Information about Sub-Saharan Africa África subsahariana
Journal Issue
Share
Article

Angola: Selected Issues and Statistical Appendix

Author(s):
International Monetary Fund
Published Date:
April 2005
Share
  • ShareShare
Information about Sub-Saharan Africa África subsahariana
Show Summary Details

I. Oil Sector and Government Revenues1

A. Introduction

1. Oil production in Angola currently accounts for about half of GDP and about 75 percent of government revenue. With oil production forecast to double over the next three years, projections for the government’s fiscal position in the medium term will be crucially dependent on both the value of oil production and the proportion that will accrue to the government. However, in addition to the usual uncertainties associated with projections of the total value of oil output, the government’s share has recently been subject to volatility. For instance, between 2000 and 2003, while both overall production levels and government tax revenue from oil were generally rising, data now available indicate that the government’s share of total oil receipts declined from 53 to 43 percent.

2. This note examines the tax system of the Angolan oil industry and recent production and tax data, to try to estimate the effects of some identifiable factors which might have caused changes in the government’s share in total oil receipts in Angola and to produce forecasts.2 Although a structural model of the Angolan oil sector– such as the oil revenue model, developed for the independent Oil Diagnostic Study, published by the government in May 2004– would provide the best way of doing this, use of such a model requires comprehensive data on oil companies’ costs and careful calibration of its results against outturn revenue data. In advance of fully articulated results from such an exercise, this note contributes a methodology to estimate the costs claimed by oil companies, using data about their combined tax liabilities, and suggests a way of using such data to forecast the average government share in oil receipts and future government oil revenues.3 Illustrative forecasts of the government share and of future government revenues from oil are also shown.

3. The remainder of this note is organized in five sections: Section B contains a brief introduction to the Angolan oil industry; Section C describes the oil taxation system in Angola; Section D examines how production shifts between different production blocks have contributed to the fall in the government’s share of total oil receipts; and Section E describes a methodology for estimating costs and deductions from tax date and hence for forecasting future oil revenues for the government. Section F concludes.

B. Oil Production in Angola

4. Angola’s oil output is expected to double to 2 million barrels per day (bpd) around mid-2007. Moreover, with proven reserves of over 5 billions barrels,4 Angolan oil production could continue at very high levels into the 2020s. In addition to oil, recent plans to extract natural gas and to build a LNG plant, as well as improved extraction of liquid petroleum gas (LPG), are expanding the range of activities of the Angolan natural resource sector.

5. Angolan oil has been flowing predominantly from offshore fields since production from shallow waters started off the cost of the Province of Cabinda in 1968. Although the development of shallow water fields sustained a doubling of output during the 1980s, it was not until the 1990s, with the application of new techniques of deepwater drilling, that a dramatic boost in production occurred; the newly discovered deep water fields pushed production from under 0.5 million bpd in 1990 to the current level of 1 million bpd, and has added considerably to reserves.

6. In 1976, the government set up a national oil company, the Sociedade Nacional de Combustivel de Angola (Sonangol) and in 1978 promulgated a law that made the government the sole owner of Angolan oil and Sonangol the sole concessionaire for its exploration and extraction. Subsequently, the continental shelf has been divided into a total of 35 blocks, most of which have been offered for bidding by international oil companies for the development of extraction activities. Seven production blocks are currently active: the onshore blocks FS and FST (the only onshore blocks now in service), the shallow water blocks off Cabinda (called “Block 0”), the shallow water block 2, deep water block 3, and the newer deep and ultra-deep water blocks 14, 15, and 17. Other blocks (such as blocks 31 and 32) are currently being explored and another (block 18) is expected to start production in 2007 (see Table I.1 for a summary).

Table I.1.Angola: Summary of Oil Industry, 2004
BlockTypeProduction

Started/Planned in
Production, 2004

(thousands of barrels per

day)
First Commercial

Discovery
CabindaShallow water1973393
FS - FSTOn-shore13
2Shallow water39
3Deep water120
14Deep water2000611997
15Ultra-deep waterlate 20031341998
17Ultra-deep water20022351997
18Ultra-deep water20071997
Source: Upstream, September 17 2004.
Source: Upstream, September 17 2004.

7. Future growth in output is expected to come primarily from new deep and ultra-deep water blocks, as production has peaked, or is close to peaking, in the more mature onshore and shallow-water fields (see Figure I.1). These newer blocks have provided the major contribution to production growth in the past three years, and their profitability has also been the major drive for investments to the industry (see Table I.2).

Figure I.1.Angola: Oil Production by Major Blocks, 2000 - 2008

(Millions of barrels per day)
Table I.2.Angola: Major Current Investment Projects
BlockProjectTotal Expected Investment

(billions of U.S. dollars)
Expected Contribution to

Production

(thousands of barrels per day)
Declared Recoverable Reserves

(millions of barrels)
14Benguela/Belize2.2120 (late 2005)
15Kizomba A Project3.4250 (2005)1,070
15Kizomba C Project3.0250 (2007)
17Dalia Project3.9225 (mid 2006)1,590
18Great Plutoniobetween 2 and 3250 (late 2007)
Source: Upstream, 17 September 2004, and EIA.
Source: Upstream, 17 September 2004, and EIA.

C. Oil Taxation in Angola

8. There are two distinct oil tax regimes in Angola: (i) a tax and royalty regime and (ii) production sharing agreements (PSA). The Angolan government also raises revenue from signature bonuses and exploration rights. The tax and royalty regime applies to operations in the area of Cabinda and in the onshore blocks FS-FST. In 2003, company operations in these blocks accounted for almost 47 percent of total production and for almost 48 percent of oil revenues accruing to the government. PSAs apply in all other blocks.

9. The Angolan tax and royalty regime provides revenue to the government through three taxes: the production tax (commonly known as royalty), the income tax (commonly referred to as Petroleum Income Tax, PIT), and the transaction tax (commonly referred to as Petroleum Transaction Tax, PTT).

Box I.1.Angola: Tax and Royalty Regime

To compute royalties, transaction, and income taxes the following rules and concepts apply:

1) Output value. As a first step, to compute the tax base towards all taxes (production, income and transaction taxes), oil output is valued monthly at an agreed average market price (preço de referência fiscal).

2) Production tax. The production tax (royalty) is an ad valorem tax levied on the value of production. The rate is 20 percent in the Cabinda blocks and 16.67 percent in blocks FS and FST. The tax is payable in kind or in cash.

3) Taxable income. Taxable income is computed as the value of production minus:

  • i) operating costs (including rents paid to third parties for exploration and development activities);
  • ii) amortization of exploration, development and installation costs.

The full costs of exploration, development, and installation of production sites are amortized over a period of six years.

4) Transaction tax (petroleum transaction taxes, PTT). Transaction taxes apply only to the Cabinda blocks and are levied on taxable income with a fixed 70 percent rate. Production and investment premiums are deductible towards this tax.

Production premiums equal the value of output computed at a reference price. The reference price grows by 7 percent a year and differs from block to block (in 2004 the price for Cabinda A was US$ 18.217). Investment premiums are equal to 50 percent of the amounts invested and capitalized each year for the block Cabinda A, and they are calculated according to an agreed formula for the blocks Cabinda B and C.

5) Income tax (petroleum income tax, PIT). This is levied on taxable income and has a flat rate of 65.75 percent. Production and transaction taxes are deductible toward this tax.

Income taxes are payable within the same framework as transaction taxes.

10. The production tax is levied on the value of output produced by joint ventures. This tax guarantees revenue to the government even if production is not profitable for the ventures. Income and transaction taxes are instead levied on net income, which is total production minus operating costs and amortization. Because the production tax increases the marginal cost of extracting oil, income and transaction taxes are designed in part to offset the distortionary effects of the production tax. This is effected by allowing companies to deduct “production and investment premiums” (also called “production and investment incentives”) in proportion to the level of production reached and the level of investments incurred (see Box I.1).

11. Companies operating under Production Sharing Agreements function as contractors to Sonangol (which operates as government concessionaire), by either forming consortiums or acting individually. If a consortium is formed, exploration and production are carried out through an operator, which may or may not be the company with the largest share in the consortium. Sonangol can itself participate in the consortium, which it now does through a production subsidiary (Sonangol Pesquisa & Produção).

Box I.2.Angola: PSA Regime

In Angola, cost oil, profit oil and income taxes are computed in the following way:

1) Determination of cost oil. In any given year companies are allowed to recoup their costs up to a fixed proportion of gross revenues from a given field (this cap is normally set at 50%). Allowable costs include:

  • i) operating costs;
  • ii) amortization of exploration costs, development expenses, and costs for abandoning the field;
  • iv) production premium.

Operating costs are recoverable on a recurrent basis whereas exploration and development costs are recoverable over a period of four years. Companies are allowed to increase the initial value of amortizable costs by a certain percentage (“uplift”), which is, on average, 50 percent.1 Recoverable costs in excess of the cap are rolled over to subsequent years. In most blocks, if recoverable costs are not fully amortized within five years since they were incurred, the cap increases.

2) Government’s share of profit oil. Profit oil is computed as:

Profit Oil is split between the government and the oil companies according to formulas that change from block to block and that generally depend on either the water depth of the wells, or the cumulative production within a block, or the rate of return of the block.2 Most commonly, the government’s share varies from 20 percent of profit oil when the rate of return is below 25 percent, to 90 percent of profit oil when the rate of return is above 40 percent.

3) Income tax. Under a PSA, oil companies pay income taxes on their share of profit oil at a rate of 50 percent. No other deductions are allowed towards this tax.

1 The practice of allowing an “uplift” on amortizable costs is a feature of PSA contracts in many countries; but Angola’s 50 percent provision is unusually high.2 The rate of return is based on accounting profits and total accounting costs.

12. The companies participating in a consortium finance all the necessary investments and operating costs and then, when production starts, they recoup these costs by retaining a share of the oil produced (“cost oil”). What remains (“profit oil”) is shared between the government and the companies (including Sonangol whenever participating as an operator) in proportions that depend, from block to block, on the cumulative production of the block, its internal rate of return, or the depth of the wells under the sea level. Sonangol, operating as “concessionaire”, is responsible for marketing and remitting the government’s share of total profit oil. Finally, companies pay income taxes on their share of profit oil (see Box I.2).

13. The main effect of implementing a PSA rather than a tax and royalty regime is that ownership of the oil and control of oil activities remains with the government; the tax provisions can be adjusted to be largely equivalent between the two systems. By setting a cap on recoverable cost oil, revenues are guaranteed to the government under a PSA even if extraction is not profitable for the consortium. From this point of view, the cap works as a royalty and creates the same distortions to incentives as a royalty.

D. Government Revenue Take: Past Estimates of TRR

14. The government’s reported share of total oil receipts fell from 53 percent in 2000 to 43 percent in 2003 (see Table I.3). This section explores and quantifies possible explanations. Because no fully consistent disaggregated data are currently available for 2000 or 2001, the analysis focuses on the period from 2002 to the first half of calendar year 2004.

Table I.3.Angola: Government Revenue Take, 2000 - 2004(in US$ millions)
20002001200220032004 1/
Value of oil production7,4146,1447,7399,0075,242
Total government oil revenue3,9453,1943,3043,8922,257
Implied goverment share (in percent)53.252.042.743.243.1
Sources: Ministry of Finance web site http://www.minfin.gv.ao/dni/petroleo/exportacao_2004.pdf; Sonangol; and IMF staff estimates.

First half of calendar year 2004 only.

Sources: Ministry of Finance web site http://www.minfin.gv.ao/dni/petroleo/exportacao_2004.pdf; Sonangol; and IMF staff estimates.

First half of calendar year 2004 only.

15. The government’s share in gross oil revenues has varied appreciably from block to block as well as over time. In 2002, for example, the share varied from 24 percent in the relatively new deep-water block 17 to 62 percent in the more mature deep-water block 3. By the first half of 2004, the government share had risen to 26 percent in Block 17 and fallen to 59 percent in Block 3. Possible factors behind these changes include differences in tax regimes, the precise structure of the profit-sharing agreements, the production costs and maturities of fields (which affects the level of investment and profitability), as well as variability in oil prices and quality.

16. The most frequently cited explanation of the large decline in the government’s share of oil receipts, which subsequently will be called the Tax / Receipts Ratio (TRR), is the shift in production towards newer, deep-water fields. In addition to the effect of the PSA regime itself, lower TRRs in these fields may reflect:

(i) High start-up costs. In the early production life of new fields (or when re-investment is needed), there will be heavy investment costs to be amortized. These will contract taxable income. Moreover, because of the taxation system, the government’s share of net revenues is lower at the early production life of new fields.

(ii) Higher operating costs and tax allowances. Deep-water exploration requires more extensive investment and higher operating costs than shallow-water or onshore fields, so that costs are likely to absorb a greater part of the gross revenue accruing to oil companies and to allow companies to claim higher tax allowances.

(iii) Oil quality. Oil from the newer blocks is heavier and of poorer quality than the oil extracted from more mature fields (see Table I.3).

17. In order to explore further the decline in the aggregate TRR, it is helpful to consider this as the weighted average of the TRRs that originate in different blocks, using as weights the percentage contribution of each block to total oil receipts. Based on changes in the composition of the oil sector in recent years, changes in the TRR can be then be decomposed into:

i) a compositional effect arising from differences in TRRs between blocks and changes in the relative importance of different blocks over time because of changes in production or relative prices;

ii) changes in the TRR within each block because, for example:

  • a) when oil receipts increase in a block, deductions or cost oil become proportionally lower with respect to oil receipts (or stay constant if they are subject to a ‘cap’). Hence, for a given level of operating costs and amortization, when production and/or oil prices increase in a block, the TRR of that block increases as well (or, in the limit, stays the same);
  • b) under a PSA, the tax structure depends on the internal rate of return, so that a change in the oil receipts in a PSA block has the potential to affect the TRR within that block.

18. Table I.5 analyses in a simple way how the decomposition explains the drop in the government’s share of oil receipts between 2002 and 2003, and between 2003 and the first half of 2004. The table shows the TRR in each oil block for the periods and the share of each block in total oil receipts. It then calculates (in the penultimate column) the effect of changes within blocks on the overall total, assuming a fixed weight between blocks and (in the final column) the effect of changes between blocks, assuming a roughly constant TRR in each block.

Table I.4.Angola: Price Differentials from Price of Brent Crude (Smoothed), 2004
BlockFieldDiscount to price of Brent

crude (in percent)
CabindaAverage4
Cabinda ANembaparity
FS, FSTSoyo1
Block 3Canuku1
Block 14Benguela7
Kuito15
Landana5
Negage5
Block 15Hungo7
Xikomba5
Block 17Dalia9
Girassolparity
Block 18Plutonio5
Source: World Bank.
Source: World Bank.
Table I.5.Angola: Government Revenue Take (TRR) and Share of Production by Block, 2003 - 2004
20022003Contribution to Change in TRR
TRRShare of blocks in

total oil receipts
TRRShare of blocks in

total oil receipts
Weighted

changes in

TRR
Weighted changes in

share of oil receiptss
Cabinda47.048.452.046.52.35-0.94
FS-FST46.01.444.11.6-0.030.12
Block 247.75.650.34.80.13-0.42
Block 3 & Canuko61.814.959.614.6-0.33-0.17
Block 1427.27.720.26.6-0.50-0.25
Block 150.00.021.10.70.070.07
Block 1724.222.022.725.2-0.340.75
Total Angola42.7100.043.2100.01.36-0.85
20032004 (until June) 1/Contribution to change in TRR
TRRShare of blocks in

total oil receipts
TRRShare of blocks in

total oil receipts
Weighted

changes in

TRR
Weighted changes in

share of oil receiptss
Cabinda52.046.551.442.1-0.25-2.30
FS-FST44.11.645.41.90.020.15
Block 250.34.851.74.50.07-0.13
Block 3 & Canuko59.614.658.613.8-0.14-0.48
Block 1420.26.622.94.80.16-0.38
Block 1521.10.732.18.00.481.96
Block 1722.725.225.924.80.80-0.09
Total Angola43.2100.043.1100.01.14-1.29

Data from the the web site of the Ministry of Finance; http://www.minfin.gv.ao/dni/petroleo/exportacao_2004.pdf

Sources: Ministry of Finance; and IMF staff estimates.

Data from the the web site of the Ministry of Finance; http://www.minfin.gv.ao/dni/petroleo/exportacao_2004.pdf

Sources: Ministry of Finance; and IMF staff estimates.

19. Between 2002 and 2003, the overall TRR increased by 0.5 percentage points. As shown in the last two columns of Table I.4, this reflected two opposing factors:

i) the weighted average of the TRR within blocks (assuming constant share between blocks) increased by 1.36 percentage points; while

ii) the change in the share of value added between blocks resulted in a decrease in the overall TRR of 0.85 percentage points.

Similarly, between 2003 and the first half of 2004, the overall TRR increased by 0.1 percentage points, reflecting the net effect of:

i) the weighted average of the TRR within blocks (assuming constant share between blocks) increased by 1.14 percentage points due to higher production and higher oil prices; while

ii) the change in the share of value added between blocks resulted in a decrease in the overall TRR of 1.29 percentage points.

20. The implication of these calculations is that the shift in production toward blocks with a lower TRR did substantially reduce the overall TRR: the change in the share of value added between blocks reduced the increase in the TRR by over 2 percentage points between 2002 and the first half of 2004.

21. The effect of changes in production share was however more than offset by the impact of increasing TRR within blocks. As indicated below, this rising TRR within blocks was itself largely caused by the rising average price of oil.

22. In the earlier years of the decade, the oil price was much lower. As a result, the impact of changes in the share of value added between blocks between 2000 and 2002 was likely to have been much larger, and may therefore explain a large part of the observed sharp decline in the overall TRR. However, insufficient data are currently available to test this proposition.

E. Government Revenue Take: Projections

23. Both changes in the TRR within blocks and changes in the composition of output need to be taken into account when forecasting the share of oil taxes in gross receipts. But the major element of indeterminacy comes from the need to project operating costs and amortization and their relationship to the selling price of oil from individual fields. This derives from the fact that only in the case of production taxes (royalties) is revenue dependent on gross rather than net receipts. Moreover, because of the cap on cost oil, the amount of costs that companies operating under PSA can deduct in a fiscal year depends on the amount rolled over from the previous year.

24. The most straightforward way to construct projections of government revenues would be to forecast future costs from data on past costs and incorporate them within a structural model reflecting detailed provisions of individual tax and PSA, as embodied in the tax revenue model designed for the oil diagnostic study. However, this can only be effective if recorded data on government revenues match estimates produced for the recent past by the structural model. An alternative approach adopted here is to take the observed past relationships between oil revenues and the government’s take to derive estimates of costs and the major revenue parameters and hence to project costs and revenues. In particular, since the marginal cost of extraction is likely to be very low, and since amortizations have a large component of history dependence, estimates of past costs can be readily extracted; using the formulas described in section B and data related to tax liabilities and production, it is then possible to project future tax liabilities or calibrate projections of the TRR.

25. In order to extrapolate information about costs and amortization, it is necessary to know the amount which companies paid under each category of tax. These tax data are then manipulated as described in Box I.3 using inversions of the formulae described in section B. The analysis is based on different and somewhat incomplete sources of data for the fiscal years: 2002, 2003, and 2004.5 The implied levels of costs in each year are shown in the first three columns of Table I.6 and I.7.

Table I.6.Angola: Costs and PTT as Percentage of Total Production; Cabinda Blocks, and FS-FST Blocks, 2002 - 2008
Projections
200220032004 1/2005200620072008
Prices 2/23.728.236.230.230.831.432.0
Cabinda A
Production (millions of barrels)1009792104114120114
Implied Costs (in millions of US $)9809631,1291,4521,5931,2021,179
Implied Costs ($/barrel)9.89.912.313.9141010.3
Implied Costs / Value of Prod. (percent)42.335.234.346.646.032.232.6
Implied Inv. Premium (in millions of US $)0-1802,178000
TRR (in percentage points)45.150.855.342.042.351.451.2
Cabinda B
Production (millions of barrels)49433847495250
Implied Costs (in millions of US $)357234360385407436430
Implied Costs ($/barrel)7.35.59.48.28.38.48.6
Implied Costs / Value of Prod. (percent)30.019.225.126.226.025.825.9
Implied Inv. Premium (in millions of US $)00.26.31.51.81.31.0
TRR (in percentage points)53.060.258.255.956.156.155.9
Cabinda C
Production (millions of barrels)10836433
Implied Costs (in millions of US $)13517870122776357
Implied Costs ($/barrel)13.821.321.021.121.221.321.4
Implied Costs / Value of Prod. (percent)55.575.464.177.276.074.973.8
Implied Inv. Premium (in millions of US $)0000000
TRR (in percentage points)36.323.130.421.922.623.324.0
FS-FST
Production (millions of barrels)5565544
Implied Costs (in millions of US $)41617170656250
Implied Costs ($/barrel)8.812.412.213.213.814.114.2
Implied Costs / Value of Prod. (percent)39.041.832.241.942.943.042.5
TRR (in percentage points)46.044.150.543.943.243.243.5
Sources: DNI; IMF staff estimates and projections.

Estimates based on data for the period January - September 2004 published on the web site of the Ministry of Finance; http://www.minfin.gv.ao/dni/petroleo/exportacao_2004.pdf.

Prices are used to compute cost oil / value of production and the TRR.

Sources: DNI; IMF staff estimates and projections.

Estimates based on data for the period January - September 2004 published on the web site of the Ministry of Finance; http://www.minfin.gv.ao/dni/petroleo/exportacao_2004.pdf.

Prices are used to compute cost oil / value of production and the TRR.

Table I.7.Angola: Cost Oil and Government Profit Oil; Blocks from 2 to 18, 2002 - 2008
Projections
200220032004 1/2005200620072008
Prices 2/23.728.236.230.230.831.432
Block 2
Production (in millions of barrels)18161515151412
Implied Cost Oil (in millions of US$)172150186195204184166
Implied Costs ($/bbl)9.39.212.713.013.413.513.6
Implied Cost Oil / Value of Prod.39.335.032.840.340.740.339.8
Gov. Prof. Oil / Implied Prof. Oil71.568.269.973.073.073.073.0
TRR (in percentage points)47.750.352.250.352.252.447.3
Block 3 & Canuko
Production (in millions of barrels)51464339332925
Implied Cost Oil (in millions of US$)313370365385367348322
Implied Costs ($/bbl)6.28.18.410.011.012.013.0
Implied Cost Oil / Value of Prod.27.128.022.632.034.637.039.3
Gov. Prof. Oil / Implied Prof. Oil86.982.085.285.085.085.085.0
TRR (in percentage points)61.859.665.157.155.054.252.2
Block 14
Production (in millions of barrels)24221820506386
Implied Cost Oil (in millions of US$)3324013463869611,2151,651
Implied Costs ($/bbl)13.717.919.519.219.119.219.3
Implied Cost Oil / Value of Prod.55.967.660.771.569.768.867.8
Gov. Prof. Oil / Implied Prof. Oil29.230.535.335.035.040.040.0
TRR (in percentage points)27.220.225.228.828.829.729.7
Block 15
Production (in millions of barrels)33636113197231
Implied Cost Oil (in millions of US$)346332,4862,8543,2403,054
Implied Costs ($/bbl)9.817.422.014.514.012.0
Implied Cost Oil / Value of Prod.57.948.673.647.645.137.9
Gov. Prof. Oil / Implied Prof. Oil0.031.221.021.021.040.0
TRR (in percentage points)21.132.129.229.233.339.8
Block 17
Production (in millions of barrels)7179808078119161
Implied Cost Oil (in millions of US$)1,0171,3931,5651,8052,3803,3733,543
Implied Costs ($/bbl)14.417.719.423.020.021.022.0
Implied Cost Oil / Value of Prod.59.761.354.076.565.267.269.1
Gov. Prof. Oil / Implied Prof. Oil24.921.920.920.030.035.040.0
TRR (in percentage points)24.222.726.926.127.928.829.7
Block 18
Production (in millions of barrels)7373
Implied Cost Oil (in millions of US$)1,4561,463
Implied Costs ($/bbl)20.020.0
Implied Cost Oil / Value of Prod.63.762.5
Gov. Prof. Oil / Implied Prof. Oil20.020.0
TRR (in percentage points)26.126.1
Sources: DNI; IMF staff estimates and projections.

Estimates based on data for the period January - September 2004 published on the web site of the Ministry of Finance; http://www.minfin.gv.ao/dni/petroleo/exportacao_2004.pdf.

Prices are used to compute cost oil / value of production and the TRR.

Sources: DNI; IMF staff estimates and projections.

Estimates based on data for the period January - September 2004 published on the web site of the Ministry of Finance; http://www.minfin.gv.ao/dni/petroleo/exportacao_2004.pdf.

Prices are used to compute cost oil / value of production and the TRR.

Box I.3.Angola: Estimation of Past Costs

The formulas used for the estimates are based on the formulas described in Section B. In particular:

1) For Tax and Royalty regime; First of all, taxable income was estimated. By inverting the formula for the petroleum income tax (PIT), we have that:

This allows operating costs and amortization to be derived by identity:

Inverting the formula for the petroleum transaction tax (PTT), we have that:

Finally, recalling that production premiums in Cabinda A equal the value of production computed at a reference price, it is possible to compute both production and investments premiums separately.

2) For PSA. Recalling that for companies operating under PSA the PIT is levied on profit oil with a rate of 50 percent, information about income tax liabilities can be used to compute the level of profit oil accruing to companies. Specifically:

Total profit oil can then be computed from the identity:

In this formula, the government’s share of profit oil must be the amount accruing to the government before the 10 percent retention by Sonangol. Cost oil is then:

26. From these estimates can be derived ratios for the “total cost per barrel”, “production and investments premiums relative to oil receipts”, and “cost oil relative to total oil receipts”. Moreover, for blocks 14, 15 and 17, estimates can be constructed of the internal rate of return.

27. In using these estimates to derive projections of costs for the period 2005 - 2008 the following assumptions were made:

(i) Because the blocks in the Cabinda Province and blocks FS-FST are mature blocks, costs were projected so as to maintain cost per barrel and total costs in line as much as possible with past figures. In this respect, in Cabinda A, because of the development of the Banzala field, costs are projected to first increase and then stabilize to levels estimated for 2002 - 2004. Similarly, costs are exceptionally high in 2005 in Cabinda C, because of the development of a condensate gas utilization scheme and of the Bomboco oil project.

(ii) Block 2 is a mature block: considerations similar to those in point (i) above apply for this block.

(iii) Block 3 Canuku is a relatively recent block. Investment costs are expected to have an impact in the near future.

(iv) Production in Block 14 and in Block 17 is expected to increase significantly in 2007. For this reason costs are projected to increase from 2006 to 2007 as a reflection of investments incurred before and during 2006.

(iv) Block 15 is a new block. Projected amortization for the period 2003 - 2007 implies an initial investment costs (before the uplift) of about US$8 billions. As a consequence, cost per barrel is projected to decrease.

28. The detailed projections of costs and the TRR in Table I.6 and I.7 are based on illustrative forecasts of production levels by block. Prices are set at their expected long term levels. Projections were also considered under different price scenarios.

29. The results are summarized in Table I.8 below. Scenario 1 assumes that oil prices will fall to their expected long-term level in real terms in 2005 and that they stay at that level.6 Scenario 3 is based on the interim WEO projections, and Scenario 2 incorporates an intermediate projection of oil prices.

Table I.8.Angola: Projections of TRR, 2005-2008
2004 1/2005200620072008
Scenario 1
Prices (US$ / barrel)36.230.230.831.432.0
TRR (percent)45.738363739 2/
Scenario 2
Prices (US$ / barrel)36.235.034.033.032.0
TRR (percent)45.740373938 2/
Scenario 3
Prices (US$ / barrel)36.239.236.734.733.7
TRR (percent)45.741393939

These figures are based on the stylized assumption that the outturns for the fourth quarter of 2004 are identical to the third quarter data. Data are published on the web site of the Ministry of Finance; http://www.minfin.gv.ao/dni/petroleo/exportacao_2004.pdf.

The TRRs differ as a consequence of different past profiles of oil prices.

These figures are based on the stylized assumption that the outturns for the fourth quarter of 2004 are identical to the third quarter data. Data are published on the web site of the Ministry of Finance; http://www.minfin.gv.ao/dni/petroleo/exportacao_2004.pdf.

The TRRs differ as a consequence of different past profiles of oil prices.

30. Table I.8 shows that, under all of these price assumptions, the decline in the TRR (partially arrested in 2004 by the rise in the price of oil) will resume in 2005 and 2006, but may then level out or reverse. Because of the possibility of rolling-over unrecovered costs, the profile of the TRR will depends on the entire sequence of prices.

31. Three factors explain the pattern displayed in Table I.8: the composition of production, the amortization of costs, and the level of oil prices. Large changes in the composition of production, which will continue to decrease the aggregate TRR, will diminish around 2007 as the growth in production from blocks 14, 15, and 17 begins to stabilize. At the same time, large investment costs in these newer blocks will have been almost fully amortized, leading to TRRs which are nearer (although still well below) those in the mature fields. These two factors will raise the overall TRR in 2007 after the declines of 2004-06. At this point, however, prices will play a role. In scenarios 2 and 3 the price is assumed to fall after 2007, and the TRR does not increase. Only in scenario 1, where there is an increase in price, is the overall TRR expected to increase.

32. With the same methodology, it is possible to project overall government revenue from oil under alternative oil price assumptions. Table I.9 shows projections for the government’s oil revenues for the period 2004 – 2008 for two different price scenarios: at the oil prices projected in the interim WEO and at the long-term oil price.

Table I.9.Angola: Projections of Government Revenue, 2005 - 2008
2004 1/2005200620072008
Average Price of Angolan Crude (US$ / barrel)36.239.236.734.733.7
Value of oil production (US$ millions)12,01816,76321,50826,03426,352
Total government oil revenue (US$ millions)5,4956,8738,38810,15310,277
Implied goverment share (in percent)45.741393939
Average Price of Angolan Crude (US$ / barrel)29.430.230.831.432
Value of oil production (US$ millions)9,75712,87617,87523,31224,677
Total government oil revenue (US$ millions)4,0144,9126,4228,6359,542
Implied goverment share (in percent)41.138363739

These figures are based on the stylized assumption that the outturns for the fourth quarter of 2004 are identical to the third quarter data. Data are published on the web site of the Ministry of Finance; http://www.minfin.gv.ao/dni/petroleo/exportacao_2004.pdf.

These figures are based on the stylized assumption that the outturns for the fourth quarter of 2004 are identical to the third quarter data. Data are published on the web site of the Ministry of Finance; http://www.minfin.gv.ao/dni/petroleo/exportacao_2004.pdf.

33. This same methodology can be used to compute the sensitivity of revenues and TRR to changes in prices or quantities. For example, if the projected average price of oil in 2005 is US$38.2 instead of US$39.2, the realized average TRR would be 40.9 percent instead of 41.4 percent. This implies that tax revenue for the government would decrease by US$60 for every US$100 decrease in total oil sales. Similarly, if realized production in 2005 was 1 percent higher than expected (in all blocks), the government take would increase by US$41 for every US$100 increase in total oil sales.

34. Another use of the methodology is to demonstrate graphically (Figure I.2) how the TRR in each of the years from 2005 to 2008 varies with the level of the price of oil. Because the costs that can be rolled over from one year to the next depend on oil receipts and hence on prices (the lower the price the higher the amount rolled over), the position of the line for the TRR in 2007 –say– depends on the prices realized in 2005 and 2006. The assumption made in the figure is that the prices realized are those of the Scenario 2 of Table I.8, which means that:

Figure I.2.Angola: TRR and Prices, 2005 - 2008

  • the curve for TRR in 2006 is conditioned on a price of US$ 30.2 per barrel in 2005;
  • the curve for TRR in 2007 is conditioned on a price of US$ 30.2 and 30.8 in 2005 and 2006 respectively;
  • the curve for TRR in 2008 is conditioned on a price of US$ 30.2, 30.8, and 31.4 in 2005, 2006, and 2007 respectively.

F. Conclusions

35. Angola’s oil sector is expanding at an unprecedented pace. Together with the recent surge in oil prices, its expansion has led to a dramatic increase in government revenues from oil, from US$3.2 billion in 2001 to an estimated US$4.5 billion at an annual rate in the first half of 2004. However, in the same period the government share of oil receipts declined from 52 percent to an estimated 43 percent. Estimates in this section suggest that one major reason for this decline was a shift in production toward newer deep-water fields, where amortization costs are large, although the impact of this development was diminished in 2004 by the effect of rising oil prices.

36. Data on recent production and revenue by block have been used to derive projections for the government’s share in revenue from oil for the period 2005 - 2008 using basic tax formulas and parameters from PSAs. These imply that, under the price sequence considered, the government share of oil revenue will fall until 2006. Depending on oil prices, this share might then start to recover. Nevertheless, given the projected doubling in oil production over the next three years, total government revenues from oil are expected to rise strongly, under the most likely scenarios for oil prices.

References

    Alvesson, Magnus, ShishirBhattarai, and GonzaloPastor, July2003, “Angola: Sources and Uses of State Oil Revenues”, Angola: Selected Issues and Statistical Appendix (International Monetary Fund).

    • Search Google Scholar
    • Export Citation

    Energy Information Administration, January2005, Angola, Country Analysis Brief.

    Energy Information Administration, 2005, International Energy Outlook 2004.

    Hodges, Tony, 2001, Angola: Anatomy of an Oil State (Indiana University Press).

    Ministry of Finance, January2005, Dados de Receitas de Petróleo Jan-Dez 2004 (Previsão vis-a-vis Execução (Oil Receipts Data Jan-Dec 2004 – Expected vs Realized), http://www.minfin.gv.ao/economia/exppetromenu.htm.

    • Search Google Scholar
    • Export Citation

    Ministry of Finance, January2005, Exportaçóes de Petróleo Bruto e Valores dos Impostos e Taxas 2003 (Crude Oil Export and Value of Revenues and Taxes 2003), http://www.minfin.gv.ao/economia/exppetromenu.htm.

    • Search Google Scholar
    • Export Citation

    The Economist Intelligence Unit, 2004,Angola, Country Report.

    Upstream, 17September2004, Angola Focus.

1

Prepared by Paolo Dudine.

2

For a broad account of oil taxation in Angola, together with a detailed description of the role and activities of the national oil company Sonangol, see “Sources and Uses of State Oil Revenue” in Angola: Selected Issues and Statistical Appendix, IMF (2003).

3

Initial outturns from an exercise using the oil diagnostic revenue model with provisional cost data indicate growing over-predictions of government revenue over time. The government has indicated that it will support further work to update the model and generate projections.

4

Recent reports from oil companies suggests that actual reserves could be much higher than this figure for proven reserves which comes from Angola Country Analysis Brief, EIA (2005).

5

International oil companies make payments on a monthly base in respect of production, income, and transaction taxes, generally with a delay of a month, directly to the National Bank of Angola. The Ministry of Finance, through its Tax Division (DNI) keeps a record of the taxes paid by the international companies, together with assessments made by Sonangol of its own liabilities for tax and of the remittances it is due to make to the government for the government’s share in total profit oil. These data are shown on the Ministry of Finance web site by block, company, and type of tax for 2003 and January-September 2004. Data for 2002 were derived from internal documents.

6

The long-term oil price is derived from the interim WEO assumption for reference crude oil prices in 2010, adjusted for the average discount of Angolan oil and the expected inflation rate in the USA. This implies an Angola oil price of about US$30 per barrel in today’s prices.

Other Resources Citing This Publication