Chapter 5: Issues in Medium-Term Management of Oil Resources
- International Monetary Fund
- Published Date:
- March 2007
Under the Third Five-Year Development Plan (2000/01–2004/05), Iran introduced a number of important fiscal reforms aimed at reducing the dependency of public finances on oil revenue and containing expenditure growth. These included a tax reform to strengthen the non-hydrocarbon revenue base, the establishment of the Oil Stabilization Fund (Appendix 3) to cushion the impact of fluctuations in oil prices on expenditure, and actions to improve expenditure management and transparency.
Despite these efforts, fiscal reform in general has been limited, and measures have often been implemented under pressing domestic and external political circumstances. Moreover, increases in oil prices have provided temporary respites from underlying long-term fiscal problems and have given policymakers more leeway to increase expenditure. As a result, Iran’s public finances continue to face a number of key challenges: heavy dependence on hydrocarbon revenue, low non-hydrocarbon revenue, procyclical fiscal policy, and excessive subsidization (Appendix 4). These issues need to be addressed in a medium-term framework, which would also help smooth the macroeconomic impact of fluctuations in oil prices, reduce inflation, and preserve hydrocarbon wealth for future generations.
This chapter outlines long-term objectives for preserving hydrocarbon wealth for future generations and discusses a medium-term fiscal outlook that is consistent with these objectives and with the need to bring down inflation and reduce vulnerability to a possible decline in oil prices. The discussion is focused on central government operations consolidated with the OSF, mainly because data limitations preclude an analysis of general government operations, including municipal budgets and the Social Security Organization.1 This chapter does not address the issues of quasi-fiscal activities in the banking system (Chapter 2), fiscal dominance over monetary policy (Chapter 3), or the relationship between public enterprises and central government financial operations.
The next section of the chapter derives three sets of long-term simulations of consumption from hydrocarbon wealth. Then the results of these simulations are compared with a baseline medium-term fiscal projection in order to assess whether a transition from the projected baseline scenario toward a fiscal position suggested by the simulations is feasible.2 The chapter concludes with recommendations for incorporating long-term considerations into the design of the medium-term fiscal framework, with an emphasis on maintaining real per capita hydrocarbon wealth constant in the medium term.
Analytical Framework for Long-Term Analysis
In oil-producing economies, fiscal policy should aim to accumulate substantial net assets during the period of oil production to cover non-oil deficits in the post-oil period (Barnett and Ossowski, 2003). This section draws on an intertemporal optimization framework3 to estimate optimal government consumption and savings from hydrocarbon wealth.4 In this context, intergenerational equity considerations are given prominence, while fiscal sustainability issues are not explicitly examined, assuming that the government’s intertemporal budget constraint is always met.5 The theory stipulates that the optimal consumption path is a function of the net present value of oil revenue, the initial net debt of the government, the real rate of return on assets, an intertemporal discount factor, the rate of population growth, and the degree of risk aversion (Box 4).
Box 4.Optimal Consumption from Hydrocarbon Wealth
The purpose of this model is to determine an optimal rule on how to distribute hydrocarbon wealth across generations. The optimal solution to the government’s consumption level is derived from equation (1):
U is welfare function;
β is a discount factor of the social planner in the welfare function; β = 1/(1 – γ) where γ is a discount rate; and
CG,t is government consumption at time t.
Equation (2) defines net wealth WG,0 as the starting net wealth FG,0 and the present discounted value of future oil revenues YG, s:
r is a real rate of return on assets, and
YG,s is oil revenue in period s.
Equations (3) and (4) define the optimal path of government current spending from hydrocarbon wealth:
1/ρ denotes the coefficient of intertemporal substitution of consumption between two periods or the coefficient of relative risk aversion in a stochastic framework; and
n is the population growth rate.
(a) β(1 + r) > 1. If the welfare of future generations is discounted at 1 percent (implying β = 0.99), consumption grows at CG, t+1 / CG,t = [β(1 + r)]1/p. It is straightforward that about 3 percent growth of consumption is consistent with ρ = 1.5 if β = 0.99 and R = 0.04.
(b) β(1 + r) = 1. Because the discount rate is equal to the real rate of return, the right-hand side of equation (3) is simplified as:
(c) If β(1 + r) < 1, there will be a decline in consumption growth no matter what ρ might be.
The paper presents three sets of simulations for consumption from hydrocarbon wealth. The first simulation assumes that the discount rate in the welfare function is below the real rate of return. This is consistent with positive long-term growth of per capita output and consumption and implies that the society is patient and saves enough to ensure real per capita growth in consumption from hydrocarbon wealth. The second simulation assumes that the discount rate is equal to the real rate of return. Under this assumption, there is no long-term real per capita growth and, therefore, real per capita consumption from hydrocarbon wealth will remain constant indefinitely, which is akin to the conclusions of the permanent income hypothesis. In the third simulation, a preservation of real hydrocarbon wealth is targeted. This is equivalent to assuming that the discount rate is higher than the real rate of return, given the positive rate of population growth.
The issue of whether savings are invested in financial or real (physical) assets does not affect the general conclusions of the optimal consumption theory. Public investment in infrastructure and human capital financed by savings from oil revenue can contribute to an increase in the long-term growth rate of the non-oil sector. Fiscal sustainability, however, requires that public investment be sufficiently productive to generate tax revenue higher than or equal to the prevailing return on financial assets of the equivalent amount.6 Assuming that this rule is observed, the following analysis does not make a distinction between investment in financial and physical assets.
As a share of oil revenue is saved and invested, the return on these investments becomes an important source of additional non-oil revenue for the budget. Accordingly, the government’s consumption from hydrocarbon wealth can be measured by the non-oil current deficit, including implicit energy subsidies as current expenditure (Table 16), minus net interest income.7
|Domestic sales prices in rials|
|Fuel oil (rial/liter)||88|
|Gas oil (rial/liter)||160|
|Natural gas (rial/m3)||71|
|Border prices in rials (at market exchange rates)|
|Fuel oil (rial/liter)||676|
|Gas oil (rial/liter)||1,131|
|Natural gas (rial/m3)||410|
|Implicit subsidy (in trillions of rials)||117|
|Implicit subsidy in percent of GDP||10.4|
|Implicit subsidy in billions of dollars||14.2|
|Average exchange rate (rials per U.S. dollar)||8,282.0|
Estimates of the path of optimal consumption from hydrocarbon wealth are subject to significant uncertainties. They are highly sensitive to several factors, including long-term oil and gas prices, the volume of proven reserves, the extraction rate, and the real rate of return. Of these factors, oil and gas prices are the most difficult to predict. Some empirical research suggests that oil prices do not revert to a long-term mean, whereas other research finds only a very slow reversion and high persistence of shocks (Cashin, Liang, and McDermott, 2000; and Engel and Valdés, 2000). Uncertainties regarding the oil revenue outlook may also stem from the discovery of additional oil and gas reserves or the development of alternative energy sources. Thus, the estimated consumption path needs to be frequently revisited, because new information may lead to large variations in estimates.
Implications for Iran of Long-Term Analysis
The estimate of hydrocarbon wealth hinges on a number of key assumptions. The assumed oil price of $22 per barrel is equal to a 10-year average oil price for Iranian crude (1993–2003) in 2003/04 U.S. dollars. Because of the market segmentation for natural gas, measuring border prices for Iranian gas is a major challenge. A conservative assumption of $41 per 1,000 cubic meters was used in this scenario; this figure is somewhat lower than the 10-year average prices in Europe or North America. On the basis of proven reserves, the country’s oil and gas resources are estimated to last for 75 and 78 years, respectively, assuming that extraction of oil grows at 1 percent per year and extraction of gas accelerates in the next decade and then levels off. A real rate of return of 4.0 percent is assumed, broadly in line with long-term real U.S. Treasury bond rates of 2 percent (60 percent weight) and real stock returns of 7 percent (40 percent weight).8 Government domestic and external debt, net of OSF foreign exchange deposits, is estimated at about $10 billion. Given the assumptions in Table 17, the overall oil and gas wealth, net of government debt, is about $861 billion in 2003/04 dollars (Table 18).
|Proven oil reserves (in billions of barrels)||130.7|
|Proven gas reserves (in trillions of m3)||26.7|
|Oil production in 2002/03 (in millions of barrels per day)||3.2|
|Gas production in 2002/03 (in billions of m3)||100.0|
|Long-term real price for Iranian crude oil (at constant 2003/04 dollars per barrel)||22.0|
|Long-term real price of Iranian gas (at constant 2003/04 dollars per m3)||41.0|
|Long-term annual (real) return on capital (percentage)||4.0|
|Expected average annual rate of population growth (percentage)||1.1|
|Government external net debt (in billions of dollars) at end-2003/04||7.3|
|Preliminary Projected||2003/04 2004/05|
|Initial debt, net of OSF foreign exchange deposits||10.0||…|
|Estimated consumption from oil wealth if:|
|Oil wealth constant in real terms||34.4||35.1|
|Oil wealth constant in real per capita terms||25.2||25.9|
|Optimal consumption from oil wealth||8.0||8.4|
|Estimated consumption from oil wealth (in percent of GDP) if:|
|Oil wealth constant in real terms||25.2||22.2|
|Oil wealth constant in real per capita terms||18.4||16.4|
|Optimal consumption from oil wealth||5.9||5.3|
|In percent of GDP||21.6||21.4|
|Non-oil primary balance, excluding capital expenditure||15.3||17.2|
Simulation of optimal consumption from oil wealth. According to equation (4) in Box 4, the optimal consumption from hydrocarbon wealth is estimated at $8 billion in 2003/04, compared with the estimated consumption of about $29 billion. Under this framework, the optimal per capita consumption from hydrocarbon wealth should increase every year, which would help maintain the amount of consumption from hydrocarbon wealth roughly constant in terms of GDP at about 5 percent indefinitely (assuming 3 percent real GDP growth in the long run, as in Figure 14).
Figure 14.Optimal Path for Consumption from Oil Wealth, 2003–83
Source: IMF staff estimates and projections.
Simulation of consumption from oil wealth consistent with maintaining constant real per capita wealth. To maintain hydrocarbon wealth constant in real per capita terms, the limit on consumption from hydrocarbon wealth is estimated at $25 billion in 2003/04. Under this framework, such consumption will increase in constant dollar terms every year at the rate of population growth but will decline rapidly as a share of GDP, because the rate of population growth of 1 percent is well below the assumed long-term GDP growth rate of 3 percent (Figure 15). Thus, to maintain the intertemporal budget constraint in the long run, the current primary deficits would need to decline in the future, which would require some additional fiscal measures.
Figure 15.Constant per Capita Real Consumption from Oil Wealth, 2003–83
Source: IMF staff estimates and projections.
Simulation of consumption from oil wealth consistent with maintaining constant real wealth. To maintain hydrocarbon wealth constant in real terms, the real return ($34 billion) on the total wealth could be consumed from 2003/04 onward. The long-term implication of this assumption is that consumption from hydrocarbon wealth both in real per capita terms and relative to GDP would decline steeply, potentially warranting sharp increases in taxation or reductions in expenditure to maintain the intertemporal budget constraint. Another concern associated with this scenario (which is not captured by the framework) is a possible substantial REER appreciation that could result from the large spending out of foreign-currency-denominated oil revenue. If sustained, such appreciation could cause Dutch disease which would hinder the development of the nonoil sector.9
Sensitivity to oil price assumptions. Only under the simulation of the optimal policy scenario can consumption from hydrocarbon wealth be indefinitely maintained constant relative to GDP. The corresponding current primary non-oil deficit, estimated at about 5 percent of GDP, is sensitive to assumptions about oil prices. Should the assumed real oil prices be higher (lower) during the projection period, the long-term optimal consumption from hydrocarbon wealth would increase (decrease) by about ¼ percent of GDP for each dollar in excess of (below) the baseline oil price of $22 per barrel. For instance, if the assumed real oil price is $30 per barrel, the optimal level of consumption from hydrocarbon wealth is 6.8 percent of GDP, or about 2 percent of GDP above the level of consumption consistent with the baseline price of $22 per barrel. The government could indefinitely maintain consumption from hydrocarbon wealth at its 2003/04 level relative to GDP (22 percent) only if the assumed real oil price were about $88 per barrel. In the scenario to preserve real per capita wealth, consumption from oil wealth is estimated to increase (decrease) by $1 billion in 2003/04 for each dollar in excess of (below) the baseline price. With respect to maintaining total real oil wealth, consumption from oil wealth is estimated to increase (decrease) by almost $2 billion in 2003/04 for each dollar in excess of (below) the baseline price.
Consistency of the Baseline Medium-Term Framework with Long-Term Parameters
A stylized baseline medium-term fiscal scenario is developed as a reference for comparison with the three simulations presented above (Figure 16). This scenario is based on a number of assumptions on reforms contained in the draft Fourth Five-Year Development Plan, including fiscal reforms. This scenario is also based on a conservative assumption of a steady decline in oil prices to about $24 per barrel at the end of the plan period, as well as a gradual fiscal adjustment that will be achieved mainly through additional revenue measures, including energy subsidy reform10 and value-added tax (VAT) implementation in 2006/07. A gradual reduction in explicit subsidies is also assumed.
Figure 16.Scenarios for Consumption from Oil Wealth, 2003/04–2009/10
Source: IMF staff estimates and projections.
All three sets of long-term simulations point to the need for a change in the fiscal stance under the baseline scenario. This scenario is somewhat tighter than is needed to preserve hydrocarbon wealth in real terms over the next five years. Maintaining real per capita hydrocarbon wealth constant calls for additional moderate fiscal tightening relative to the baseline scenario (see Figure 15). Shifting the fiscal stance toward the optimal consumption path will necessitate a drastic fiscal tightening relative to the baseline scenario, which is clearly not feasible over the medium term.
Although a transition toward an optimal consumption path would result in maintaining consumption from hydrocarbon wealth constant relative to GDP indefinitely,11 such a transition can only be accomplished in stages over the long run. As a first step, the preservation of real per capita hydrocarbon wealth could be achieved over the medium term; subsequently, consideration could be given to moving closer to maintaining constant consumption from oil wealth relative to GDP.
To maintain oil wealth constant in real per capita terms, consumption from oil wealth under the baseline scenario would have to be reduced by about 4 percentage points of GDP during 2004/05–2005/06, with a smaller adjustment of about 0.5 percentage point of GDP during 2007/08–2009/10. Given the need to reduce external and fiscal vulnerability to a potential decline in oil prices, an up-front fiscal adjustment is preferable, because it is less costly to implement from a position of strength when oil prices are high. Indeed, the current fiscal impulse12 has contributed to a significant decline in the current account surplus despite higher oil prices, making the external position more vulnerable to a possible decline in oil prices in the medium term. However, a fiscal tightening of about 4 percentage points of GDP per year at the beginning of the projection period, relative to the baseline scenario, appears excessive, because it could have an unacceptably high output cost.
This chapter relies on a long-term analytical framework to determine criteria for preserving hydrocarbon wealth for future generations. It presents three possible long-term criteria in comparison with a medium-term baseline fiscal scenario: (1) maintaining total real hydrocarbon wealth constant, (2) preserving real per capita hydrocarbon wealth, and (3) targeting optimal consumption from hydrocarbon wealth. The baseline fiscal scenario is broadly in line with maintaining total real hydrocarbon wealth constant. However, additional fiscal adjustment relative to the baseline is needed to move toward preserving real per capita hydrocarbon wealth or converging with the optimal path for consumption from hydrocarbon wealth. Short- and medium-term factors are likely to play a crucial role in choosing the appropriate medium-term path of fiscal adjustment. Given the current fiscal stance, it is likely that a feasible option would lean toward the preservation of hydrocarbon wealth in real per capita terms rather than to the optimal path. This means that additional fiscal measures might need to be considered beyond the medium term to deal with the projected decline in consumption from hydrocarbon wealth relative to GDP.
According to preliminary studies, the Social Security Orgaization is not expected to face any cash flow problems owing to positive demographic dynamics for the next 10 to 15 years. Beyond this period, deficits are expected to emerge, but the exact amount of unfunded liabilities is not available.
The baseline scenario is provided for illustrative purposes only and does not represent the staff’s current projections. The medium-term scenarios discussed in this chapter do not take into account the changes in oil market conditions and prospects that have occurred since 2004.
Engel and Valdés (2000) provide an overview of the application of optimal consumption models to the analysis of fiscal sustainability in oil-producing countries.
The intertemporal budget constraint requires that hydrocarbon wealth be equal to the present value of future current primary balances. All simulations in this paper assume that any estimated decline in consumption from oil wealth relative to GDP will be matched by either revenue or expenditure measures to maintain the intertemporal budget constraint. The issue of medium-term fiscal sustainability is not addressed in this paper, because it is a complex concept involving the analysis of the composition of government net debt by maturity, currency and instruments, and financing constraints.
This principle is valid regardless of the presence of oil resources.
In this chapter, this definition is used for non-oil primary current deficit.
These weights are chosen for illustrative purposes only. The optimal weights should be determined based on a capital asset pricing model.
Dutch disease describes a situation in which, as the real exchange rate appreciates, the tradable goods sector contracts compared with the nontradables sector because the currency appreciation makes tradable goods less competitive and leads to an increase in imports. The result is a shift of resources away from the production of tradable goods and toward nontradables.
On the basis of broad indications regarding the planned energy price reforms, it is assumed that energy prices will increase by 25 percent per year and that two-thirds of the additional revenue generated by the reform will be spent on social protection and public investment.
Barnett and Ossowski (2003) recommend this criterion as a benchmark of long-term fiscal sustainability in oil-producing countries.
The non-oil fiscal deficit increased from 10.8 percent of GDP in 1999/2000 to 18.7 percent in 2002/03; it declined to 16.4 percent of GDP in 2003/04.