Debt reality versus perceptions

Macro-credit risk and associated volatility in 2023

Just as April showers bring May flowers, 2022’s historic rise in real interest rates likely has sown 2023’s credit traumas.  Amid high Uncertainty, this may be one of the few clear predictions one can make for 2023: that credit events – defaults, liquidity scares and debt-driven inflation – likely will be the dominant source of macro volatility.  Importantly, this is not a forecast for another Global Financial Crisis (GFC), or even for credit to suffer significant losses as an asset class.  Indeed, credit may be the best place to seek alpha (excess returns) as the difference between borrowers who pay at higher real rates and those who do not will be stark.

Rather, I expect 2022’s real interest rate rise to unleash a Schumpeterian destruction of unsustainable borrowers that were only able to survive amid the last decade’s aberration of sustained negative real rates.  Just as lions cull the old and infirm from herds of wildebeest, the historic jump in real interest rates is about to do the same in the global economy.  The fantasy earnings of cryptoland were the first to succumb, but the real victims of a shakeout that has been a decade in the making are yet to come.  As discussed in Clash of the Themes, the global economy is undergoing a dramatic change in sources and means of production as it moves from globalization to automated Localization.

But debt perceptions are not always reality in a world of structural change.  Countries, businesses and individuals dependent on globalization are like aging wildebeest, with their slowing potential income growth leaving them vulnerable to the sharp teeth of higher real interest rates.  This is particularly true of those that incurred high debts to fund either capital tied to increasingly obsolete dispersed global supply chains or, worse, to extend consumption as growth slowed in the last decade.  In contrast, the economies and entities that are driving real interest rates sustainably higher through accelerated demand to invest in high-return capex to localize automated production will continue to thunder across the economic plain, keeping the herd alive.

I provide a guide to the macro-credit risks in the global economy: which ones are most likely to manifest in 2023 and what form they are most likely to take.  Because these events are likely to be the main source of volatility this year, my analysis will be of interest not just to credit managers, but also investors in equities, rates, FX, commodities, and cross-asset portfolios.  For credit specialists, while I do not tackle single-name credits (beyond sovereigns), I do identify corporate credit markets that may be at greatest risk of significant individual credit events.

Key Observations:

• 2022’s historic rise in real interest rates will hasten the unsustainability of many globalisation-dependent economies amid ongoing Localization.  I expect a sharp uptick in credit events among both companies and countries to be the key source of macro volatility this year.

• Credit beta likely will underperform as a result. But the differentiation of globalization’s “zombies” from Localization’s winners implies credit may offer the best opportunities for alpha.

• I identify 2023’s most significant macro-credit risks through an original framework that (1) accounts for total economy sustainability (including private and domestic debt), (2) isolates economies that face immediate payments difficulties; and (3) predicts the type of macro-credit event: default, selective default, inflation, or hyperinflation.

• Five economies make my watchlist for the greatest potential to drive broad macro volatility in 2023: (1) Brazil; (2) Turkey; (3) Colombia; (4) Hungary; and (5) Egypt. Attractive expression of these risks are in currency, rates and relative-value credit, and possibly short equity exposures.

• Regionally, Latin America and CEEMEA are at the highest risk of contagion from macro-credit risks. FX likely offers the best risk-reward expression of these risks.

• Within the dollar bloc countries, mortgage debt in Canada and New Zealand looks more risky than Australia where household deleveraging is more advanced. As with other regional risks, cross-FX likely is its best expression, and long AUDNZD looks particularly compelling.

• Corporate debt in Brazil, Canada, Chile, France, Hungary, and Turkey appears at greater relative risk for significant credit events and may be a differentiating source of alpha for single-name credit managers given relatively benign pricing in some of these markets.

• Markets may have overestimated sustainability risks in the UK and euro area, while spreads on offshore debt in Nigeria and South Africa appear too wide.

• Aside from obvious war-related risks in Russian debt, portfolio insurance should take into consideration serious tail risks from China and Japan. Implied correlations via FX options and swaptions may offer sufficiently cheap leverage to hedge.

Who, when & how?

Sorting the macro-credit wheat from the chaff is a three-part question: who is most at risk of a credit event, when will risks become acute, and how (or in what form) will they manifest?  I start by introducing and explaining a total-economy sustainability framework for identification of macro-credit events that I developed and favor over traditional debt sustainability analyses (DSAs).  Unsustainability is a necessary but not a sufficient condition, so I then develop a set of indicators to isolate when concern is likely to transition to panic.  I then add a framework of political economy to determine the form of event given macro-credit events’ element of sovereign choice due to the variety of debtors — households, firms and government — and the ability of the last one to rewrite the rules and pursue inflationary finance.  With the who, when and how answered, I close with my watchlist of most significant risks, a round-up of other risks and opportunities, and a few low-delta, high-impact tail risks to consider.  Readers pressed for time or with a TikTok attention span may want to skip to the final section.

Who: Defining & determining sustainability

Credit events arise from unsustainable debts: those that the borrower – or its creditors – perceive cannot be paid.¹  The most basic definition of sustainability is where the net present value of current and future income equals or exceeds the net present value of current and future expenditures and debt servicing costs.  But because countries, at least theoretically, live forever, the necessary condition for debt sustainability distils to a constant (or falling) ratio of debt to income.  For this to be true, income must grow at least as fast as new borrowing and debt servicing costs combined.  This condition is at the heart of the Debt Sustainability Analysis (DSA) developed by International Monetary Fund and used by both international policy institutions and macro analysts.²

It also is at the heart of the broader definition of macro sustainability that I use.  The IMF’s DSA, consistent with the Fund’s mission to preserve international payments and government finance, focuses exclusively on government and external debt.  In the modern global economy, that focus is too narrow as it excludes both domestic debt and private debt crises, both of which affect a wide range of creditors and macro stability as illustrated by the GFC whose roots lay in unsustainable household liabilities.  Indeed, for most countries private debts are best seen as contingent claims on the government as few governments, democratically elected or not, are willing to stand idly by amid a private-sector debt crisis.

National, or total, debt sustainability starts by recognizing that all debts, public and private, are paid from the same national income stream: GDP.  The flow of national income must meet households’ consumption and debt servicing, firm’s investment and debt servicing, and the taxes of both that pay for government services, investment, and debt servicing.  If debt servicing for any of the three sectors grows too large, either consumption, investment or government spending must fall.  Or new debt must be taken out to pay existing debts, until creditors refuse to lend and a crisis ensues.

This focus on national income also highlights why the world is ripe for widespread credit events.  As discussed in Clash of the Themes, Localization is up-ending global production, undermining the engine of growth for billions of people and dozens of nations that were dependent on the era of Hyperglobalization for growth.  The winners, those economies with the attributes necessary for Localization, are fueling a capex boom that is pushing real interest rates sustainably higher.  Yet the winners’ consequent potential growth gains simultaneously undermine potential in countries whose capital stock and work forces are geared to globalized production.  The combination of rising real rates and falling potential growth for the losers leads to an unsustainable combination, even with relatively low levels of debt.  Unfortunately, in many cases, debts have ballooned as the globalisation-dependent have borrowed to maintain prior levels of consumption or doubled-down on unproductive investment to extend growth.

The approach that I take to national debt sustainability is equivalent to that of the IMF’s long-run payments approach with a couple tweaks to incorporate household and non-financial corporate debt.  First, I segment debt to recognize that private debts (household and corporate) generally pay a spread over the interest rate on government debt.  Second, while the IMF DSA only considers net new borrowing/saving of the government (the primary balance), I account for net new borrowing/saving of the whole economy (the current account balance).³

A couple other notes on debt measures are necessary.  First, consolidated, net government debt matters for payments sustainability, not gross government debt.  If the central bank, a subsidiary of the government, owns a significant share of government debt and remits its profits – interest on the debt – back to the government the payments net to zero.  The same holds for government pensions.  But the debt service to private owners of state-owned enterprises' and local governments' obligations does matter and is included in consolidated net government debt.⁴

Second, for private debts – household and corporate – I use non-financial debt.  Financial obligations (i.e. banks’ liabilities) effectively “double count” debt from a payments perspective: household deposits at a bank (a financial liability) are used by the bank to fund a loan to a business (the final non-financial obligor).  Financial institutions merely act as “passthrough” vehicles for non-financial debt, adding an intermediation fee that I account for in the spread over government yields that I assume for households and non-financial corporates in the sustainability calculation.  Thus, adding financial debt to non-financial debt of the economy would overstate the payments constraint.⁵

This is not to say that financial debts are unimportant.  Indeed, weak banks and associated liquidity crunches often are the precipitant of debt crises.  Bank weakness is exacerbated by reliance on non-resident deposits that can flee and is less subject to government financial repression in a financial crisis.  Hence, I explicitly consider financial sector soundness and sources of borrowing in the next section, but it is not a part of the payments-based sustainability calculation.

Figure 1 compares debts by sector and measures of sustainability across a panel of 57 countries.  For each economy, total debts of each sector, stacked columns, are presented as shares of GDP on the left axis, while two measures of sustainability (also expressed as GDP shares) are shown as markers read on the right axis.  I have presented both the standard government-only DSA (orange discs) and my own total-economy measure of sustainability (green diamonds).  Both are expressed as the deviation of annual savings from that required for a stable debt-to-GDP ratio; i.e. a positive deviation (net saving) means that the country’s debt is shrinking relative GDP by that percentage per year; a negative deviation (net borrowing) means the country’s debt is growing by that magnitude in percentage points of GDP per year.⁶ Table A1 in the Appendix replicates these data – with a country-code key – for those seeking precise estimates.

A couple observations are immediately clear from the graph and illustrate how debt perceptions may be quite different from reality.  First, the level of debt is not a good measure of sustainability, even when expressed relative to national income as in Figure 1.  Indeed, most of the “below the line” (unsustainable) markers, both for government and national sustainability metrics, are in countries with average or below ratios of debt to GDP.

As explained in Clash of the Themes – and mathematically shown in the equations of footnotes 1 and 2 – potential growth rates and real interest rates are the key drivers of debt sustainability. Countries that grow fast or benefit from low real interest rates can sustain larger debt burdens. Japan and Singapore, countries with the largest debt burdens in Figure 1, at 360 and 343⁷ percent of GDP, respectively, have total debts more than double the median 171 percent of GDP yet appear sustainable due to low real interest rates (and fast growth in Singapore’s case).  Conversely, Argentina’s national sustainability (-35%) and government sustainability (-29%) are literally off the bottom of the chart (like Sri Lanka, another country already in default) despite a debt-to-GDP ratio that is little more than a quarter of Japan’s.  The difference is astronomical real interest rates.

The second observation is that government debt sustainability, the focus of traditional DSA frameworks (and most media and analysts), can be extraordinarily misleading in assessing credit risks.  Japan, again, presents a positive example: according to traditional metrics, Japan’s government debt sustainability has been “unsustainable” for nearly three decades, beginning when national unsustainability in the early 1990s precipitated Japan’s financial crisis (Figure 2).  Japan’s lost decade repaired private savings – private sector non-financial debt fell by 50% of GDP from 1994 to 2004 – and, combined with high trust in government, allowed Japanese residents to fund the nation’s extraordinary public debt for more than 30 years.  While there has been a worrying dip in the total sustainability metric since Covid, it still suggests that Japan can continue to fund itself if real interest rates remain low.  This presents its own Being is believing risks in Japanese policy, as discussed in the final section, but illustrates well the potential for traditional DSAs to overstate debt risks.

But the more worrying problem with traditional DSA metrics is their tendency to vastly understate macro-credit risks.  Indeed, the most significantly underappreciated macro risk of 2023 serves as a clear example: Latin America’s largest and the world’s 12th largest economy, Brazil, is on a deeply unsustainable debt path despite government finances that look only moderately concerning.  Brazil’s national savings deficit relative to a stable debt path is a whopping 11.8% of GDP, the third highest among countries in Figure 1 that are not in default.  In contrast, its government savings shortfall is a worrying but still manageable 2.2% of GDP.  While Brazil’s national sustainability has been worse – briefly in 2015 and most seriously in the late 1990s, notably both periods of trauma for Brazilian and Latin American asset prices – circumstances were quite different from today (Figure 3).

In both cases the rise in real interest rates was local, caused by a deterioration in market confidence in Brazilian policy credibility that steadfast orthodoxy fixed.  Now, Localization and the associated US capex boom is driving a global rise in real interest rates – that also threatens Brazilian policy credibility – while simultaneously undermining Brazil’s globalization-dependent growth and repayment capacity.  As a result, President Lula faces a very different global and domestic economic landscape than in his first term.  Then, global real interest rates were falling amid the “dot.com” bust and Brazil’s potential growth was 3.0%; now, the US capex boom is in full stride while Brazil’s trend growth has fallen to an anæmic 1.2%.

“If something cannot go on forever, it will stop.” – Herb Stein, aka Stein’s Law

When does Stein’s Law bite?

Unsustainable, as illustrated by Japanese government debt, is a vague concept.  Almost by definition any debt is unsustainable when it is incurred, or the borrower would not need it.  The key question for credit analysis is when borrowers (and their creditors) have misjudged repayment capacity to the extent that a reorganization or repudiation of those obligations is both necessary and imminent.  I.e. when does unsustainable mean stop?

There are four ways that debt goes bad: (1) overestimation of returns on the investment the debt was used to fund (or, relatedly, mismanagement); (2) liquidity events; (3) fraud or misallocation of funds; and (4) a change in repayment terms or income.

Fundamental credit analyses like DSAs, including the national payments approach I use, aim to identify the first case, but they are imprecise and generally unhelpful in assessing when a credit event will happen.  Liquidity crises can precipitate bankruptcies in sound credits, but more typically initiate one in a known unsustainable credit when investors’ worries turn to fear.  Predicting the turning point from worry to fear is the difficulty.  Frauds often accompany macro credit events – e.g. subprime fraud within the GFC – but are rarely their cause and often are not even acknowledged until after the fact as Theranos and, allegedly FTX, illustrated.

One of the central themes of Clash of the Themes is Localization’s ongoing historic change in global production which is radically changing income and payment terms with its differential effects on growth rates and sustained boost to real rates.  But until Localization and its effects become conventional wisdom, both debtors and creditors can “whistle past the graveyard”, blissfully unaware of the changes in repayment capacity of the former.  That is, until they are confronted with the hard reality of repayment, which is the focus of this piece.  The historic rise in real interest rates in 2022 is likely to force that confrontation at the weakest links of the global debt structure in 2023.  Once it does, Being is believing can turn worry to fear with shocking rapidity and with the potential to spread from debtor to debtor as panicked markets look for like credits that may be next.

Confrontations with reality

Confrontation with reality can make Stein’s Law – that which is unsustainable will stop – bite in three ways.  The first is the “hard” reality of interest rate passthrough to both the stock of debt and its servicing: how quickly does it impair borrowers’ ability to meet payments or to reduce creditors’ willingness to lend?  The second is the “indirect” stresses on the financial system that create payments difficulties or liquidity events when financial conditions tighten.  The third is Being is believing effects: shifts in market perceptions of debt sustainability that become self-fulfilling.  For selected countries, Figure 4 presents a table of metrics related to each type of confrontation with reality, sorted by national unsustainability (column 2); the full panel of countries is available in Appendix Figure A2.  For each metric, cells within the column are color coded from deep red for the least sustainable to dark green for the most sustainable.

Hard realities

Columns 3-7 of Figure 4 present five ways that the sharp rise in real interest rates in 2022 can force a “hard confrontation” with reality as they measure the speed with which rates passthrough (or can) to the stock of debt, debt service or repayment ability.  Passthrough is immediate for floating rate and maturing, rolled debt.  Higher proportions of debt with floating rates (column 3) or maturing in the next year (column 4) thus hasten servicing unsustainability directly.  Both the stock of debt to be repaid and the costs of servicing it can increase rapidly with currency depreciation when the debt is denominated in a foreign currency.  Official FX reserves can be used to either pay foreign currency obligations or slow currency depreciation, but reserves can be encumbered by other policy priorities or by hidden forward books as in 1990s South Africa and present-day Turkey, so columns 5 and 6 report both gross and net (of reserves) foreign currency debt.  Note, importantly, that columns 3-6 refer only to the outstanding stock of government and government guaranteed debt.  Private debt may worsen the picture for some countries.

Rapid increases in debt servicing costs can impair repayment ability and swift changes in the stock of debt may lead to a “sudden stop” in new financing by lowering perceived creditworthiness.  But more subtly, passthrough into servicing costs also increases the borrowing requirement of the debtor further hastening unsustainability. For instance, Figure 5 presents a history of quarterly changes in US national savings/borrowing need due to changes in real interest rates, after accounting for the speed with which they passthrough to servicing costs.  Consistent with 2022 being the worst year on record for bond returns, the additional draw on national savings is one of the highest since World War II.  Column 7 of Figure 4 estimates the additional savings (or borrowing) needed to meet the new servicing costs on the existing stock of national debt – government and private – for 2023.  In many countries, the estimate exceeds annual potential income growth.

Indirect realities

Financial system weaknesses also can bring forward unsustainability as an impaired financial system gives debtors fewer degrees of freedom to manage short-term payments difficulties, leading to possible liquidity crises that trigger credit events.  Columns 8 and 9 of Figure 4 present two measures of financial system frailty: the share of loans in the banking system that are nonperforming (NPLs), and net foreign lending of the banking system.  A banking system that is struggling with impaired assets is less able to extend new credit to domestic borrowers that experience difficulty.  Similarly, a banking system that is reliant on foreign banks for funding (negative net international lending) is less stable than one reliant on domestic deposits.  Domestic banks more at risk of a run or sudden stop from foreign depositors or lenders.

Being is believing triggers

Credit events often begin with a simple shift in beliefs: Being is believing effects.  The trigger for the change may be the hard and indirect measures listed above even when they do not force a payments problem.  But perceptions also can shift with changes in commonly monitored metrics of sustainability. These metrics also can form a narrative channel for contagion to economies with similar profiles (e.g. “The Fragile Five”) that spread amid market fears when one country has a credit event.

The final seven columns of Figure 4 list several metrics that are commonly monitored in the press and analyst research as sources of credit trauma.  As discussed in the previous section, while I think total economy sustainability is a more appropriate approach, the standard IMF, government-only model for debt sustainability is reported in column 10.  Current account deficits (column 11) are more readily available – especially to the press – than DSAs and play the villain in every macroeconomics textbook, so are a common trigger of macro sustainability fear.  While one-year rollover risk (column 4) is more pertinent, once debt fears begin to fester, investors begin to look further ahead for signs of trouble, so two-year rollover risk (column 12) begins to come into focus.  Foreign ownership share (column 13) is not a malignancy per se, but is a clear channel for Being is believing contagion.  Domestic creditors are less spooked by foreign credit events, but international investors, particularly if their portfolio has just suffered a loss in a “similar” country, are much more likely to pull funding.

The perceptions of market peers also form a feedback loop for Being is believing fear, accelerating unsustainability incidents. Agency ratings, while produced as advance indicators of credit events, have a sad history of lagging market pricing and have spawned a multitude of “predictive ratings” models from analysts as a result. In fact, both agency ratings and market pricing appear to contain information about future credit events but also both demonstrate herding behavior that is more consistent with coincident adjustment than early warning.⁸ Column 14 of Figure 4 presents the three main credit rating agencies’ ratings for each country and cells are shaded to indicate the outlook for revision (at least one positive outlook is shaded green, red for at least one negative outlook).  Column 15 offers a market alternative, the default probability implied by 5-year credit default swap spreads, albeit on foreign currency debt only.  While the usefulness of predictive macro credit models is unproven, column 16 presents Refinitiv’s macro variable-based StarMine one-year default probability.

Big picture observations

I will turn to focus on country-specific risks in the final section, but it is worth taking a moment to glean several macro insights from Figure 4 and the patterns revealed by its color coding.

First, as illustrated by the shading applied to the country names, Latin America and CEEMEA⁹ stand out as worrying and thus particularly prone to contagion once a first domino falls.  Perhaps surprising, given that the economy’s coincident debt-banking crisis was never fully solved but instead “smoothed” by European Central Bank policies that now are coming under greater stress, euro area countries seem to be only modestly worrying.  Even the worst euro area credit in Figure 4, Greece, appears safe for the immediate future given its near perfect “green strip” of hard payments metrics.

Second, as illustrated by Greece, the inclusion of hard payments risks can radically change the perceived risk profile of a country.  The United Kingdom – only months ago roiling global markets with concern over its government debt sustainability – is one of the best-looking countries on the table in terms of near-term constraints.  Indeed, only four countries have a smaller share of government debt falling due within the next year and only five are better within the next two years.  The UK may have a long-run sustainability problem, but it has little near-term payments risk.

The reverse also is true and reveals potential “hidden” sustainability problems.  Turkey appears sustainable on both national and government-only measures due solely to negative real interest rates engineered by strong-arming of the central and commercial banks, and adroit use of geopolitics to eke out new funding.  Yet high shares of debt that are floating rate (15%), FX denominated (58-67%) or due in 2023 (25%) and 2024 (20%) highlight how quickly Turkey can go from “getting by” to default. Similarly, even countries with solid finances like Czech may be vulnerable to contagion given large rollover needs in 2023.

Figure 4 also illustrates some clear differences between my broader approach to debt sustainability and market pricing or agency ratings.  Everyone agrees on the worst credits: the two defaulted countries in the panel, Sri Lanka and Argentina, have the egregious national unsustainability and exhibit among the poorest metrics in each column of Figure 4.  But there are significant differences elsewhere.

Latin America looks far worse under my framework than either market pricing or ratings agencies indicate.  CDS markets assign 3-7 times the default risk to Egypt as to Colombia, Brazil, Chile, and Mexico, while ratings agencies rate the last two investment grade and the former a notch or two above Egypt.  Egypt clearly is among the most acute macro credit risks with the 10th worst national sustainability and terrible metrics for rollover risk and interest rate passthrough.  But Colombia, Brazil and Chile have even worse national sustainability, while Mexico is only moderately better, and Brazil’s near-term risk metrics are worrying.

Indeed, comparing the two worst regions for sustainability under my total debt framework, Latin America and CEEMEA, markets seem to have a positive bias towards the former while ratings agencies’ bias bends to the latter.  Hungary, Romania and South Africa all have higher CDS-implied default rates than any Latin American country other than the defaulted Argentina, yet have similar or better long-run sustainability scores and, with the exception of Hungary, generally better near-term risk metrics.  Conversely, ratings agencies appear to unduly reward CEEMEA countries despite their absolute level of risk.

Yet markets and ratings agencies also are picking up risks that DSAs, even one based on a total debt framework, miss.  As noted above, Turkey’s management of its real interest rates allows it to appear sustainable on both a national and government-only basis even with high debt and low savings.  But its ability to continue to manage real interest rates is not sustainable and high near-term repayments and foreign currency debt imply that risks around its debt are high, something correctly priced into 5-year CDS spreads and reflected in its ratings only a notch or two above distress.  Similarly, markets rightly assign a high probability of default to Russia due to its war in Ukraine and associated sanctions, despite its ample payment capacity and low debt.

How: Credit event type and the political economy of choice

There is one more joist we need to add to the framework before focusing on specific risks for 2023: the sources and political economy of macro-credit events.  Sovereigns – the clue is in the name – are different from private creditors as they can rewrite the domestic rules for their own debts and those of their citizens, they can legislate purchases of debt by domestic entities (banks, pensions), and most importantly, they can issue the currency in which domestic debts, public and private are paid.  Some sovereigns also have sufficient geopolitical influence to bend the rules on debt issued in foreign jurisdictions or currencies, and to extort new official lending.

Thus, a sovereign faced with acutely unsustainable national debts, in theory, has a wide range of possible types of credit event to choose from, not only for its own debts but for those of its citizens.  For its own debts it can choose default, selective default, restructuring, financial repression, inflation, or hyperinflation.  For its citizens, firms and banks, it can choose to allow systemic defaults, restructure debts through legislative fiat, use financial repression or inflation to reshuffle losses among its citizenry, or bail out borrowers selectively or completely (potentially creating its own debt problems in the process).

Because the sovereign’s choice of type of credit event largely determines who among debtors and creditors takes the loss, it is a political rather than economic choice.  As a result, sovereign choice can seem unpredictable or irrational from a “national interests” perspective.  But the choice usually is rational from a political perspective.  Sovereigns – even in autocracies – are beholden to their power bases: general citizenry; ethnic, religious, tribal, and political factions among them; elites’ social and political cliques or hierarchies; bureaucracies; and business interests.  The economic incentives of a government’s power base(s) may not align with either the national interest or rival factions, but is likely to be reflected in the government’s choice.

Hierarchies of choice: who will take the loss?

Predicting the type of credit event thus requires an analysis of which domestic sector it is likely to arise in and what political incentives will drive the government’s choice of resolution.  Determination involves a five-step process: (1) Is a specific economic sector – government, corporate or household –the source of unsustainability or is it jointly created?  (2) Are there constraints on sovereign’s choice of resolution?  (3) If so, which choices are available?  (4) From that menu, whose economic loss is least politically costly to the government?  And (5), is the government sufficiently stable to enforce its choice?

The first few questions are financial and economic analysis, while the last two are questions of political economy.  In both cases, full answers required detailed, country-by-country analysis that is both beyond my scope and expertise, or data availability in a 57-country study.  But there are common macro indicators available across countries than can narrow the probable source and type of credit event an unsustainably indebted country may face.  My selection of macro credit risks in the next section is based on these higher-level analyses, but I close this section with some strategic advice on further diligence for readers with interest or portfolio exposures in specific countries.

Finding the source of unsustainability or weakest link

Clues to the likely source of a macro-credit event can be gleaned from sectoral debt statistics, indicators of the capital efficiency with which debt is being used, and some common sense.  The size and growth rate of debt in each sector are available for most of the countries in my panel and offer prima facia evidence of sector-specific imbalances.  But recall from the first section that sustainability is a function of debt, real interest rates and real income growth, the latter two of which are more difficult to break down sectorally due to externalities between them.  Further, if the underlying source is the national growth model, unsustainable debt accumulation could occur in any sector depending on social and cultural choices.

My solution to this problem is to compare the relative size and growth of sectoral debt with capital efficiency of debt in the economy.  The relative size and growth rate of debt in each sector can give some indication of the source of the unsustainability or at least the likely weakest link.  The capital efficiency of debt helps to identify an underlying malady in the economy.  Capital that grows faster than debt implies productive use of the debt that at least partially pays for itself.  Conversely, debt that outpaces capital growth implies it is being used for unsustainable consumption or wasted on non-productive capital.  Comparing the relative rates of sectoral debt accumulation to total-economy productivity of debt then gives us a rough idea of whether it is specific sectors are driving the unsustainability or the whole economy.

For the same countries as Figure 4, again ordered by national unsustainability, Figure 6 pulls together all the relevant metrics for the sectoral analysis. Each sector’s debt stock and 10- and 5-year debt growth are displayed in columns 3-11. The 10-year and 5-year capital efficiency of debt is estimated in columns 12 and 13.¹⁰

The US serves as an interesting, illustrative example since I claimed in both Solved: Drivers of the dollar cycle and Clash of the Themes that its unheralded investment boom is a key driver of the rise in global real interest rates.  The last two columns of Figure 6 show that the growth of US debt has outpaced capital by nearly 25 percentage points of GDP in the last decade and almost 18 in just the last five years.  By my calculations the capital-to-income ratio was flat from 2012 to 2017 but rose by just over 3 percentage points since (not shown).  A sectoral breakdown of US debt shows clearly where problem lies.  Corporate debt to GDP compares favorably with other countries, albeit has grown somewhat faster in the US.  Household’s debt remains elevated versus other countries but has fallen continuously over the last decade.  In contrast, the 23 and 18 percentage point gains in US government debt to GDP over 10 and 5 years, respectively, nearly perfectly match the excess growth of debt over capital.

More pertinent to the focus of this piece – 2023 macro-credit risks – are some of the countries highlighted in Figure 4 as potential acute credit concerns.  Consistent with my underlying thesis around the disruptive effects of Localization, the underlying problem in the most worrying countries appears to be one of growth model with both households and governments, to varying degrees, using debt to maintain prior, unsustainable rates of consumption growth.  Brazil, Chile, Colombia, and South Africa have among the worst capital efficiency numbers in Figure 6, while Mexico’s numbers are more middling to poor.  In all cases, the primary driver is an estimated drop in capital-to-income, most severely in South Africa and Brazil where it fell 13 and 12 percentage points, respectively, in the last decade, most of which has occurred in just the last five years.  The trends in capital accumulation in these countries mirror significant declines in their potential growth rates over the same period, pointing to a problem of growth model.

The sectoral breakdown of debt accumulation highlights the differing ways each economy has dealt with the collapse in potential growth that have, except in the case of Mexico, exacerbated debt unsustainability.  With capital falling, new debt was used to maintain the unsustainable consumption growth these countries experienced in the Hyperglobalization era.  In Chile and Colombia, both governments and households have ramped up debt from low levels, while in South Africa the government has sought to cushion the fall in expected living standards.  Brazil’s debt expansion has been broad based, and combined with the retrenchment in its capital stock, has put it on one of the least sustainable paths.  It is perhaps no surprise that in all these countries popular discontent with government has risen.  Only in Mexico has the government of President López Obrador been able to take a more pragmatic path.

Figure 6 also flags worrying trends in a handful of other countries that, while likely not acute given their respective short-run metrics in Figure 4, nonetheless bear monitoring.  Unlike the economies discussed above, in France, Japan, and most frightfully China, it is wasteful capital investment that appears to be driving worsening debts.  In all three countries capital-to-GDP ratios are rising, yet France, Japan and especially China display among the worst estimates for capital efficiency on Figure 6.  Looking at the sectoral breakdown of debt accumulation, while government plays a large role, corporates, the sector most responsible for investment is a primary driver of debt growth, alarmingly so in France.  Although total sustainability and the short-term risk metrics for these countries indicate none are a near-term macro risk, the implied poor investment of their corporate sectors suggest that they may face significant single-name corporate credit events – as has already begun in China – with the rise in real rates.

New Zealand illustrates a potential shortcoming of macro proxies relative to detailed, bottom-up financial analysis.  Its capital efficiency of debt is positive, but the accumulation of debt in the household sector – mostly mortgage debt – is among the worst in the 57-country.  The actual capital efficiency of this debt is questionable given that housing is a non-traded good and New Zealand’s long-run unsustainability is driven by its international borrowing, as illustrated by New Zealand’s 7.7 percent of GDP current account deficit and massive international bank borrowing (Figure 4).  Short-run risk indicators in Figure 4 are not flashing red, but the effect of rising real rates on homeowners likely will slow New Zealand’s economy in relative terms, and combined with its large international borrowing requirement, put downward pressure on NZD.

The political economy of choice

The source of a national credit event may place constraints on the sovereign’s choice of solution.  For instance, financial repression and restructuring by fiat are more difficult for many, perhaps millions of individual debtors than for the sovereign itself.  But in cases where the sovereign debt is the source of trouble or weakest link – which may occur because the sovereign has bailed out private sector debtors first – the sovereign still may face constraints.  For example, a government that has issued only foreign currency debt in foreign jurisdictions has no more choice than a typical private creditor: pay, default or restructure.  Similarly, a government with only inflation-indexed domestic debt cannot use inflation or hyperinflation to erase its debt.  Remaining choices are likely to be constrained by complex political realities.  As noted above, there is no substitute for detailed, bottom-up country-by-country diligence, but there are several broad macro indicators that can help us narrow the probabilities around sovereign choice.  Figure 7 presents 13 metrics that influence sovereign choice at a high level.

The clearest political bifurcation for any government is foreign versus domestic creditors.  The former neither vote in a democracy nor (typically) form a core source of support for a ruling autocracy or oligopoly.  As a result, foreign creditors will nearly always be de facto if not de jure subordinated creditors to domestic debtors.¹¹ Columns 3-6 of Figure 7 display four variables related to foreign creditor status that influence a sovereign’s choice in a credit event.  Shading is from the perspective of a foreign creditor with dark red indicating a greater likelihood of being affected by default.

Foreign creditors are exposed to default or restructuring losses in two ways: selective default on debt issued in foreign jurisdictions – typically foreign owned, and well proxied by FX debt – or losses on domestic debt that is owned by foreigners.  A high share of gross foreign currency debt (column 3) incentivizes a sovereign to selectively default on foreign debt while protecting its own citizens from loss.  Note that while net FX debt (column 4) is better used to calculate payments sustainability, once a sovereign chooses default it is strategically advantaged by defaulting on all FX debt and hoarding its official reserves.  A high foreign ownership share of overall debt issuance (column 5) instead encourages a sovereign to default on all debt as it again disproportionately saddles foreign creditors with losses.  A potential constraint on the sovereign’s incentive to inflict losses on foreigners is the extent to which it requires ongoing funding from them, hence a large current account deficit serves as a disincentive to default on foreign creditors unless no other options are available.

If a sovereign cannot shift pain to foreigners, it can still shield domestic debtors and potentially moderate the pain for domestic creditors by electing to inflate away the real value of debts through monetary expansion.  However, this option is often misunderstood as a single option when sovereigns in fact have two distinct choices: higher inflation and hyperinflation.

Higher inflation will erode the principal value of existing debt, particularly if it is longer maturity.  But it helps little if debts are short maturity or if ongoing borrowing needs remain significant.  In the latter cases, only hyperinflation can solve the debtor’s sustainability problems.  If the debts are short maturity the higher rate of inflation does not have a chance to work its compound magic on the principal due, and any refinancing – or required new borrowing – will face higher nominal and (likely) real interest rates as inflation expectations rise and markets charge a risk premium for increased inflation uncertainty.  Repeated attempts to raise inflation will engender exponential Being is believing effects as market expectations race to keep ahead of the government and a hyperinflation ensues.  The hyperinflation will obliterate existing debt values and will only end once the government budget deficit is brough to zero and new guarantees of central bank independence are issued.

Columns 7-11 of Figure 7 present several metrics of constraints and motivations related to sovereigns’ choice of inflation as a solution, with shading again from the perspective of a foreign (FX-hedged) creditor: darker red implies a higher chance of default, while darker green implies higher chance of inflation/hyperinflation.

A sovereign’s choice for inflation or hyperinflation will be crucially limited by its prior issuance choices.  For instance, a sovereign gains little benefit from inflation if its debts are issued in another currency (Figure 7, column 3).  Similarly, a high share of floating-rate debt (column 7), undermines a sovereign’s ability to use inflation to provide payments relief, and a high share of index-linked debt (column 8) removes any benefit, including principal relief, from either inflation or hyperinflation.

For a sovereign that is unconstrained by its prior issuance decisions, the age distribution of its citizens plays an important role in the politics of its choice of inflation versus default or restructuring. In the economics of the life-cycle, the young typically are debtors, the middle-aged build their savings while rearing children, and the old live off the investments of their youth. The young, as debtors, benefit from inflation, whereas the middle-aged suffer most from its frictional costs and the old bear the greatest “credit” loss from inflation.¹² Thus, median voter age (Figure 7, column 9) and the ratio of older-to-younger voters (column 10) play an important role in influencing government choice, specifically between domestic default, inflation and hyperinflation.  Inflation is an attractive choice in Nigeria because its population is the youngest of countries on Figure 7; for Japan, a country where those over 45 years of age outnumber those under 35 by more than three times, inflation would be a very unpopular sovereign choice.

Finally, the government’s refinancing need will influence its choice for inflation.  A large primary deficit (column 11) that requires new financing is more likely to push a government towards hyperinflation as default will shut it off from credit markets and higher inflation can quickly devolve to hyperinflation if the government cannot cut its real borrowing.

Sovereigns also have the option of financial repression. Because financial repression typically targets domestic savers,¹³ it has a strong political odor of class warfare.  Pension and retirement assets (Figure 7, column 12) are an attractive pool of savings for sovereigns because they are generally immobile and defined by government legislation.  Elites’ wealth, in contrast, is a wildcard for financial repression.  Although elites are a tempting political target in a democracy, they also tend to form the governing class (or have a disproportionate influence upon them) and their wealth often is both less transparent and more mobile.  As a rough proxy for ability to tax or financial repress elites, I include a metric of corruption (column 13), where more corrupt societies are presumed to be less able or willing to confiscate elites’ assets.  Again, columns 12 and 13 are shaded to indicate higher likelihood of default as dark red, with higher likelihoods of financial repression dark green.

The last two columns of Figure 7 present another consideration: the political stability necessary to enforce or stick to a sovereign choice.  Any macro-credit event is going to involve significant domestic economic hardship that can initiate a negative feedback loop between the public and potentially evolving sovereign choices.  The existing government may be forced to backtrack, may fall, or could even be removed.  But each change in sovereign choice likely constrains subsequent choices, increasing the likelihood of default as the only option.  Column 14 reports the average level of five different World Bank indicators of political stability and rule of law, and column 15 reports the five-year change to indicate the recent direction of travel.

As noted above, macro indicators are a blunt instrument and detailed, bottom-up political and financial analysis are necessary to appropriately handicap specific macro-credit risks.  As flagged regarding France, Japan and China detailed single-name corporate credit analysis likely will bear fruit.  In New Zealand – and similarly Australia and Canada (see Figure A3 in the Appendix) – the focus of micro financial analysis likely will be more fruitful in housing finance.

But detailed political diligence by country also is necessary, especially against the backdrop of the Politics of Rage.  The political economy of sovereign choice is likely to be strongly influenced by consideration of the widening trust gap between elites and average citizens discussed in Clash of the Themes.  Do governments attempt to “buy off” disaffected citizenry by dispersing losses elsewhere through populist policies?  Or do they shore up their support with elites by saddling their political rivals and their angry masses with the losses?

The Politics of Rage also is likely to influence the path to credit events.  Populist forces remain strong across advanced economies and are growing in some.  In recent years centrist parties have erred on the side of increased deficit spending to split populist voters.  This was enabled in large part, especially in G10 economies, by low or even negative real interest rates.  Pulling back from that strategy, or from previous promises, will be difficult and politically risky, but in a world of sustained higher real interest rates it is a fiscal inevitability.

The watchlist

With the full framework in hand, I now turn to the specific macro-credit risks that I believe have the greatest potential to roil global markets this year. I follow that with a broader analysis of macro-credit risks that markets misjudge in my view.  Unsurprisingly, emerging markets dominate both lists, but advanced economies also offer risks and opportunities.  The order of my “watchlist” is intentional and signals my subjective judgement regarding the likelihood of the event causing significant market volatility in 2023 (albeit not necessarily an actual credit event).

1. Brazil, the 5th least sustainable economy, is at the top of the watchlist because its weakness in nearly every dimension leaves it highly vulnerable and, as the 12th largest economy in the world, its potential to create broad macro volatility is high.  Reviewing the metrics in Figures 4, 6 and 7, the only dimension in which Brazil looks good is its low share of FX-denominated and foreign owned debt.

Rapid passthrough of its long-run unsustainability – roughly a third of its government debt matures in 2023 and another third in 2024 – will rapidly sour Brazil’s finances, economy or both.  The implied 1.8% of GDP increase in needed borrowing (or saving), a third of which is the government’s, offsets the government’s primary balance even before new President Lula’s ill-timed spending plans have a chance to be passed.  The large rollover need also creates ample opportunity for failed auctions given a market already nervous over political instability in a divided country, erosion of fiscal rules and the 2024 retirement of central bank President Campos Neto.  The Brazilian financial system’s large dependence on foreign bank lending (47% of GDP) creates both potential for auction failures to cascade into a liquidity crisis and a — two-way —conduit for regional and global contagion.

If Brazil can avoid contagion from abroad or an own goal from unstable politics and poorly conceived fiscal policy in 2023, its near-term payments needs are manageable.  But an eventual credit event appears inevitable.  Its longer-term sustainability — a savings gap of 11.8 percent of GDP — is too large and broad based (government and private) to be addressed without serious economic reforms.  The falling capital stock and potential growth rate point to a broken growth model and an inability outgrow problems as Brazil did in Lula’s first presidency.  Yet, its young but rapidly aging and undereducated population calls into question realistic capacity to reform, even in the medium term.  Low levels of gross foreign debt (and foreign debt ownership) limit options for selective default, while moderately high indexation (26%) and floating debt (8%) cut incentives for inflation, though a still young population suggests the possibility cannot be excluded.  Few pension assets and pervasive corruption limit the government’s ability to enact financial repression, but high domestic debt ownership and wealth inequality may make domestic default an attractive option, particularly for left-wing government.

Relative to my assessment, markets appear to price in too much likelihood of an offshore default and too little of an onshore default or inflationary event.  The 16% probability of default priced into 5-year CDS seems too high.  As a non-credit specialist, I am hesitant to suggest receiving outright Brazilian CDS – which reference USD debt – especially given that they would likely trade wider in a domestic crisis.  But they likely represent good relative value versus either Brazilian corporate credits or Colombian sovereign CDS (see below) given the lower probability of an external default by Brazil.

Meanwhile it is hard to disentangle how much of the 2-percentage point rise in onshore real rates in the last two years reflects beta to the rise in global rates and how much reflects increased concern about repayment.  But in contrast to 2016, there is little indication in the flat real and nominal yield curves of significant concern over either an onshore default or an inflation event over the next few years.  This suggests to me that belly of the curve, specifically the 2-5 year sector, is quite vulnerable.  But long USDBRL is probably the best expression for Brazilian macro-credit risks.  High carry is an impediment, but currency weakness is likely to follow from disappointing growth, erosion of fiscal prudence and possible difficulty rolling onshore debt.  Further, with real rates in Brazil already unsustainable, growth poor and gross FX debt tiny, it is questionable how firmly the BCB would stand in the way of a renewed bout of BRL depreciation if it was triggered by debt sustainability concerns.

2. Turkey looks sustainable both under my national sustainability calculation and a standard government-only DSA, but as noted this is entirely due to financial repression and coercion of onshore interest rates.  This has been enabled by adroit use of geopolitical wrangling to obtain financing and heavy leaning on Turkish banks.  But upcoming elections, significant rollover needs in both 2023 and 2024, and a big, sustained current account deficit (5.7% of GDP) suggest that recent stability in TRY is coming to an end.  But more troublingly, degrees of freedom in policy choice are almost certainly closing behind the scenes.  High shares of FX-denominated (67% gross), floating (15%) and inflation-indexed (8%) debt mean that 90% of government debt is immune to rampant inflation.  Financial repression, too, likely is reaching its limits. Turkey has negligible pension assets, high corruption and pervasive multi-currency accounts, and years of pressure on banks likely have pushed them to their limits (and saddled them with large contingent liabilities like FX-protected TRY deposits).  If President Erdogan fails in his bid for re-election or runs out of geopolitical rabbits to pull from his fez, few choices beyond default will remain.

The election in June is a natural breakpoint after which incentives to delay the inevitable will abate.  TRY sell-offs generally are swift, but the build-up of pressure for this event may lead to an even more precipitous drop.  The ability to manage interest rates, especially cross-currency swap rates is likely to deteriorate sharply, making payers attractive.  5-year CDS already price in nearly a one-in-four chance of default on FX debt, which ultimately may prove attractive given the declining choice of sovereign options, but uncertainty over President Erdogan’s ability to extract new official funding or extend financial repression suggest better risk-reward in short TRY, paying TRY cross-currency swaps, or short Turkish financials (equity and credit).

3. Colombia has the least sustainable debt of any non-defaulted economy, but it is at little risk of a near-term sui generis credit event due to low rollover risk and low passthrough of real rates into 2023 (and 2024) payments.  Colombia makes the near-term watch list because it lives in a troubled neighborhood – Latin American is dense in unsustainable economies with declining growth prospects and unstable politics – and has large external vulnerabilities that make it highly susceptible to contagion.  While little of Colombia’s debt comes due in 2023, a large share is FX-denominated (45% gross, 24% net) or foreign owned (41%) and Colombia has a large current account deficit (5.1% of GDP).  As with many of its Latin America peers, political and policy uncertainty add to Colombia’s vulnerabilities.

While Colombia’s sustainability gap of 13.1% of GDP is huge and it shares the growth-model problem of its Latin American peers, unlike Brazil, the maturity structure of Colombian government debt is supportive of consolidation through savings: 60% matures beyond five years and a third is more than 10 years’ maturity.  This suggests that the peso will bear the brunt of any contagion, though the loss of Colombia’s sole investment grade rating, Moody’s, which is more likely than not, may add to the downside of Colombian government bonds.  Colombian 5-year CDS spreads also appear skinny relative to Brazil, given that its default risks, while lower overall, are skewed towards external default and Brazil’s are skewed to internal default or inflation.

4. Hungary faces one of the largest hits to savings (2.8% of GDP) from rates passthrough of any country in 2023 due to the rapid deterioration of its financial conditions in late 2022.  Combined with large external borrowing dependency – a current account deficit of 6.7% of GDP, 80% foreign ownership of its debt and net international bank borrowing of 61% of GDP – new borrowing is likely to increase pressure on both the forint and HUF rates despite the carry attraction of the latter.  The potential to fall below investment grade – it is two notches above junk and on watch from S&P – will add pressure given high international participation in its debt market.

However, Hungary is far from default, has low rollover risk in coming years and relatively low levels of debt overall.  CDS pricing of a 5-year default risk of nearly 13% seems rich — especially versus its Latin American peers — given that Hungarian potential growth remains robust, its capital stock and capital efficiency are high, and its finances are manageable.  However, a rapid increase in nonfinancial corporate borrowing in recent years, late-2022’s sharp tightening of financial conditions, and a European economic slowdown point to potential problems for Hungarian corporate debt.

5. Egypt is a well acknowledged credit risk and rightly so given its deeply unsustainable finances, the 10th worst in my panel, and a need to roll half of sovereign debt outstanding this year.  The passthrough of the high rollover implies an additional borrowing (or saving) need of 2.7% of GDP in a country already running a current account deficit of 3.6% of GDP.  A 45% devaluation of the Egyptian pound both exacerbates debt sustainability – 40% of gross government debt and 32% of net is FX denominated – and signals a potential change in central bank policy.  So far, Egypt has been able to rely on official creditors and GCC patrons to provide hard-currency support but it may be reaching the end of its tether.  With a young population and little inflation-indexed debt, a turn towards inflation financing may become attractive if Egypt can get through 2023’s massive refinancing need without a credit event.  Given the 2023 risk, the 41% default probability priced into 5-year CDS appears fair and may even understate the risk.  Better opportunities lie in USDEGP upside and the small index-linked government bond market, which should outperform if Egypt avoids default and turns instead to inflation financing.

Other opportunities, risks and counter-consensus views

Several regional or cultural clusterings stand out as particularly exposed to contagion, especially if any of the watchlist countries fall into trouble.  In Latin America, as noted, Colombia is likely to avoid macro-credit volatility if Brazil’s problems or other regional volatility doesn’t trigger it.  But it is not the only Latin American country at risk from contagion.  Mexico and Chile both are high on the unsustainability list and suffer from the broader LatAm malady of failing growth models in a world of Localization.  But their near-term finances and savings profiles should shield them from serious concerns over their near-term sustainability.  Rather, Mexico’s need to roll almost a quarter of government debt and poor bank balance sheets make it vulnerable to regional or global contagion, especially given its status as the “liquid” emerging market in times of panic.  Chile’s problems are likely to manifest in Chilean corporate credit given their borrowing binge of the last decade and poor metrics for capital efficiency of debt.

Eastern Europe, too, looks vulnerable to contagion from Turkey, Hungary and the ongoing war in Ukraine.  Romania’s huge national and government borrowing requirements (current account and primary deficits of 8.4 and 4.7 percent of GDP, respectively), and large share of FX-denominated debt (67% gross, 37% net) leave it particularly vulnerable to regional contagion.  Poland also may be exposed to problems in Hungary given the high correlation of its markets, an upcoming national election, and significant current account and primary balance deficits of its own (4.0 and 2.9 percent of GDP, respectively).  Perhaps surprisingly, even the normally staid Czech Republic may be exposed to regional volatility.  Though not a real credit risk, the rapid deterioration of its current account and fiscal balances resulting from the Russia-Ukraine war may leave it and markets unprepared to roll nearly two-thirds of government debt this year.  Debt is small relative to GDP (27%) and reserves are large (44%), but the novelty of Czech’s needs may add unusual volatility to Czech rates and the koruna.

Though not a regional bloc, the dollar bloc economies all have attracted attention for their high levels of household – primarily mortgage – debt.  New Zealand just misses the top 10 most unsustainable economies on a national basis, while Australia and Canada only register as fiscally unsustainable given the energy-price driven swing to surplus in their respective current accounts (see Figure A4 in the Appendix).  New Zealand’s yawning current account deficit (-7.7% of GDP) leaves the NZD more exposed than its anglo-dollar peers, but Appendix Figure A3 reveals more interesting differences within the bloc.  Household debt accumulation has been a primary driver of debt in the last decade for all three economies (columns 9-11), but in Australia household retrenchment began five years ago (column 11) and in Canada corporate debt uptake has outpaced households.  The capital efficiency of debt also has been poorer in the two larger dollar bloc economies than in New Zealand.  As noted previously, greater capital efficiency in a non-traded sector like housing is not necessarily helpful to an international debtor.  But the more rapid accumulation of corporate debt in Canada combined with poorer capital efficiency of debt suggests that Canadian corporate debt bears greater scrutiny amid higher real servicing costs.

There also are some debtors on which markets may be overly pessimistic.  A rapidly abandoned fiscal expansion plan in the United Kingdom led to an historic surge in gilt yields that roiled global bond markets and led to the quick exit of Prime Minster Truss.  The UK does have the third worst national sustainability of any advanced economy due to relative high combined government and household debt, and large current account deficit.  But the UK’s government-only DSA (-0.5% of GDP) barely registers concern.  More importantly, the maturity structure of UK government debt is truly enviable: just 17% is due within the next two years, only a third within five years and almost half is longer than ten years’ maturity.  This suggests that L’affaire de Truss had more to do with politics – massive pro-cyclical spending plans in more-indebted countries like the US received little protest – and the duration exposure of UK pensions than realistic debt sustainability concerns.

Concern over euro area debts, too, is similarly ameliorated by consideration of payment profiles.  Long-run debt sustainability of the EMU’s most troubled countries, Greece, Italy – and perhaps a surprise to many – France remains a concern, both on a national and government-only basis, as shown in Figure 4.  But none of these countries have near-term payment concerns (note: data on Greek debt characteristics in Figure 4 do not include official loans that represent more than half of total government debt).  On a national basis, even Portugal and Spain now appear sustainable.  Indeed, the largest credit risks in the euro area in 2023 may reside with French corporates.  The deep red shading of France’s corporate debt accretion and poor capital efficiency stand out in Figure 6.

Despite Ghana’s recent default, debt of the two African countries analyzed may offer upside.  CDS markets price a 50% chance of default on Nigerian foreign debt and a 16% chance in South Africa.  Both seem excessive given the relative lack of triggers for a credit event.  In Nigeria’s case the primary near-term risk is associated with the February election and end-May change in government.  The key driver of fiscal – and national – unsustainability is a national fuel subsidy paid in local currency that has exploded with naira depreciation.  All political parties have pledged to end the subsidy and the 2023 budget does not include funds for it.  Even if it is only phased out (as in the ruling APC’s plan) or worst case, extended, inflationary financing is far more likely than default.  Nigeria’s 2023 maturing foreign-currency debt and coupon payments represent just 8% of debt outstanding, or 3% of GDP.  Amid high oil prices a foreign-debt default makes little strategic sense for an incoming government, whereas already high inflation, no indexed debt and the youngest population among the 57 countries in my panel all augur for inflationary finance.  Similarly, South Africa has little 2023 rollover risk, a tiny primary budget deficit and suffers mostly from a large current account deficit that is more likely to close through rand depreciation.

Tail risks to mention

There also are some important tail risk macro-credit events to mention.  All three that I consider – the implications of state collapse in Russia and sustainability crises in China or Japan – are textbook examples of Uncertainty: non-quantifiable risks.  But none are far-fetched scenarios and thus should be considered in designing portfolio insurance.

The war in Ukraine brings obvious macro-credit risks that I have elided here.  I excluded Ukraine from the analysis because it is impossible to even estimate its potential growth, payments needs or real interest rates amid an ongoing war of unknown length.  Russia and other former Soviet states are included in the study and come with war-related tail risks.  While Kyrgyzstan and to a lesser extent Uzbekistan have long-run sustainability questions, there are few near-term payments triggers for either.  Russia itself has little debt and massive payments surpluses from oil revenue, despite the expense of war.  Beyond its sanctions-related technical default on coupon payments last year, default is a question of will not ability.  But as highlighted in Clash of the Themes, state stability is a concern for Russia.  A state collapse or regime change in Russia could, potentially for an extended period, impair Russia’s ability to pay its obligations.  Just as pertinently, given their dependence on Russian trade and infrastructure, its land-locked client states could face the same problem.  Kyrgyzstan, with a 12.5% of GDP current account deficit is particularly worrying in this regard.

Two low-probability, but far higher impact risks to the global economy are sustainability events in China and Japan.  Predictions of debt crises in both economies have been widow makers for at least two decades and there are good reasons not to expect that to change.  Both economies are massive external creditors that consistently self-fund their extraordinary debts.  But when leverage is large, small changes in gross returns or costs can lead to massive net payments changes.  And in both countries, subtle changes imply substantial changes to sustainability.

For China, the risk revolves around its potential growth rate.  My calculations assume a rate just above 6% based on my Kalman filter estimation of its pre-Covid trend.  But years of overinvestment – China’s capital efficiency of debt is the worst of any economy in Figure 6 – workforce growth that is turning negative, a possible “middle-income growth trap”, and a broad-based retreat from China by multinationals incentivized by Localization, political pressure and concerns over intellectual property, call into question whether China’s actual potential growth may be half or even a third of that rate.  If one assumes 2% potential growth for China, national sustainability swings from a savings plus of 11.5% of GDP to a deficit of 1.6% of GDP.  Time will tell, but this is less a tail risk than a potential reality not yet revealed to us.

A realistic tail risk is that China’s long-anticipated debt crisis already has begun.  A debt crisis in China is unlikely to involve an acute financial and economic meltdown and instead look more like Japan’s “Lost Decade” because China owns its own debt, has capital controls and the government directs much of the economy.¹⁴  If one assumes that China’s potential growth is zero in this scenario, the national sustainability gap widens -7.7% of GDP.  While China’s mechanisms of state control would buy it time, the primary adjustment would be a substantial fall in investment.  Zero-growth China already would be a major shock but if achieved through sharply reduced infrastructure and residential investment the negative shock growth in globalization- (and thus China-) dependent economies would rapidly bring forward unsustainability of most of the countries in the top half of Figure 4.

For Japan, debt sustainability is a question of Being is believing risk that could change instantly:  How strongly Japanese residents believe in the safety of their own government’s debt and in Bank of Japan (BoJ) policy?  At 360%, Japan’s total debt is the largest in the world as a share of national income, yet it remains sustainable due to negative real interest rates.  Without the BoJ’s yield-curve control program, real interest rates in Japan would be higher, though by how much is anyone’s guess.  Net new JGB issuance is almost wholly absorbed by the BoJ.  Yet Japanese residents continue to hold their substantial wealth in unusually high proportions – by comparison to peers – of their own country’s government bonds and bank deposits backed by the BoJ.

This is a Being is believing trap.  Any loss of faith by Japanese residents in either the BoJ or in Japanese government debt would rapidly undermine Japan’s debt sustainability.  At 1% real JGB yields, national sustainability would fall to -3.4% of GDP; at 2% real rates, -7%; and at 3% real rates, -10.6%.  Unlike in the Chinese scenario conjectured, a Being is believing crisis would lead to a rapid and calamitous drop in the yen, Japanese asset prices and Japan’s economy with serious repercussions for global financial markets.

It is impossible to quantify the likelihood of these tail risk other than to say they are “low”. But responsible portfolio hedging strategies should consider them because they are realistic and have massive potential impact. Long-dated FX-option and swaption-based correlation trades with high leverage and limited downside likely offer the best means of insurance.¹⁵

Appendix: Full data tables

1. Credit events also can arise from unwillingness to pay even sustainable debts, a topic that I will elide here.

2. This sustainability condition, applied to government debt, is PB = [(r – g)/(1 + g)]*D, where PB is the primary balance of the government (revenue less non-interest spending), D is its stock of debt, both as shares of GDP, r is the real interest rate, and g is the potential growth rate of the economy.  Or, in words the net borrowing or lending of the government equals the autonomous rate of change in real government debt given its debt service burden and the sustainable growth of the economy (and taxes).

3. I augment the standard DSA equation to NS = {[(r – g)/(1 + g)]*NGD} + {[(r + c – g)/(1 + g)]*(NHD + NCD)}, where NS is net national savings (rather than the government’s primary balance), the first term on the right, representing the evolution of net government debt, NGD, is identical to the standard DSA, but I add to it the second term representing the evolution of net household debt, NHD, and net nonfinancial corporate debt, NCD.  Government debt pays an average real interest rate of r, while private debts, both household and corporate, are assumed to pay an additional credit spread of c, and the whole economy grows at a potential growth rate of g. is the potential real growth rate, r is the sovereign real interest rate, and c is the average credit spread on private debt.  I use my own calculations for potential growth, five-year government index-linked yields where available for r, and five-year nominal government rates less my own estimate of inflation expectations for countries without a government index linked market, and national BBB/Baa credit spreads where available for c or a composite average of advanced/emerging economy BBB/Baa spreads as applicable.  Please see Clash of the Themes for notes on my calculation of potential growth and expected inflation (“Missingflation” estimates).

4. I use the IMF’s measure of consolidated net government debt.

5. I use the BIS’s measures of non-financial debt by sector for each economy.

6. Note: the calculation assumes that all debt is floating or continuously refinanced at current real interest rates and the economy grows steadily at potential growth.  As a result, the measure overstates the actual net new borrowing/lending of the government.  This difference is addressed in the next section where we consider explicitly shares of floating debt and rollover rates.

7. Singapore’s total debt is, however, overstated relative to other countries as net government debt was not available and hence gross government debt was used instead.  Given Singapore’s massive sovereign wealth funds, net government debt is actually negative.  The point remains as the Figure 1 amply demonstrates.

8. See “On the information content of credit ratings and market-based measures of default risk,” Oleg R. Gredil, et alia, Journal of Financial Economics, vol. 146, no. 1, pp. 172-204, October 2022; and “The stability and accuracy of credit ratings,” Paulo Viegas de Carvalho et alia, manuscript, October 2014.

9. Central and Eastern Europe, Middle East, and Africa.

10. Few countries have reliable, timely measures of capital stocks.  I estimated each country’s desired growth of capital stock and recent deviations from that path using total investment shares of the economy with the help of a few assumptions.  First, I assumed that countries adhere to the “Golden Rule” of growth economic: they maintain a desired capital-to-income (K/Y) ratio, or rate of ratio growth.  In the long-run, that implies countries invest at a rate that replaces depreciation (d) and grows the capital stock at least as fast as the economy grows (potential, g) or faster (g + q) if the society desires and increasing ratio.  Hence, in equilibrium the investment share of GDP should equal (d + g + q)*K/Y.  I assume that a 10-year average is roughly equal to equilibrium and thus any deviation of the investment share from a trailing 10-year average is a desired change in capital accumulation.  To cumulate over five and ten years, I discounted past years’ accumulations geometrically by GDP growth and an assumed 10% annual rate of capital depreciation.

11. There are exceptions to this “rule”: (1) where a debtor is the client state of its creditor; and (2) when factionalism within domestic politics is severe enough that a ruling party may choose to punish its rivals by saddling them with losses rather than foreign creditors it might need to fund itself.

12. Again, there are exceptions: in less developed countries without social safety nets or financial savings mechanisms, the “investments” of the aged typically are in their children who then pay for their retirement, in which case their loss from inflation is lower or even negligible.

13. Though not usually thought of as such, capital controls, Tobin taxes and domestic capital or asset requirements for foreign banks (e.g. for high-quality liquid assets) are forms of financial repression directed at foreigners.

14. Russia is not the only country I suggested may suffer from state instability in Clash of the Themes: the other country was China.  Last month’s Covid-related unrest across the country appeared to put an exclamation point on that warning.  In the event of regime instability, a Chinese debt crisis could be violently disorderly.

15. For illustrative examples, please see “Three Questions: Knight time for vol,” Marvin Barth, Nikolaos Sgouropoulos, and Hitendra Rohra, Barclays Macro Research, 24 February 2020.