Meld. St. 26 (2016–2017)

The Management of the Government Pension Fund in 2016 — Meld. St. 26 (2016–2017) Report to the Storting (white paper)

To table of content

Part 2
Thematic articles

7 The national wealth and the investments in the GPFG

Financial theory suggests that the composition of financial wealth should be considered in the context of the investor’s liabilities and other wealth. There is a comprehensive research literature on portfolio choices for households, and the Mork Commission examined how such research can be applied to Norway as a financial investor. This thematic article outlines how the investments in the GPFG can be considered in such a framework. In this report, this perspective is also discussed in section 3.1 on the equity share.1

Theoretical framework

The financial wealth in the GPFG places Norwegian government finances in a special position internationally. The financial wealth held by central government corresponds to almost three years’ GDP in the mainland economy. Few others countries are in a similar position. This may be one of the reasons why there is little discussion of portfolio choices for countries in financial literature. However, there is an extensive literature studying corresponding issues for households.

Classic contributions to the literature on portfolio choices for households were focused on purely financial assets. Later on, the literature was expanded to also take into account assets that cannot be traded in a market. Another key part of the literature examines the relationship between preferences for stable consumption and the compensation expected for carrying risk. See chapter 6 of the report of the Mork Commission for a detailed review of the literature.

Some contributions also look at other economic agents than households. Merton (1993) seeks to model the preferences of a university endowment. Bodie and Briere (2013) examines sovereign wealth funds, and note that one must, in addition to looking at central government’s aggregate wealth and liabilities, also look at the country’s human capital and natural resources. Bremer (2016) examine portfolio choices for countries that also have petroleum in the ground, but do not consider all aspects of the national wealth.

The literature starts out from the premise that risk can be categorised into risk that is tradable, in the form of securities trading, and risk that is not tradable. Listed equities, which confer a right to the profits of a company after other liabilities have been paid, are an example of tradable, risky income. An individual’s future labour income is also uncertain, but cannot be traded in the same way as a security. Consequently, it is an example of non-tradable, risky income. It is also usual to consider the economic rent from natural resources as the net present value of non-tradable capital income.

A decision as to how much financial risk a household shall take and how it shall compose its portfolio of tradable, financial assets must, according to financial theory, be considered in connection with the quantities and characteristics of other, non-tradable income and assets, as well as what the resources are going to be used for. Such a balancing of the portfolio serves to maximise the welfare of the household, by optimising the ratio between expected risk and return in total wealth.

For a household it is appropriate to start out from an overall asset and liability balance sheet. The asset side will comprise capital and the estimated net present values of all future income (total wealth). Typical assets of a household will be real estate, financial assets such as equities and bank deposits, as well as future labour income. For a young household, future labour income will usually represent the main wealth component. For an older household, its home and financial assets will tend to be the main wealth components. The liability (debt) side of the balance sheet is the net present value of future liabilities, for example payment for the purchase of goods and services. Such net present values can be difficult to calculate in practise and with a high degree of precision. There must over time be assumed a balance between income and expenses. This implies that the net present value of future liabilities, in the form of purchase of goods and services, cannot exceed the net present value of future income.

The quantities and qualities of the assets and liabilities affect the ability to absorb risk. For most households, their labour income will be less risky than their financial income. Early in working life, the value of a household’s future labour income is relatively high. The theory then suggests that the household should hold a large share of equities in its financial portfolio since the large, non-tradable future labour income has a risk profile corresponding to that of a fixed-income portfolio.2 When a household is nearing the retirement age, the value of future labour income is relatively low. The household should then hold more fixed-income securities and less equities in its financial portfolio, such as to keep the risk in its overall assets more or less unchanged.

It can be challenging to change consumption habits from year to year. It is reasonable to assume that most households prefer relatively stable consumption of goods and services. High financial risk taking may conflict with the ability to maintain stable consumption. Stable future consumption may be considered a liability item, or as a negative holding of a low-risk asset, such as fixed-income securities. This should be matched by low-risk income. The more stability is desired, the more will a household need to be compensated in the form of a higher expected return in order to carry risk. Over a longer time horizon there may, at the same time, via economic growth, be a link between expected return and household consumption expectations. A desire to maintain the relative funding contribution from the financial wealth may then work in the opposite direction.

The asset allocation in the financial wealth should, moreover, be adapted to the implicit holding of various assets, such as non-tradable natural resources. All else being equal, more risky non-tradable wealth components will imply that it is optimal to have a less risky asset allocation in the financial wealth. This insight implies that it is not necessarily optimal to keep the asset allocation in the financial wealth fixed over time, as the relative magnitude of the various wealth components may change over time.

Is the investor the Norwegian government or the nation of Norway?

Public sector finances have distinctive characteristics compared to the finances of private agents. Households and businesses have a limited lifespan and income that is largely market-determined. Government is an agent with an infinite lifespan, in principle, and whose most important source of revenue is its right to tax private sector economic activity. Its lifespan means that government may adopt a very long time horizon for its financial investments.

For households and businesses, a financial balance sheet of assets and liabilities can be prepared, which in addition to the wealth components requires an estimate of the net present value of future cash flows. For government, however, it is difficult to estimate the net present value of its key asset, the right to tax households and businesses. Furthermore, it is difficult to estimate future expenditure, since government’s obligations to provide public services are predominantly political rather than contractual. Consequently, public sector revenues and expenditures are not directly comparable to the income and expenses of a household or business, but rather a policy tool for using the real resources of society in a political desirable manner.

Taxes on private sector economic activity is the dominant source of public sector revenues. Tax revenues differ from the non-financial income of households in that these are not determined in a market – in principle, government can itself decide what proportion of economic value to channel into the public coffers via the tax system. At the same time, the structure of the tax system may affect both labour input and productivity in the private sector, and thus also the tax base. Major parts of government revenues will in practice take the form of so-called distortive taxes, which are assumed to have a negative impact on private sector economic activity. This is also referred to as the marginal cost of public funds. The negative effects are assumed to be particularly high when tax rates are high or vary significantly over time. This will in practice limit what proportion of economic value can be taxed.

The Ministry of Finance prepares, on the basis of projections for the Norwegian economy and population, estimates for long-term developments in public revenues and expenditures. The purpose of these estimates is to analyse the sustainability of government finances, i.e. whether current welfare schemes can also in future be funded at the current tax level. The analyses in, inter alia, the 2017 white paper on Long-Term Perspectives on the Norwegian Economy shed light on this issue.

Another key difference between government tax revenues and the income of private parties is that the primary function of taxation is to channel purchasing power away from the private sector in order to manage the allocation of the overall real resources of society in the form of manpower and production equipment. For households, income is, irrespective of whether it comes in the form of wage income, capital income or transfers, the basis for purchasing goods and services. Resource use in society can also be managed through government orders and regulations that directly affect resource allocation. Government can, for example, introduce conscription instead of taxing households to employ professional soldiers. The limitation on government activity, the budget constraint, is not primarily financial, but determined by which real resources are available.

The revenues from the GPFG are foreign currency revenues, which are invested abroad and can only be used to pay for goods and services produced abroad. The alternative would have been to fund such imports through export revenues.3 This would have required man-hours and productive capacity that can instead be used in production for domestic consumption or investment. Hence, the foreign currency revenues from the GPFG do not directly increase the number of man-hours available in the Norwegian economy, but changes the composition of production. A decline in such revenues would reduce consumption opportunities over time, but can be partially countered by increasing domestic labour supply and production.

The Mork Commission noted that the distinctive characteristics of the government balance sheet and the GPFG imply that it is inappropriate to consider the Fund in the context of other public sector revenues for purposes of assessing the investment strategy of the GPFG. Instead, the Commission recommended that the Fund be considered in the context of overall economic activity in Norway.

Norway’s national wealth

The national accounts specify the value and composition of overall economic activity in Norway in a single year. A weakness of the calculations in the national accounts is that these do not reflect whether natural resource revenues are sustainable over time. Petroleum activities mean that oil and gas are converted into revenues that are included in GDP, while the reduction in petroleum resources on the Norwegian continental shelf is not reflected in the national accounts. Overexploitation of renewable natural resources will also be reflected in a temporary revenue increase in the national accounts. Since Norway’s ongoing revenues are influenced by large, but temporary, petroleum revenues, the national wealth provides a better illustration of the factors that will contribute to Norway’s future revenues.

Figure 7.1 Net national wealth. Percent

Figure 7.1 Net national wealth. Percent

Source Ministry of Finance.

The national wealth can be calculated in different ways, but the wealth is generally defined as the net present value of the future consumption opportunities it provides. Such calculations are presented in the 2017 white paper on Long-Term Perspectives on the Norwegian Economy; see Figure 7.1. These comprise calculations for four components included in overall national wealth:4

  • Human capital. The human capital is estimated as the net present value of future labour input.

  • Natural resources. The petroleum wealth estimate is calculated as the net present value of future economic rent in the petroleum sector. As a simplification, other natural resources are disregarded.

  • Fixed assets. The national accounts use the estimated value of fixed assets, calculated at the replacement cost of such assets.

  • Financial wealth. Estimate of Norway’s net financial wealth abroad from the Statistics Norway financial balance sheets. The GPFG accounts for most of the financial wealth.5

The net present value of future labour input is estimated to represent about three fourths of overall national wealth. This implies that long-term welfare developments will primarily be determined by labour input and how much one gets from each man-hour (the productivity of labour). Both the part of the wealth that has been extracted in the form of oil and gas reserves, and invested in the GPFG, and the part remaining on the seabed in the form of petroleum resources, are estimated to represent much smaller portions. Fixed assets, financial wealth and future economic rent in petroleum activities are estimated at 14 percent, 9 percent and 3 percent, respectively, of the national wealth.

Estimates from such calculations are uncertain. The value of the petroleum wealth depends, inter alia, on the future oil price. Moreover, the value of the human capital depends on future productivity growth. If long-term economic growth is significantly weaker than anticipated, the net present value of future production – and thus the net present value of the human capital – will also be less than estimated. If long-term productivity growth is lower, it is likely that the return on the financial wealth, including the GPFG, will also be lower. In addition, the value of the human capital depends on the labour input per capita. If an average worker works longer hours in future, the value of the human capital will also increase. Hence, future generations’ trade-offs between consumption and leisure will be decisive.6

The uncertainty does not only pertain to the magnitude of these components, but also to the relationships between them. There may be a correlation between certain parts of the assets and liabilities, because both of these are influenced by the same underlying causes, for example future productivity growth. Such interrelationships are difficult to estimate and may vary over time.

A national wealth perspective is one potential approach to certain choices…

The theoretical framework for the composition of a decision-maker’s financial wealth involves comparing the net present value of future income to the net present value of future liabilities. The composition of the financial wealth shall then be such as to optimise the ratio between expected risk and return for the overall wealth, considering the future liabilities. For the GPFG, it will in such a framework be appropriate to adopt a national wealth perspective, in which the financial wealth is considered in the context of society’s other revenues.

Such a perspective may shed light on key choices in structuring and investing the financial wealth in the GPFG. The equity share largely determines the level of risk in the GPFG, and is an example of a choice that may be approached in this manner. For government, the economic rent from underground oil and gas reserves is uncertain. The oil price has historically been highly volatile. The conversion of such economic rent into well-diversified financial wealth abroad serves, all else being equal, to reduce the risk associated with the overall petroleum wealth. One may therefore, for a given level of risk, carry more risk in the financial part of the petroleum wealth as the value of the oil remaining in the ground is reduced over time.

Both the Mork Commission and Norges Bank pointed to the significant conversion of wealth that has taken place in recent years, from underground oil and gas reserves to financial wealth abroad, which, when taken in isolation, is an argument in favour of a higher equity share in the GPFG; see section 3.1. At the same time, the framework implies that other considerations must also be taken into account in choosing the level of risk. The guidelines for fiscal policy and the Fund’s explicit objective facilitate a long time horizon and the ability to withstand short-term fluctuations in value. In addition, it needs to be taken into account how large fluctuations in overall consumption the nation can tolerate over time. A revenue shortfall will have a negative impact on consumption opportunities. It is, as noted by the Mork Commission, undesirable to have large fluctuations in the tax level and the standard of public services over time. Whether a revenue shortfall shall result in scaled back public services for households, or higher taxes and lower private consumption, will be a political choice. Such choice also needs to take the costs of tax financing into account.

Such an analytical framework can also shed light on the role of the Fund in the funding of future expenditure. The guidelines for fiscal policy imply that the non-oil deficit in the fiscal budget shall be fully funded through a transfer from the Fund. The objective of this mechanism is to avoid accumulating financial debt while at the same time accumulating savings on the part of government in the GPFG. A national wealth perspective makes it clear that this would not be meaningful on an aggregate basis.

A national wealth perspective implies that future liabilities need to be taken into account in the composition of the financial wealth. Projections indicate that government finances will come under considerable pressure in future years, as the result of an aging population. This is because public benefits are predominantly paid for through taxes on income generated by the population of working age, while children, young people and older people are net recipients of such benefits. The savings in the Government Pension Fund will make it easier to meet such expenditure. At the same time, there will be little scope for meeting the increased expenditure resulting from aging of the population by increasing current savings. Temporary oil and gas revenues cannot fund a permanent and growing gap between public revenues and expenditures.

Both the increase in age-related expenditure and the limited contribution of the Government Pension Fund were known at the time of the introduction of the fiscal policy guidelines in 2001. It was emphasised, at that time, that this challenge must primarily be met through reforms that make government finances more robust in relation to such aging, such as measures to increase labour supply and public sector productivity. Subsequent reports from the Ministry of Finance, as well as the report from the Thøgersen Commission, have also adopted such an approach.

…but of little relevance to other and more detailed choices

In principle, many issues are important for assessing the strategy for the investments in the GPFG from a national wealth perspective. The above framework suggests that it is feasible to adapt the composition of the financial part of the wealth to reflect changes in the value of the non-tradable parts of the national wealth. Trond Døskeland, Associate Professor, Norwegian School of Economics (NHH), highlights, in a memorandum written in connection with the 2017 white paper on Long-Term Perspectives on the Norwegian Economy, the human capital, which largely reflects productivity and future labour input, and asks whether there are investments that would generate relatively higher returns if productivity growth in Norway and the remainder of the OECD were to decline significantly. In a scenario in which prices decline for other important export products like fish and metals, are there any counter-cyclical investments? Another example is the discussion of oil and gas equities in the GPFG; see the discussion in the 2017 white paper on Long-Term Perspectives on the Norwegian Economy. A permanent decline in oil and gas prices will have a negative impact on the Norwegian economy and future revenues. Could and should such vulnerability be reduced by divesting the GPFG’s holding of equities in oil and gas companies?

In principle, one can analyse the risk associated with each of the various national wealth components and the correlation between these, thereby arriving at an asset allocation that is theoretically optimal. In practice, however, it is challenging to perform such analyses because they require precise empirical knowledge about the relationships between various prices and developments in global stock and bond markets. Historical data show that the correlation between the various components is uncertain and varies over time, as also noted by the Mork Commission. Furthermore, any attempt at calculating an optimal allocation will serve to complicate and fragment the strategy for the investments in the GPFG. It will also have to be modified on an ongoing basis, such as when the interrelationships change in nature, for example as the result of time-variable compensation for carrying risk in the financial markets. This would entail considerable transaction costs.

More fundamental objections may also be raised against the use of such a framework. The Mork Commission notes that the tobacco consumption of the Norwegian population reduces the value of the human capital as a national wealth component. From a national wealth perspective, this might be countered by investing in tobacco companies, thereby having a stake in the revenues from the sale of tobacco. This would be contrary to the Storting’s stand that investment in such companies shall be avoided on ethical grounds.

Another objection is that government may in many cases have other, more targeted policy tools for addressing some of these issues than tilting the composition of the investments in the GPFG. An example of this is the vulnerability of the Norwegian economy to a permanent decline in oil and gas revenues, which is discussed in the 2017 white paper on Long-Term Perspectives on the Norwegian Economy. It is noted in that discussion that a good capacity for economic restructuring, supported by economic policy and labour market flexibility, will be much more important in such a situation than whether or not the GPFG has divested its oil and gas equities.

Human capital measured by the net present value of future labour input is the dominant revenue component of national wealth. This implies that lower labour input is a key risk factor for consumption opportunities. Should the composition of the investments in the GPFG be changed if future generations have other leisure preferences than the current generation? It was noted, upon the establishment of the GPFG, that petroleum revenues represent a non-renewable natural resource, and that savings in the Fund are intended to distribute such resource evenly across generations. Consequently, savings in the Fund are motivated by the fair distribution of the petroleum wealth across generations, but not by evening consumption out across the same generations.

The investments in the GPFG reflect the long time horizon of the Fund and the desirability of broad diversification of risk across countries, sectors and companies. There will over time be major changes in industrial structure and production technology worldwide. There is reason to believe that technological developments will continue, but it is nonetheless difficult to predict which sectors and companies will benefit or suffer from such developments. Dimson (2015) notes large variations in long-term returns between sectors. Broad diversification of the investments in the GPFG protects us against loss as the result of concentrated positions in countries or industries with low long-term returns. It is of decisive importance for the ability to carry the long-term risk associated with equity investments. If one sought to make detailed modifications to the composition of financial wealth on the basis of national wealth considerations, it might involve significant costs in the form of a less diversified platform for the long-term returns on the GPFG. The Ministry of Finance does not know of any other, similar fund that makes this type of modifications in its investment strategy.



The thematic article is based on chapter 6 of the report of the Mork Commission (NOU 2016: 20 green paper).


The relationship is less unambiguous if there is a correlation between labour income and long-term equity returns, see e.g. Collin-Dufresne and Goldstein (2006).


In the short run, imports may also be funded by borrowing abroad.


The method used, as well as alternative calculation methods, is outlined on the Ministry of Finance’s website.


Klaus Mohn (2016) finds few indications that increased savings on the part of government is offset by increased borrowing on the part of the private sector. Nor is it reflected in significant private sector debts owing to creditors abroad, as measured here.


For the estimates in Figure 7.1, it is assumed that the number of man-hours in the base year remains unchanged. This is a common approach in preparing such estimates. It is also assumed that the productivity of labour is maintained at the current level. Findings based on other assumptions are found in the documentation memorandum on the Ministry of Finance’s website.

To front page