Resource funds have been funded from a variety of sources. In some cases, large trade surpluses have been added to the foreign-exchange reserves of some countries, so as to substantially exceed the normal needs for foreign-exchange stabilisation. In other cases, government pension systems have accumulated substantial reserve funds. The currently most typical, however, are resource funds which are funded by revenues from harvesting non-renewable natural resources, such as oil and gas.
However, these funds are governed by fiscal rules which are often motivated by the permanent-income theory in macroeconomics (Hall 1978). Its great weakness is that it ignores risk. Simple extensions to the risky case can be found in the classical analyses of an individual’s optimal spending and portfolio allocation. Nevertheless, the case of a natural resource fund must take into account the fiscal need for smooth tax rates and government services, which is referred to as backward smoothing. Thus, a resource fund investing should be analysed in the broader framework of asset-liability management (Choudhry (2007). As yet another complication, the substantial movements in risk-free interest rates in recent decades raise the question of how such movements should influence the rules for drawing from a resource fund.
These considerations contrast starkly with the challenge of sovereign debt management for countries with net debt positions. Although a net asset position may seem obviously preferable, the simultaneous tasks of investing the fund and using it as a source of fiscal revenue are far from trivial.
Once a resource fund has been established, policymakers have to decide on at least three important issues: first, how much risk to assume in the asset portfolio, and second, how much to draw from the fund to support current spending. A third decision will be equally important, namely, how to distribute the draws from the fund over time. The goal must be to address these issues and how they fit into the overall fiscal policy framework. The issues are related. Although risk premia may motivate high risk-taking, the observations suggest that policymakers have a low tolerance for non-stochastic, or planned, variation over time in tax rates and public services as well as a strong desire to preserve value for future generations. This combination of attitudes cannot be reconciled within the standard expected utility framework (Lindset and Mork 2018).
A side effect of this smoothing of public spending and taxes is that the portfolio risk for the long run increases. Increased portfolio risk spills over into public spending and can increase the long-run spending risk considerably. As the owner of a resource fund, the government will want to use it to enhance government services and/or keep a lid on taxes. Barro (1979) and others have presented convincing arguments that both the tax system and the stream of government services ought to be smooth. This smoothness should work backwards as well as forward. That is, policymakers should not only plan for smoothness in future services and tax rates, but they should also avoid sudden changes from past patterns in response to unexpected shocks. In practice, policymakers often have only limited leeway when it comes to changing government services from one period to the next.
Forward smoothing is ensured by a low value of the elasticity of intertemporal substitution. Backward smoothing is provided to some extent by risk aversion because low risk-taking limits the effects of negative random shocks. However, in the fiscal model, the degree of (relative) risk aversion is independent of the level of consumption and wealth. A more natural assumption would be that this aversion becomes stronger the more strained the government’s finances are compared to recent experience. This assumption can be approximated by introducing habit formation in the style of fiscal governance. Naturally, it does not mean that policy decisions are governed by habits in a literal sense, but that models of habit formation offer a suitable technique for modelling variations in risk aversion and hence backward smoothing.
If the government wants to maintain a smooth flow of taxes and government services, the rules for resource fund portfolio rebalancing after asset price changes should be formulated so as to safeguard the funds needed to secure this smoothness. As a response to price changes in the risky part of the asset portfolio, the portfolio has to be rebalanced to obtain the optimal portfolio weights.
Using the proceeds of a resource fund as a regular supplement to other government revenues makes sense in an emerging economy with a sizeable resource fund. However, this practice offers greater complications than it may seem at first. A rule permitting annual draws corresponding to the expected real return is optimal only under highly restrictive conditions.
A fiscal rule derived from the expected real return is optimal if policymakers’ elasticity of intertemporal substitution is small. However, unless the risk is negligible or risk aversion extremely low, the annual draws should be considerably lower than the expected real return as an allowance for risk. Furthermore, in the presence of risk, such a rule makes fiscal policy bumpy as either tax rates or government services (or both) must move in response to the vagaries of financial markets. This problem can be mitigated by introducing the responses to market returns gradually over time, which is referred to as backward smoothing and the model as habit formation. However, backward smoothing requires lower risk-taking, which in turn implies lower return on average. Thus, smoothing can be bought at the expense of lower average resource fund draws. Moreover, risk-taking must not only be lower but should move countercyclically relative to the financial markets. Rebalancing should be modified as well so that, in some instances, the resource fund should sell rather than buy equity after a stock market decline.
Ignoring this implication may lead to premature depletion of the fund. As a final implication, backward smoothing, although dampening short-term uncertainty about resource fund draws, will typically increase long-term uncertainty about future draws. These insights make it critical of the fiscal rule for emerging economies. By allowing draws corresponding to the expected real return it ignores risk. By permitting temporary deviations from the real-return rule, it does allow backward smoothing; but it ignores its implications for risk-taking and rebalancing. When this rate declines, the resource fund draws should decline as well; however, in this case, smoothing does indeed make sense.