Chapter

Chapter 1 Managing Public Wealth

Author(s):
International Monetary Fund. Fiscal Affairs Dept.
Published Date:
October 2018
Share
Show Summary Details

Introduction

Public sector balance sheets (PSBSs) provide the most comprehensive view of public wealth, yet they are little understood, poorly measured, and only partly managed. Standard fiscal analysis focuses on flows—revenues, expenditures, and deficits—with assessments of stocks largely limited to gross debt. The focus on debt misses large swaths of government activity and can fall victim to illusory fiscal practices.

Broadening the focus to public wealth sheds light on the assets that governments control, as well as on nondebt liabilities that receive scant attention in standard analysis (Figure 1.1).1 The systematic assessment of PSBSs increases transparency and accountability by examining the entirety of what a state owns and owes, its evolution over time, how it is being managed, and where the risks lie. Most governments do not provide their citizens such transparency, thereby avoiding the additional scrutiny it brings.

Figure 1.1.Public Sector Balance Sheets

(Percent of GDP 2016)

Source: IMF staff estimates.

Note: *Based on a single year of data, in most cases compiled as part of a Fiscal Transparency Evaluation: Albania, 2013; Austria, 2015; Brazil, 2014; Colombia, 2016; The Gambia, 2016; Guatemala, 2014; Kenya, 2013; Peru, 2013; Portugal, 2012; Russia, 2012; Tanzania, 2014; Tunisia, 2013; Turkey, 2013; Uganda, 2015.

Measures of balance sheet strength add information relative to indicators based solely on debt in explaining macroeconomic outcomes. New empirical analysis finds that financial markets consider governments’ asset positions in addition to debt levels in determining borrowing costs, expanding the results in the literature for emerging markets (Hadzi-Vaskov and Ricci 2016; Henao-Arbelaez and Sobrinho 2017) and advanced economies (Gruber and Kamin 2012). Moreover, countries with stronger balance sheets pay lower interest on their debt. Empirical evidence also shows that countries with strong balance sheets experience shallower and shorter recessions compared with those with weaker balance sheets (Box 1.2).

Applying the balance sheet approach to fiscal policy is long overdue (IMF 2015a). During the global financial crises policymakers provided fiscal stimulus and monetary and financial support to cushion the economy. The resulting loss of public wealth since the crisis makes it especially important to take a balance sheet view on public finances today. Population aging in many countries adds to the urgency of taking a long-term view on net worth. And recent experience gathering PSBSs, obtained in part through fiscal transparency evaluations, makes their broader compilation feasible (IMF 2018a).

Taking a balance sheet approach enriches fiscal policy analysis in three key ways:

  • First, it reveals the full scale and nature of public assets and nondebt liabilities. Current benign neglect of public assets suggests that there is considerable scope to boost returns (Box 1.1).

  • Second, it improves the identification and management of risk. Looking at both sides of the balance sheet reveals mismatches. Taking a long-term view through the intertemporal balance sheet allows a comparison of current wealth against future fiscal pressures. And applying fiscal stress tests gauges the resilience of public finances.

  • Third, it can improve fiscal policymaking. The balance sheet approach allows for a more systematic evaluation of the impact of policies on public finances by recognizing their short- and long-term effects on both the asset and liability sides of the ledger. Although the comprehensive approach advocated in this report is new, fiscal policy analysis has often looked beyond deficits and debt. Government balance sheets have been used in fiscal analysis (Buiter 1983; Allen and others 2002; Traa and Carare 2007), although these efforts were hampered by data limitations.2 Individual asset categories have also been analyzed—natural resources in IMF (2012a), nonfinancial assets in Bova and others (2013), and financial asset returns in Seiferling and Tareq (2015)—and stock-flow adjustments are discussed in Jaramillo, Mulas-Granados, and Kimani (2017). Existing approaches to fiscal and debt sustainability, fiscal space, and fiscal risks use some of these insights. The balance sheet approach brings these elements together to provide a comprehensive assessment of the impact of policies on public finances and facilitate risk management across the entire public sector.

Recognizing assets on the balance sheet does not negate the vulnerabilities associated with high public debt. Many assets are illiquid or not marketable and would not be available to meet rollover or deficit financing needs in the short term. Asset valuations are also more volatile than debt and can be highly correlated with the economic cycle—meaning their values can be at their nadir when financing needs are most pressing. Therefore, the assessments of gross debt, deficits, and financing needs remain important for fiscal policy.

The analysis of PSBSs has several limitations. First, data quality can be an issue, especially when looking at the broader public sector. Second, valuation can be a challenge, particularly for nonfinancial assets that are rarely traded, and with differing approaches taken for different components of the balance sheet across countries. These limitations have been addressed to the extent possible in making international comparisons, although some residual issues remain (see Annex 1.2 for details). Third, the public sector consists of many different entities, with each facing its own constraints and risks, often requiring analysis of specific entities. Fourth, the conclusions depend on the robustness of assumptions—issues that gain prominence when projecting over the very long term in an intertemporal balance sheet, as evidenced by sensitivity analyses. More broadly, while the balance sheet enriches the assessment of public finances, it cannot be interpreted in isolation from other factors, such as institutional quality, access to markets, and the monetary and exchange rate regime.

In addition to presenting balance sheet estimates for a wide range of countries, this report provides a conceptual framework for analyzing them. It uses this framework and the data to address three questions, with case studies throughout the chapter to illustrate specific points:

  • 1. What do PSBSs look like, and how have they evolved? The report shows the size and composition of PSBSs across a large range of countries, detailing developments over time. While deficits have been brought under control, net (financial) worth remains significantly below precrisis levels, leaving lower buffers to respond to future risks.

  • 2. How can the PSBS approach improve risk analysis and promote resilience? The report applies a range of measures of risk to the PSBS and highlights the critical nature of balance sheet effects when assessing risks. Fiscal stress tests for Finland, The Gambia, and the United States demonstrate that those balance sheet effects on net worth can be larger than the impact of increased fiscal deficits.

  • 3. How can the PSBS approach strengthen fiscal policy? Several case studies show how the PSBS approach can be used to evaluate fiscal policies, analyzing the effects of demographics, natural resource exploitation, and public investment.

Conceptual Framework

The PSBS brings together all of the accumulated assets and liabilities that the government controls. It extends the scope of fiscal analysis beyond the standard measures of debt to include all assets, whether financial, infrastructure, or natural resources, as well as liabilities that are rarely included in government debt, such as pension obligations to public sector employees. It extends the perimeter of coverage from general government to the entire public sector, bringing in public corporations, including the central bank.

The static balance sheet is then extended through time in two ways (Figure 1.2). First, the evolution of the balance sheet is explained using the integrated stock-flow framework embodied in the Government Finance Statistics Manual 2014 (IMF 2014). This allows the changes in net worth to be decomposed into fiscal deficits, investments, and valuation changes. Second, the balance sheet is used to determine the long-term intertemporal net worth under current policies, combining discounted future flows of revenues and spending with the static balance sheet.

Figure 1.2.The Balance Sheet Framework

Source: IMF staff.

Note: Blue boxes denote incremental additions to the framework.

Composition of the Public Sector Balance Sheet

The PSBS consists of the assets and liabilities of general government and public corporations, including the central bank.3 Liabilities consist of (1) debt securities and loans, (2) pension obligations owed to public sector employees, and (3) currency and deposits, payables (including those in arrears), and some guarantee schemes. Debt securities and loans are the main stock indicator of standard fiscal analysis, worth 95 percent of GDP at the general government level in the sample of 31 countries with full PSBSs. Existing pension obligations to public servants embody a stream of contractually required payments, yet are rarely reported in standard analysis; they amount to 46 percent of GDP in these countries (Figure 1.3). These refer to pension obligations already owed to public sector employees, and do not include pension obligation to private sector employees.4 Government assets comprise financial and nonfinancial assets, including natural resources. Financial assets (99 percent of GDP) are often marketable and relatively liquid, with the exception of direct loans and nonlisted equity holdings in public corporations, which may be less reliably valued.5 Nonfinancial assets include buildings, infrastructure, and land. Many of these assets comprise the public capital stock and play an integral role in delivering economic and social outcomes; they are typically illiquid and nonmarketable, or marketable only over the medium to long term (for example, privatizations). Natural resource reserves can represent the largest asset on the state’s balance sheet in commodity producers. Annex 1.2 provides details on definitions, coverage, and compilation methodology.

Figure 1.3.Additional Elements of the Public Sector Balance Sheet

(Percent of GDP)

Source: IMF staff estimates.

Note: Data labels use International Organization for Standardization (ISO) country codes.

Including the assets, liabilities, and operations of financial and nonfinancial public corporations in the balance sheet shows the full scale of wealth under the government’s control.6 It also allows for a stronger understanding of risk factors across the balance sheet, providing opportunities for better asset and liability management. Including public corporations requires a consolidation of cross holdings of assets and liabilities, which can be a channel through which fiscal risks spread, as demonstrated for The Gambia later in the report. These cross holdings are country-specific, with the largest typically between government and the central bank and other financial public corporations.

Consolidations can be large and have the potential to change the picture. For example, in Japan, while gross outstanding public sector debt securities and loans were worth 283 percent of GDP in 2017, the majority were held by other public sector units, leaving 134 percent of GDP in the hands of private creditors. The same is true in the United States, where the equivalent figures are 164 and 110 percent of GDP.7 These differences are partly the result of quantitative easing, which has led to an unprecedented expansion of the asset holdings of many advanced economy central banks. From the perspective of the consolidated public sector, however, quantitative easing did not lead to a significant expansion of public sector asset holdings, since central banks implemented quantitative easing mainly by purchasing claims on other public sector units.

Assessment of Balance Sheets over Time

The PSBS can explain changes in public wealth—a stock variable—in the recent past or projected near future. The fiscal deficit adds to debt and decreases net worth, although this decrease is offset partly by public investment.8 The operations of public corporations and the net impact of valuation changes on assets and liabilities may either reduce or add to public wealth. Asset valuations are significantly more volatile than liability valuations, so it would be imprudent to react to these changes on a year-to-year basis. However, over the course of decades, ignoring secular trends in valuation misses a large part of the change in public wealth. The importance of the valuation channel for public wealth is illustrated by gains in financial asset values in a sample of European countries, which, since 2000, have added 12 percentage points of GDP to their net worth, offsetting almost a quarter of their cumulative issuance of debt over the same period.9 Net worth developments can also be decomposed for fiscal projections. This helps avoid the fiscal illusion that arises when governments on face value improve the fiscal position by lowering the immediate debt and deficits but reduce net worth over time. For instance, privatizations increase revenue and lower deficits but also reduce the government’s asset holdings. Similarly, cutting back maintenance expenditure reduces the deficit and lowers debt, but also reduces the value of infrastructure assets, which could cost more in the long term.10

Balance sheet analysis can also be used to look into the longer term. A striking aspect of public wealth estimates shown in Figure 1.1 is that one-third of countries in the sample have negative net worth. However, the static balance sheet does not recognize the government’s largest “asset”: its power to raise revenue in the future.11 Intertemporal net worth combines the static net worth with projections of future revenue and expenditure flows. These projections rely on long-term assumptions, with countries having weaker institutions and less stable revenue streams subject to higher discount rates to account for greater risk. The extent to which intertemporal net worth differs from zero thus provides a sense of how far current policies deviate from the government’s intertemporal budget constraint, with negative numbers indicating adjustment needs.

Examination of Balance Sheet Strengths and Risks

Balance sheet strength is not an end in itself, but rather a tool to support the objectives of public policy. The long-term aim of government is not to maximize net worth, but to provide goods and services to its citizens and possibly to create a buffer against uncertainty about the future. Current net worth should be seen in this context. Governments that believe their net worth is too low to ensure their current objectives of public policy may choose to improve their net worth as an operational goal, as Australia has done.

In addition to net (financial) worth, a range of other indicators provide information on the state and resilience of the balance sheet. These include the standard measure of gross debt, as well as measures that explore risk mismatches and the degree of hedging present in the balance sheet. These measures provide a dashboard of indicators to consider when assessing fiscal health.

Fiscal stress tests assess the resilience of public finances to a large macroeconomic shock. They often draw on information from sources outside the balance sheet, such as financial system assessments to inform the size of possible financial sector contingent liabilities, and estimates of sovereign-bank feedback loops or the link between macroeconomic shocks and growth. Stress tests can reveal risks that the standard debt and deficit framework misses. Examples include exposure to entities outside the general government perimeter, valuation changes to government assets, and contingent liabilities emanating from the private sector. Once identified, actions can be taken to mitigate these risks, drawing on the fiscal risk management toolkit (IMF 2016a). Stress tests also provide guidance on the size of the buffers necessary to absorb a large shock, so that procyclical policy can be avoided.

Evolution of Public Wealth

Balance sheets expanded rapidly during the financial crisis, on both the asset and liability sides, accompanied by a sharp decline in net worth, as governments allowed countercyclical fiscal policies to operate. Modest declines in public wealth have continued after the global financial crisis, even as fiscal deficits have been reined in. This section explores central and general government balance sheets for a broader set of 69 countries and territories covering 87 percent of global GDP, and developments over time of the PSBSs for 17 countries comprising 54 percent of global GDP.

The State of Balance Sheets13

Balance sheet size, composition, and net worth vary considerably across the sample of 69 countries and territories (Figure 1.4, panel 1).14 Based on general or central government data excluding natural resource assets and pension liabilities, assets average 102 percent of GDP, ranging from 398 percent of GDP in Norway to 21 percent of GDP in India, roughly evenly split between financial and nonfinancial assets.15 Against these assets stand average liabilities of 70 percent of GDP. As a result, static net worth in the sample varies from –111 percent of GDP in Greece to 348 percent of GDP in Norway, with an average positive net worth of 32 percent of GDP. Net financial worth averages –22 percent of GDP, with Greece and Norway again at the extremes.

Figure 1.4.State of General Government Balance Sheets, 2016

(Percent of GDP)

Source: IMF staff estimates.

Notes: In all panels, the data exclude land and natural resource assets and pension liabilities. Data labels use International Organization for Standardization (ISO) country codes.

1 Central government data.

2 Norway’s total assets are 398 percent of GDP, while its net worth is 348 percent of GDP.

3 Bhutan’s total assets are 272 percent of GDP.

Mismatches in the balance sheet and other risks beyond net worth show a similarly heterogeneous picture. For a subsample of (mainly European) countries, the ample data provide insight into balance sheet riskiness, using measures of liquidity and foreign exchange mismatches, risk-weighted assets and liabilities, and comovement between assets and liabilities.

Liquidity. General government liquid assets average 16 percent of GDP across the sample (Figure 1.4, panel 2), ranging from Moldova (5 percent of GDP) to Japan (62 percent of GDP). Combined with short-term liabilities of 14 percent of GDP on average, countries’ net liquid positions vary from –30 percent of GDP to 21 percent of GDP, with The Gambia, Italy, and Barbados exhibiting the largest mismatches.

Foreign exchange. Many countries borrow in foreign currency and thus have significant foreign exchange liabilities. Against these liabilities, some have significant foreign exchange assets that need to be taken into account when assessing exchange rate risk.16 Net foreign exchange exposure can reveal significant mismatches, showing, for instance, that Barbados, The Gambia, Kenya, Tanzania, and Uganda all have significant foreign exchange debt with little compensating foreign exchange assets (Figure 1.4, panel 3). In contrast with foreign exchange debt data, data on foreign exchange assets are scarce, which limits the analysis.

Risk-adjusted assets and liabilities. This indicator provides a guide to the volatility (and hence inherent risk) of both sides of the balance sheet. All categories of assets and liabilities are weighted by their volatility, with total assets and liabilities adjusted down by their aggregate risk weight to provide a risk-adjusted measure (Figure 1.4, panel 4). Financial assets are more volatile than liabilities for almost all countries in the sample. This is primarily because financial assets include inherently volatile components such as equities and other investment, often held in social security funds, whereas many liabilities are government debt securities that are repaid at maturity.17 Thus, a country like Norway, with high investments in financial markets through its sovereign wealth fund, features a high average risk weight on its assets and hence a relatively large difference between total assets and risk-adjusted assets, while the risk adjustment for liabilities is small. The combination of high exposure to volatile assets and relatively stable liabilities can result in rapid changes in net worth and liquidity.

Natural hedge. Many countries in the sample see significant comovement between the valuation changes of assets and liabilities. In many cases, these comovements dampen the valuation changes of net financial worth, providing a natural hedge in the balance sheet. In some countries, valuation changes in assets and liabilities reinforce each other, amplifying the impact on net financial worth.

The Evolution of Balance Sheets over Time

During the global financial crisis policymakers provided fiscal stimulus and monetary and financial support. While these policy actions reduced static public sector net worth, they contained the propagation from bank and financial market failures, thereby supporting prices, economic activity, and employment. By doing so they protected the future tax base, preserving intertemporal public wealth.

Public sector balance sheets expanded during the global financial crisis, while net worth declined sharply. In the 17 countries for which full PSBS time series data have been compiled, liabilities increased by about 39 percentage points of GDP between 2007 and 2016. However, a concomitant expansion of public sector assets occurred, with assets increasing by 22 percentage points of GDP during 2007–09 in the immediate wake of the crisis, partly because of financial sector interventions; in subsequent years, assets retreated slightly to remain 14 percentage points of GDP above precrisis levels. Both sides of the PSBS remain significantly larger than they were precrisis (Figure 1.5, panel 1).

Figure 1.5.Public Sector Balance Sheets, 2000–16

(Weighted average of 17 countries, percent of GDP)

Source: IMF staff estimates.

Note: The data exclude land and natural resource assets and pension liabilities.

Net worth remains well below precrisis levels, even as fiscal deficits have been reined in. Overall, public sector net financial worth deteriorated by US$11 trillion or 28 percentage points of GDP during the postcrisis decade, with a modest decline continuing even in the later years (Figure 1.5, panel 2). Net worth declined by a similar, although slightly lower, 25 percentage points of GDP, with the difference attributable to public investment. This average marks a wide dispersion, with net worth declining by as much as 49 percentage points of GDP in the United Kingdom, while increasing by 167 percentage points of GDP in Norway, much of this because of strong valuation gains from its equity holdings. While fiscal deficits in the advanced economies most affected by the crisis have largely been brought back to moderate levels (see the April 2018 Fiscal Monitor), the deterioration in net worth caused by the crisis still needs to be addressed.

The postcrisis deterioration in public wealth was driven by deficits, but balance sheet effects significantly cushioned the decline. For the 17 countries with public sector time series data, a decomposition of postcrisis developments shows the relative roles of debt accumulation, public investment, operations in the public corporation sector, and valuation changes. Among these countries, net worth fell from 42 percent of GDP in 2007 to 17 percent in 2016 (Figure 1.6). Fiscal deficits were the largest factor, contributing 38 percentage points of GDP to the overall decline. Together with the 9 percentage point of GDP denominator effect, net worth dips into negative territory.18 However, some of the deficits were used to invest rather than to consume, raising net worth by some 8 percentage points of GDP. While valuations fell during the crisis, reflecting falling asset prices, they rebounded in subsequent years, adding another 16 percentage points of GDP to net worth. Such large balance sheet effects are common across countries, and emphasize the usefulness of a PSBS approach.

Figure 1.6.Decomposition of Changes in Net Worth, 2007–16

(Weighted average of 17 countries, percent of GDP)

Source: IMF staff estimates.

1 Expressed as percent of 2007 GDP.

Further analyzing the effects of the crisis on balance sheets requires a look at individual countries. This point is illustrated by looking at the evolution of the PSBSs of the United Kingdom and Finland below and the general government balance sheet in China (Box 1.3).

The United Kingdom balance sheet expanded massively during the crisis, with balance sheet effects driving most of the movement in net debt—the main fiscal measure used in the United Kingdom.19 Most of the expansion in the balance sheet was the result of large-scale financial sector rescue operations that resulted in reclassification of the rescued private banks into the public sector and increased (non–central bank) public financial corporation liabilities from 0 in 2007 to 189 percent of GDP in 2008, with similar movements in financial assets (Figure 1.7, panel 1).20 These balance sheet effects drove most of the movements in net debt during the crisis period, as the government borrowed to inject funds into the banks. In the early crisis years when the major financial sector operations occurred, the contribution to net debt from balance sheet effects was comparable to that from the fiscal deficit (Figure 1.7, panel 2). Even in subsequent years, the balance sheet effects contributed significantly to net debt changes, both positively and negatively.21

Finland was also hit hard by the crisis. Yet the channels through which the crisis affected its public wealth differed considerably from the United Kingdom, with valuation changes playing a major role (Figure 1.8). Between 2000 and 2007, static net worth increased steadily from 20 to 59 percent of GDP, as the government reduced debt and experienced large net positive valuation changes, stemming mainly from the equity asset holdings of its partially funded pension schemes. Increasing public pension liabilities partly offset these positive effects. At the start of the crisis, large negative valuation changes of 17 percentage points of GDP occurred in a single year, mainly because of the decrease in the value of the government’s asset holdings. In contrast, the impact of the crisis on debt was felt more slowly, with fiscal deficits decreasing net worth by 17 percentage points of GDP between 2008 and 2016. During this postcrisis period, the recovery of financial markets increased asset valuations once again. This was partly the result of ultra-low interest rates, which, however, increased the discounted value of pension obligations. Overall, these countervailing developments on assets and liabilities provided a natural hedge to the Finnish PSBS, resulting in broadly stable net worth over the postcrisis period (Brede and Henn 2018).

Figure 1.7.The United Kingdom: Balance Sheet Developments, 2000–16

(Percent of GDP)

Sources: Her Majesty’s Treasury 2018a; and IMF staff estimates.

Figure 1.8.Finland: Changes in Net Worth, 2000–16

(Percent of GDP)

Sources: Finnish authorities; and IMF staff estimates.

Note: GG = general government.

Using the Balance Sheet to Identify Fiscal Risks

The evolution of balance sheets highlights the large and long-lasting implications that the materialization of fiscal risks can have on public wealth (IMF 2012b). Case studies illustrate how to assess those risks using fiscal stress tests, with a focus on three specific components of the PSBS that fall outside traditional fiscal analysis: (1) valuation changes in the general government in Finland, (2) financial public corporations in the United States, and (3) nonfinancial public corporations in The Gambia.

Stress Testing Finland’s Balance Sheet

Finland’s PSBS features large financial assets, most of them assets of its partially funded pension schemes, against which stand large pension liabilities. A fiscal stress test examines the resilience of Finland’s public finances against a large but plausible macroeconomic shock that includes considerable falls in asset prices. The shock’s impact on net worth is far greater than the increase in debt. The analysis concludes that the fiscal consolidation currently under way, combined with planned health and social service reforms, will provide sufficient buffers to avoid procyclical consolidation after a shock.

Finland’s PSBS is relatively healthy with a net worth of 30 percent of GDP and a positive intertemporal net worth of 114 percent of GDP (Figure 1.9).22 The latter reflects projected future primary surpluses that result from ongoing fiscal consolidation and implementation of planned health and social service reform (IMF 2017b).23 However, the size and composition of Finland’s balance sheet, which includes pension funds that are heavily invested in equities, leaves the balance sheet exposed to asset price valuation risks. Furthermore, Finland’s economy has been subjected to large macroeconomic shocks in the past, and these have permanently lowered real GDP levels (see the October 2018 World Economic Outlook).

Figure 1.9.Finland: Intertemporal Balance Sheet

(Percent of GDP)

Source: IMF staff estimates.

In light of these sensitivities, a fiscal stress test is applied to determine whether the balance sheet provides sufficient buffers to withstand a large future macroeconomic shock. The stress test applies a macroeconomic shock similar to previous crises, including the Nordic banking and the global financial crisis, although slightly less severe. In this test, real and potential GDP fall by a cumulative 10 percent over two years, remaining permanently lower. At the same time, equity prices and housing prices fall by 40 percent and 15 percent, respectively.24 The stress test thus targets the exposures in Finland’s balance sheet.

In this stress scenario, fiscal deficits increase as revenues decline while expenditures increase owing to the operation of automatic stabilizers. Over the longer term, expenditures remain elevated relative to GDP, as a result of some expenditures (for example, health) that remain constant in nominal terms when GDP decreases. Debt rises about 20 percentage points above the baseline in the first two years after the shock. The deterioration in net worth is significantly larger, falling by 45 percentage points of GDP by the second year, because of the impact of asset prices and increased pension liabilities (driven mainly by interest rate effects, Figure 1.10). The long-term impact of the stress scenario is even larger, with permanently higher fiscal deficits translating into an 85 percentage points of GDP decrease in intertemporal net worth.

Figure 1.10.Finland: Net Worth

(Percent of GDP)

Source: Statistics Finland; and IMF staff estimates.

Comparing the impact of the shock to the intertemporal net worth suggests that Finland’s public finances have sufficient buffers to withstand a large macroeconomic shock while avoiding costly procyclical fiscal consolidation. A valid question is to what extent nonfinancial assets of the state could be used to fund future primary balances.25 If they cannot be used, it may be more prudent to focus on net intertemporal financial worth, and then slightly higher buffers would be advisable.

A Fiscal Stress Test for the United States

Like Finland, the United States PSBS features large financial assets, although most of them are held outside the general government in pension funds and government-sponsored enterprises (GSEs) such as Fannie Mae and Freddie Mac. They would not be incorporated in traditional fiscal analysis but are brought out when looking at the PSBS. This case study presents the size of those assets and subjects the PSBS to a macroeconomic stress scenario. Such a shock results in a loss of net worth of about 26 percent of GDP by 2020, far larger than the direct impact of the fiscal deficits alone.

Public sector net worth in the United States has been falling since the early 1980s. The trend was exacerbated by the global financial crisis, during which a range of risks within the balance sheet materialized (Figure 1.11). Overall, net worth deteriorated to –17 percent of GDP in 2016, with net financial worth standing at –101 percent of GDP. With financial assets of 112 percent of GDP, the financial public corporations sector is large relative to the general government.26 The resilience of public finances in the United States therefore cannot be assessed without considering the wider public sector that includes these public corporations.

Figure 1.11.United States: Public Sector Balance Sheet

(Percent of GDP)

Sources: US Federal Reserve Board of Governors; and IMF staff estimates.

Note: GSE = government-sponsored enterprise; MBS = mortgage-backed securities.

The largest, although not the most volatile, class of financial assets on the PSBS are loans to the private sector. These include 44 percent of GDP in mortgages, mostly held by GSEs.27 They also include federal holdings of student loans (6 percent of GDP), which account for most of the increase in the public sector’s loan portfolio since 2007, and are typically unsecured.

Among financial assets, the portfolio of state and local government pension funds has historically been the largest source of risk, as these funds are exposed to large equity price fluctuations (Figure 1.12). When stock prices decline, the resulting increase in the unfunded portion of pension liabilities is explicitly backed by local governments. Shoag (2013) shows that fluctuations in asset returns have a direct impact on state government spending, with large consequences for local economic activity. Many state and local government pension funds are currently underfunded, with a total shortfall of 8 percent of GDP. In addition, the federal defined benefit pension fund faces a similar shortfall of almost 10 percent of GDP.28 The aggregate shortfall of state and local pension funds masks substantial heterogeneity in funding status across states, ranging from a surplus of 4.3 percent of state GDP in Wisconsin to a gap of 27 percent of GDP in Illinois (Figure 1.13). In most cases, the funding status has deteriorated considerably since 2007, driven by large negative returns during the global financial crisis.

Figure 1.12.United States: State and Local Government Retirement Fund Assets and Liabilities

(Percent of GDP)

Sources: US Federal Reserve Board of Governors; and IMF staff estimates.

Figure 1.13.United States: State and Local Government Retirement Funds

(Percent of GDP)

Source: IMF staff estimates.

Applying a fiscal stress test to the United States PSBS identifies and assesses fiscal vulnerabilities associated with these holdings. The scenario is based on the Federal Reserve’s severely adverse supervisory scenario.29 The scenario involves a severe global recession, combined with a steeper yield curve and a rapid drop in equity and real estate prices. The asset price drop particularly affects the large exposure to financial assets, suggesting the stress scenario is well-suited to analyze these potential risks.

This stress scenario leads to an estimated 26 percent of GDP decrease in United States public sector static net worth by 2020 (Figure 1.14; for details, see Gonguet and others, forthcoming). In the scenario, tax revenue falls sharply, which leads to a rapid accumulation of fiscal deficits, increasing debt by a cumulative 9 percent of GDP within three years. But the decline in net worth from balance sheet effects is even larger, at about 17 percent of GDP. These effects include a 6 percent of GDP drop in the value of the government’s nonfinancial assets, mainly because of the revaluation of publicly owned structures as a result of lower real estate prices.30 An additional 7 percent of federally held student loans would not be paid back (0.3 percent of GDP).31 However, the effects on financial public corporations are larger still. Equity price falls lead to state and local pension liabilities being underfunded by an additional 7 percent of GDP, with substantial differences in how the stress scenario affects individual states because of the risk-taking behavior of their pension funds. The relatively limited losses on the mortgage loan portfolio held by GSEs (0.6 percent of GDP) reflect the fact that the portfolio has shrunk by 11 percent of GDP since the crisis, as well as the cushioning of losses by real estate collateral.32 The composition of assets has also become less risky as the GSEs have disposed of most of their private asset-backed securities—a large source of losses during the crisis. Only some of these balance sheet losses would require immediate additional debt issuance to finance them, whereas others can remain on the balance sheet for a long time.

Figure 1.14.United States: Effects of a Severe Stress Scenario on Static Net Worth, 2020

(Percent of baseline GDP)

Source: IMF staff estimates.

Note: GSE = government-sponsored enterprise.

Assessing Fiscal Risk in The Gambia

While the United States stress test focused on the role of financial public corporations, The Gambia case study looks at nonfinancial public corporations, another important component of the PSBS. It illustrates how macroeconomic stress can propagate through the public corporation sector and eventually to the budget through the realization of contingent liabilities. In addition to the immediate macroeconomic impact, a severe macroeconomic shock would cause cascading problems in public corporations, which would push the financing needs of the government into unsustainable territory.

The Gambia case study underscores that a PSBS can be estimated even in a very constrained data environment. The Gambia’s PSBS shows liabilities exceeding assets by a large margin, with net (financial) worth estimated at –46 (–82) percent of GDP (Table 1.1). Most of the financial assets are nonmarketable and so would not be readily available to meet outstanding obligations. The balance sheet is highly exposed to refinancing, interest rate, and exchange rate risks, because of the large amount of short-dated domestic debt (27 percent of debt falls due within one year) and large (concessional) foreign exchange loans. However, unlike Finland and the United States, there is relatively little exposure to valuation risk, as the public sector holds few tradable securities.

Table 1.1.The Gambia: Public Sector Balance Sheet, 2016

(Percent of GDP)

The Gambia’s public sector balance sheet features large nonfinancial public corporations.

Central GovernmentPublic CorporationsPublic Sector
Total assets47.361.961.0
of which: Nonfinancial assets13.422.435.8
Financial assets33.939.425.2
Total liabilities93.561.9107.2
of which: Debt securities78.42.061.4
Net financial worth-59.5-22.4-82.0
Net worth-46.20.0-46.2
Source: IMF staff estimates.

The stress test examines the impact of a natural disaster—a combined drought and pandemic. It affects both agriculture and tourism, two mainstays of the economy.33 The direct macroeconomic impact of the stress scenario on government and public corporation finances is considerable, with the deficit increasing by 8 percent of GDP. Realizations of contingent liabilities from the public corporation sector (largely utilities with precarious finances) would increase the deficit by an additional 10 percent of GDP. This pushes gross financing needs from an already high 25 percent of GDP to 49 percent of GDP (Figure 1.15). With little absorptive capacity in the domestic market, low offsetting assets, and limited availability of additional foreign financing, the only way the government would be able to meet its financing needs would be through central bank financing, as occurred during a previous shock in 2014.

Figure 1.15.The Gambia: Gross Financing Needs

(Percent of GDP)

Source: IMF staff estimates.

The stress scenario exposes public sector cross holdings as a key shock transmission channel. About 20 percent of GDP in loans is consolidated in public sector accounts. Many of these loans are from one public corporation (often the pension fund) to another (for example, the electricity and telephone companies). If financing needs resulting from the realization of contingent liabilities in the stress scenario are not addressed, they could quickly cause cascading defaults in the public corporation sector. In addition, public corporations owe each other arrears equivalent to 4 percent of GDP (an amount likely to increase during a crisis), further increasing the cascading effects of defaults. By identifying these channels early, the government can plan ahead to identify where in the chain the government and donors can best intervene to avoid losses cascading through public corporations and onto the budget.

Using the Balance Sheet to Evaluate Fiscal Policies

This section evaluates a range of policies through the prism of the PSBS, focusing on the largest public assets: natural resources, the public capital stock, and future revenue. By converting natural resources into financial assets, Kazakhstan mitigated the impact of the 2014 oil price shock. The Indonesia study shows how a tax-financed infrastructure push can have large positive impacts on the public capital stock and net worth. An intertemporal balance sheet analysis for Finland and Norway finds that policy reforms have strengthened Finland’s fiscal position, while the continuation of current policies in Norway would eventually imply a drawdown of its large assets.

Balance Sheet Effects of an Oil Price Shock in Kazakhstan

The balance sheet approach recognizes natural resources as assets. Kazakhstan has converted a portion of its large natural resource assets into a diversified and liquid sovereign wealth fund, the National Fund of the Republic of Kazakhstan (NFRK), which helped cushion the economic impact of macroeconomic and oil price shocks in 2014.

Like many oil exporters, natural resources form the largest asset on Kazakhstan’s balance sheet (Table 1.2). In 2016, these assets were estimated to be worth 219 percent of GDP. Once extracted and sold, the balance sheet approach records the conversion of one asset (resources) into another (cash). This aligns the treatment of natural resource assets with other nonfinancial assets; in other words, the sale of oil is treated in the same way as the sale of a building or public land.34

Table 1.2.Kazakhstan: Public Sector Balance Sheet, 2016

(Percent of GDP)

Kazakhstan’s public sector balance sheet features large natural resources and financial assets.

General GovernmentPublic CorporationsPublic Sector
Total assets348.699.9399.0
of which: Nonfinancial assets263.427.3290.7
Financial assets85.272.6108.3
Total liabilities16.999.967.2
of which: Debt securities11.58.910.5
Net financial worth68.3-27.341.0
Net worth331.70.0331.7
Source: IMF staff estimates.

The ultimate impact of natural resource extraction on net worth is determined by what the government does with its cash receipts. If they are used to fund ongoing expenditures, such as salaries or transfers, public wealth declines. If, on the other hand, revenues are used to purchase alternative (financial or nonfinancial) assets or reduce liabilities, net worth remains unchanged, although the nature of the asset has changed. Kazakhstan is an example of a resource producer that has taken the latter path, converting part of its significant oil assets into financial assets in the NFRK. The fund was worth 46 percent of GDP at the end of 2016, primarily in the form of foreign currency bonds (about 80 percent) and equities (20 percent).

The conversion of natural resource assets into the NFRK has reduced fiscal risks. It has helped diversify Kazakhstan’s assets away from a single, highly volatile resource asset, into a more diversified portfolio of financial assets, which has improved the state’s risk-return position. It has also converted illiquid natural resource assets into highly liquid financial assets, which can be drawn on relatively easily in the event of a crisis.

The NFRK has played a key role as a shock absorber, as the Kazakh economy is subject to frequent large economic shocks. The most recent was a 2014 external shock, where a 60 percent fall in oil prices, combined with an external demand shock from Russia and China, led to a sharp depreciation of the national currency and a slowdown in growth. The fiscal balance deteriorated from a surplus of 5 percent of GDP in 2013 to a deficit of 6 percent of GDP in 2015, and public debt increased (due to both the large deficit and the depreciation).

The overall effects on Kazakhstan’s PSBS were large. First, higher fiscal deficits (in part caused by lower oil revenue) increased liabilities by a cumulative 31 percentage points of GDP between 2013 and 2016 through higher borrowing and an increased drawdown on existing financial assets. Second, the decline in oil prices lowered the valuation of the country’s remaining natural resource assets. Third, natural resource exploitation depleted oil reserves, lowering the value of remaining natural resource assets. Fourth, although increasing the value of foreign debt, the exchange rate depreciation also prompted a significant positive valuation effect (in local currency terms) as a result of the high amount of US-dollar-denominated financial and natural resource assets. These positive currency effects dominated, resulting in an increase in net worth (Figure 1.16). However, the persistence of these effects may differ considerably. Currency valuation effects, for instance, were quite persistent in Kazakhstan, but in general may be short-lived.

Figure 1.16.Kazakhstan: Evolution of Net Worth

(Percent of 2016 GDP)

Source: IMF staff estimates.

Combined with the large buffers in the sovereign wealth fund, these balance sheet effects provided room for the government to undertake countercyclical fiscal policy. Between 2014 and 2017, the government undertook fiscal stimulus of over 10 percent of GDP, largely financed by increased transfers from the NFRK. In addition, the authorities provided support of about 4 percent of GDP to the financial sector in 2017—a contingent liability that materialized as a result of the macroeconomic shock—funded partly from the NFRK.

Assessing the Long-Term Impact of a Public Investment Surge in Indonesia

The public capital stock is another large asset in a country’s PSBS. Public investments in nonfinancial assets have a distinct balance sheet impact that is missed if looking at debt and deficits alone. This is illustrated for Indonesia, where a tax-financed infrastructure investment surge boosts public sector net worth, both immediately and in the longer term.

Indonesia’s PSBS has positive net worth (Figure 1.17). Public sector assets are large, exceeding 160 percent of GDP in 2016, with natural resources accounting for half of nonfinancial assets. On the liability side, currency and deposit obligations are significant, reflecting a large public banking sector, while pension liabilities are relatively small. Static net worth stood at 93 percent of GDP in 2016, despite a steady decline since 2010 or earlier, owing to falling natural resource wealth.

Figure 1.17.Indonesia: Public Sector Balance Sheet, 2010–16

(Percent of GDP)

Source: IMF staff estimates.

Indonesia is considering embarking on an aggressive extension and upgrade of its public infrastructure, to be implemented in the next few years (Shin 2018), financed in part by raising its low tax take. Indonesia’s fixed public capital stock is low compared with its neighbors’, and public investment is insufficient to maintain it. Tax revenues have fallen over the past decade to about 11 percent of GDP in 2017—well below its peers’—as revenue from a shrinking oil and gas sector decreased. In response, the authorities are considering implementing a Medium-Term Revenue Strategy (MTRS)—a comprehensive plan that integrates revenue mobilization and tax policy reform—to raise revenue to finance infrastructure investment (Jin 2018).

The balance sheet approach provides a comprehensive assessment of the investment plan in three ways. First, when the investment occurs, the approach recognizes the creation of an asset. Second, it includes public corporations, which are responsible for about 40 percent of net public investment in Indonesia. Third, by considering the intertemporal aspect, it recognizes the impact on growth and future revenues that can come from an increase in investment (see the October 2014 Fiscal Monitor). While the first and third aspects are standard in macroeconomic models, they are absent from the budget documentation in most countries.

A tax-financed investment surge could boost Indonesia’s intertemporal net worth by some 6½ percent of GDP and raise potential GDP.35 In the scenario, tax revenue increases by an incremental 1 percentage point of GDP per year for three years, reaching 3 percent of GDP above baseline by 2022, broadly in line with the MTRS. The additional tax proceeds finance additional public investment; however, in line with findings for emerging markets worldwide (IMF 2015b), only two-thirds of the public investment surge is converted to physical capital. After three years, both tax revenue and the stock of public nonfinancial assets remain at their higher levels.36 The improvement in the public sector’s financial position is substantial, with static net worth increasing by more than 4 percent of baseline GDP, as a result of the creation of infrastructure assets. The long-term impacts are even larger. Although the additional taxation depresses GDP, this is offset by the impact of a higher public capital stock, resulting in a permanent 1⅓ percent level increase of both potential and real GDP, increasing revenues and the primary balance.37 The public sector’s intertemporal net worth—which combines the static balance sheet with the net present value of future revenue and expenditure flows—improves by 6½ percent of baseline GDP (Figure 1.18).38

Figure 1.18.Indonesia: Intertemporal Net Worth

(Percent of 2023 baseline GDP)

Source: IMF staff estimates.

Raising investment efficiency would increase the benefits even further and potentially improve intertemporal net worth by as much as 10 percent of baseline GDP—highlighting the benefits of strengthened infrastructure investment efficiency. The economywide impacts could be greater still, as positive spillovers to private wealth, outside of the growth impact, are not captured in the PSBS.

Assessment of Long-Term Fiscal Outcomes in Two Nordic Countries

An application of the intertemporal balance sheet approach to Finland and Norway highlights the public sector’s largest assets and liabilities in the form of future revenue and expenditure. Although both countries have strong fiscal positions, a continuation of Norway’s current policies would eventually lead it to eat into its natural resource wealth. Norway’s large assets, however, provide considerable buffers to smooth the adjustment of current policies, which will happen under the current fiscal rule. Finland’s past, ongoing, and planned reforms have already led to a major improvement in intertemporal net worth, illustrating the benefits of modest but sustained reform.

Finland and Norway are Nordic neighbors whose balance sheets feature some key differences. Both countries are wealthy advanced economies with aging populations. Both have manageable levels of debt—Finland at 57 percent of GDP and Norway at 31 percent of GDP—and relatively high pension liabilities (Figure 1.9 and Table 1.3).39 The major difference lies in Norway’s natural resource wealth, comprising its sovereign wealth fund and remaining subsoil natural resources, which together are worth more than 400 percent of GDP.

Figure 1.19.Norway and Finland: Intertemporal Net Worth

(Percent of GDP)

Source: IMF staff estimates.

Table 1.3.Norway: Public Sector Balance Sheet, 2016

(Percent of GDP)

Norway’s public sector balance sheet features large pension liabilities.

General GovernmentPublic CorporationsPublic Sector
Total assets563.5119.6644.9
of which: Nonfinancial assets230.635.6266.2
Financial assets332.983.9378.7
Total liabilities142.7119.6224.1
of which: Debt securities20.710.030.7
Net financial worth190.1-35.6154.6
Net worth420.70.0420.7
Net present value of primary balances-225.9
Intertemporal net worth194.8
Source: IMF staff estimates.

Norway’s intertemporal net worth is lower than this large asset base would suggest. Current policies imply large primary deficits into the future, which, cumulated over the next 50 years, result in an intertemporal net worth of 195 percent of GDP (Figure 1.19).40 While this is a robust number by any standard, it is considerably lower than the country’s static net worth. To look at this another way, if Norway maintains its current policies, its oil wealth would be at least partly consumed by future aging-related expenditures, going against its fiscal rule. However, continued adherence to the fiscal rule would bring about sufficient policy change to prevent this outcome. Specifically, policy adjustment—further reviewing, for instance, costly disability schemes, and instituting systematic public expenditure reviews—would reduce future primary deficits and improve intertemporal net worth to avoid a depletion of the sovereign wealth fund. Norway’s vast wealth implies that any such policy adjustment can be pursued in a very gradual way, smoothing the transition.

In contrast with Norway, Finland’s intertemporal net worth exceeds its static net worth, reflecting a series of reforms that include postcrisis fiscal consolidation and pension reform. Finland also plans to further reform its health and social services sectors. Collectively, these reforms permanently reduce demographic-related expenditures, improving intertemporal net worth to 114 percent of GDP—an example of the impact of modest but sustained reform on long-term public wealth.

Balance Sheet Analysis in Practice

Australia, New Zealand, and the United Kingdom manage public wealth using balance sheets. All three countries produce PSBSs that inform high-level policy and day-to-day fiscal management. First, they use the aggregate data to set overall fiscal policy objectives. Second, they improve asset management to maximize the efficiency of use and returns on public assets, something also done in Uruguay. Third, they identify, analyze, and manage fiscal risks emanating from within the balance sheet as well as from external shocks.

Using Balance Sheets to Guide Fiscal Policy

Both Australia and New Zealand focus on strengthening their balance sheets over time, to improve national saving and provide a buffer against external shocks. Their fiscal policy objectives explicitly include improving net (financial) worth in addition to reducing net debt and achieving or maintaining surpluses. To operationalize this, both countries project their balance sheets forward to demonstrate that policies are consistent with fiscal objectives.41 The balance sheet projections extend between 6 and 10 years and cover all key aggregates: assets, liabilities, and net (financial) worth. The authorities have used these projections to demonstrate the impact of pension reforms, tax changes, and public investment surges.

Both countries also consider the long-term evolution of net (financial) worth. Australia’s 40-year projections estimate the effects of demographic change on health and pension expenditure, based on previous, current, and proposed policies (Figure 1.20). New Zealand estimates intertemporal net worth, finding that despite a strong static net worth of 41 percent of GDP, large projected deficits over the coming 40 years result in intertemporal net worth of –57 percent of GDP, making it clear that adjustment is needed (Table 1.4). Both countries use these findings to motivate policy changes.

Figure 1.20.Australia: Net Financial Worth Projections

(Percent of GDP)

Improving Balance Sheet Management

Both New Zealand and the United Kingdom have strengthened their focus on balance sheet management, while Uruguay has introduced a balance sheet approach to debt management. They aim to improve the use of public assets, make sure they are being used to meet high-priority policy needs, and raise financial rates of return (Box 1.1).

The 2018 New Zealand Investment Statement (New Zealand Treasury 2018) provides an assessment of the use of all public assets, regardless of whether the assets are used for commercial or policy objectives. A common criticism of the balance sheet approach is that many public assets are held for policy reasons (such as schools and hospitals), are not marketable, and should not be included in fiscal analysis or be expected to provide a financial return. Thus, undertaking rigorous balance sheet assessments is of little use. However, because the investment statement presents the balance sheet in terms of use, distinguishing between social, financial, and commercial assets (Table 1.4), the government can set performance benchmarks by the use of the asset. Are social assets being used effectively and efficiently for high-priority purposes? Are financial assets securing a high enough return relative to risk? And are commercial assets generating sufficient shareholder returns? To answer these questions, the investment statement assesses each sector, company, or financial holding against a range of criteria. It finds that social assets are aging and unlisted commercial companies are under-performing. In contrast, listed companies and financial investments have benefited from rising equity markets and have performed well.

Table 1.4.New Zealand: Intertemporal Balance Sheet

(Percent of GDP)

New Zealand classifies its public assets by their use.

AssetsLiabilitiesNet Worth
Social57.57.150.5
Financial33.250.5-17.3
Commercial20.112.08.1
Static balance sheet110.869.541.3
Future flows11,381.91,480.0-98.1
Intertemporal balance sheet1,492.71,549.5-56.8

The United Kingdom authorities are at an early stage in the process of balance sheet management. They recently initiated a balance sheet review, intended to improve balance sheet management and fiscal outcomes by:

  • Improving returns on assets. This could include, for instance, pooling investment fees on various government financial assets.

  • Improving the compensation to government for bearing risk. In several cases, the government acts as an insurer of last resort to the private sector. The balance sheet review is an opportunity to assess whether it is adequately compensated for bearing such risk, and to renegotiate contracts in cases where it is not.

  • Reducing the costs of liabilities. Liabilities take many shapes and sizes, but reducing their costs could entail, for example, reducing building lease costs by better using assets the government already owns.

In the short term, evaluating the United Kingdom’s stock of assets along with its stock of liabilities will facilitate their integrated management. This will support fiscal outcomes and release resources to reinvest in the public sector. The review will also assess balance sheet indicators and evaluate interest rate, credit, foreign exchange, and liquidity risks. In the long term, the review can become the foundation for embedding balance sheet management into ongoing decision making.

Government debt managers in Uruguay are realizing costs savings by taking a public sector balance sheet approach to the management of risks and costs. Debt managers follow a well-defined mandate where they try to minimize expected debt servicing costs and the opportunity cost of holding liquid assets—subject to an acceptable level of risk—over the medium to long term. They do this for the entire public sector, including public corporations and the central bank. The approach has uncovered interest rate, currency, and maturity mismatches between assets and liabilities, and flow mismatches related to ongoing operations of public corporations. In particular, it has identified net foreign currency liability exposure and revealed capital market bottlenecks. In response, the authorities have further developed the domestic debt market and promoted the development of risk management products, which will, over time, improve the debt manager’s ability to match characteristics of public sector financial assets and liabilities.

Fiscal Risk Management

All these countries carefully examine risks within their PSBSs. Australia publishes a qualitative assessment of balance sheet risks (Commonwealth of Australia 2018). New Zealand and the United Kingdom have both performed detailed balance sheet risk assessments, including fiscal stress tests, and have taken active steps to address identified risks.

The New Zealand investment statement examines fiscal risks in a comprehensive way. It analyzes aggregate fiscal risks through fiscal stress tests for a range of scenarios. The stress tests examine the direct fiscal costs on spending, as well as valuation effects, discretionary support, and the costs of replacing asset losses from an earthquake (one of three scenarios). In addition, the stress tests evaluate the impact on the discounted value of future revenues. Although the results of stress tests show the fiscal position is robust, they point to opportunities to mitigate risk, and inform the target level of government debt with sufficient buffers. Last, financial risks are assessed against a range of measures, including a value-at-risk analysis, and find that while losses of 2–4 percent of GDP could occur, the balance sheet is generally robust.

The 2017 Fiscal Risk Report provides a comprehensive scan of risks facing the United Kingdom’s public finances, including macroeconomic, spending, revenue, and balance sheet risks (Office for Budget Responsibility 2017a). The report assesses the entire PSBS, and the fiscal stress test finds that interest rate and inflation risk are among the key exposures. First, the increasing share of inflation-linked debt has increased the exposure to inflation. Second, because of the quantitative easing program of the Bank of England, the average maturity of public sector debt has declined, increasing interest rate risk in the Bank’s balance sheet as well as the public sector. This is also true for other countries in which central banks have undertaken quantitative easing.42 The UK government has acted to mitigate these risks, by changing the debt issuance policy away from inflation-linked bonds (Her Majesty’s Treasury 2018a). It also revised the financial relationship between the Treasury and the Bank of England, so that capital transfers can be made to the Bank in the event of large valuation losses (Her Majesty’s Treasury 2018b).

Conclusion

Analyzing public wealth brings a range of benefits by offering a broader fiscal picture beyond debt and deficits. It provides transparency to markets and accountability to citizens, squarely drawing attention to what the government owns, in addition to what it owes. This matters as governments with stronger balance sheets face lower financing costs and are better placed to weather recessions.

While there are considerable challenges in compiling reliable balance sheets, basic balance sheet estimates can be compiled even in low-capacity countries. Initially, it may require drawing on third-party sources and using assumptions to make informed estimates. Subsequent improvements to accounting and statistical capacity can, over time, provide more reliable valuations and improve consistency. Once governments produce these estimates, basic balance sheet analysis can be done using the framework presented in this report.

Comprehensive balance sheets allow for better informed assessments of fiscal policies and risks, and can raise the tenor of the policy debate. Governments should consider the effect of policies on assets and nondebt liabilities, in addition to their effects on debt. Current levels of public wealth should be compared with long-term fiscal pressures to assess how governments can meet building demographic pressures. Analyzing both sides of the public sector balance sheet is also necessary for effective risk management, where valuation changes, particularly on the asset side, have large impacts on public wealth. Identifying these risks allows governments to take action early, rather than dealing with the consequences after problems occur. Last, balance sheet analysis enriches the policy debate, by increasing transparency and asking how public wealth can be better used to meet society’s economic and social goals.

Box 1.1.Potential Revenue Gains from Better Asset Management

Many governments can improve returns on public sector assets. While recognizing that these assets often have operational objectives, there is still considerable room to improve asset management. Given the scale of public assets and the existing poor quality of asset management, Detter and Fölster (2015) argue that a small increase in yield could provide significant increases in fiscal revenues. Governments should at a minimum expect a reasonable rate of return from the large commercial and financial assets they control. Benchmarking the returns that governments receive from their nonfinancial public corporations and financial asset holdings across countries, this box provides estimates of the potential revenue gains if performance is increased to the 75th percentile of sampled countries.

First, the analysis looks at the return on assets among nonfinancial public corporations across a sample of 14 countries from our PSBS database.1 The country-specific sectorwide return on assets of nonfinancial corporations is defined as the net operating surplus as a share of total assets. For the sample, the average return on assets during 2010–16 was 1.9 percent (median 0.6 percent; see Figure 1.1.1). This compares with the equivalent rate of return of 8 percent for United States private nonfinancial corporations over the same period (Osborne and Retus 2017). Raising the return on asset performance from the 25th to the 75th percentile of the cross-country distribution of returns would bring average yields to 4.3 percent—still well below the comparable private sector rate of return—and increase profits by an average of about 1 percent of GDP.

Figure 1.1.1.Nonfinancial Public Corporations Returns

Source: IMF staff estimates.

Second, the analysis looks at the returns obtained on general government financial assets in a number of European countries.2 It constructs a time series of the returns on these assets by subtracting transactions (sales and acquisitions) from the total change in the value of the assets and adding income accruing from these assets.3 The capital asset pricing model is used to decompose these returns into compensation for risk (βy) and a measure of performance (αy), using the country median as the benchmark index. It estimates the following regression:

where ROAy is the return on assets in country y, and ROAb denotes the cross-country median return on assets.4 Ranking the countries on their performance measures α suggests that revenue gains from an improvement in asset management performance from the 25th to 75th percentile of the cross-country distribution of this measure would generate a further 2 percent of GDP in returns.

This overall revenue gain from improved management of nonfinancial public corporations and government financial assets of 3 percent of GDP per year is equivalent to corporate income tax revenue in advanced economies. Still, it leaves out the potential gains from better management of government nonfinancial assets. For example, Detter and Fölster (2015) argue that governments’ real estate portfolios are heavily underestimated, and that there is considerable scope for both better management and higher returns. However, estimating the potential gains from better management of nonfinancial assets is beyond the scope of this report.

1 Australia, Canada, El Salvador, Finland, France, Georgia, Indonesia, Japan, Kazakhstan, Korea, Lithuania, New Zealand, Norway, and the United Kingdom. Across these countries, public corporations are in different industries and for that reason may have different return on asset profiles. Robustness analyses using return on equity—which captures the difference in capital intensity across industries—and risk-adjusted returns yield much the same results.2 Austria, Belgium, Denmark, Estonia, Finland, France, Germany, Ireland, Italy, Latvia, Lithuania, the Netherlands, Norway, Poland, Portugal, the Slovak Republic, Slovenia, Spain, Sweden, and the United Kingdom.3 The results are robust when looking at valuation changes only, that is, when income from the assets is excluded. They are also robust to excluding the crisis years 2008–09, although the prospective gains in return on assets would be somewhat smaller.4 This approach is similar to that used by Samphantharak and Townsend (2009), who apply the capital asset pricing model framework to household balance sheets.

Box 1.2.Balance Sheet Strength and the Macro Economy1

Public sector balance sheet strength is a measure of the health of public finances. But are governments with stronger balance sheets better able to engage in countercyclical fiscal policy during recessions? If so, would that allow them to shield the broader macro economy better from the impact of recessions? And lastly, do financial markets take account of assets and balance sheet strength?

To answer these questions, this box estimates the response of real per capita government spending and real per capita GDP in the aftermath of recessions. It uses the local projection method developed in Jordà (2005) and Jordà, Schularick, and Taylor (2016) on a sample of 17 advanced economies. To gauge the differential effect of balance sheet strength, the box divides the sample into countries entering a recession with a strong or weak balance sheet, defined as net financial worth above or below the sample median. It distinguishes balance sheet effects from debt effects by including private and public debt as control variables. The analysis builds on the October 2016 Fiscal Monitor, which estimated the impact of private and public debt on the pace of economic recovery after financial crises and regular recessions.

Figure 1.2.1 depicts the conditional cumulative changes in government expenditure and GDP from the start of recessions. It distinguishes countries that entered the recession with a strong initial balance sheet (blue line) from those entering the downturn with a weak balance sheet (red line). The figure suggests that countries entering a slump with a strong balance sheet used the greater flexibility it provided to increase real per capita expenditure to respond to the crisis.2 There is some indication that, as a result, these countries faced shallower recessions and faster returns to growth. The differences in government spending are statistically significant starting in the second year, whereas the differences in economic growth are significant in years 4 and 5.

Figure 1.2.1.Fiscal Policy and Recovery in the Aftermath of Economic Recessions

Source: IMF staff estimates.

Note: Bands represent 90% confidence intervals.

Financial markets also seem to recognize public sector assets. To gauge the link between balance sheet strength and financial markets, this box estimates three separate regressions of sovereign bond yields on (1) debt, (2) debt and assets, and (3) net worth. The results suggest that debt, assets, and net worth all matter for yields (Figure 1.2.2). The significant coefficient estimate on assets indicates that balance sheet strength adds information to an analysis solely based on debt numbers.

Figure 1.2.2.Impact of a 10 Percent of GDP Change on Yields

(Basis points)

Source: IMF staff estimates.

1 Details of the analyses in this box can be found in Annexes 1.3 and 1.4.2 Expenditure also reflects other factors besides countercyclical policy choices, such as changes in interest rates.

Box 1.3.China—Revisiting the General Government’s Balance Sheet

Compiling China’s general government balance sheet is a challenge. The perimeter of general government should include all public entities that are government-controlled and nonmarket producers. But the numerous public entities and complex layers of government make it difficult to delineate the perimeter precisely. Significant government holdings of state-owned enterprises (SOEs) and financial institutions and the widespread subnational off-budget borrowing further blur the classification. This box updates the estimates in the October 2016 Fiscal Monitor, broadens the coverage by extending the time horizon to 2010–17, and covers separately the central and local governments.1

  • Financial assets (75 percent of GDP) consist of government deposits in banks, equity holdings of the national social security fund, and public corporations. Official government equity holdings of nonfinancial SOEs at nominal values were about 56 percent of GDP in 2017, but this is subject to uncertainty, as many SOEs are not listed and their profitability has fallen since 2010 (Figure 1.3.1). A more conservative estimate using the net present value of SOEs’ expected future net profits puts the valuation at about three-quarters of the headline value. In addition, the government share of equity in financial institutions is estimated at 11 percent of GDP in 2017 (Yang, Zhang, and Tan 2017). Deposits include fiscal budget deposits (5 percent of GDP), and deposits held by government organizations (another 15 percent of GDP net of estimated accounts payable).

  • Financial liabilities (67 percent of GDP) include official government debt (37 percent of GDP). The analysis uses the broader “augmented” concept to include off-budget borrowings, which raise the debt in the balance sheet by an additional 30 percent of GDP (IMF 2018b).

Figure 1.3.1.Weak Financial Performance of State-Owned Enterprises

Sources: China Public Finance Statistics Yearbooks; and IMF staff estimates.

The general government’s net financial worth remains positive, at 8 percent of GDP in 2017, although it has deteriorated in recent years.2 Net financial worth has declined, notably at the subnational levels driven mainly by rising local government debt and underperforming SOEs (Figure 1.3.2). This points to rising vulnerabilities from a balance sheet perspective.

Figure 1.3.2.Government Net Financial Worth

(Percent of GDP)

Sources: China Public Finance Statistics Yearbooks; and IMF staff estimates.

These estimates are subject to caveats. First, subnational governments own land resources and invest in infrastructure, which could provide buffers and generate revenue to service their debt. Firm-level data on local government financing vehicles, however, suggest that liabilities of those loss-making ones have risen (Li and Mano, forthcoming) and that returns on new infrastructure have fallen, in some cases below interest costs (Lam and Moreno-Badia, forthcoming). Second, the government’s holdings of SOE equity could be higher than the conservative estimates presented here.

The Chinese authorities are taking steps to compile balance sheets, including accounting reforms and pilot programs for seven provinces and two central ministries. These contribute to the commitment to complete a consolidated accrual-based balance sheet by 2020. At the same time, the authorities have reiterated the ban on off-budget borrowings and committed to raise SOE efficiency. These measures should be complemented with aligning data compilation with the Government Finance Statistics Manual, which will help assess the overall impact of fiscal policy and increase international comparability (Mano and Stokoe 2017).

1 The estimates include the government’s equity holdings of financial institutions, the national security fund, and the financial assets of government departments and organizations. For details, see Lam and Moreno-Badia (forthcoming).2 Estimates are lower than those by Li and Zhang (2017) and Yang, Zhang, and Tan (2017) because they include nonfinancial assets and contingent liabilities of the financial sector.
Annex 1.1. Public Sector Balance Sheet Database Coverage
Annex Table 1.1.1.Public Sector, General Government, and Central Government Coverage
Public Sector (31)General Government (31)Central Government (7)
Albania*BelgiumBarbados
AustraliaBhutanMalawi
Austria*BulgariaMarshall Islands
Brazil*Hong Kong SARMicronesia
CanadaChinaPalau
Colombia*CroatiaSerbia, Republic of
El SalvadorCyprusSolomon Islands
FinlandCzech Republic
FranceDenmark
Gambia*Estonia
GeorgiaGreece
GermanyHungary
Guatemala*Iceland
India*Ireland
IndonesiaItaly
JapanKyrgyz Republic
KazakhstanLatvia
Kenya*Lithuania
KoreaLuxembourg
New ZealandMoldova
NorwayNetherlands
Peru*Poland
Portugal*Romania
Russia*San Marino
South AfricaSlovak Republic
Tanzania*Slovenia
Tunisia*Spain
Turkey*Sweden
Uganda*Switzerland
United KingdomUkraine
United StatesUruguay

Public sector estimates drawn from the Fiscal Transparency Evaluations cover only a single year. PSBS time series are available for 17 countries.

  • Public sector and general government data can also be presented at lower levels of coverage. Thus, central government can be shown for 69 countries and territories, and general government for 62.

Annex 1.2. Public Sector Balance Sheet Methodology

This annex describes the methodology used to construct the database developed in this report. It explains the definitions used in the compilation of the public sector data, the main data sources, and the methodology used. It also describes the main concepts involved in the estimation of the intertemporal balance sheet and details of the balance sheet strength indicators developed in the report. Last, it provides an overview of the countries and variables covered in the database. The PSBS database, together with country-specific documentation on sources and methods is intended to be published in the near future.

Public Sector Balance Sheet

Definitions

The PSBS database is compiled using the conceptual framework of the IMF’s Government Finance Statistics Manual 2014 (GFSM 2014). This section presents that framework’s definitions in terms of coverage of institutions, stocks, and flows.

Coverage of Institutions

The public sector consists of all resident institutional units that are deemed to be controlled by the government. This includes all government units, such as departments, agencies, and nonprofit institutions controlled by the government, as well as corporations controlled by a government unit or another public corporation. Control of a corporation is established when the general corporate policy is determined by government. These public corporations comprise government controlled market producers that operate in both the financial and nonfinancial sector of the economy.

The database presents the data for the consolidated public sector as well as for its different subsectors, as follows (see Annex Table 1.2.3 for specifics on country coverage):

  • General government, with data for the central government level also available;43

  • Nonfinancial public corporations; for analytical purposes, natural resource corporations are presented separately from other nonfinancial public corporations; and

  • Financial public corporations, split to identify separately the central bank, sovereign wealth funds (where they operate as financial corporations), and other financial public corporations.

Annex Table 1.2.1.Composition of the Public Sector Balance Sheet
AssetsLiabilities
Nonfinancial assetsSpecial drawing rights
Fixed assetsCurrency and deposits2
LandDebt securities
Mineral and energy resources1Loans
Other nonfinancial assetsEquity and investment fund shares3
Financial assetsInsurance, pension, and standardized guarantee schemes
Monetary gold and special drawing rightsPension entitlements
Currency and depositsClaims of pension funds on pension managers
Debt securitiesOther insurance, pension, and standardized guarantee scheme liabilities
LoansFinancial derivatives and employee stock options
Equity and investment fund sharesOther accounts payable
Insurance, pension, and standardized guarantee schemes
Financial derivatives and employee stock options
Other accounts receivableNet Worth (= Assets – Liabilities)
Annex Table 1.2.2.Risk Weights of Assets and Liabilities, by Instrument
Weight
Financial assets, by instrument
Monetary gold and SDRs0.000
Currency and deposits0.000
Debt securities0.049
Loans0.064
Equity and investment fund shares0.564
Insurance, pension, and standardized guarantee schemes0.000
Financial derivatives and employee stock options0.049
Other accounts receivable0.049
Liabilities, by instrument
SDRs0.000
Currency and deposits0.000
Debt securities0.000
Loans0.122
Equity and investment fund shares0.000
Insurance, pension, and standardized guarantee schemes0.000
Financial derivatives and employee stock options0.014
Other accounts payable0.090
Sum of weights1.000
Source: IMF staff estimates.Note: Risk weight of each instrument is the standard deviation of valuation changes in that instrument relative to the sum of standard deviations of all asset and liability components. SDRs = special drawing rights.
Annex Table 1.2.3.Time Series Availability in the Public Sector Balance Sheet Database
CountryCentral GovernmentGeneral GovernmentPublic Sector
NFAxFA,LxLNRPENSLevelNFAxFA,LxLNRPENSNFAxFA,LxLNRPENS
AlbaniaNA2011–162011–16NACGin2011–162011–162011–162011–162013201320132013
AustraliaNA2000–162000–162000–16CGin2000–162000–162000–162000–162000–162000–162000–162000–16
AustriaNA2000–16NANACGex2000–162000–162000–162000–162015201520152015
BarbadosNA2000–162006–16NABCGNANANANANANANANA
BelgiumNA2000–16NANACGexNA2000–16NA2014–15NANANANA
BhutanNA2010–14NANACGinNA2010–14NANANANANANA
BrazilNA2006–162014–162010–14CGin2014–162006–162014–162010–142014201420142014
BulgariaNA2000–16NA2000–2016CGexNA2000–16NA2000–16NANANANA
Canada2000–162000–162000–162000–16CGex2000–162000–162000–162000–162000–162000–162000–162000–16
ChinaNA2010–16NANACGinNA2010–16NANANANANANA
ColombiaNA2008–162008–16NACGin2008–162008–162008–1620162016201620162016
CroatiaNA2002–16NANACGexNA2002–16NANANANANANA
CyprusNA2000–16NANACGexNA2000–16NANANANANANA
Czech RepublicNA2000–16NA2000–16CGex2000–162000–162000–162000–16NANANANA
DenmarkNA2000–16NANACGexNA2000–16NA2014–16NANANANA
El Salvador2003–162006–162003–162006–16CGex2003–162006–162003–162006–162003–162006–162003–162006–16
EstoniaNA2000–16NANACGexNA2000–162000–142014–15NANANANA
Finland2000–162000–162000–162000–16CGex2000–162000–162000–162000–162000–162000–162000–162000–16
FranceNA2000–16NANACGex2000–162000–162000–162000–162008–162008–162008–162008–16
Gambia, The2016201620162016BCGNANANANA2016201620162016
GeorgiaNA2012–16NA2012–16CGin2012–16NA2012–162012–162012–16NA2012–162012–16
GermanyNA2000–16NANACGex2000–162000–162000–162000–162001–162001–162001–162001–16
GreeceNA2000–16NANACGexNA2000–16NANANANANANA
GuatemalaNA2014NA2014CGin20142014201420142014201420142014
HungaryNA2000–16NANACGinNA2000–16NANANANANANA
Hong Kong SAR2000–162000–162000–162000–16CGin2002–162002–162002–162006–16NANANANA
IcelandNA2000–16NA2016CGinNA2000–16NA2013–16NANANANA
India2003–162003–162003–162003–16CGinNANANANA2004–162004–162004–162004–16
IndonesiaNA2008–162008–162010–16CGin2008–162008–162008–162010–162010–162010–162010–162010–16
IrelandNA2000–16NANACGinNA2000–16NA2014–15NANANANA
ItalyNA2000–16NA2000–16CGexNA2000–16NA2000–16NANANANA
JapanNA2000–16NA2000–16CGex2000–162000–162000–162000–162000–162000–162000–162000–16
KazakhstanNA2012–16NA2010–16CGin2012–162012–162012–162010–162012–162012–162012–162012–16
KenyaNA2013NA2013BCG20132013201320132013201320132013
Korea2012–162012–16NA2002–16CGin2000–162002–162000–162002–162002–162002–162002–162002–16
Kyrgyz RepublicNA2014–16NA2014–16CGin2014–162014–162014–16NANANANANA
LatviaNA2000–16NANACGexNA2000–16NA2014–15NANANANA
LithuaniaNA2000–16NANACGexNA2000–16NA2012–15NANANANA
LuxembourgNA2000–16NANACGexNA2000–16NANANANANANA
MalawiNA2009–16NANABCGNANANANANANANANA
Marshall IslandsNA2008–16NANABCGNANANANANANANANA
MicronesiaNA2008–16NANABCGNANANANANANANANA
MoldovaNA2005–162006–16NACGinNA2005–162006–16NANANANANA
NetherlandsNA2000–16NANACGex2001–152000–162001–152011–12NANANANA
New ZealandNA2006–162006–162006–16CGin2006–162006–162006–162006–162006–162006–162006–162006–16
NorwayNA2000–16NA2000–16CGex2000–162000–162000–162000–162000–162000–162000–162000–16
PalauNA2008–16NANABCGNANANANANANANANA
PeruNA2006–16NANACGin2006–152006–162000–1620132013201320132013
PolandNA2000–16NANACGexNA2000–16NA2014–15NANANANA
PortugalNA2000–16NANACGex2000–152000–152000–16NA2012201220122012
RomaniaNA2000–16NA2000–16CGexNA2000–16NA2000–16NANANANA
Russian FederationNA2001–162014–162012CGin2014–162001–162014–1620122012201220122012
San MarinoNA2002–16NANACGinNA2002–16NANANANANANA
SerbiaNA2007–12NANABCGNANANANANANANANA
Slovak RepublicNA2000–16NA2000–16CGexNA2000–16NA2000–16NANANANA
SloveniaNA2004–16NA2000–16CGexNA2004–16NA2000–16NANANANA
Solomon IslandsNA2012–16NANABCGNANANANANANANANA
South AfricaNA2000–16NA2000–16CGex2000–162000–162000–162000–162001–162001–162000–162001–16
SpainNA2000–16NA2000–16CGinNA2000–16NANANANANANA
SwedenNA2000–16NA2000–16CGexNA2000–16NA2000–16NANANANA
Switzerland2000–162000–162000–162000–16CGin2000–162000–162000–162000–16NANANANA
Tanzania2014201420142014CGin20142014201420142014201420142014
TunisiaNA2013NA2013CGexNA2013NA20132013201320132013
Turkey2014–162008–162014–162013CGin2014–162008–162014–1620132013201320132013
Uganda2015201520152015BCG20152015201520152015201520152015
UkraineNA2008–16NANACGinNA2008–16NANANANANANA
United Kingdom2000–162000–162000–162000–16CGin2000–162000–162000–162000–162000–162000–162000–162000–16
United States2001–162001–162001–162001–16CGin2001–162001–162001–162001–162001–162001–162001–162001–16
UruguayNA2001–16NA2001–16CGinNA2001–16NA2001–16NANANANA
Note: “Level” indicates the institutional coverage of central government data in the database, where CGin = central government, including social security funds; CGex = central government, excluding social security funds; and BCG = budgetary central government. NFAx = nonfinancial assets excluding land and natural resources; FA = financial assets; Lx = liabilities, excluding pension-related liabilities; LNR = land and natural resources; PENS = pension-related liabilities; and NA = not available.

Following the GFSM 2014 criteria to delineate market producers from nonmarket producers, some legally incorporated units have been reclassified to the general government. These criteria are based on the analysis of whether the corporations provide all or most of their output at economically significant prices or not.44

Central banks are included within the public sector. They are separately identified, recognizing the fact that their monetary liabilities (currency on issue) are irredeemable, and have no ongoing financing costs. The equity liability of the central bank (equivalent to its individual net worth) is reported on a book value basis—the difference between the value of its assets and nonequity liabilities. An alternative approach would recognize the discounted value of its seigniorage profits (Buiter 1983). Here, this is implicitly incorporated in the intertemporal balance sheet as part of the present value of dividend revenue flows, which boost the government’s primary balance. Central bank dividend flows are assumed to remain stable as a share of GDP and are not separately modeled.

Coverage of Stocks

The PSBS database includes all assets (financial and nonfinancial) owned and liabilities owed by the public sector or the relevant subsector at the end of each reporting period. Following the standard approach in macroeconomic statistics, economic ownership rather than legal ownership is used as a reference. Net worth is a balancing item representing the extent to which liabilities are covered by assets.

The composition of the balance sheet that is used in the analysis is summarized in Annex Table 1.2.1, which shows how assets and liabilities are disclosed in the database, broken down by type of asset or financial instrument.45 For analytical purposes, financial assets and liabilities are further broken down by currency of denomination and residual maturity, where available.

The PSBS data allow the calculation of several indicators, which are useful from an analytical perspective to measure balance sheet strength, namely: net worth, net financial worth, net liquid assets, net foreign exchange assets, risk-weighted assets and liabilities, and the degree of natural hedging (see the “Balance Sheet Strength” section).

The coverage of categories of assets and liabilities in balance sheets that are compiled by statistical authorities vary significantly from country to country. Some categories are often not recognized in the published balance sheets, and the PSBS database has therefore covered these categories by IMF staff estimates where data sources permitted. Most notably, these estimates include: nonfinancial assets—particularly land and mineral and energy resources—and public sector employment-related pension liabilities. The latter refer to pension entitlements of civil servants and public corporation employees under specific employment-related schemes, thus excluding other social security pension entitlements, which are of a contingent nature.

When these types of assets and liabilities could not be estimated, the relevant main aggregates of the balance sheet items were marked as “not available.” To ensure a correct cross-country comparability, alternative main aggregates were calculated, used in some cross-county empirical analysis, and disclosed as memorandum items, as follows:

  • Nonfinancial assets, excluding land and mineral and energy resources;

  • Total assets, excluding land and mineral and energy resources;

  • Liabilities, excluding pension-related liabilities (pension entitlements and claims of pension funds on pension managers);

  • Net financial worth, excluding pension liabilities; and

  • Net worth, excluding land, mineral and energy resources, and pension liabilities.

Coverage of Flows

The database includes the main flow aggregates, separating transactions and other economic flows. It also includes some more detailed categories of flows, which are directly related to assets and liabilities, such as interest receivable and payable, and rent, as well as those related to the relationship between government and public corporations—such as dividends, subsidies, or capital transfers payable and receivable.

Transactions correspond to interactions between units by mutual agreement or through the operation of the law. They are presented in the PSBS database in an abbreviated statement of operations, with the following main aggregates disclosed:

  • Revenue and expense—which are transactions that increase or decrease net worth, respectively; and

  • Net acquisition (acquisitions less disposals) of both nonfinancial and financial assets, and net incurrence (incurrence less repayment) of liabilities—which are transactions that change the composition of assets and liabilities but not net worth.

These aggregates allow the calculation of the following balancing items:

  • Net operating balance (NOB) is the difference between revenue and expense, with the latter including consumption of fixed capital; and

  • Net lending or borrowing (NLB) is the difference between revenue and expenditure; the latter corresponds to the sum of expense and net acquisition of nonfinancial assets.46 NLB is often also referred to as the “fiscal balance” or the “deficit/surplus.”

The PSBS database also includes other economic flows (OEFs) that result from revaluations (changes in prices and exchange rates) and other changes in the volume of assets and liabilities. The latter category can include: the economic recognition or derecognition of produced assets, such as valuables (or public monuments, if these are included in the balance sheet); entry and exit from the asset boundary of natural resources, as a result of changes in prices that make the exploitation of those resources viable or unviable; destruction of assets from large-scale, discrete events, such as earthquakes, volcanic eruptions, floods, or other natural disasters; or the reclassification of units (for example, a government unit that is transformed into a public corporation).

The database allows a full integration of stocks and flows, where source data permit. Therefore, the stock at the end of the reference period corresponds to the sum of the stock at the beginning of the reference period plus transactions and OEFs occurring during the reference period. For the net worth indicator, this accounting identity can be illustrated as follows:

By the definitions for net operating balance and net lending/borrowing, these can be denoted as follows:

in which Rev corresponds to revenue, Exp corresponds to expense, and Inv corresponds to net investment in nonfinancial assets.

This allows us to rearrange equation (1) as follows:

This is the approach followed in the analysis of the evolution of balance sheets in the report, where the change in net worth is explained by the sum of the fiscal balance, investment, and valuation effects.

Data Sources

Data for the central and general government generally replicate data reported by country authorities in the IMF’s Government Finance Statistics (GFS) database. Where these data fail to cover all categories of assets and liabilities listed above, they are complemented by other data reported by statistical authorities at the national level or other international organizations, such as Eurostat or the Organisation of Economic Co-operation and Development (OECD). Where data on fixed assets are not readily available, they are sourced from the IMF’s capital stock database (IMF 2017a). Any remaining data gaps are addressed, where possible, through IMF staff estimates (see the “Methodology” section).

Data for the central bank generally replicate stock data reported by country authorities in the IMF’s Monetary and Financial Statistics database through the standardized report forms. For transactions and other economic flow data, and for those countries that do not submit standardized report forms, data are compiled through the conversion of the central banks’ financial statements to the PSBS database template.

Data sources for other public corporations are country-specific and are captured in country-specific database documentation. The preferred data sources are statistical estimates produced by country authorities for the aggregate subsector, often compiled as a component of the sectoral accounts. Where these are not available, IMF staff estimates (either calculated specifically for this report or in fiscal transparency evaluations) are used. In these estimates, aggregate financial statements’ data from major state-owned enterprise ownership or annual reports (adjusted for unit reclassifications) are converted to the PSBS database template. When aggregate data are not available, the conversion of individual financial statements for the major state-owned enterprises is used. The latter option considers materiality: a sample of the largest public corporations, representing a significant share of total public corporation assets (covering about 80–90 percent of the total sector) is used and the aggregate result of the financial statements’ conversion factored up to account for the nonsample units.47

The public sector data are calculated by aggregating the estimates for general government, nonfinancial public corporations, and financial public corporations, and by identifying and consolidating (or eliminating) the most significant cross-holdings of assets and liabilities or intrapublic sector transactions.48 A nonexhaustive list of the most relevant items identified for consolidation in the public sector includes:

  • General government units’ deposits at the central bank or other public banks;

  • Central bank and other public corporations’ holdings of securities issued by government units;

  • General government units’ equity stakes in public corporations;

  • Loans provided by general government units to public corporations;

  • Loans provided by public banks to government units or other public corporations;

  • Property income such as interest and dividends paid or received on the aforementioned items; and

  • Subsidies and other capital transfers provided by government units to public corporations.

Methodology

Valuation of Assets and Liabilities

In accordance with the GFSM 2014 guidelines, assets and liabilities are valued at market value, where possible. This is normally the case for assets and liabilities in the form of debt securities and equity of listed corporations, whose values can be observed in the markets.49 Other financial assets and liabilities are often reported at nominal value. Nominal value reflects the value of the financial instrument at creation plus any subsequent flows, such as transactions (for example, accrual of interest or repayment of principal) or other economic flows such as exchange rate and valuation changes other than market price changes.50 It is considered a good proxy for market value in cases where financial instruments are not traded.

Where market values are not available for produced nonfinancial assets (fixed assets, inventories, and valuables), they are usually reported on a written down (or depreciated) replacement cost, that is, the current acquisition price of an equivalent new asset minus the accumulated depreciation (consumption of fixed capital), amortization, or depletion.

Public corporations’ assets and liabilities are generally reported based on fair value, following accounting standards such as International Financial Reporting Standards.51 However, the equity of these corporations, both in their balance sheets and as assets of the government, is often reported at its book value, which may be different from the market value. The equity value of public corporations in the PSBS database is set equal to their net asset value. This includes reserves, and is adjusted for provisions and deferred tax assets, which are not recognized in macroeconomic statistics. Because of data limitations, no adjustment is done to reflect the difference between the book and market values of listed shares.

Nonfinancial assets include land under buildings or other structures as well as stewardship land like that where national parks or other heritage sites are located. Because of the underlying difficulties in valuing such stewardship land, or historical heritage buildings, national estimates of nonfinancial assets normally do not include an estimation for these types of assets.52 In the absence of any alternative data sources for these estimations, the PSBS database does not attempt to value them.

The detailed methodology used to estimate specific categories of assets and liabilities, is as follows:

Fixed Assets. Existing government estimates for fixed assets other than historical/heritage assets are used where available, relying on authorities’ application of the perpetual inventory method on detailed asset-level information. With this method, the value of the stock is based on estimates of acquisitions and disposals that have been accumulated (after deduction of the accumulated consumption of fixed capital, amortization, or depletion) and revalued over a long enough period to cover the acquisition of all assets in the category. However, data are often missing or poorly reported, with serious valuation issues (Bova and others 2013).

Where there are gaps, estimates of fixed assets (for example, infrastructure, buildings) are provided based on the IMF’s capital stock and investment database (IMF 2017a), which includes estimates for the public capital stock also compiled through the perpetual inventory method, and calculated for the overall level of investment, rather than for detailed asset-level investment.53

Mineral and Energy Resources. Country estimates for mineral and energy resources are often based on various estimation techniques. Not many countries disseminate such data. To attain consistency, the PSBS database follows the GFSM 2014 valuation guidelines to estimate these values. Estimates for the stock of mineral and energy resources in the PSBS database correspond to the net present value of the expected pretax cash flows resulting from their commercial exploitation. Sources and methods for these estimates differ by type of commodity, and the choice of estimation method was largely determined by the availability of source data, and attempts to consider country-specific economic conditions in these estimations.54

The value of stocks of oil and gas were estimated using the following data sources: (1.1) production over the lifetime of the asset, from the Rystad database (Rystad Energy 2018); (1.2) prices (in US dollars) from World Economic Outlook (WEO) forecasts available at the end of the reference year; (1.3) costs of production (in US dollars), from the Rystad database; and (1.4) exchange rates, from WEO forecasts available at the end of the reference year.

Sources 1.1, 1.2, and 1.3 were used to calculate future US dollar cash flows over an 85-year horizon. These US dollar cash flows were converted to domestic currency using the WEO exchange rate forecasts (source 1.4). The net present value of the domestic currency cash flows was calculated using a discount rate equivalent to the average (2000–22) long-term (10-year) government bond yields in WEO plus a risk factor (1 percentage point for advanced economies, 3 percentage points for emerging economies, 6 percentage points for low-income developing countries). When WEO government bonds were not available, the central bank policy rate plus 5 percentage points was used.

The value of stocks of coal, metals, and other minerals were estimated using the following data sources: (2.1) estimates (in constant 2014 US$ prices), from the World Bank’s The Changing Wealth of Nations 2018 (Lange, Wodon, and Carey 2018); (2.2) United States Geological Survey data on 2016 reserves and 2014–16 production by commodity and by country (Wilburn, Bleiwas, and Karl 2016), where available; (2.3) prices (in US$) from WEO commodity prices for 2000–16; and (2.4) exchange rates, from the current vintage of WEO exchange rates.

Estimates for 2015 and 2016 are based on the changes in reserves in those years, for those commodities for which reserve data are available (source 2.2). Where these are not available (usually cases where reserves for a particular commodity are relatively small), the assumption was that the value of the stocks is unchanged from 2014 onward. The obtained estimates based on the constant 2014 US$ prices were converted to current US$ prices using the price index obtained through WEO commodity prices (source 2.3), and subsequently converted to domestic currency using WEO exchange rates (source 2.4).

For countries where subsoil assets can be owned by units other than government, the calculated estimates were prorated using alternative (country-specific) indicators on ownership of land under which the mineral and energy resources lie. Where such country-specific adjustments occurred, it is revealed in the database documentation.

Annex Box 1.2.1.The Statistical Treatment of Natural Resource Assets

In line with Government Finance Statistics (GFS) guidelines, government receipts from natural resources are generally treated as revenue, and therefore as an improvement in net worth, even though the funds come from the sale of a nonrenewable asset. This is despite the GFS recommendation that natural resources be recorded as nonfinancial assets on the government’s balance sheet, and even though each extracted barrel of oil or ton of iron ore reduces the remaining stocks of those assets. This treatment differs from the sale of all other nonfinancial assets, whose receipts are not recorded as revenue.

Treating natural resource receipts as revenue overstates government revenues and implies that government is running a better net operating balance than would be the case if those receipts were instead treated as the sale of a nonfinancial asset. The GFS framework allows for the reduction in net worth through an adjustment to other economic flows, but this still overstates the net operating balance. In practice, however, few governments record the value of natural resources on their balance sheet, and so do not record the depletion-related other economic flows.

Government balance sheets including estimates for natural resource assets and accounting for its depletion provide a clearer picture of government net worth and its developments over time. An alternative statistical treatment, as suggested by Traa and Carare (2007) excludes natural resource proceeds from revenues, and instead records those proceeds in the same way as receipts from sales of other nonfinancial assets, such as government buildings or public lands. While this does not impact the fiscal deficit, it reduces revenue and worsens the net operating balance, making clear the extent to which governments are running down public assets. As an example, applying this treatment to Kazakhstan would reduce the net operating balance by an average of 10 percent of GDP from 2010–16 (Figure 1.2.1.1). A similar approach is taken in the World Development Indicators, which adjust net national savings for the depletion of natural resources (Lange, Wodon, and Carey 2018).

Figure 1.2.1.1.Kazakhstan: Net Operating Balance

(Percent of GDP)

Source: IMF staff estimates.

Taken further, advocates of environmental accounting suggest that the current treatment of proceeds from the extraction of natural resources overstates not only government revenues but also economic activity (Obst and Vardon 2014; Coremberg 2015). They argue that the current treatment of sales of natural resources as giving rise to output and value added is incorrect, and that the portion of the sale related to the implicit value—or economic rent of the natural resource—should instead be treated as the sale of a nonproduced asset, rather than value added.

Pension Liabilities. Public sector pension entitlements are the claims that current and past public sector employees hold against their employers—they represent contractual payments that are established as part of the compensation agreement and must be paid, even in the event of future policy changes (representing accrued-to-date entitlements of existing beneficiaries). It is important to note that these employment-related pension liabilities exclude implicit obligations to households under general social security arrangements, as these are potentially subject to policy changes.55

The ideal data source for the employment-related pension liabilities are estimates produced by the country authorities, disclosed in the government’s financial statements, in statistical estimates of the sectoral accounts balance sheets, or in supplementary tables on pensions (as is the case for most EU members).56

When authorities’ estimates are not available, an estimate is produced using a model developed by IMF staff to calculate the accrued-to-date pension entitlements of civil servants and other public sector employees. This model uses actuarial projections of pension expenditure of these employment-related pension schemes.57 The estimate of the accrued benefit assumes that the share of the benefit accrued declines with age: in 2015, from 100 percent for those ages 55 and older to 0 for those ages 25 and younger. The population covered by the pension system is assumed to match the structure of the overall population (projections for population use the 2017 UN World Population Prospects—United Nations 2017). The discount rate is assumed to be 1 percentage point above the rate of GDP growth.58

Where the aforementioned estimates are available for only a single year because of data limitations, it is assumed that the entitlements as percentage of GDP remain constant over time.

Maturity and Currency Breakdowns

Where the national data sources include no breakdowns of financial assets and liabilities by maturity and currency, these breakdowns are estimated by IMF staff as follows:

  • Liquid assets include “currency and deposits” and “other accounts receivable,” while short-term liabilities are defined as the sum of “currency and deposits,” “other accounts payable,” and “current debt” (debt securities and loans issued with less than one-year maturity, and long-term debt securities and loans, with a remaining maturity of less than one year).

  • The current and noncurrent breakdown of debt securities and loans is obtained through three sources: World Bank’s Quarterly Public Sector Debt database, Eurostat, and the Dealogic database on debt securities. Repayments of outstanding IMF loans (where applicable) in the year after the reference period are subtracted from the short-term loans.

  • Foreign and domestic currency breakdowns of the debt securities are extracted from the Dealogic database and general government gross debt in foreign currency from the WEO database is used as a proxy of the total liabilities in foreign currency. These data are cross-checked against the outstanding amount of IMF loans (denominated in special drawing rights, SDRs, that is, foreign currency).

Intertemporal Balance Sheet

Intertemporal net worth is defined as follows:

where A0 and L0 are current assets and liabilities, Rt and Gt are future primary government revenues and expenditures at time t, and r is the discount rate. The intertemporal budget constraint states that intertemporal net worth should at least be equal to 0, a condition that should hold in a world where real interest rates are above real growth rates. In the very long term, it should equal 0 exactly, as no utility is derived from positive net worth at the end of time. To avoid double counting, flows associated with current assets and liabilities are excluded from future primary balances—hence, where there are resource assets, future resource revenues are excluded, and similarly the flows associated with accrued pension liabilities are excluded from primary spending. This approach draws on earlier work on intertemporal balance sheets, including Buiter (1983), Blanchard (1990), IMF (2016a), and Traa and Carare (2007).

The intertemporal balance sheet includes the estimates of assets and liabilities of the static balance sheet, combined with the discounted future revenue and primary expenditures flows for the next 50 years, on a no-policy change basis. Estimates of future flows are based on (1) a combination of medium-term fiscal forecasts out to the year 2022, as presented in the IMF WEO, and (2) from 2023 onward, long-term economic and fiscal projections, following the methodology presented in IMF (2016a).59 The long-term projections are unconstrained (so they do not require that the intertemporal budget constraint is met), and are based on an extension of current policy beyond 2022, with the following assumptions:

  • Nominal GDP projections assume inflation, productivity increases and the participation rate follow long-term averages, with changes in working-age population—under the United Nations’ medium-fertility scenario—driving any changes. For some countries, long-term average age cohort participation rates are used, which allows for variation in participation rates.

  • The fiscal projections follow the approaches developed over recent years (for a survey, see Anderson and Sheppard, 2009; for specifics, see Commonwealth of Australia 2015, Canada Department of Finance 2016, New Zealand Treasury 2016, and Office for Budget Responsibility 2017b). Primary revenues are generally assumed to remain constant as a share of GDP. Primary expenditures are split between age-related pension and health expenditures, which grow in line with demographic trends (see Clements and others 2015); and other primary expenditures, which are held constant as a share of GDP. Interest expenditures are forecast assuming a normalization of interest rates over the medium term.

  • The discount rate for long-term fiscal projections is set according to the implicit interest rate on government debt. This is in the mid-range of discount factors used in the balance sheet analysis: riskier natural resource assets are assumed to have a higher discount rate (10-year bond yields plus a risk factor), whereas more certain pension flows are assumed to have a lower discount rate (nominal GDP growth plus 1 percentage point). The 50-year horizon of the projections means that results are sensitive to discount rate assumptions. To isolate the impact of policy changes on flows, variations in fiscal projections because of policy changes or shocks are compared with the baseline using the baseline nominal GDP denominator and discount rates.

Balance Sheet Strength

Balance sheet strength measures can be grouped into three categories: those derived solely from the assets side; those derived solely from the liabilities side; and those derived from both sides of the balance sheet. The specific measures used in the analytical chapter are discussed subsequently.

Size of Balance Sheet

The size of balance sheet is defined as the average of the size of assets and liabilities, in percent of GDP. Balance sheets with larger assets or liabilities are normally exposed to large valuation changes. Valuation changes may expose the economy to macroeconomic risks, depending on the source of vulnerabilities and the nature of valuation changes. For instance, exposure to valuation changes in equity markets and pension liabilities may amplify crisis impacts on public finances (see Brede and Henn 2018).

Solvency: Net (Financial) Worth

Net worth is a measure of solvency. It is calculated as total assets minus total liabilities, expressed in percent of GDP. While providing a snapshot of solvency, it suffers from the various valuation issues that accompany the constituent parts of the balance sheet, particularly stemming from nonfinancial assets. Furthermore, it does not distinguish between assets that can be sold to meet financing needs, and assets that are not marketable.

Net financial worth is calculated as total financial assets less liabilities, expressed in percent of GDP. In general, financial assets and liabilities are more reliably valued and more readily marketable than nonfinancial assets. Given that pension-related liabilities are based on estimates, which could affect cross-country comparability, a measure for net worth excluding pension-related liabilities is also introduced.

Risk-Adjusted Assets and Liabilities

Risk-adjusted assets and liabilities provide a measure of the assets and liabilities corrected for their riskiness or underlying volatility. This measure is based on estimates of the volatility of each asset (liability) class relative to the sum of the volatilities of all asset and liability components.

First, a measure of valuation changes in each of the asset and liability items is constructed.60 To do so, transactions are deducted from total changes in the value of these items. Next, the relative volatility of valuation changes of individual items is defined as their riskiness, and labeled as the item’s risk weight (RW):

in which i is the indicator for a specific item of assets or liabilities.61 These risk weights are calculated on a sample of European countries for which detailed data on transactions and valuation changes of individual general government balance sheet items are available.62 The resulting risk weights are in Annex Table 1.2.2. Using these risk weights and the size of individual balance sheet items, a comprehensive measure of the riskiness of the asset and liability side of the balance sheet are constructed, which are denoted as ∑i RWi Ai and ∑i RWi Li . Last, these values are deducted from total assets and liabilities to get risk-adjusted assets (RAA) and liabilities (RAL):

Liquidity and Currency Mismatch

The liquidity mismatch is measured using the “net liquid assets” indicator, which is calculated as current assets less current liabilities—that is, assets or liabilities that are maturing within one year—expressed in percent of GDP to reflect the materiality of the mismatch.63 It is a measure of whether the public sector has sufficient liquid assets to support its short-term financing needs.

Similarly, currency mismatches are assessed using the “net foreign exchange assets” indicator, which shows the net impact of exchange rate fluctuations on the balance sheet. It is calculated as foreign exchange denominated assets less foreign exchange denominated liabilities, expressed in percent of GDP to reflect the materiality of foreign exchange mismatches.64

Natural Hedge

The natural hedge is a measure of volatility calculated as the variance of valuation changes in net financial worth (NFW) relative to the variance of valuation changes in financial assets and liabilities. It measures the covariance between the valuation changes in assets and liabilities, both expressed in percent of GDP, normalized by the size of the movements in assets and liabilities. The measure can be decomposed into two parts: how correlated the financial assets and liabilities are; and whether there is a mismatch between the sizes of financial assets and liabilities.

As net financial worth is defined as financial assets net of liabilities, valuation changes in net financial worth (that is, changes resulting from other economic flows) can be represented as follows:

in which OEFFA,OEFL,andOEFNFW denote other economic flows in financial assets, liabilities, and net financial worth, respectively, all expressed in percent of GDP. Then:

in which FA denotes financial assets and L denotes liabilities. The equation shows how the volatility of net financial worth is dampened by the covariance between financial assets and liabilities.

To come up with a normalized measure of the volatility in net financial worth, the volatility of net financial worth is divided by the standard deviations of financial assets and liabilities, resulting in a unit-less measure—similar to the measure of correlation. The relative volatility of NFW to the volatility of financial assets and liabilities is presented as σn:

This is the natural hedge measure. It can be rewritten by plugging equation (1) into (2):

in which x=σFAσL, and CorFA,L represents the correlation between financial assets and liabilities.

The relative standard deviations (xand1x) are proxies for the contribution of size mismatch between financial assets and liabilities to the variation in net financial worth—if one side of the balance sheet is much bigger than the other side, its variations will dominate the variations in net financial worth. CorFA,L represents how valuation changes in financial assets and liabilities move together.

Fiscal Stress Tests

A fiscal stress test applies a large but plausible macroeconomic shock to the fiscal accounts. It can combine the direct impact on growth and revenue with effects on asset prices and realizations of contingent liabilities to assess the full fiscal impact of the stress event. Following the methodology outlined in IMF (2016a), fiscal stress tests contains three key elements:

  • A macro-fiscal shock: identifying an extreme macro scenario (including changes to asset prices), and applying it using a fiscal forecasting model, which allows accounting for nonlinearities and budget rigidities;

  • A contingent liability shock, based on an assessment of contingent liabilities that might be realized in the event of a macro crisis and their cost; and

  • An assessment of the impact of the macro-fiscal shock and contingent liability realization on the government’s comprehensive balance sheet, incorporating the value of future revenues and expenditures to provide a fuller picture on fiscal solvency.

A fiscal stress test can provide three summary outputs, depending on its focus. These can be used in assessing fiscal risks and providing guidance on the channels through which a macroeconomic crisis might impact public finances:

  • Public wealth, as assessed against the change in the government’s net worth or net financial worth, incorporating future fiscal flows;

  • Government liquidity needs, as assessed against gross financing needs; and

  • The financing burden, in the form of interest expense against revenue collections.

Annex 1.3. Balance Sheet Strength and Sovereign Bond Yields

This annex describes how the estimates for the impact of balance sheet strength measures on government bond yields in Box 1.2 are derived. It performs the estimation for the full sample of countries as well as for advanced economies and emerging markets separately.65

It estimates the following fixed effects panel specification:

in which yjt is the long-term government bond yield of country i in year t, extracted from the Thomson Reuters Datastream Economics database;66 and xit a balance sheet variable, the main variable of interest. These indicators include general government gross debt, total assets, financial assets, net worth, and net financial worth, all lagged to minimize the bias originating from reverse causality.67 All balance sheet indicators are based on general government data from the PSBS database introduced in this Fiscal Monitor, except for gross debt, which is extracted from the World Economic Outlook database. All variables are expressed in percent of GDP. The set of variables zit controls for the possible channels through which macro-fiscal conditions may affect sovereign bond yields. The control variables include the growth rate of real per capita GDP, the US 10-year bond yield, the average inflation rate in country i, the short-term interest rate, and the general government primary balance.68 Last, ci and λt represent country and time fixed effects, respectively. The sample period is 2001–16.

The estimation results show that financial markets seem to account for government assets and net worth when pricing sovereign bonds. Phrased differently, balance sheet indicators beyond gross debt matter for sovereign yields. Specifically, total or financial assets are highly significant variables, both as stand-alone balance sheet variables and in regressions together with gross debt. Similarly, net (financial) worth are highly significant stand-alone explanatory variables for the pricing of sovereign bonds (Annex Table 1.3.1). These results are most clear in the full sample and the sample consisting of advanced economies, while significance is generally lower in the much smaller sample consisting solely of emerging markets. The results are robust to using a different time period, excluding the crisis years.

Annex Table 1.3.1.Government Balance Sheet and Sovereign Bond Yields
Dependent Variable: Long-Term Government Bond Yields
Full Sample
Lagged net worth-0.007***
Lagged net financial worth-0.006**
Lagged gross debt0.014***0.013***0.010***
Lagged total asset-0.009***-0.009***
Lagged financial assets-0.007***-0.010***
Observations409415445447685448454
Number of countries31313333333333
Advanced Economies
Lagged net worth-0.005***
Lagged net financial worth-0.006***
Lagged gross debt0.015***0.014***0.012***
Lagged total asset-0.003**-0.003**
Lagged financial assets-0.004***-0.007***
Observations328334348350579351357
Number of countries24242525242525
Emerging Markets
Lagged net worth-0.025***
Lagged net financial worth-0.013
Lagged gross debt0.041**0.0080.006
Lagged total asset-0.031***-0.024***
Lagged financial assets-0.046*-0.041**
Observations818197971069797
Number of countries7788988
Note: The table represents the fixed effects estimation results investigating the impact of balance sheet indicators on long-term government bond spreads. Total assets exclude land and natural resources, and liabilities exclude pension liabilities for reasons of cross-country comparability. For the same reason, net worth excludes all of the aforementioned items, and net financial worth excludes pension liabilities. Control variables include real per capita GDP growth, US 10-year bond yield, average inflation rate, short-term interest rate, general government primary balance, country and time fixed effects not reported in the table for brevity. The sample period is 2001–16.*, **, and *** represent statistical significance at 10, 5, and 1 percent, respectively.

The magnitude of the impact of net (financial) worth on yields is comparable but somewhat smaller than the impact of gross debt. In the whole sample, a one percent of GDP increase in government net (financial) worth lowers yields by some 0.7 (0.6) bps, compared with a 1 bps increase in yield when gross debt increases by the same amount. The effect of net worth is more pronounced in emerging markets, where a 1 percent of GDP increase in net worth can lower yields by some 2.5 bps, which is consistent with higher and more variable yields in emerging markets. These results are consistent with Hadzi-Vaskov and Ricci (2016) and Gruber and Kamin (2012), in finding that financial markets seem to account for government assets and net worth when pricing sovereign bonds, and that the effect of fiscal variables of interest (gross/net debt, assets) on bond yields/spreads is larger for emerging market economies than advanced economies. The emerging market regressions should, however, be interpreted with caution given the small sample size.

Annex 1.4. Balance Sheet Strength and the Macro Economy

This annex provides a summary of the econometric specification to study the impact of balance sheet strength on the macro economy (Box 1.2). The analysis is based on the local projection method introduced by Jordà (2005) and Jordà, Schularick, and Taylor (2016) using a sample of 17 advanced economies for which time series data are available. The baseline regression is as follows:

in which the dependent variable yi,p+h – yi,p is the cumulative growth rate (log difference) in real GDP or real government spending, both in per capita terms in country i, h years after the business cycle peak. Peak years are identified as the start of the recession; they are the last year in which real per capita GDP grows, that is, the year followed by a year in which it declines (Bry and Boschan 1971). The dummy variables di,pSanddi,pW denote, respectively, strong and weak balance sheets in the peak year. Strong (weak) balance sheets are defined as those with net financial worth above (below) the sample median. Following the analysis in the October 2016 Fiscal Monitor, the variables xi,pPrandxi,pPu present the average annual change in the five years before the peak of private debt, and the level of public debt as a percent of GDP at the peak, respectively. Yi,p-l is the set of lagged control variables. Controls include two lags of real per capita GDP growth rates, government expenditures, public debt, and private debt.69 Last, αi,h are country-year fixed effects, and εi,p+h denotes the error term. Standard errors are computed using the Driscoll and Kraay (1998) method to correct for heteroskedasticity, cross-sectional dependence, and serial correlation.

The data cover the period 1970–2015 and come from various sources. Data on net financial worth are taken from the World Inequality Database, which provides a long time series for 17 advanced economies.70,71 Data on public debt and private credit are sourced from the database compiled in the October 2016 Fiscal Monitor. Real per capita GDP is extracted from the World Economic Outlook and Penn World Table, whereas government spending data are sourced from Mauro and others (2015). In these time series we observe 53 recessions.

The regressions support the view that countries with a strong balance sheet face shorter and shallower recessions (Annex Table 1.4.1). The results on expenditure show a statistically significant difference between the coefficient for countries with a strong and weak balance sheet, with p-values below 5 percent starting from the second year. Results in the GDP regression are also significant, although the p-values for the test of the difference between the coefficients for countries with a strong and weak balance sheet are only below 5 percent in years 4 and 5. This lower significance is likely due to the limited number of observations in the sample. The findings are robust whether or not the xi,pPrandxi,pPu variables and their interactions are included, as well as for using net worth instead of net financial worth as indicator of balance sheet strength.

Annex Table 1.4.1.Recovery and Fiscal Policy in the Aftermath of Economic Recessions
Real Government Expenditure per CapitaReal GDP per Capita
Year 1Year 2Year 3Year 4Year 5Year 1Year 2Year 3Year 4Year 5
θs3.90***

(1.05)
8.77***

(1.24)
14.69***

(3.04)
24.39***

(4.02)
33.46***

(3.80)
-1.60***

(0.32)
-0.77*

(0.62)
1.23*

(0.60)
4.29***

(0.73)
9.30***

(0.95)
θw1.31

(2.21)
0.30

(1.91)
1.92

(2.24)
-11.31**

(4.41)
-2.81*

(1.99)
-2.78***

(0.96)
-2.84**

(1.13)
-0.70

(1.28)
-0.06

(1.31)
2.67*

(1.56)
R20.800.840.850.850.910.830.740.760.840.91
θs = θw (p-value)0.420.010.010.000.000.340.120.260.030.01
Peaks53535252425353525242
Source: IMF staff estimates.Note: The table reports the estimations using the local projections model. The first five columns present the coefficients for real per capita GDP, and the second five represent those for real per capita government expenditure as dependent variables (both cumulative changes starting from the peak before economic recessions). The regressions also include fixed effects and control variables that are not reported. Robust standard errors are reported in the second row of each line where *, **, and *** denote p-values less than 0.32 (1 standard deviation), 0.05 (2 standard deviations), and 0.01 (3 standard deviations), respectively.
References

    Allen, M.,C.Rosenberg,C.Keller,B.Setser, andN.Roubini. 2002. “A Balance Sheet Approach to Financial Crises.” IMF Working Paper 02/210, International Monetary Fund, Washington, DC.

    Amante, A.,P.Anderson,T.Jonasson,H.Kamil, andM. G.Papaioannou. Forthcoming. “Implementation of Sovereign Asset and Liability Management in Emerging Market Countries: The Case of Uruguay.” IMF Working Paper, International Monetary Fund, Washington, DC.

    Anderson, B., andJ.Sheppard. 2009. “Fiscal Futures, Institutional Budget Reforms, and Their Effects: What Can Be Learned?OECD Journal on Budgeting3: 7117.

    Appleby, M.,J.Harris,S.Hida,M.Norat, andA.Soler. 2018. “Fiscal Stress Test for The Gambia.” IMF Country Report 18/100, International Monetary Fund, Washington, DC.

    Bernardini, M., andL.Forni. 2017. “Private and Public Debt: Are Emerging Markets at Risk?IMF Working Paper 17/61, International Monetary Fund, Washington, DC.

    Blanchard, O.1990. “Suggestions for a New Set of Fiscal Indicators.” OECD Economics Department Working Papers No. 79, OECD, Paris.

    Board of Governors of the Federal Reserve System. 2018. “Dodd Frank Act Stress Test 2018: Supervisory Stress Test Methodology and Results.” Washington, DC.

    Bova, E.,R.Dippelsman,K.Rideout, andA.Schaechter. 2013. “Another Look at Government’s Balance Sheets: The Role of Nonfinancial Assets.” IMF Working Paper 13/95, International Monetary Fund, Washington, DC.

    Brede, M., andC.Henn. 2018. “Finland’s Public Sector Balance Sheet: A Novel Approach to Analysis of Public Finance.” IMF Working Paper 18/78, International Monetary Fund, Washington, DC.

    Bry, G., andC.Boschan. 1971. Cyclical Analysis of Time Series: Selected Procedures and Computer Programs. Cambridge, MA: National Bureau of Economic Research.

    Buiter, W. H.1983. “Measurement of the Public-Sector Deficit and Its Implications for Policy Evaluation and Design.” IMF Staff Papers30 (2): 30649.

    Cabezon, E., andC.Henn. 2018. “Counting the Oil Money and the Elderly: Norway’s Public Sector Balance Sheet.” IMF Working Paper 18/190, International Monetary Fund, Washington, DC.

    Canada Department of Finance. 2016. “Update of Long-Term Economic and Fiscal Projections.” Ottawa.

    Clements, B. J.,K.Dybczak,V.Gaspar,S.Gupta, andM.Soto. 2015. “The Fiscal Consequences of Shrinking Populations.” IMF Staff Discussion Note 15/21, International Monetary Fund, Washington, DC.

    Commonwealth of Australia. 2015. 2015 Intergenerational Report: Australia in 2055. Canberra.

    Commonwealth of Australia. 2018. “2018–19 Budget Paper No. 1.” Canberra.

    Coremberg, A. 2015. “Accounting for Natural Resources Wealth and Productivity from SNA and Beyond.” Paper presented at the International Association for Research in Income and Wealth Conference, Dresden.

    Detter, D., andS.Fölster. 2015. The Public Wealth of Nations. Basingstoke, UK: Macmillan.

    Di Maggio, M.,A.Kermani, andC.Palmer. 2016. “How Quantitative Easing Works: Evidence on the Refinancing Channel.” NBER Working Paper 22638, National Bureau of Economic Research, Cambridge, MA.

    Driscoll, J. C., andA. C.Kraay. 1998. “Consistent Covariance Matrix Estimation with Spatially Dependent Panel Data.” Review of Economics and Statistics80 (4): 54960.

    Easterly, W.1999. “When Is Fiscal Adjustment an Illusion?Economic Policy14 (28): 5686.

    El Rayess, M.,A.Halstead,J.Ralyea,J.Harris, andA. F.Tieman. Forthcoming. “Indonesia’s Public Wealth: A Balance Sheet Approach to Public Wealth.” IMF Working Paper, International Monetary Fund, Washington, DC.

    Escolano, J.2010. “A Practical Guide to Public Debt Dynamics, Fiscal Sustainability, and Cyclical Adjustment of Budgetary Aggregates.” IMF Technical Notes and Manuals, International Monetary Fund, Washington, DC.

    Federal Housing Finance Agency. 2017. “Dodd-Frank Act Stress Tests Results—Severely Adverse Scenario.” Washington, DC.

    Finanzministerium. 1933. “Finanzieller Überblick über den Reichshaushaltsplan.” Berlin, Germany.

    Frame, W.,A.Fuster,J.Tracy, andJ.Vickery. 2015. “The Rescue of Fannie Mae and Freddie Mac.” Staff Report No. 719, Federal Reserve Bank of New York, New York, NY.

    Gonguet, F.,K.Hellwig,J.Harris, andA. F.Tieman. Forthcoming. “Public Wealth in the United States.” IMF Working Paper, International Monetary Fund, Washington, DC.

    Greenwood, R.,S. G.Hanson,J. S.Rudolph, andL. H.Summers. 2014. “Government Debt Management at the Zero Lower Bound.” Working Paper 5, Hutchins Center on Fiscal and Monetary Policy, Brookings Institution, Washington, DC.

    Gruber, J. W., andS. B.Kamin. 2012. “Fiscal Positions and Government Bond Yields in OECD Countries.” Journal of Money, Credit, and Banking44 (8): 156387.

    Hadzi-Vaskov, M., andL. A.Ricci. 2016. “Does Gross or Net Debt Matter More for Emerging Market Spreads?IMF Working Paper 16/246, International Monetary Fund, Washington, DC.

    Henao-Arbelaez, C., andN.Sobrinho. 2017. “Government Financial Assets and Debt Sustainability.” IMF Working Paper 17/173, International Monetary Fund, Washington, DC.

    Her Majesty’s Treasury. 2018a. “Managing Fiscal Risks: Government Response to the 2017 Fiscal Risks Report.” Policy Paper, London.

    Her Majesty’s Treasury. 2018b. “Financial Relationship between HM Treasury and the Bank of England.” London.

    Irwin, T.2012. “Accounting Devices and Fiscal Illusions.” IMF Staff Discussion Note 12/02, International Monetary Fund, Washington, D.C.

    International Monetary Fund. 2012a. “Fiscal Frameworks for Resource-Rich Developing Countries.” IMF Staff Discussion Note 12/04, Washington, DC.

    International Monetary Fund. 2012b. “Fiscal Transparency, Accountability and Risk.” IMF Policy Paper, Washington, DC.

    International Monetary Fund. 2014. Government Finance Statistics Manual. Washington, DC.

    International Monetary Fund. 2015a. “Balance Sheet Analysis in Fund Surveillance.” IMF Policy Paper, Washington, DC.

    International Monetary Fund. 2015b. “Making Public Investment More Efcient.” IMF Policy Paper, Washington, DC.

    International Monetary Fund. 2016a. “Analyzing and Managing Fiscal Risks—Best Practices.” IMF Policy Paper, Washington, DC.

    International Monetary Fund. 2016b. “United Kingdom Fiscal Transparency Evaluation.” IMF Country Report 16/351, International Monetary Fund, Washington, DC.

    International Monetary Fund. 2017a. “Estimating the Stock of Capital in 170 Countries (Jan 2017 Update).” Washington, DC.

    International Monetary Fund. 2017b. “Finland. Article IV Staff Report.” IMF Country Report No.17/370, Washington, DC.

    International Monetary Fund. 2018a. Fiscal Transparency Handbook. Washington, DC.

    International Monetary Fund. 2018b. “People’s Republic of China. 2018 Article IV Staff Report.” IMF Country Report No. 18/240, Washington, DC.

    Jaramillo, L.,C.Mulas-Granados, andE.Kimani. 2017. “Debt Spikes and Stock Flow Adjustments: Emerging Economies in Perspective.” Journal of Economics and Business94: 114.

    Jin, H.2018. “Prioritizing Fiscal Reforms to Enhance Productivity and Equity in Indonesia.” IMF Country Report 18/33, International Monetary Fund, Washington, DC.

    Jordà, O.2005. “Estimation and Inference of Impulse Responses by Local Projections.” American Economic Review95: 16182.

    Jordà,M.Schularick, andA. M.Taylor. 2016, “Sovereigns versus Banks: Credit, Crises, and Consequences.” Journal of the European Economic Association14 (1): 4579.

    Krishnamurthy, A., andA.Vissing-Jorgensen. 2011. “The Effects of Quantitative Easing on Interest Rates: Channels and Implications for Policy.” NBER Working Paper 17555, National Bureau of Economic Research, Cambridge, MA.

    Lam, W. R., andM.Moreno-Badia. Forthcoming. “What Does China’s Government Balance Sheet Tell Us about Fiscal Policy?IMF Working Paper, International Monetary Fund, Washington, DC.

    Lange, G.-M.,Q.Wodon, andK.Carey (eds.). 2018. The Changing Wealth of Nations 2018. Washington, DC: World Bank.

    League of Nations. 1938. “Questionnaire on Public Debt.” Document C.L.51.1938.II.A, Geneva.

    Li, B. G., andR.Mano. Forthcoming. “The Debt of Local Government Financing Vehicles: New Findings Using Firm-Level Analysis.” IMF Working Paper, International Monetary Fund, Washington, DC.

    Li, Y., andX. J.Zhang. 2017. China’s National Balance Sheet: Theories, Methods, and Risk Assessment. New York: Springer Press.

    Li, Y., andX.Chang. 2015. China’s National Balance Sheet (2015): Leverage Adjustment and Risk Management. New York: Springer Press.

    Luck, S., andT.Zimmermann. 2017. “Employment Effects of Unconventional Monetary Policy: Evidence from QE.” SSRN Electronic Journal. https://papers.ssrn.com/sol3/papers.cfm?abstract_id=3002047.

    Mano, R. C., andP.Stokoe. 2017. “Reassessing the Perimeter of Government Accounts in China.” IMF Working Paper 17/272, International Monetary Fund, Washington, DC.

    Mauro, P.,R.Romeu,A.Binder, andA.Zaman. 2015. “A Modern History of Fiscal Prudence and Profligacy.” Journal of Monetary Economics76 (C): 5570.

    Milesi-Ferretti, G. M., andK.Moriyama. 2004. “Fiscal Adjustment in EU Countries: A Balance Sheet Approach.” IMF Working Paper 04/143, International Monetary Fund, Washington, DC.

    New Zealand Treasury. 2016. “Demographic, Economic, and Fiscal Assumptions and Modelling Methods in the 2016 Long-Term Fiscal Model.” Wellington.

    New Zealand Treasury. 2018. “He Puna Hao Patika, 2018 Investment Statement.” Wellington.

    Obst, C., andM.Vardon. 2014. “Recording Environmental Assets in the National Accounts.” Oxford Review of Economic Policy30 (1): 12644.

    Office for Budget Responsibility. 2017a. Fiscal Risks Report. London.

    Office for Budget Responsibility. 2017b. Fiscal Sustainability Report. London.

    OsborneS., andB.Retus. 2017. “Returns for Domestic Business.” Survey of Current Business, US Bureau of Economic Analysis, Suitland, Maryland.

    Rystad Energy. 2018. Upstream Database.

    Samphantharak, K., andR. M.Townsend. 2009. “Households as Corporate Firms.” Cambridge University Press, Cambridge, UK.

    Seiferling, M., andS.Tareq. 2015. “Fiscal Transparency and the Performance of Government Financial Assets.” IMF Working Paper 15/9, International Monetary Fund, Washington, DC.

    Shin, J.2018. “Indonesia’s Growth Strategy: Boosting Potential Growth with Structural Reforms.” IMF Country Report 18/33, International Monetary Fund, Washington, DC.

    Shoag, D.2013. “Using State Pension Shocks to Estimate Fiscal Multipliers since the Great Recession.” American Economic Review103 (3): 12124.

    Shoag, D.2014. “Past and Present High-Risk Investments by States and Localities.” Brookings Institution Press, Washington, DC.

    Traa, B., andA.Carare. 2007. “A Government’s Net Worth.” Finance and Development44 (2): 4649.

    Turner, D., andF.Spinelli. 2012. “Interest-Rate-Growth Differentials and Government Debt Dynamics.” OECD Journal: Economic Studies10322.

    United Nations. 2017. “World Population Prospects: The 2017 Revision.” New York.

    Wilburn, D. R.,D. I.Bleiwas, andN. A.Karl. 2016. “Global Stocks of Selected Mineral-based Commodities.” In Scientific Investigation Report 2016–5152, US Geological Survey, Reston, VA.

    Yang, Z. Y.,B.Zhang, andL. M.Tan. 2017. “Chinese Government Balance Sheet.” Social Sciences Academic Press, Beijing.

Public wealth refers to public sector net worth. The two terms are used interchangeably throughout this report. See Annex 1.1 for details on the sample.

In fact, the move toward compiling government balance sheets started much earlier, as evidenced by the publication of the central government balance sheet in Weimar Germany (Finanzministerium 1933) and a questionnaire on government balance sheets from the League of Nations (1938).

Annex 1.2 provides details on these elements, how they are valued, and how these estimates were compiled.

Following the Government Finance Statistics Manual (GFSM), pension obligations to private sector employees under pay-as-you-go social security schemes, such as the US Social Security or Japan’s National Pension System and Employee Pension Insurance, are not included in the static balance sheet. They are instead incorporated in future expenditure flows in the intertemporal balance sheet.

Some financial assets may be earmarked to specific uses or liabilities, such as deposits associated with grants for specific projects or assets tied to pension obligations. These encumbered assets are therefore unavailable for other financing needs under current institutional arrangements. However, examining these financial assets in a consolidated way may reveal potential benefits from improvements in asset management.

Because of data limitations, for many countries the analysis includes only central government public corporations.

This includes Treasury holdings held by federal trust funds, including Old Age and Survivors and Disability trust funds (which are classified inside general government) as well as holdings of public corporations.

Public investment refers to net acquisition of nonfinancial assets, which is part of the fiscal deficit (see Annex 1.2). Traditional accounting of the deficit does not take account of the assets built up by such investment.

Weighted average of Austria, Belgium, Denmark, Estonia, Finland, France, Germany, Ireland, Italy, Latvia, Lithuania, the Netherlands, Norway, Poland, Portugal, the Slovak Republic, Slovenia, Spain, Sweden, and the United Kingdom.

Easterly (1999) and Irwin (2012) provide fuller account of these practices. Milesi-Ferretti and Moriyama (2004) distinguish between structural reductions in debt that improve net worth, and nonstructural reductions that reduce debt by decumulating assets.

Governments also carry a “liability” in expectation that they will provide future goods, services, and transfers.

This section uses simple averages.

These estimates cover a broader range of countries and territories but are less comprehensive than those presented in Figure 1.1. For seven countries, the data are available only for central government. To make the data comparable across countries, the figure excludes land and natural resource assets and pension liabilities.

Based on central government data for India, which may partly explain the small number.

Central bank foreign exchange reserves are excluded from this analysis.

For this analysis, debt securities are measured at face value, as they are almost always repaid at maturity. Using market prices for debt securities increases their volatility by 0.6 percent of GDP on average.

This denominator effect displays the impact of moving from 2007 to 2016 GDP in the denominator. The 2007 bar is expressed in percent of 2007 GDP, while all other bars are expressed in 2016 GDP.

Public sector net debt records the public sector’s gross debt minus its holdings of liquid financial assets. Balance sheet measures published by the United Kingdom show the government’s equity holdings of public sector banks, rather than the entirety of their balance sheets, as presented here (see IMF 2016b).

Much of the decline in net worth over the period reflects increasing debt and pension liabilities.

These effects mainly comprise gains in other accounts receivable and payable, other reserve revaluations, and net premiums or discounts of gilt issuance.

For details on how these data are compiled and consolidated, see Annex 1.2.

To avoid double counting, future pension payments to civil servants that are already recorded as a liability in the static balance sheet are excluded from the primary balances. In line with GFSM 2014, private sector pension liabilities are excluded from the static balance sheet but are included in intertemporal net worth. However, public assets related to these private sector pension obligations are included in the static balance sheet.

See Brede and Henn (2018) for details of the shock and its impact.

While some public nonfinancial assets could be sold without large repercussions on economic activity and tax revenues (for example, converting a public highway to a private toll highway), some might be difficult or impossible to sell (like in-city roads, sewage infrastructure, and land in remote areas).

In the United States statistics, nonfinancial public corporations are included in general government.

GSE-held mortgages are financed through GSE-issued debt and agency-backed securities. With the formal federal takeover of Fannie Mae and Freddie Mac in 2008, these previously implicit government liabilities were made explicit. Cumulative draws by Fannie Mae and Freddie Mac on the treasury between 2008 and 2011 amounted to US$187.5 billion (1.3 percent of 2007 GDP). See Frame and others (2015).

The federal government employee pension fund—formally, the Civil Service Retirement System—has been closed to new entrants since 1983 and holds only treasury securities, so it is less volatile.

See Board of Governors of the Federal Reserve System (2018). The scenario is more severe than the shocks associated with the 2008–09 global financial crisis. The stress test assumes that no countercyclical fiscal policy measures are taken, so the increase in the deficit is smaller than during 2009, when expansionary measures were taken to dampen the effects of the crisis. In the scenario, GDP growth is –6 percent, while the unemployment rate increases to almost 10 percent, real estate prices decline by one-third, and equity prices by almost two-thirds.

Historically, about 20 percent of variations in the national real estate price index are reflected in the valuation of the public sector’s structures.

This relatively limited impact on the federal student loan portfolio is consistent with the strong recovery power of the federal government in case of default, the rarity of discharge cases, and rules in place allowing for temporary relief (deferment, forbearance, grace periods, etc.).

The loan loss estimates do not directly compare with the potential treasury drawdowns, which would include the consequences of the shock on other assets, the impact of valuation allowances on deferred tax assets, and the effect of provisioning rules. However, they are of similar magnitude to the Federal Housing Finance Agency (2017) estimates under the severely adverse scenario that estimates a potential incremental treasury draw by Fannie Mae and Freddie Mac of about US$100 billion (0.4 percent of baseline GDP) over two years.

See Appleby and others (2018) for details, as well as the interaction of the public sector with the banking sector, illustrating the sovereign-banking feedback loop. While no contingent liabilities arise from the banking system as a result of high loss-absorbing buffers, these are heavily eroded. The stress test uses the June 2017 vintage of data. Since then, the debt position has deteriorated further.

The implications of this statistical treatment of natural resources are explored in detail in Annex Box 1.2.1.

The effect of the tax-financed investment surge on GDP is modeled using the IMF G20Mod model. For details see El Rayess and others (forthcoming). These long-term projections are subject to considerable uncertainty.

After three years, public investment remains at a level that maintains the higher capital stock in perpetuity. The remaining tax revenue is assumed to be spent on priority current expenditure categories, such as health, pension, and education expenditures.

This scenario allows for monetary policy accommodation, and therefore a constant discount rate. If instead monetary policy were tightened, the impact of the investment surge on intertemporal net worth would be about 5 percent of GDP. Details and further sensitivity analysis can be found in El Rayess and others (forthcoming).

Note that Indonesia’s intertemporal net worth is –18 percent in 2023 compared with +1.8 percent of GDP in 2016. The difference is due mainly to a further decline in natural resource wealth.

These are the consolidated debt securities and loans of the public sector. General government gross debt is 63 percent and 37 percent in Finland and Norway, respectively.

This analysis is based on Cabezon and Henn (2018) and the long-term projections are subject to considerable uncertainty. In the absence of any adjustment, Norway’s non-oil primary fiscal deficit in year 50 would surpass 10 percent of GDP. In an infinite horizon version of the model that takes this large primary deficit into account beyond year 50, intertemporal net worth is –82 percent of GDP.

In Australia, the balance sheets are projected by each level of government independently.

Maturities decline significantly when moving from the general government to the consolidated public sector level, by about 3 years to 11 years in the United Kingdom in 2016 (Office for Budget Responsibility 2017a), about 1½ years to below 3 years in the United States in 2014 (Greenwood and others 2014), and almost 3 years to 6 years in Japan at the end of 2017.

Central government data includes social security funds. When these data are not available, they were proxied with, by order of preference, data on central government excluding social security funds, or budgetary central government. The specific choice for each country is available in Annex Table 1.2.3.

Chapter 2 of the GFSM 2014 presents more detailed guidance on institutional unit and sector classification. Details on specific adjustments are captured in the country specific database documentation.

Chapter 7 of the GFSM 2014 presents definitions and valuation methods for each type/instrument of assets and liabilities. The valuation methods that were used in the analysis for specific types of assets and liabilities are summarized below.

This corresponds to the “above-the-line” approach for calculating net lending or borrowing. Since double-entry recording is used for recording all flows in the GFSM 2014 conceptual framework, that balancing item can also be calculated from “below-the-line,” as the difference between the net acquisition of financial assets and the net incurrence of liabilities.

Because of source data limitations, data for public corporations were in most cases limited to those corporations under control of the central government. Data for public corporations under the control of state and local governments were generally not available in aggregate formats.

These eliminations do not change the balancing items of the balance sheet or the statement of operations, but have an influence on the levels of assets and liabilities or revenue and expense reported by the public sector.

Because of the lack of source data, for some countries the PSBS database presents debt securities at valuations other than market, such as nominal or face value (the latter corresponding to the amount to be paid at maturity).

Should market price changes be included, the price will represent a market value.

Fair value is akin to market value. International Financial Reporting Standard 13 defines it as the price that would be received to sell an asset or paid to transfer a liability in an orderly transaction between market participants at the measurement date (an exit price).

As discussed in Bova and others 2013, because of their nature, location, or attached regulations, they may not be sellable and therefore are excluded from the governments’ balance sheets, or valued at one unit of local currency, even though they may create revenue (for example, tourism receipts) and generate maintenance costs.

A detailed description of the sources and methods of the capital stock and investment database can be found at https://www.imf.org/external/np/fad/publicinvestment/pdf/csupdate_jan17.pdf.

PSBS database estimates differ from the World Bank’s The Changing Wealth of Nations 2018 because the World Bank uses a discount rate of 4 percent for all countries and constant value data for prices, whereas the PSBS database uses different vintages of commodity-specific prices from WEO reports.

Expense for social security benefits payable to households are instead picked up in the intertemporal analysis, as they are embodied in future expenditure.

If no actuarial projections are available, they are built using current year (2015) pension spending of those pension schemes in percent of GDP, and they assume it grows in line with the old age dependency ratio (this is consistent with a naïve projection model under which the benefit ratio and pension eligibility remain constant over time).

This difference of 1 percentage point corresponds to the average observed in the advanced economies over the past 25 years (Escolano 2010; Turner and Spinelli 2012).

In some cases, adjustments are made to align projections with authorities’ existing estimates.

For reasons of cross-country comparability, we analyze total assets excluding land and natural resources and total liabilities excluding pension liabilities.

Note that we use one index for assets and liabilities to indicate that we look at a balance sheet item’s volatility relative to all other balance sheet items, be they assets or liabilities.

Countries included in the analysis include Austria, Belgium, Bulgaria, Croatia, Cyprus, Czech Republic, Denmark, Estonia, Finland, France, Germany, Hungary, Ireland, Italy, Latvia, Lithuania, Luxembourg, Malta, the Netherlands, Norway, Poland, Portugal, Romania, the Slovak Republic, Slovenia, Spain, Sweden, and the United Kingdom.

A more nuanced definition of liquidity would also account for the ability of the government to sell the assets without an adverse impact on price. Data limitations at present preclude reporting on this basis.

Where available, foreign-exchange-linked assets and liabilities are included.

Advanced economies in the sample are Australia, Belgium, Canada, Czech Republic, Cyprus, Denmark, Finland, France, Germany, Greece, Iceland, Ireland, Italy, Japan, Luxembourg, the Netherlands, New Zealand, Norway, the Slovak Republic, Slovenia, Spain, Sweden, Switzerland, the United Kingdom, and the United States. Emerging markets included in the sample are Croatia, Hungary, India, Indonesia, Kazakhstan, Kyrgyz Republic, Moldova, Poland, and South Africa.

Long-term bond yields used here are as defined by Thomson Reuters Datastream. These are the yields for 10-year bonds for most countries, excluding Belgium and Cyprus (6 years), Kazakhstan (up to 5 years), Kyrgyz Republic and Moldova (2 years), Slovenia (11 years), and the United Kingdom and the United States (20 years).

Assets excluding land and other natural resources, and liabilities excluding pension liabilities, both for reasons of cross-country comparability.

Foreign buyers of emerging market sovereign debt in particular may also care about public foreign exchange assets. Ideally these would be included in the set of control variables, but they are not because of data limitations.

We use a standard set of control variables from Bernardini and Forni (2017). This specification does not account for possible collinearity or nonlinear relations between the control variables and balance sheet strength dummies.

The data are annual and cover the following countries: Australia, Canada, Czech Republic, Denmark, Finland, France, Germany, Greece, Italy, Japan, Republic of Korea, Netherlands, Norway, Spain, Sweden, the United Kingdom, and the United States.

The World Inequality Database is available at https://wid.world/. The database is based on a collaborative effort of various scholars. Some of the series are based on official statistics, while others are estimates based on different data sources available (fiscal data, survey data, and national accounts). Although the public sector database introduced in this Fiscal Monitor is more detailed and comprehensive than the World Inequality Database, the local projections model is data intensive and requires long time series. Therefore, the empirical estimations in this annex use data from World Inequality Database, which goes back to 1970 (compared with the PSBS database that covers the period from 2000). The correlations between (changes in) net financial worth in the World Inequality Database and the PSBS database introduced in this Fiscal Monitor are positive and significant at the 1 percent level.

    Other Resources Citing This Publication