Selected Issues


Selected Issues

Oil Revenues in Angola: Tradeoffs and Opportunities1

Angola is facing a stark trade-off between declining oil fiscal revenues over the medium term and increasing social and public investment needs. Opportunities do exist to make the most of Angola’s remaining oil reserves, whilst reducing its debt burden and building fiscal buffers. However, a sound fiscal framework for the use of oil revenues that includes a well-designed fiscal stabilization fund may be needed. This chapter assesses the macroeconomic implications of different fiscal rules for managing oil revenues in Angola, and discusses policy options to address related tradeoffs. Under a spend-as-you-go (SAYG) fiscal rule, falling oil production and volatile oil prices lead to declining revenues, volatile public investment, and rising debt. Under a more active fiscal rule, public investment can be scaled up gradually, whilst at the same time building fiscal buffers and insulating the non-oil economy from volatile oil price movements.

A. Introduction

1. Whilst Angola remains the second largest oil producer in Africa, oil production has leveled off and could decline in the medium term. Aging oil fields and years of under-investment due to lower oil prices could lead to a steady decline of oil production over the coming decades. Meanwhile, oil prices are expected to remain soft at US$50–55 per barrel over the medium term. With limited fiscal and external buffers and significant infrastructure and social spending needs. Angola needs to find a more sustainable mix of policies than those pursued in the past. It is facing trade-offs between declining oil revenues and increasing social and public investment needs. However, opportunities do exist to make the most of Angola’s remaining oil reserves, whilst reducing its debt burden and building fiscal buffers. This chapter assesses the macroeconomic implications of different fiscal rules for managing oil revenues in Angola and discusses policy options to address related tradeoffs.2

2. The challenge for resource-rich economies like Angola is how to transform subterranean assets (oil wealth) into productive assets above the ground, including financial, physical and human capital assets. There are essentially two fundamental questions that need to be answered regarding the use of natural resource revenues. First, how much should be consumed and how much should be saved? And second, out of savings, how much should be allocated to foreign vs. domestic assets. Venables and Wills (2016) argue that for capital abundant economies, all resource revenues should be used to accumulate foreign assets and only the annuity value of these assets should be consumed in any period—this is the permanent income hypothesis (PIH) rule. For a country like Angola, however, which is relatively capital scarce, has high infrastructure and social needs and incomplete access to international capital markets, a substantial proportion of resource revenues may be better allocated to paying down existing debt and investing in domestic assets and human capital formation.

3. The typical way in which resource rich economies have attempted to apply these principles is through commodity-based sovereign wealth funds (SWFs).3 According to the Santiago Principles (2008), there are five distinct types of SWF, designed to support different policy objectives of the government given the specific economic challenges the country faces. In the case of Angola, with substantial development needs and public finances that depend to a significant extent on oil revenues, some combination of a ‘stabilization’ fund and a ‘development’ fund would seem most suitable.

B. Angolan Oil Sector: Developments and Relevance

4. Angola is the second largest oil producer in Africa, but production has begun to decline of late. Angola produced, on average 1.632 million barrels per day (MMbbl/day) in 2017, second in Africa only to Nigeria.4 Proven reserves stood at 9.5 billion barrels in 2016, compared to 37.4 billion in Nigeria and 48.4 billion in Libya, making Angolan oil reserves the 3rd largest in Africa and the 17th largest in the world.5 Since the discovery of onshore oil in the Cuanza Basin in the 1950s, production levels gradually increased between the 1960s and 1990s, reaching almost 0.75 MMbbl/day by 2000. Deepwater exploration began in the 2000s leading to oil production levels soaring to nearly 2 MMbbl/day in 2008. However, since then production remained flat and, more recently, even begun to decline. Currently, most of the country’s oil production comes from offshore fields in the northern province of Cabinda and deep-water fields in the Lower Congo basin (Figure 1). Current production levels are around 10 percent lower than their two-year average of 1.75 MMbbl/day since August 2014, and below Angola’s OPEC quota of 1.673 MMbbl/day.

Figure 1.
Figure 1.

Oil Fields in Angola

Citation: IMF Staff Country Reports 2018, 157; 10.5089/9781484360255.002.A001

5. The oil sector in Angola is regulated by Sonangol, the state concessionaire and national oil company. Sonangol has stakes in 32 out of the 36 oil producing blocks in Angola, and is the 4th largest oil producer in the country, after BP, Total and Exxon. The largest oil producing block in Angola is block 17. It is located offshore in the Lower Congo Basin and operated by Total, Statoil, Exxon and BP, producing over 0.6 MMbbl/day of crude oil in 2016.

6. The Angolan oil sector is facing challenges due to underinvestment and the oil price collapse of 2014/15. Structural factors include the natural decline of oil fields, which has not been offset by entry into production of new oil wells, years of underinvestment in the sector, and (at least until recently) coordination problems between sector players and Sonangol.

7. Angola is heavily dependent on oil, both as a source of fiscal revenues and foreign exchange earnings. The oil sector accounted for about 64 percent of tax revenues and over 95 percent of exports in Angola in 2017. High dependence on oil revenues and volatile oil prices make budgeting difficult and increases the volatility of public spending. The track record has shown a systematically conservative oil price forecast but an over optimistic production forecast (Figure 2).

Figure 2.
Figure 2.

Oil Revenues in Angola, 2002–17

Citation: IMF Staff Country Reports 2018, 157; 10.5089/9781484360255.002.A001

Sources: OPEC, MINFIN and Fund staff calculations.

8. Angola needs to reduce the procyclical nature of government expenditure whilst at the same time building fiscal buffers and allowing room to increase public investment. It needs to address three key fiscal issues going forward. First, achieve a sustained scaling up of public investment to support growth and reduce poverty. Public investment in Angola is currently low relative to sub- Saharan African peers, at 5 percent of GDP in 2017. Second, entrench public debt sustainability, including by reducing public debt (projected to reach 72.9 percent of GDP at end-2018) to lower levels. Third, reduce its heavy dependence on oil revenues and better shield public spending from oil revenue volatility. The latter could be achieved through: (i) finding a more stable revenue source for the budget, for instance, by adopting a VAT on January 1, 2019, as planned, and (ii) building fiscal buffers to support countercyclical fiscal policies and savings for future generations. Fiscal buffers could be managed by a well-designed fiscal stabilization fund with clear deposit and withdrawal rules and within a coherent medium-term fiscal framework (MTFF). This chapter discusses the tradeoffs associated with these fiscal objectives and identifies policies that would allow scaling up public investment and reducing public debt through use of a stabilization fund and adequate management of non-oil revenue taxation.

C. Model

9. We use the IMF’s Debt, Investment, Growth and Natural Resources (DIGNAR) model to assess the macroeconomic impact of different fiscal rules of oil revenues in Angola. DIGNAR is a small open economy general equilibrium model, developed by Melina, Yang & Zanna (2016). It can be used to analyze the macroeconomic and debt sustainability effects of scaling up public investment in resource-rich developing countries. The model allows for increasing returns to scale in public infrastructure, but in a setting of inefficient public investment and absorptive capacity constraints. It also allows for flexible fiscal specifications for the use of oil revenues, including a spendas-you-go (SAYG) rule and an exogenous investment scaling up rule, as well as a resource fund that may be used as a fiscal buffer.

10. The model consists of three production sectors, two types of households, and a government that spends, taxes and issues debt. On the production side, the oil sector has exogenous paths for production and oil prices, the non-tradable sector uses labor, private capital and public infrastructure to produce output using Cobb-Douglas technology, and the tradable sector uses an identical production function but also exhibits a learning-by-doing externality on technology. On the consumption side, agents are either perfect foresight optimizers with access to financial markets or hand-to-mouth consumers. The government raises revenues via distortionary taxation (labor, capital and consumption taxes), receives payments from the oil sector and issues domestic and foreign debt. It then uses these revenues for recurrent expenditure, investment in public infrastructure and debt service.

11. The choice of a fiscal policy rule will determine how volatile oil revenues will affect economic performance. The model allows exploring several fiscal policy rules. The first is a SAYG rule in which all oil revenues in each period are used to fund public infrastructure investment, keeping recurrent spending fixed. Any endowment at the stabilization fund remains at its initial level and taxes adjust endogenously to keep debt on a sustainable trajectory. Another option is a delinked approach under which the path for public investment is determined exogenously, and financed by oil revenues, debt issuance and non-oil taxes. Any surplus (deficit) revenues are saved in (withdrawn from) the stabilization fund. Under the delinked approach, the government needs to make explicit choices related to the path for public debt and non-oil taxation.

12. Given the exogenous paths for oil production and the oil price, the choice of fiscal policy determines how much oil revenues are spent, and on what. The government can build a fund to insulate spending from volatile oil revenues. Choosing to invest in public infrastructure boosts growth by increasing the returns to private factors but also crowds out private investment, and is subject to inefficiencies—that is, only a fraction of public investment turns into public capital. Finally, the use of distortionary non-oil taxation to maintain debt sustainability has implications for the non-oil economy, by affecting the after-tax marginal return to capital and labor as well as the steady-state level of consumption.

13. The DIGNAR model is calibrated to the Angolan economy using macroeconomic data up to and for 2017. The calibrated parameters are shown in Table 1.

Angola: Table 1.

Calibrated Parameters

article image

D. Simulations

14. This section describes the simulations carried out to assess the macroeconomic impact of different fiscal rules for managing oil revenues in Angola. It begins by describing the oil production and price scenarios, before setting out the fiscal policy scenarios and simulation results.

Oil Production and Price Scenarios

15. We project a gradual decline in oil production by 2023 and an annual decline of 5 percent per year thereafter. As described in Section B, oil production in Angola peaked in 2008 but has remained relatively flat since the oil price collapse of 2014/15. We assume that oil production would remain roughly constant at 1.65 MMbbl/day until 2023.6 This would be consistent with the government implementing proposed reforms to improve the operational efficiency and financial soundness of Sonangol and encourage foreign investment in the sector to slow the decline in production levels.7 From 2023 onwards, our baseline scenario assumes an annual decline in production levels of 5 percent per year.

16. We use the IMF’s Spring 2018 WEO projections for oil prices out to 2023 and then simulate an AR(1) process for oil prices out to 2040.8 Oil prices are inherently volatile. Since the early 1990s oil prices have ranged between lows of US$12 per barrel in 1998 and highs exceeding US$110 in 2012. Since the collapse between 2014 and 2016, the oil price has recovered somewhat to just over US$50 per barrel in 2015. Using the IMF’s Spring 2018 World Economic Outlook (WEO) projections, we forecast a pickup in oil price for the Angolan basket to US$62.5 per barrel in 2018, followed by a gradual decline to US$54 per barrel by 2023. We then simulate two autoregressive processes for oil prices from 2023 onwards as follows:


where pt is the Angolan oil price at time t, ρ is an autoregressive (AR) term, and εt~N(0, σε) are i.i.d. shocks. Using data over the period 2016–2023, we estimate a persistence parameter, ρ = .93 and a standard deviation of oil price shocks of σε = 0.1. We use this parameterization to simulate the baseline oil price scenario. Whilst the period used is relatively short, absent any other information, this offers the ‘best guess’ regarding the dynamics of the oil price in the coming years. For the alternative scenario we increase the standard deviation of the shocks series threefold, to σε = 0.3, roughly the size of the shock to oil prices following the global crisis of 2007–08. The oil production and price scenarios are presented in Figure 3.

Figure 3.
Figure 3.

Oil Production and Price Scenarios, 1990–2040

Citation: IMF Staff Country Reports 2018, 157; 10.5089/9781484360255.002.A001

Sources: Angolan authorities and Fund staff calculations.

Simulation 1: SAYG Under Different Oil Price Scenarios

17. The first fiscal rule considered in this chapter is a spend-as-you-go (SAYG) policy rule. Under SAYG all oil revenues are used to finance public investment in every period, and the stabilization fund remains at its initial level of about 3 percent of GDP.9 We use the baseline oil production scenario and focus on the macroeconomic performance of the economy under two oil price scenarios: (i) a baseline path, and (ii) an adverse (more volatile) oil price path. The simulation results are presented in Figure 4.

Figure 4.
Figure 4.

SAYG Under Baseline and Alternative Oil Price Scenarios

Citation: IMF Staff Country Reports 2018, 157; 10.5089/9781484360255.002.A001

Sources: Fund staff calculations.

18. Under the SAYG rule, falling oil production and volatile prices lead to declining revenues, volatile public investment, and rising debt. Under the baseline scenario of falling oil production, government oil revenues, as a share of total revenues, decline from around 60 percent in 2017 to around 35 percent by 2040. As expected under the SAYG rule, the volatility in oil prices is translated into volatile resource revenues. The decline in revenues are transmitted directly into a decline in public investment which falls from 5 percent of GDP in 2017 to under 3 percent of GDP by 2040. Debt levels, after falling in the first year due to a temporary increase in oil prices, rise steadily to around 75 percent of GDP by the end of the simulation period. To prevent public debt from rising even faster, non-oil taxes must also increase over the simulation period, and remain highly volatile.

19. Higher (distortionary) non-oil taxation and falling public investment results in falling private consumption and investment. Together these cause a severe contraction of the economy, both in the short and medium term, with non-oil output falling by between 10–12 percent compared to the initial level.

Simulation 2: Delinked Approach with a Gradual Scaling Up of Public Investment

20. The second set of simulations assume that public investment is scaled up gradually over time to 8.5 percent of GDP. As noted in Section B, public investment in Angola is low compared to its peers. The second set of simulations assume that public investment is scaled up gradually over time from 5 percent of GDP in 2017 to 8.5 percent of GDP—the sub-Saharan average—by 2040. We focus on the baseline scenarios for oil production and prices to abstract from any direct income effects. After peaking at US$62.5 per barrel in 2018 the oil price gradually declines out to 2023, after which it is then subject to shocks out to 2040, as in the previous simulations. The simulation results are presented in Figure 5 where, compared with Figure 4, the path for resource revenue is replaced with that of the stabilization fund. Under the delinked approach, public investment rise and becomes much less volatile than under the SAYG rule.

Figure 5.
Figure 5.

Gradual Scaling Up of Public Investment Under the Delinked Approach

Citation: IMF Staff Country Reports 2018, 157; 10.5089/9781484360255.002.A001

Sources: Fund staff calculations.

21. We then assume three variants of a scaling-up fiscal rule:

  • 1. Passive tax policy—non-oil taxes respond as needed to keep debt on a non-explosive path (Figure 5, solid black lines).

  • 2. Active tax policy—non-oil tax rates are increased immediately from their initially low levels to create enough fiscal space and support building a stabilization fund (Figure 5, dotted blue lines).

  • 3. Debt paydown—the proceeds from gradual increases in non-oil taxes are used to prioritize debt paydown (Figure 5, dashed red lines).

22. Under the passive tax policy debt rises as public investment is scaled up, eventually requiring sharp non-oil tax increases and depleting the stabilization fund. Debt is on an upward trajectory from a very early stage due to the gradual scaling up of public investment and the saving of excess oil revenues in the stabilization fund. In the medium term, as oil revenues fall, non-oil taxes must be hiked sharply from 2025 onward to keep debt on a non-explosive path. Yet, debt exceed 70 percent of GDP towards the end of the simulation horizon. The public investment scaling up leads to a depletion of the stabilization fund in just five years. Private consumption—a major non-oil tax base in the model—is stable in the initial years of the simulation, but declines sharply as non-oil tax rates increase. It is worth noting however, that the decline in consumption under all three scenarios is less than under the SAYG rule scenarios. Higher public investment turns into higher public capital, which crowds in private investment. Together, higher private and public investment boost non-oil output growth. The stable paths for consumption and investment are due to the fact that fiscal policy is insulating the economy from the only source of volatility in the model—the oil price.

23. We now consider an alternative scenario in which the government pre-empts the declining oil revenues and begins to mobilize non-resource taxation immediately.10 Non-oil tax rates are assumed to rise immediately but smoothly over the simulation period. The key fiscal policy choice now is how to use the fiscal space created by mobilizing non-oil tax revenues.

24. Under the active fiscal policy, higher non-oil tax rates allow the government to build up a stabilization fund before transitioning to the financing of public investment. Excess government revenues in the early years, when public investment is still low, are used to build up a stabilization fund, which peaks at just over 7 percent of GDP by the mid-2020s, and debt remains relatively stable over the simulation period. Consistently higher non-oil taxes lead to declining private consumption from the outset, and the decline in the outer years is comparable to that under the passive fiscal policy rule. Stronger non-oil revenue mobilization also allows public debt to be reduced to about 50 percent of GDP towards the end of the simulation horizon.

25. Prioritizing debt paydown leaves less resources available for the stabilization fund in the near term but creates fiscal space to rebuild it when oil prices rise in the future. Under the debt paydown scenario, the initial excess oil revenues are used to pay down debt instead of accumulating in a stabilization fund. This reduces the debt service burden substantially with debt falling to around 40 percent of GDP by 2040. The government is now able to manage future fluctuations in the oil price, building a stabilization fund in years of high oil prices and running it down when prices fall. As under the active fiscal policy rule, the gradual increase in non-oil taxes leads to lower consumption, but higher private and public investment boosts non-oil output growth.

E. Concluding Remarks

26. Dwindling oil reserves combined with oil price volatility and significant spending needs would benefit from a sound MTFF for managing oil revenues. Given the volatile nature of oil prices, and the high dependence of the budget on oil revenues, the establishment of a stabilization fund within a MTFF would allow the government to achieve the dual objective of (i) building a fiscal buffer to reduce Angola’s vulnerability to a decline in oil revenues, and (ii) insulating public investment spending from oil price volatility.

27. Prioritizing growth-enhancing infrastructure investment and improving the business environment could yield significant growth benefits. The fiscal policies pursued under the delinked approach suggest that scaling up public investment could have a strong impact on non-oil GDP growth. However, public investment should increase gradually, in line with absorptive capacity. Moreover, Angola’s public investment management framework should be improved to reduce inefficiencies (which have been estimated to be high), increase the quality of public infrastructure, and maximize the crowd in of private investment.

28. Using part of an oil windfall to reduce public debt could be prioritized, as well as reforms to mobilize additional non-resource tax revenues. The use of a stabilization fund to manage excess oil revenues raises the important question of portfolio allocation for such a fund. Given the level of public debt—projected to reach 72.9 percent of GDP in 2018—priority could be given to using some of the oil revenue windfalls to pay down debt. Structural fiscal reforms would need to be implemented alongside the fiscal framework for oil revenues. These include mobilizing domestic non-oil tax revenues, including by implementing a VAT and increasing public investment efficiency as mentioned above.


  • Hamilton, J.D. 2009. “Understanding Crude Oil Prices.” The Energy Journal 30 (2): 179206.

  • International Working Group of Sovereign Wealth Funds. 2008. Sovereign Wealth Funds: Generally Accepted Principles and Practices, “Santiago Principles”.

    • Search Google Scholar
    • Export Citation
  • Melina, G., S.C.S. Yang, and L.F. Zanna. 2016. “Debt Sustainability, Public Investment, and Natural Resources in Developing Countries: The DIGNAR Model.” Economic Modelling 52: 630649.

    • Search Google Scholar
    • Export Citation
  • Richmond, C. J., I. Yackovlev, and M.S.C.S. Yang. 2013. “Investing Volatile Oil Revenues in Capital-Scarce Economies: An Application to Angola.” IMF Working Paper No. 13–147. International Monetary Fund.

    • Search Google Scholar
    • Export Citation
  • Venables, A. J., and S.E. Wills. 2016. “Resource Funds: Stabilizing, Parking, and Inter-Generational Transfer.” Journal of African Economies 25 (suppl. 2): ii20ii40.

    • Search Google Scholar
    • Export Citation

Prepared by Thomas McGregor (AFR).


A similar study was conducted for the 2012 Article IV Consultation, see Richmond, Yackovlev & Yang (2013). The model used here is that of Melina, Yang & Zanna (2016).


Some examples of SWFs include: Chile – structural balance rule; spending = difference between structural balance target and estimate structural revenue; Norway – spending rule; non-oil budget averages 3 percent of SWF; Russia—mechanism to reduce budget’s dependence on oil and gas revenues (abandoned after 2009 crisis until 2015); Timor-Leste—3 percent of fund + oil reserves transferred to budget; Nigeria—price-based fiscal rule; revenue above a predefined oil price is saved in fund.


OPEC statistics show that Angola was the largest oil producer in Africa and the 12th largest in the world in 2016.


OPEC Annual Statistical bulletin 2017.


We use Sonangol’s current production projection for the short-run, 1.650MMbbl/day, which constitutes a decline of 5.2 percent between 2017 and 2023 and is a conservative assumption for the decline in production.


The government is committed to revitalizing the investment climate in the oil and gas industry, by implementing the recommendations of the five task forces that were convened by President João Lourenço early in his administration to stem the risk of declining oil production over the medium term.


Real oil prices are typically thought to follow a random walk process without drift—that is, an AR(1) process with constant and persistence coefficient equal to 1. Here we explicitly estimate this relationship for the Angolan oil price. It turns out that the persistence term is estimated to be ρ = .93, and statistically different from 1, which we use in our simulations. This process behaves almost identically to a random walk process.


The SWF of Angola (FSDEA) was initially capitalized with USD 5 billion in 2013/14. In the simulations we assume the initial size of the stabilization fund at 3 percent of GDP. In addition, the government has liquid assets deposit at the BNA but we ignore these additional funds in the simulations.


Like many other resource-rich countries, including in the SSA region, Angola’s tax potential is considerably higher than its actual tax-to-GDP ratio (e.g., Regional Economic Outlook for sub-Saharan Africa, Chapter 1, October 2015).

Angola: Selected Issues
Author: International Monetary Fund. African Dept.