* With apologies to John Donne: “Debt be not proud, though some have called thee / Mighty and dreadful, for thou art not so / For those, whom thou think’st, thou dost overthrow / Die not, poor debt, nor yet canst thou kill me.”
We live in a world profoundly addicted to debt-financed consumption.
For most of us, our first car and our first home were financed with debt. We borrowed with intent to repay, and most of us did just that. We were, of course, no richer because we’d just borrowed to buy a house or a car: We had a new asset, exactly offset by a new liability. Our expected future consumption was reduced, not advanced, by this borrowing. While we were realigning our lifestyle to improve the subjective mix (with a nice house and a car), our lifestyle was improved in some ways and reduced in others (fewer restaurants and holidays), with no objective net difference.
Today, many people, companies and countries borrow to fund current consumption, with no evident intent to repay. As it comes due, our debt is something we intend to replace with new (and often larger) debt. We’re not just borrowing from Peter to pay Paul; we’re borrowing a bit more from Peter, to pay Paul … and to finance additional consumption with the difference. How naive of us, as young adults, to have once thought we might never have to pay back the principal!
Greece recently hit a wall, and had to break a lot of promises to its citizens, notably the retirees and prospective retirees from government employment. Iceland’s banks hit their wall a couple of years ago. Many people who were late buyers during the U.S. housing bubble hit a wall and are in default. Italy, Spain, Portugal, Ireland, Illinois, California and New Jersey are all fast-careening toward their respective walls.
The nature of that wall is generally the same: We cannot find a lender willing to lend us more, to pay off our old debts, and so those debts truly come due. Our choice, in each case, is either to reduce our consumption, in order to pay down that debt, or default.
Of course, with each default, the failed borrower suffers damage, not least being a string of broken promises to trusting stakeholders. But the lenders suffer reciprocal damage. While debt is extinguished for some, so too are assets for others. It is in this fashion that wealth is destroyed in a financial crisis.
Is the U.S. the lead junkie in a world addicted to debt-financed consumption? Are we careening toward perhaps the biggest array of sovereign defaults in world history? Time will tell, but the sheer magnitude of global sovereign debt is not reassuring.
Why Are Bond Indexes Capitalization Weighted?
Bond investors are lenders. As creditors, why should we deliberately choose to lend more to those who are most deeply in debt?
Bond indexes are mostly capitalization weighted. Consider Table 1a. Greek debt is nearly three times the debt of Australia, meaning cap-weighted sovereign bond investors have loaned three times as much money to Greece as to Australia. If Greece has three times the debt service capacity of Australia, this should be fine, because Greece is just as able to service its debt—ceteris paribus—as Australia. But Australia has three times the GDP of Greece. Therefore, on this simple measure, Greece has about nine times as much debt, per dollar of GDP, as Australia. If the yields are similar, as they were a year ago, one might reasonably prefer to own more Australian debt than Greek debt.
Consider an efficient markets perspective. In an efficient market, what does it matter if Greece owes more and is less able to service its debt? If Greek debt is more risky than Australian debt, we should garner exactly the right amount of incremental yield to provide the same risk-adjusted expected return for both countries’ debt. However, we see little evidence of this sort of market efficiency.
If we do not believe that prices are always and everywhere correct, then we should be curious about debt levels, as measured against a borrower’s ability to meet their debt obligations. In other work, we’ve examined this very question. We find that a Fundamental Index methodology applied to bonds—weighting companies according to the size of their business, or weighting countries according to the size of their economy—adds considerable value, relative to cap weighting.1 It is not our intent in this paper to explore the correct weighting scheme for bonds in any detail. Rather, we want to examine the debt loads themselves.
Measuring Sovereign Capacity To Service Debt
How might we estimate a country’s ability to produce goods and services—and eventually wealth—that might be accessed for debt service? There is no direct measure. However, we can consider the factors of production in a capitalist economy. Economics literature typically identifies two or three factors of production: capital, labor and sometimes resources (a subsector of capital). We take it one step further by breaking out energy (normally a subsector of resources, but we think it’s large enough to merit its own category).
We have identified four factors that crudely proxy for these factors of production; hence, for a country’s ability to service its debt.
- Capital: GDP is imperfect, equally crediting the creation of consumables (e.g., auto production and car wash services), alongside destruction of wealth (e.g., litigation expenses and wars) and expenditures that do not enhance wealth (e.g., regulatory compliance). Still, it’s the most widely used gauge of the size of an economy.
- Labor: A nation’s population is the simplest gauge.2
- Resources: A nation’s landmass is a very rough gauge of access to resources.3
- Energy: The aggregate energy consumption of a nation is a measure of the energy that goes into production of goods and services. One caveat is that this may be sourced externally, through petroleum imports.
In Figure 1a, the “Bond Cap Weight” column measures the capitalization-weighted exposure of a country’s bond market debt, as a percentage of global sovereign bond issuance, spanning the developed economies of the world.4 These data include local-currency bonds, as well as debt denominated in dollars, euros or other benchmark currencies. As a quality control check, we also include a column labeled “Public Net Debt,” which measures the aggregate 2009 public debt—less gold and foreign currency reserves—as a percentage of the world total, as drawn from the 2010 CIA Fact Book.
The next four columns compare the fundamental scale of these economies, using the above four metrics as proxies for the four factors of production for goods and services. On the far right is the equal-weighted average of the four fundamental weights (or three measures, for those instances where there is no energy consumption data).
We should reflect on what’s missing from this table. We exclude countries with no tradable bond debt. This includes solvent countries like Kuwait, Lichtenstein, Monaco and Saudi Arabia, as well as insolvent countries like Zimbabwe. Based on the CIA World Fact Book data, these countries collectively owe barely 5 percent of net world sovereign debt and 3.6 percent of sovereign bond debt. Wouldn’t it be nice if we could choose to own the debt of Monaco or Saudi Arabia, rather than Greece or Belgium!
We also exclude any debt that is not in the form of publicly traded bond debt. There are several categories, some of which can dwarf the sovereign bond debt.
- Unfunded entitlement programs: The unfunded portions of Social Security and Medicare are vivid examples, as are the unfunded pay-as-we-go pension obligations of Western Europe. These shortfalls are huge in the U.S. and most of Europe. But in Japan, Australia, Sweden, the Netherlands and New Zealand, such programs are largely prefunded.
- Off-balance-sheet debt: A domestic example would be the modest prefunded portion of Social Security and Medicare, in the form of “trust funds,” which own nonmarketable U.S. Treasury Bonds.5 While several countries have replaced these entitlement programs with national defined contribution pension funds, with individuals owning their share of these funds, others pursue a pay-as-you-go approach. Outside the U.S., trust funds for prefunding national entitlement programs are not the norm. In any event, our own “trust funds” don’t come close to fully prefunding for the projected entitlements.
- Government-sponsored entities: GSEs—such as Fannie Mae, Freddie Mac and others of their ilk—are backed by the full faith and credit of the government; hence, by future tax receipts. In the U.S., they’re bigger than our external national debt. Japan has much smaller GSEs; most other developed economies have none of any consequence.
- State and local debt and unfunded pension obligations: These are excluded for the U.S.; they’re roughly half as large as the direct public debt. No other country in the world has as much state and local debt—a consequence of U.S. tax policy that allows local and state debt to remain exempt from federal tax and also allows prospective obligations to remain unfunded.
- Bank-owned sovereign debt: Sovereign debt, owed to banks, is not uncommon in the emerging markets. In the column labeled “Public Net Debt,” which measures the aggregate 2009 public debt—including bank debt, while subtracting gold and foreign currency reserves—we can see that this is not a major “missing link.”
In the U.S., the combination of GSE debt, state and local debt, unfunded pensions and entitlements all add up to just under $60 trillion, roughly 10 times the official U.S. public debt. By contrast, none of these hidden forms of debt, apart from bank debt, is consequential in the emerging markets. We’ll come back to this topic shortly.
Figure 1a color-codes the debt burden for the developed economies, with purple indicating better debt coverage ratios and red indicating possible debt service trouble spots. If any Fundamental weight for a country, as a share of the world economy, exceeds the cap weight by more than 100 percent, it’s flagged in dark purple with white text. If it exceeds the cap weight by at least 25 percent, it’s flagged in pale purple with dark purple text. Reciprocally, if a country’s cap-weighted bond market debt exceeds any Fundamental weight for that country, as a share of the world economy, by more than 100 percent, it’s flagged in dark red with white text. If it exceeds the fundamental weight by at least 25 percent, it’s flagged in pale red with dark red text.
In the developed economies of the world, there’s a lot of red ink in Figure 1a. Many countries carry debt—not even counting often-vast off-balance-sheet debt—which is out of proportion with their scale in the world economy.
Still, there are pockets of discipline. Australia, Poland and Slovakia show no “red” at all, meaning that the sovereign bond debt isn’t 25 percent greater than their economic factors of production, on any of the four metrics. Canada, Finland, New Zealand, Norway, Slovenia and Sweden are each “out of bounds” on only one of the four measures.6
Collectively these “Prudent Nine” comprise less than 4 percent of world sovereign bond debt, and over 6 percent of world GDP, 17 percent of world landmass, 7 percent of world energy use and 8 percent of world capacity for sovereign debt, as approximated by the RAFI weighting methodology, which combines the four previously mentioned factors of production.7 Furthermore, several of the “Prudent Nine” have less hidden debt. For instance, Australia, New Zealand, Norway and Sweden largely prefund their future pension obligations. A cynic might suspect that those who have too much explicit debt will begin to pursue hidden debt, either off-balance-sheet or unfunded entitlements, as was revealed in the case of Greece.
Greece is looking to Germany to save it from the gaping maw of debt. So, let’s consider Germany. Germany has a reputation for prudence and probity in the eurozone. But that’s only true by comparison with the Mediterranean rim; Germany is strained on all four of our measures. Germany has 5 percent of the world GDP (proxying for available capital), 1.6 percent of the population (available labor), 1.1 percent of the land area (available resources) and 3.3 percent of the world’s energy consumption (the energy factor of economic production). Its share of world sovereign bond debt exceeds all four of these measures.
Germany’s capacity for carrying debt—approximated by averaging these four fundamental measures of the scale of Germany’s economy—is 2.7 percent of the world total. This is barely half as large as Germany’s share of world sovereign bond debt. Greece has a RAFI weight of 0.4 percent, about one-third of its sovereign bond cap weight. These German and Greek debt coverage ratios are not dissimilar. The perceptions of German prudence, contrary to this objective evidence, illustrate why we think the debt addiction of the developed markets is so very dangerous.
Worse, Germany has:
- an aging population and a flood of prospective retirees in the coming 20 years
- daunting off-balance-sheet obligations, mostly in the form of pay-as-you-go pensions
- a young and growing immigrant population that was not consulted in creating these long-horizon entitlement programs, and that will not benefit proportionally from these programs
If one were to mark these obligations to market, as if they were prefunded with debt, and compare the total with Germany’s ability to service the debt and future entitlements, it’s possible that Germany’s total debt burden exceeds any credible exit strategy. This would mean that Germany—a bulwark of fiscal prudence in Europe—is probably near-bankrupt, on a mark-to-market basis.
In effect, one might argue that Portugal, Ireland, Italy, Greece and Spain (derisively—and unfairly—characterized as “PIIGS”), are bankrupt states seeking shelter from larger near-bankrupt states. The collective bond debt of PIIGS is 2.6 times its collective RAFI weight in the world economy, which relates to its ability to service debt. That’s an acknowledged problem. Belgium serves as the governance center for the EU, yet its debt burden is near-identical to this figure … as is the ratio for the G-5 in aggregate! Isn’t it a sad irony that the G-5 economies have a debt burden—relative to the scale of these economies based on the four factors of production—as the so-called PIIGS. And yet we have the temerity to label the Mediterranean rim countries the “PIIGS”?!
Finally, it bears mention that the cap-weighted sovereign debt indexes are happy to include nations that carry hefty debt burdens, until the ratings agencies catch up with reality and downgrade their debt. Then, the index providers apparently don’t know what to do with these newly fallen angels. After being downgraded to BB in June, and after the bond prices had cratered, Greece was removed from the developed world sovereign indexes and not added to the emerging markets indexes. As far as we can tell, Greek bonds no longer have a home in the major international fixed-income indexes.
The Emerging Markets Debt Conundrum
Emerging markets debt commands a premium yield. And yet, by objective measures, their debt coverage ratios are far better than the developed markets.
On June 30, the Merrill Lynch Global Emerging Markets Sovereign Plus Index, which spans the dollar-denominated debt of the emerging markets, was priced to yield 6.0 percent. This was 3 percent higher than U.S. 10-year Treasurys. This 3 percent “risk premium” rewards us for bearing the incremental default risk and political risk associated with serving as a lender to the emerging markets. In 2008, President Rafael Correa of Ecuador repudiated his nation’s debt, despite ample financial resources to pay the debt. This kind of disrespect for international law—and for the integrity of a nation’s agreements—prompts investors to fear investing in emerging markets, perhaps particularly emerging markets debt.
How precarious are the debt burdens in the emerging economies? Surprisingly benign! Consider the so-called BRICs.8 As we can see on Figure 1b, they collectively comprise 22 percent of world GDP, and yet have only 5 percent of world bond debt (and, according to the 2010 CIA World Fact Book, net of gold and foreign currency reserves, just 2.5 percent of the world’s total public debt). India and China have issued only local currency debt, which is difficult or impossible for foreign investors to access. India’s debt is held in part by the IMF and/or World Bank and otherwise not traded or investable. Most cap-weighted indexes exclude these two countries, because their debt is not investable.
Even this overstates the debt picture, from a global investor’s perspective: The second column of Figure 1b shows that Chile, China, Hong Kong, Russia and Taiwan have gold reserves, foreign currency reserves and/or investments in the developed economies’ stocks and bonds, amply exceeding their total debt. No wonder, then, that Greater China is on a roll: They’re the bankers; we’re the debt-addled consumers, who can’t stop consuming on borrowed funds!
Similarly, Saudi Arabia, Kuwait, Qatar, the Emirates, as well as tax havens like the Cayman Islands, Monaco and Liechtenstein, all have no net debt. Most such countries, as with China and India, have no bond debt that any foreign investor would be permitted to buy. These “net creditors” would have a significant collective “fundamental weight,” if only there were bonds to buy!
If the BRICs—especially Greater China—are carrying less debt than they can comfortably support (based on their GDP, their population, their resources or their energy consumption), then surely there must be trouble spots in the emerging markets. Otherwise, why should investors demand a substantial risk premium for emerging markets debt?
Indeed, there are some pockets of “red” on Figure 1b: Across all four factors of production, Singapore and Taiwan each have a share of world bond markets rivaling their fundamental economic footprint in the world economy.9 Of course, many investment professionals would consider these to be part of the developed world—belonging in Figure 1a, not Figure 1b. For example, FTSE includes Singapore in the Developed World indexes. But we’re using the United Nations definition of emerging markets; according to the UN, Taiwan and Singapore are emerging markets.
Let’s consider the rest of the emerging markets list. Not one of the other 40 emerging markets in this list—which spans all countries that are included in any of the major EM debt indexes—has as much debt as any of the G-5 countries, whether measured relative to GDP or relative to the RAFI fundamental economic footprint of these countries. The emerging markets are bathed in purple ink in Figure 1b, because in almost all cases, their debt is modest relative to their evident ability to carry debt, based on the four factors of economic production.
The developed markets comprise 62 percent of the world’s GDP and owe 89 percent of the world’s sovereign bond debt (and 92 percent of total world public debt). The emerging markets collectively produce 38 percent of the world’s GDP and owe just over 10 percent of world sovereign bond debt. Do hidden debt and off-balance-sheet debt change this picture? Yes. As with the role of gold and currency reserves, these factors skew the picture in the “wrong” direction.10 In many instances, the developed economies have vast off-balance-sheet debt, while most of the emerging markets have little off-balance-sheet debt, and often have substantial gold or foreign currency reserves.
Given that emerging market stocks are now priced at valuation ratios (price-earnings ratios, price-book ratios, dividend yields) similar to the developed economies, we might wonder why the stocks get a “free pass” on the feared political risk of these markets, while the sovereign debt does not. Similarly, when we saw a “flight to quality” in the fall of 2008 and spring and summer of 2010, why did this imply a shift in investment preferences away from the emerging markets, toward the U.S., Germany and Japan, and not the opposite?
One might reasonably argue that—absent political risk—emerging markets are collectively more creditworthy than U.S. Treasurys. Which invites a provocative question: When will U.S. Treasurys be priced to offer a “risk premium”—a higher yield—than the most stable and solvent of the so-called emerging markets?
Appendix: Debt Burden And GDP Growth
It is beyond the scope of this short paper to explore the wisdom of our surging public debt, though our views on the topic are self-evident. Still, we might pose the question: Which countries have skated through the “global financial crisis” largely unscathed? Again, we might turn to the CIA World Fact Book for some simple evidence.
If we regress 2009 GDP growth against debt burden—defined as the size of a country’s debt relative to the fundamental RAFI scale of its economy—and against the 2008-09 average deficit, we find the results on Figures 2 and 3. The bivariate regression results, across the 75 countries, are as follows:
2009 Growth = 3.33% [t-Stat is 10.2]
- 0.005% x ln (Debt / RAFI Weight) [t-Stat 5.3]R2 = 0.453
- 0.18% x (2009 Fiscal Deficit / GDP) [t-Stat 3.7]
Every 1 percent increase in the ratio of a country’s debt, relative to its RAFI-weighted share of the world economy (proxying for the country’s ability to service its debt), reduced GDP growth in 2009 by 5 basis points (7 basis points in a univariate regression). If the real cost of sovereign debt is 2 percent (i.e., if the yield that the country must pay the bondholders is 2 percent above inflation), then the damage that debt inflicts on GDP growth would appear to be roughly three times as large as this direct cost. The univariate correlation is -49 percent; this result is significant at the 0.1 percent level.
Figure 3 shows that every 1 percent of deficit spending, as a percentage of GDP, reduced a country’s 2009 GDP growth by 18 basis points (22 basis points on a univariate basis). The univariate correlation is -59 percent; this result is also significant at the 0.1 percent level.
Neo-Keynesians will argue that our causality is confused: They would argue that it’s the plunging GDP that triggers additional debt and deficit spending, not the other way around. Causality is difficult to prove in either direction. But, it merits mention that Keynes himself never argued for structural deficits. That seems to be the war cry of the neo-Keynesians. Keynes argued for budget surpluses in most years, affording a nation an opportunity for deficit spending to soften the impact of economic downturns.
While the sample period is only one year and one financial crisis, and therefore must be taken with a grain (or even a shaker-full) of salt, both results are highly statistically significant. However, since we do not have access to data from multiple “global financial crises,” we should perhaps take heed of the implications of this admittedly limited result.
While Figures 2 and 3 examine the economies of the world for one year (2009), Figure 4 examines one economy (the U.S.) for over 50 years. Milton Friedman observed that the true tax rate is the rate of spending: Spending must be covered by current or future taxes, so deficits merely represent deferred taxation. So, how does growth in the private sector economy respond to growth in spending? Badly.
There is a 73 percent correlation between increases in federal spending and decreases in private sector GDP (the gross GDP, less public sector spending). This evidence would suggest that every 1 percent increase in federal outlays—as a percentage of GDP—reduces the private sector GDP by 1.85 percent. Again, the neo-Keynesians will argue that the causality is backward: Plunging private-sector GDP requires soaring expenditures to arrest the damage. Again, causality is difficult to prove, either way. However, the relationship is overwhelming, with a t-Statistic of 3.1.
Figure 5 updates the graph from our 2009 white paper, “The 3-D Hurricane: Deficit, Debt and Demographics.”10 As yet, there has been no material deleveraging in the U.S. economy. We’ve taken a breather on accumulating net new debt, and we’ve transferred some private-sector debt to the government. However, deleveraging has yet to begin in earnest.
Most of us know someone who has taken on debt amounting to several years of income. If it’s for a first home, and our friend’s income is rising quickly, we would not think them foolish to take on that first mortgage. But, if it’s a middle-aged friend with stable income, especially one fast approaching retirement, we would likely think it very unwise for them to take on massive debt. Most of us are unsurprised when these friends encounter serious difficulties: They’ve boosted their consumption lifestyle on borrowed funds. The creditors eventually want to get paid.
Many observers fret that, if we deleverage (indeed, even if we stop running up additional debt), we face a serious recession. They confuse credit-funded consumption with prosperity. Is the entry-level clerk who borrows to buy a Mercedes and a condo, and then finds that he cannot afford the payments, prosperous? Does he have a natural, inalienable right to continue consuming beyond his means?
As a nation, regardless of our decisions to borrow more or to reduce our borrowings, we’ll still be producing as much in goods and services as in the past. We’ll just no longer be consuming goods and services beyond what we produce as a nation. If our lifestyle has been funded in part on debt, then deleveraging will mean a reduced lifestyle for all, but only to the extent that we’ve been consuming more than we were able to produce. That consumption is unsustainable, regardless of our fiscal and monetary policies and regardless of our intentions with regard to future debt.
If we would counsel our overleveraged friends to cut their spending and start whittling down their debt, why should our counsel to nations be any different? Should we be surprised that the economies for creditor nations are soaring, while the debtor nations find their growth crippled by every economic shock?
1 See Arnott, Hsu, Li and Shepherd, “Valuation-Indifferent Indexing for Bonds,” Journal of Portfolio Management, Spring 2010. Just as we damage our returns when we weight stocks according to their popularity—i.e., cap weighting—we also hurt our bond results, if we weight bonds according to the magnitude of a borrower’s debt load.
2 The working age population might be a better gauge. We chose total population because it’s universally available for all countries.
3 We chose to use the square root of landmass, in order to avoid grossly rewarding big, sparsely populated countries like Russia, Australia and Canada, or penalizing small, crowded countries like Luxembourg, Hong Kong and Singapore. For midsize countries like Argentina or Germany, this adjustment makes little difference.
4 Based on the UN definition of developed and emerging economies.
5 One interesting “factoid” is that the 2010 CIA Fact Book shows the U.S. as having far less debt in 2009 than it did in 2007. How’s that? In 2007, the unmarketable debt held in the Social Security, Medicare and other national trust funds were correctly counted as U.S. public debt. In 2009, this $5 trillion debt was excluded. Was there political pressure to make this change? Is there a growing intent to spend the trust funds, rather than to continue even partially prefunding these obligations? We may never know! Either way, for our analyses in this paper, we added the unmarketable Treasury bonds back into the U.S. Bond and Public Debt columns.
6 Interestingly, in each case, the population is the sole outlier; it would appear that its debt is well within bounds on three factors of production: capital, resources and energy.
7 It’s interesting to note that most of these countries also breezed through the “global financial crisis” better than the countries with more debt. They enjoyed average GDP growth in 2009 of 1.7 percent, double that of the G-5 and of the eurozone.
8 We’ve long found this label puzzling: four countries with almost nothing in common but a shared acronym! Even though China shares borders with Russia and India, the three countries have less in common—culturally, economically or legally—than essentially any countries on the developed economies list. Consider it a labeling-cum-marketing coup by Goldman Sachs!
9 Note also that Singapore has a sovereign wealth fund that is larger than its aggregate debt. So, as with Chile, China, Hong Kong, Russia and Taiwan, their net debt is nonexistent.
10 See our Fundamentals white paper, “The 3-D Hurricane: Deficit, Debt and Demographics,” Research Affiliates, November 2009.