Marginalia Vol.1: Assessing the Risk of Government Debt

Marginalia are brief notes written in the margin of a book or article. John Vandenbrink, Deputy Dean of GLOBIS University, will share his notes on topics that have caught his attention in his readings on finance, business, and economics.

Media reports on government health could be misleading

Articles in the media pointing worryingly to the debt levels of the leading governments of the world began appearing during the global financial crisis in 2008, and they flourished as eyes turned to Greece this past year. Today the articles continue, often focusing on the increases in debt stemming from economic stimulus spending. Concerns about the abilities of governments to make good on their financial promises used to be limited to the developing countries, but no longer. Today the discussion includes once perceived stalwarts like the United States, the United Kingdom, and of course Japan, which has the world’s largest national debt. Are these articles to be believed? Is the analysis sound?

A standard feature in these articles is the ratio of national debt to Gross Domestic Product. This is the measure most often used to gauge the risk of debt. As a former banker and now instructor of corporate finance, I have often felt this ratio to be simplistic, so I was intrigued when I came across a blog posting by Richard Posner questioning its use. Posner, who is a US judge and commentator on economic affairs, suggests that the proper way to understand fiscal solvency is to compare assets and liabilities. This seemed inherently logical to me, though of course it would be far more complicated to calculate. In corporate finance, ratio analysis might suffice to value companies in a single homogeneous industry, but if the companies in question were as diverse in nature as the nations of the world, then forecasting cash flows to compare assets and liabilities would be mandatory. Why not also for nations? Pursuing Posner’s suggestion leads to a persuasive methodology and to calculations implying that the debt levels of advanced nations are more ominous than is acknowledged in all but the most worrisome articles.

The title of Posner’s posting is provocative: “Is the Federal Government Broke?” (The Becker-Posner Blog, August 29, 2010.) Posner, it turns out, received his inspiration from a report by Arnaud Marés of Morgan Stanley. Marés’ title is also provocative: “Ask Not Whether Governments Will Default, but How?” (Morgan Stanley Research, Sovereign Subjects, August 25, 2010.) Posner reviews Marés’ analysis and concludes that financial conditions of the United States and other western governments are more serious than acknowledged, that voters dependent on entitlements and government services may suffer major cuts, and that the resulting political pressures may threaten bondholders, leaving them facing the possibility of “outright default or default through inflation.” Marés concludes that the governments of advanced economies are, on the basis of current policies, “deep in negative equity;” that taxes will rise and government entitlements and services will be cut; and that pressures will come to bear on bondholders, who may face, while not likely outright default, rather, “financial oppression” through unanticipated inflation, taxation, regulatory requirements, revocation of contractual protections, and other such reductions in bond value, for which there are ample examples in history, he contends.

Posner’s and Marés’ conclusions seem extreme. Could they be right? The method of analysis that Marés puts forward, and that Posner affirms, for analyzing the solvency of governments coincides with the precepts in corporate finance for analyzing the solvency of corporations: capture the right cash flows, measure value correctly, and look forward, not backward. Marés’ report is for clients and has not been widely distributed. His logic is worth summarizing.

Debt/GDP ratio is too simplistic

What is wrong with the ratio of debt to GDP? Let’s first look at the numbers. Marés provides values for the ratio in 2009: United States (federal government) 53%, Spain 53%, Ireland 64%, United Kingdom 68%, Germany 73%, Portugal 77%, France 78%, Greece 115%, and Italy 116%. His figures generally correspond to those for 2009 in the “World Factbook,” an often-cited source prepared by the Central Intelligence Agency of the United States. Marés does not provide a ratio for Japan, but the “World Factbook” reports 189%.

The first problem with these numbers is, as Marés points out, that debt is gross indebtedness. This fails to adjust for debt used to purchase financial assets, of which there have been huge amounts issued in the wake of the global financial crisis. Using net debt in the numerator would improve the validity of the measure. For example, it would reduce UK’s ratio from 68% to 51%, according to Marés.

This is not the only problem in the numerator. As Marés also points out, missing are real economic liabilities that are not captured in national accounts. He cites as an example the contractual claims of UK civil servants, which in 2008 were valued by the UK Government Actuary Department at 58% of GDP—certainly a material amount relative to the UK’s reported debt ratio of 68% of GDP (2009).

There is also a problem in the denominator of the debt/GDP ratio. The cash flow that matters is tax revenue, not GDP. Switching denominators, Marés provides figures that place the US at 358% debt/revenue, well above Greece at 312%, with Spain the lowest at 153%. But switching denominators is not enough to create a useful indicator of solvency. As economist Gary Becker points out in his comment on Posner’s posting, the United States looks terrible on this measure because the federal government’s revenue in 2009 was only 15% of GDP, while in major European countries the ratio ranged from 34% (Ireland) to 48% (France). Low tax collections could mean that Americans are intolerant of high taxes, which is a negative factor for solvency. Or low tax collections could mean that the United States has much greater capacity to raise taxes than other countries, which is a positive factor for solvency. Marés suggests that analytically the best indicator would be debt divided by the maximum percentage of tax revenues that the government could extract. This is, of course, very hard to arrive at, and he does not provide estimates.

Still another problem with debt/GDP is that it reflects only accumulated debt, ignoring the additional debt that governments will need to issue if they do not change their policies radically. The most obvious examples are the so-called structural deficits that many governments are running year after year. Also omitted are the growing obligations deriving from the aging of the populations, such as pensions, social services, and medical care. These obligations will be enormous. Marés believes them to be the most significant oversight stemming from the debt/GDP view of solvency. Marés’ contention is irrefutable. To gauge the solvency of any advanced nation today, it seems essential to take into account the magnitude of such huge future cash outflows.

Is there an alternative measure?

If we are not to use debt/GDP, then how do we analyze the solvency of governments? This brings us to the comparison of assets and liabilities. Marés creates a “stylized balance sheet” with assets on the left and liabilities and “people’s equity” on the right. (See Figures 1 and 2.) First among assets is the government’s power to tax, which equals the discounted present value of the government’s future stream of tax revenues. Other assets are categorized as real (buildings, equipment, etc.) or financial (equity stakes in companies, loans, cash). First among liabilities are “social liabilities,” which equal the discounted present value of the government’s promised expenditures for defense, education, health and the like. The second liability is current gross debt. The plug figure in this balance sheet is people’s equity. If assets exceed liabilities, then people’s equity is positive, which means that the government can lower taxes without reducing commitments to beneficiaries of public services or to creditors. If assets are less than liabilities, people’s equity is negative, which means that the country faces some combination of increased taxes, reduced services, or losses for creditors.

Marés’ “stylized balance sheet” is analogous to balance sheets in normal accounting, but it is different in important ways. First, social liabilities are not legally enforceable, contractual obligations. They include also payments that are expected by the population but that may be reduced by legislation. Thus, “default” in Marés’ discussion has an expanded meaning, referring not only to defaults on payments to creditors but also including reductions of expected benefits to citizens. Also, “people’s equity” is not, as the metaphor of the balance sheet implies, the value of the ownership claims of shareholders after liabilities have been satisfied. The concept of equity does not transfer to governments. There are no owners of private property in this construct. As explained above, people’s equity is a measure of resources available to the nation given the current values of potential future revenues and promised expenditures, of real and financial assets currently held, and of debt currently owed.

The value of Marés’ construct is to remind us of the complexity of the problem of national debt. It forces analysis of all cash flows in to the government and all cash flows out. It forces capturing all real economic liabilities, regardless of the fact that accountants do not require governments to list them. It requires looking forward to capture the impact of demographic change. Admittedly, the calculations are very hard to arrive at, but that is the reality of the problem. Because there are many assumptions involved, results derived from calculating people’s equity can only be indicative. However, that is an improvement over the alternative.

What are the results? Marés did the calculations for the United States and eight European countries. (He did not provide details.) He used discount rates equal to nominal growth rate plus 1% for each country, which seems plausible. His results showed negative equity for every country. The values ranged from nearly three times GDP (Italy) to nearly sixteen times (Greece). The United States was in the middle at eight times. Hence the dismal conclusions reached by Marés and Posner. Are these gentlemen outliers? Perhaps, but their method of analysis stands up better than the ratio analysis commonly employed.