-smoothing and precautionary-saving models covered eight oil-producing SSA countries.
- Macroeconomic balance approach adds 13 non-SSA oil-producing countries to represent more “normal “optimal current account levels for oil-producing countries more generally.
- Current versus permanent income assessment uses only oil-producing SSA countries to establish their specific characteristics.

### Theoretical model, notation, and preferences
- Household/population notation: L1 = population per household, growing at constant rate n.
- Variables: Ct private consumption, Bt net foreign assets at the end of period, Yt real GDP, It investment, Gt government consumption, Tt net external transfers.
- Parameters: β real subjective discount rate, r real return on investment.
- All variables except for L, β, and r are expressed in per capita terms.
- Utility function: constant relative risk aversion (CRRA) form with θ as the consumer risk aversion parameter.
- Deterministic assumption: dynamic optimization models with CRRA preferences are assumed deterministic to simplify solution; expectation operator dropped.
- Resulting Euler equation reported in the text.

### Optimal consumption and current account expressions
- National cash flow per capita and real interest rate adjusted for population growth defined following Thomas, Kim, and Aslam (2008).
- Optimal consumption expression provided (see source text for full functional form).
- Per capita wealth Wt defined as B_{t-1} + (1+r̄)^{-1} ∑(1+r̄)^{-j} Ỹ_{t+j} ... (full definitions in source text).
- Consumption tilting factor ψ captures degree of intertemporal substitution in consumption.

### Decomposition of optimal current account
- Optimal current account can be broken into oil and non-oil current account balances.
- Optimal non-oil current account consists of four factors:
  - (i) return on the present-discounted value of current and future oil production (return on oil wealth);
  - (ii) present-discounted value of future changes in non-oil cash flows;
  - (iii) domestic consumption of oil (α, share of domestic consumption of oil in total oil production);
  - (iv) a consumption tilting factor ( (ψ^{-1}−1) C̃_{t} ), allowing trade-offs between present and future consumption.

### Parameter assumptions and quantitative results
- Real return on investment set as 2½ percent in line with long-term averages used by the US Office of Management and Budget.
- Population growth projections: UN population growth projections used for the period 2010–50.
- Using these assumptions, the return on oil wealth for the eight countries amounts to US$3 trillion with an average of 8½ percent of GDP in 2006.
- Country extremes reported: highest about 18 percent in Gabon and lowest is 1.1 percent in Côte d’Ivoire.

### Estimation of non-oil cash flow present value
- Key input: estimate of long-term growth in the non-oil sector.
- Historical movements of non-oil cash flows for the eight countries were very volatile.
- None of the explanatory variables suggested by Thomas, Kim, and Aslam (2008) turned out to be statistically significant for these movements.

*Box 1. Key Underlying Assumptions and Data Sample (source content).*

### 1.3 percent a year for Gabon to 2.0 percent a year for Angola.

### _wp11207 - 1.3 percent a year for Gabon to 2.0 percent a year for Angola.

### Key assumptions and model setup
- Consumption-smoothing framework:
  - Non-oil cash flow is assumed to be zero over the long term (in terms of its present discounted value).
  - Domestic oil consumption is assumed constant at about 4 percent of GDP.
  - Consumption tilting factor is assumed to be zero.
- Optimum precautionary savings framework (following Bems and de Carvalho Filho (2009a)):
  - Distinguishes a deterministic solution (consumption smoothing, no oil price uncertainty) and a stochastic solution (precautionary saving driven by oil price uncertainty).
  - Focus in this paper on the precautionary component only.
- Numerical implementation and parameter choices:
  - Discount rate set at 4 percent.
  - Curvature (risk aversion) of CRRA utility set at 2.
  - Log of the relative price of oil follows a multiplicative error process with mean 1, ρ < 0.9, σε = 0, and initial value p2005 = 53.5.
  - Model solved forward from 2005.
  - Non-oil sector grows at the same pace as the labor force.

### Present value of oil wealth and annual return (authors' estimates)
- Present Value of Oil Wealth in bn USD and Annual Return in percent of 2006 GDP:
  - Angola 80 18.0
  - Cameroon 30 1.5
  - Chad 55 2.1
  - Congo 128 15.1
  - Côte d'Ivoire 27 1.1
  - Equatorial Guinea 114 6.9
  - Gabon 148 18.1
  - Nigeria 2,365 15.3

### Country-specific initial values and parameters (selected)
- Labor Force Growth (percent) and Lifetime of Oil Estimate (selected entries as in source):
  - Angola: 2005 2.0; 2010 17.3; Lifetime of Oil Estimate 2028
  - Cameroon: 2005 1.5; 2010 5.7; Lifetime of Oil Estimate 2028
  - Chad: 2005 2.2; 2010 4.7; Lifetime of Oil Estimate 2032
  - Congo: 2005 1.5; 2010 14.5; Lifetime of Oil Estimate 2029
  - Côte d'Ivoire: 2005 1.8; 2010 8.2; Lifetime of Oil Estimate 2029
  - Equatorial Guinea: 2005 1.9; 2010 28.6; Lifetime of Oil Estimate 2035
  - Gabon: 2005 1.3; 2010 11.9; Lifetime of Oil Estimate 2040
  - Nigeria: 2005 1.5; 2010 27.8; Lifetime of Oil Estimate 2048
- Initial values for net foreign assets and oil revenue as shares of GDP, oil reserves (bn brls), and other parameters are based on IMF staff estimates and UN World Population Prospects for projected labor growth.

### Consumption-smoothing results (optimal external current account, percent of GDP, 2006)
- Consumption smoothing produced the largest optimal current account surpluses among methods:
  - Angola 54.4
  - Cameroon 12.7
  - Chad 48.6
  - Congo 57.5
  - Côte d'Ivoire 6.8
  - Equatorial Guinea 76.0
  - Gabon 37.4
  - Nigeria 26.0
- Interpretation:
  - Higher optimal current account balances for Equatorial Guinea, Congo, and Chad reflect high oil dependency and relatively smaller reserves/shorter production horizons, requiring more saving to smooth consumption post-exhaustion.
  - Optimal surpluses for these SSA oil producers are larger than for middle- and high-income oil-producing countries.

### Precautionary-saving results (percent of GDP, 2006)
- Estimated precautionary saving (positive for all countries; largest for high oil-dependency countries):
  - Angola 1.0
  - Cameroon 0.0
  - Chad 0.0
  - Congo 0.7
  - Côte d'Ivoire 0.0
  - Equatorial Guinea 0.9
  - Gabon 0.5
  - Nigeria 0.6
- Additional observations:
  - Countries with low oil-dependency (Cameroon, Côte d’Ivoire) have projected precautionary saving levels very low—less than 0.03 percent of GDP in 2006 (text note).
  - Precautionary saving declines over time alongside reductions in oil production; example: Equatorial Guinea’s precautionary saving peaks in 2011 at about 1½ percent of GDP and then declines steadily.

### Macroeconomic balance approach (method and determinants)
- Method summary:
  - Empirically based approach estimating an equilibrium relationship between current account balances and fundamentals via panel econometrics.
  - Steps: (i) estimate panel relationship; (ii) compute each country’s current account norm using medium-term projections of fundamentals; (iii) derive real exchange rate adjustment to close gap between norm and actual current account.
- Modifications for oil-producing countries:
  - Add non-oil fiscal variables to separate fiscal effects of oil revenue and fiscal policy.
  - Run panel regression on oil-producing countries only (sample includes Algeria, Bahrain, Indonesia, Iran, Kazakhstan, Kuwait, Libya, Norway, Oman, Qatar, Russia, Saudi Arabia, Syria, United Arab Emirates, and the eight SSA oil countries).
  - Exclude economic crisis and financial center variables as they are not relevant for oil-producing SSA countries.
- Robust determinants of the current account over the medium term (expected signs):
  - Non-oil fiscal balance (positive).
  - Demographics: old-age dependency ratio and population growth (expected signs negative).
  - NFA (ambiguous sign).
  - Lagged current account (positive, if persistence).
  - Oil current account balance (positive).
  - Economic growth: growth of real per capita GDP (negative); ratio of per capita PPP income to US level (positive).
- Additional analysis:
  - Separate regression including the return on oil wealth to assess consumption based on current versus permanent oil income, with the non-oil current account as representation of national consumption decisions.
  - The sum of oil-related coefficients should equal (minus) unity if oil revenue is fully reflected in the non-oil current account (following Thomas and Bayoumi (2009)).

### Estimates of long-term external sustainability (overall conclusions)
- Baseline estimates across the three models (consumption smoothing, precautionary saving, macroeconomic balance) broadly align with fiscal-sustainability-focused studies, suggesting governments in oil-producing SSA countries must save more.
- Acknowledged large margins of uncertainty due to strong assumptions, but results are considered robust insofar as different models produce qualitatively similar outcomes.
- Table 6 (optimal current account balances, percent of GDP, 2006) — comparison across methodologies and actual values:
  - Angola: Consumption Smoothing 54.4; Precautionary Saving 1.0; Macro Balance 9.8; Actual 25.2
  - Cameroon: Consumption Smoothing 12.7; Precautionary Saving 0.0; Macro Balance 0.5; Actual 1.6
  - Chad: Consumption Smoothing 48.6; Precautionary Saving 0.0; Macro Balance -30.7; Actual -9.0
  - Congo: Consumption Smoothing 57.5; Precautionary Saving 0.7; Macro Balance 5.9; Actual 1.5
  - Côte d'Ivoire: Consumption Smoothing 6.8; Precautionary Saving 0.0; Macro Balance 0.4; Actual 2.8
  - Equatorial Guinea: Consumption Smoothing 76.0; Precautionary Saving 0.9; Macro Balance 3.7; Actual 7.1
  - Gabon: Consumption Smoothing 37.4; Precautionary Saving 0.5; Macro Balance 13.7; Actual 15.8
  - Nigeria: Consumption Smoothing 26.0; Precautionary Saving 0.6; Macro Balance 0.1; Actual 26.5

*Source: Authors' estimates (from the supplied content).*

### 0.7 percent in 2015.

### _wp11207 - 0.7 percent in 2015.

### Overview
- The paper assesses long-term external sustainability for eight SSA oil-producing countries: Angola, Cameroon, Chad, Côte d’Ivoire, Equatorial Guinea, Gabon, Nigeria, and Republic of Congo.
- Three approaches are used: consumption-smoothing (forward-looking optimizing), precautionary saving (forward-looking with uncertainty), and macroeconomic balance (backward-looking).
- Key structural constraint: in six of the eight SSA oil-producing countries, oil reserves (not including gas) could be exhausted over the next 20 years without further exploration and development; this compares with projections of 100 years for middle- and high-income oil producers.

### Major empirical findings
- The macroeconomic balance approach (backward looking) produces a “normal” current external account most similar to actual outcomes.
- Consumption-smoothing and precautionary-saving models (forward looking) produce optimal current accounts that are substantially different from actual outcomes; the combined optimal current account in the precautionary model is even further away from actual.
- Assessment of current versus permanent oil income (Thomas and Bayoumi (2009) approach) indicates that national consumption decisions in oil-producing SSA countries are based on current oil income, with the current oil balance being the only statistically significant oil-related explanatory variable and its coefficient close to unity. This suggests these countries may not be sufficiently forward looking in consumption behavior.

### Key quantitative results from regressions and model interpretation
- Fixed effects estimates (column 3 of Table 7a): a 1 percentage point increase in old-age dependency worsens the current account balance by about 1 percent of GDP; a 1 percentage point increase in population growth worsens the current account balance by about 1 percent of GDP; a 1 percentage point increase in the oil balance improves the current account balance by about 0.4 percent of GDP; a 1 percentage point increase in relative income improves the current account balance by about 0.4 percent of GDP.
- The positive coefficient on lagged current account balance implies persistence in the current account balance for these countries.
- For the non-oil current account (fixed effects, columns 5 and 6): an unexpected negative and significant coefficient on the oil balance implies the non-oil current account deteriorates as the oil balance improves. Possible explanations:
  - Non-oil GDP is highly correlated with oil GDP; higher oil production stimulates the non-oil sector, increases imports, and worsens the non-oil trade balance.
  - Higher oil production raises imports of oil-related investment, goods, and services.
- The negative sign on the lagged non-oil current account balance suggests lack of persistence in the non-oil current account, contrasting with other countries.
- The positive sign on net foreign assets (NFA) indicates higher NFA generates more investment income, improving the non-oil current account.

### Robustness and sensitivity (consumption-smoothing and precautionary models)
- Consumption-smoothing approach:
  - The optimal current account balance is sensitive to initial conditions and model parameters such as the rate of return, oil prices, and non-oil cash flow.
  - If the value of the asset rises (e.g., higher oil prices or higher returns from exploitation), more can be consumed now, leading to deterioration of the current external account.
  - Under fairly extreme assumptions (rate of return above 4 percent, a doubling of oil prices, or significant non-oil cash flow), the estimated current account balance is roughly around levels observed in 2006.
  - Baseline parameterization: risk averseness (σ)=2 and discount rate (β)=4%.
- Precautionary-saving approach:
  - Higher risk aversion increases precautionary saving while leaving the consumption-smoothing component constant.
  - Raising the discount rate from 4 percent to 10 percent triples the precautionary-saving component in most countries while the consumption-smoothing component declined by about 34 percent (average over assessed period). The net effect is a lower overall optimal current account as the consumption-smoothing decline dominates.
  - A 1 percentage point increase in labor supply growth would halve the precautionary-saving component in some countries and increase the consumption-smoothing component by about 17 percent (average over assessed period). Because the consumption-smoothing component is much larger, the overall optimal current account rises.
- Sensitivity to initial values:
  - Increasing NFA by 10 percentage points of non-oil GDP changes outcomes.
  - Increasing the share of non-oil GDP reduces both precautionary saving and consumption smoothing, substantially lowering the estimated sustainable current account versus baseline.
  - A higher initial oil price reduces the sustainable current account.
- Overall robustness conclusion: the models’ outcomes are sensitive to initial conditions (particularly the share of the non-oil sector) and less sensitive to model parameters, but across varying assumptions the recommendation that oil-producing SSA countries should accumulate further precautionary savings is robust.

### Policy-relevant interpretations and recommendations
- The actual current account balances of the eight SSA oil-producing countries appear “close” to optimal only under backward-looking (macroeconomic balance) analysis, indicating prior backward-looking policies; countries need to shift to forward-looking policies urgently given dwindling oil production horizons.
- These countries are not sufficiently forward looking in consumption and saving; national consumption appears tied to current oil income rather than permanent oil income.
- Benchmarks for sustainable external current balances should:
  - Be forward looking and account for external shocks and sensitivity to model parameters and initial conditions.
  - Consider the potential for front-loaded public investment in economic infrastructure (transportation, energy, water, and other public utilities) that could develop the non-oil sector, raise the return on investment, spur non-oil growth, and broaden the export base beyond oil.
  - Maintain consistency between fiscal sustainability and external sustainability analyses because private-sector participation in the oil sector can create wedges between fiscal and external outcomes; private-sector saving and investment decisions matter for results.
- Given volatile world oil prices and relatively short oil-production time horizons, the countries should be more forward looking in trade-offs between public investment, saving, and consumption of exhaustible resources; accumulating additional precautionary savings is a recommended policy stance to guard against oil price uncertainty.

### Conclusions
- Long-term external sustainability estimates are inherently uncertain and model-dependent.
- The macroeconomic balance (backward-looking) approach aligns more closely with past outcomes, but forward-looking approaches (consumption smoothing, precautionary saving) suggest current accounts are far from sustainable.
- Without new discoveries and with short extraction horizons, forward-looking policy adjustments and precautionary savings accumulation are urgent.
- Front-loaded public investment to develop the non-oil sector could materially alter the sustainability outlook by raising returns and changing core model parameters.

*Source: Authors' estimates, _wp11207 - 0.7 percent in 2015.*

### References

### _wp11207 - References

### Exchange rate assessment and misalignment
- Akitoby, B., C., C. Gueye, D. Milkov, and M. Poplawski-Ribeiro, 2011, “Assessing the Real Exchange Rate in CEMAC Countries,” IMF Working Paper, forthcoming (Washington: International Monetary Fund).
- Bems, R., and I. de Carvalho Filho, 2009b, “Exchange Rate Assessments: Methodologies for Oil Exporting Countries,” IMF Working Paper WP 09/281 (Washington: International Monetary Fund).
- Edwards, S., 1998, Exchange Rate Misalignment in Developing Countries, (Baltimore: The John Hopkins University Press).
- Hinkle, L., and P. Montiel, 1999, “Exchange Rate Misalignment: Concepts and Measurements for Developing Countries,” World Bank Research Publication (Oxford: Oxford University Press).
- Lee, J., G. Milesi-Ferretti, J. Ostry, A. Prati, and L. Ricci, 2008, “Exchange Rate Assessments: CGER Methodologies,” IMF Occasional Paper No. 261 (Washington: International Monetary Fund).
- Williamson, J., 1994, Estimating Equilibrium Exchange Rates (Washington: Institute for International Economics).
- Deléchat, C. and A. Kireyev, 2008, “CEMAC: External Stability and Exchange Rate Assessment in an Oil-Dependent Region.” IMF Selected Issues Paper, (Washington: International Monetary Fund).

### Current account determinants and benchmarks
- Bems, R., and I. de Carvalho Filho, 2009a, “Current Account and Precautionary Savings for Exporters of Exhaustible Resources,” IMF Working Paper WP09/33 (Washington: International Monetary Fund).
- Chinn, M., and E. Prasad, 2003, “Medium-Term Determinants of Current Accounts in Industrial and Developing Countries: An Empirical Exploration,” Journal of International Economics, Vol.59, No.1, pp.47–76.
- Cá Zorzi, M., A. Chudik and A. Dieppe, 2009, “Current Account Benchmarks for Central and Eastern Europe: A Desperate Search?,” European Central Bank, Working Paper Series No. 995 (Frankfurt: European Central Bank).
- Debelle, G., and H. Faruqee, 1996, “What Determines the Current Account? A Cross-Sectional and Panel Approach,” IMF Working Paper WP96/58 (Washington: International Monetary Fund).
- Christiansen, L., A. Prati, L. Ricci, and T. Tressel, 2009, “External Balance in Low Income Countries,” IMF Working Paper WP/09/221 (Washington: International Monetary Fund).
- Thomas, A., and T. Bayoumi, 2009, “Today versus Tomorrow: The Sensitivity of the Non-oil Current Account Balance to Permanent and Current Income,” IMF Working Paper WP/09/248 (Washington: International Monetary Fund).
- Thomas, A., J. Kim, and A. Aslam, 2008, “Equilibrium Non-oil Current Account Assessments for Oil-producing countries,” IMF Working Paper WP/08/198 (Washington: International Monetary Fund).
- Morsy, H., 2009, “Current Account Determinants for Oil-Exporting Countries,” IMF Working Paper WP/09/28 (Washington: International Monetary Fund).

### Fiscal sustainability and resource-dependent economies
- Chalk, N., and R. Hemming, 2000, “Assessing Fiscal Sustainability in theory and Practice,” IMF Working Paper WP/00/81 (Washington: International Monetary Fund).
- York, R., and Z. Zhan, 2009, “Fiscal Sustainability and Vulnerability in Oil-Producing Sub-Saharan African Countries,” IMF Working Paper WP/09/174 (Washington: International Monetary Fund).

*References list from _wp11207 - References*

---


_Source: https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2011/_wp11207.pdf_