Deleveraging (or the lack of it) since the global financial crisis
In the immediate aftermath of the global financial crisis, a great deal of policy discussion centered on how to address leverage across the global financial system. Eight years on, the results can best be described as mixed. There is little evidence of actual deleveraging in developed economies’ aggregate debt ratios, yet memories of painful austerity are still raw. Nevertheless, there have been notable successes—bank leverage fell sharply, from a pre-crisis leverage multiple of 15.5x assets to equity to 9x today in the U.S., and from 24x to 18x in Europe. Household debt levels have also fallen across many developed markets. Taken together, the deleveraging has neutralized some of the more dangerous elements in the system. In contrast to this progress in the developed markets, leverage in emerging economies is rising and excess leverage is becoming a cause for concern (Exhibit 1).
In the U.S. and eurozone, household and corporate leverage have leveled off; in contrast, debt levels in these two sectors in China have risen markedly
EXHIBIT 1: COMPOSITION OF NATIONAL/REGIONAL INDEBTEDNESS (AS % OF GDP)
Source: Bank for International Settlements; J.P. Morgan Asset Management Multi-Asset Solutions; data as of October 7, 2016. Data is market value apart from Chinese government debt as a % of GDP, which is nominal value.
An optimistic read today would be that for the developed world a balance sheet transfer has taken place. The shrinkage in U.S. financial sector and mortgage market debt—offset by a sharp rise in the level of government debt to GDP—has significantly reduced systemic leverage risks, even if it has not alleviated the aggregate debt burden. Arguably, given the dollar’s reserve status, this may be a uniquely sustainable intermediate-term solution for the U.S.
The more pessimistic view is that attempts to reduce leverage via growth or inflation have palpably failed. Instead, the developed markets have fallen back on containing the debt overhang through a combination of austerity and financial repression—or they are merely ignoring it.
The emerging world’s debt has ballooned since the GFC. While it is likely that the debt surge gave global growth a welcome boost, we now face the start of an EM deleveraging cycle. The good news is that EM debt is unlikely to become the major systemic risk that U.S. housing debt proved to be. The bad news is that EM deleveraging will put a further brake on growth just as developed markets appear to be abandoning their deleveraging efforts and accepting the inevitability of a low-growth world.
Leverage today in developed and emerging economies
High global leverage ranks as a significant concern for investors. While we don’t share the more apocalyptic views of the risks associated with high debt levels, we acknowledge that aggregate economic growth could well suffer, and we anticipate some pockets of stress. The fear of elevated leverage tends to be associated with two factors: its longer-run effect on growth and the shorter-run possibility of a credit crisis (arising either from systemic risks connected to the debt itself or from the unintended consequences of policy measures deployed to bring debt levels down). In our view, the impact on growth from both the level and expansion of leverage will have a visible effect on asset returns over our 10- to 15-year horizon. Measures to address debt will likely vary significantly across regions, with emerging economies more likely to resort to familiar methods of deleveraging—inflation, currency adjustment, restructuring—while developed economies may need to explore new tools. However, in most cases, we do not see an elevated risk of a systemic credit crisis forming, as it did in 2008.
To better frame where leverage creates vulnerabilities in the system today, we focus on three elements of debt: its level, rate of growth and location. Recent academic studies have demonstrated that the first two of these factors—the absolute level of debt and its rate of growth—can affect future economic growth through several mechanisms, including the reallocation of capital, the propensity to increase savings and pay down debt, the pulling forward of future demand, etc. The final consideration—the location of debt on the national balance sheet—is crucial for understanding the risk of a systemic crisis and determining which policy tools to deploy. Simply put, the further the debt sits from the financial sector and from household balance sheets, the lower the risk of an acute crisis.
Early and extensive deployment of the sovereign balance sheet from 2008, notably by the U.S. and the UK, facilitated the transfer of private sector debt to the public sector, effectively defusing its explosive power. Private debt service ratios fell sharply as debt levels shrank (Exhibit 2). Yet even as sovereign debt levels soared, ultra-low interest rates kept DM governments’ debt servicing costs well under control (Exhibit 3). Since this transfer has largely neutralized the systemic risk of large DM countries’ aggregate debts, it is questionable how much incentive they have now to address them.
The U.S. private debt service burden fell along with the absolute level of debt itself
EXHIBIT 2: U.S. HOUSEHOLD AND NON-FINANCIAL CORPORATE (NFC) DEBT-TO-GDP AND DEBT SERVICE RATIO (AS % OF GROSS DISPOSABLE INCOME)
Source: Bank for International Settlements, J.P. Morgan Asset Management Multi-Asset Solutions; data as of August 5, 2016. Debt service ratio (DSR) is defined as the ratio of interest payments plus amortizations to income.
Despite a steep increase in government debt post-GFC, accommodative interest rates have kept debt service ratios well contained
EXHIBIT 3: U.S., EUROZONE, UK AND JAPAN SOVEREIGN DEBT-TO-GDP RATIOS AND GOVERNMENT NET INTEREST PAYMENTS (GNIP) (% OF GDP)
Source: Bank for International Settlements, OECD, J.P. Morgan Asset Management Multi-Asset Solutions; data as of August 5, 2016.
So long as investors remain willing to hold DM sovereign debt, we struggle to see an imminent day of reckoning for advanced economies that have shifted debt onto the government balance sheet. The persistence of a global savings glut goes a long way to reinforcing this status quo, as aging savers seek out lower risk assets. So, too, does financial industry regulation that requires ever-greater holdings of “riskless assets”—in what bears more than a passing resemblance to previous episodes of financial repression.
Patterns of leverage differ in emerging economies, as do potential paths of deleveraging. For the most part, the absolute level of debt in emerging economies remains subdued by DM standards. In most cases, debt service burdens are light (Exhibit 4). This may afford some comfort to investors scarred by the Asian financial crisis of the late 1990s and the Latin American debt crisis of the 1980s. Nevertheless, levels of EM private sector debt have risen sharply since the GFC, particularly in the corporate sector, and without an offset through a corresponding fall in interest rates, debt servicing costs for EM companies and households have increased steadily (Exhibit 5).
EM indebtedness is low by DM standards, and debt service burdens are in most cases light
EXHIBIT 4: GOVERNMENT DEBT-TO-GDP AND NET INTEREST PAYMENTS (GNIP) (% OF GDP) FOR CHINA, INDIA, BRAZIL AND RUSSIA
Source: Bank for International Settlements, OECD, J.P. Morgan Asset Management Multi-Asset Solutions; data as of August 5, 2016.
Contrary to the situation in the developed markets, rising debt service ratios have accompanied the increase in private sector EM debt
EXHIBIT 5: HOUSEHOLD AND NON-FINANCIAL CORPORATE DEBT-TO-GDP AND DEBT SERVICE RATIOS FOR BRAZIL, RUSSIA, INDIA AND CHINA (GDP-WEIGHTED AVERAGE)
Source: Bank for International Settlements, J.P. Morgan Asset Management Multi-Asset Solutions; data as of August 5, 2016.
Three qualities of EM leverage suggest that systemic risks are contained: (1) EM governments’ relatively unstressed balance sheets; (2) relatively low levels of hard currency-denominated debt; and (3) the concentration of leverage in the broader corporate sector—not in the financial sector. As such, EM countries will probably experience a more “traditional” deleveraging cycle than DM economies (discussed below). Where the buildup in EM leverage is not uniform across all channels, policymakers may have the choice to deal with leverage by shoring up one component of the national balance sheet at the expense of another.
EM economies: Greater flexibility to address leverage
History suggests that the composition of an economy’s balance sheet sets up some of the conditions under which credit crises tend to erupt. A crisis may originate when an economy’s balance sheet has deep external linkages and the policy response is simply not fast enough to prevent contagion (e.g., the U.S. mortgage crisis) or when stress in one part of an economy’s balance sheet is shifted to another, which cannot withstand the shock (e.g., the transfers between the balance sheets of eurozone sovereigns and banks). Ultimately, if policymakers have access to a relatively immune balance sheet with few externalities, they have more flexibility to deal with excess leverage. Countries with deep household savings might be in this camp—certainly, the way Japan has piled up leverage may offer high-savings-rate emerging economies (e.g., China) clues on how to cope.
In our view, emerging economies have more latitude than their DM counterparts to address excess leverage in the next few years through a combination of familiar patterns: balance sheet transfer, inflation and currency adjustment. Meanwhile, developed world efforts to inflate away debt appear to have largely failed, and attempts at currency adjustment in Japan and Europe seem to have backfired, contributing instead to a sharp dollar rally that, in turn, threatened the fragile trajectory of global growth. DM policymakers must choose to be much more inventive—or else more willing to overlook the problems of excess leverage and accept it as a permanent component of their economic framework.
Mapping the vulnerabilities arising from excess leverage
In assessing the world economy’s vulnerabilities to leverage today, we can identify three simplified paths to addressing the problem. Debt can be reduced rapidly and with inevitable pain; it can be reduced gradually, at a pace that doesn’t excessively disrupt economic growth; or it can be serviced into perpetuity. Since the first outcome causes both debtors and creditors to suffer, we can assume that leverage will be addressed—wherever possible—by the latter two means. Our first observation was that the risk of a systemic credit crisis is reasonably low. While there are some concentrations of leverage, its location—where that debt sits on the national balance sheet—reduces the risk of rapid contagion. Nevertheless, our second observation was that both the level and growth of debt have in some instances breached the thresholds where, according to academic literature, they begin creating significant headwinds to future economic growth.
Judging from the absolute debt-to-GDP levels in key EM and DM regions (Exhibit 6A), developed economies remain significantly more leveraged than emerging economies, despite the perception that over the last five years the developed world has been deleveraging while emerging world debt levels have risen. Yet even though absolute EM debt levels remain subdued, in almost all sectors, debt is now markedly above long-run averages (Exhibit 6B). A Bank for International Settlements (BIS) working paper from 2011 examined debt’s drag on growth. It found that in OECD countries, when government or household sector debt-to-GDP ratios drifted north of 85% of GDP and when corporate sector debt-to-GDP rose above 90%, those debt levels impeded growth.1 In most of the emerging world, absolute levels of debt have remained well below these thresholds—with Chinese non-financial corporates the clear exception. However, it is plausible that debt could start to become a drag on growth at a lower threshold in emerging markets. Hence, government debt in India and Brazil, along with EM corporate debt in general, deserves further scrutiny.
While compared with the developed markets, EM debt ratios seem favorable …
EXHIBIT 6A: ABSOLUTE LEVEL OF DEBT-TO-GDP, BY BALANCE SHEET SECTOR, IN KEY DEVELOPED AND EMERGING ECONOMIES
U.S.: United States; UK: United Kingdom; JP: Japan; EZ: eurozone; AU: Australia; CA: Canada; SE: Sweden; CH: Switzerland; CN: China; IN: India; BR: Brazil; RU: Russia; MX: Mexico; ZA: South Africa; TR: Turkey.
... current levels of indebtedness in many EM balance sheet sectors have risen well above long-term averages
EXHIBIT 6B: CURRENT DEBT LEVEL VS. 10-YEAR AVERAGE, BY Z-SCORE
Source: Bank for International Settlements, J.P. Morgan Asset Management Multi-Asset Solutions; data as of August 5, 2016. Non-financial sector is the sum of non-financial corporates, household and general government (market value)
Growth in global debt today seems relatively well contained …
EXHIBIT 7A: THREE-YEAR GROWTH RATE OF DEBT-TO-GDP, BY BALANCE SHEET SECTOR, IN KEY DEVELOPED AND EMERGING ECONOMIES
… and compares favorably with the historical pattern
EXHIBIT 7B: CURRENT THREE-YEAR AVERAGE GROWTH RATE IN DEBT VS. 10-YEAR AVERAGE, Z-SCORE
Source for exhibits 7A and 7B: Bank for International Settlements, J.P. Morgan Asset Management Multi-Asset Solutions; data as of August 5, 2016. Non-financial sector is the sum of non-financial corporates, household and general government (market value).
Across developed markets, aggregate debt levels remain elevated. Yet there is still reason for optimism. First, in the U.S., Europe and Japan, household sector debt has fallen meaningfully post-GFC and is now quite low vs. long-run averages. Second, while DM general government debt has risen, in many cases it remains below 90% of GDP, even after the massive balance sheet transfers of recent years. Further, with interest rates very low, the danger threshold may arguably be somewhat higher today than the levels inferred from historical studies. Finally, there is evidence that debt levels have begun to recede from their peak in the most important DM regions.
Nevertheless, there are pockets of potential stress in the developed world, in particular in the economies most exposed to the commodity supercycle: Canada, Australia and Sweden. Of concern are the levels of household debt in Australia and Canada, and corporate debt in Canada and Sweden. Generally, household debt, alongside financial sector debt, is the most economically significant pocket of leverage on any nation’s balance sheet. But while such vulnerabilities in Australia and Sweden are considerations for the local economies, they are unlikely to present a significant contagion risk due to the global banking system’s relatively small exposure to these countries.
If we change our lens from the level of debt to the growth in debt, and view it in two ways—current growth rates in debt in Exhibit 7A and growth rates in debt vs. long-term averages in Exhibit 7B—we see a more mixed pattern of vulnerabilities. For the most part, systemic risks are reasonably contained. While there is a significant increase in general government debt in the emerging world, a relatively small component of the issuance is denominated in hard currency. Indeed, with the exception of India, the growth rate in household debt over a standardized three-year rolling window is below historical averages.2
Away from the EM household sector, the rise in non-financial corporate leverage stands out. We suspect that this rise in EM corporate leverage may present more of a risk to growth than an equivalent increase in DM corporate debt.3
China, where the level of corporate leverage is commonly cited as a vulnerability, presents a unique case. Much of its corporate leverage has built up in state-owned enterprises (SOEs) and so may, in some circumstances, be thought of as quasi-sovereign debt. Even if it is not viewed as quasi-sovereign, there are arguably fewer obstacles to impede the transfer of SOE debt onto the government balance sheet than would likely be the case for private non-financial corporate debt.
In general, the developed markets score well on their growth in leverage over the last few years, but once again, those most exposed to EM economies and the commodity supercycle—Australia, Canada and Sweden—have the most elevated vulnerabilities. Even there, the growth in leverage is not greatly above the long-run trend. However, if GDP growth is now structurally slower, this measure of leverage will deteriorate in the next few years, potentially constituting a barrier to growth in the years to come.
Conclusion: Implications of excess debt and the playbook for deleveraging in developed and emerging economies
What do we see as the next phase of the global credit cycle? In our view, developed and emerging economies will follow markedly different trajectories. Developed economies have the capacity to maintain the status quo of elevated debt levels—provided interest rates remain depressed. More important, they have few appealing alternatives. Their efforts to inflate away the debt have largely failed; allowing CPI to drift a few tenths of a percentage point above targets will not be sufficient to meaningfully shrink government debt in real terms in the near future. With popular tolerance for austerity at its limits, we also do not see enough room for reducing government debt levels through further fiscal tightening. We therefore see the most likely course for developed economies in the next few years as acceptance—effectively learning to live with the consequences of persistently higher debt levels.
Eventually, making changes to tax systems, extending debt maturities or swapping debt for more equity-like structures may be plausible endgames, yet we see little appetite today from borrowers or savers for such steps. Instead, we believe debt’s drag on growth will persist and along with it, downward pressure on interest rates. It will likely take a future economic contraction for governments to find it politically feasible to take the necessary steps to bring down the stock of debt.
Emerging economies have rather more latitude to follow a more traditional deleveraging path. Where acute issues exist—notably, in China’s non-financial corporate sector—policymakers, by and large, have the luxury of an unstressed government balance sheet that could be tapped to provide relief. We think it is possible that in the coming years a sizable share of China’s corporate debt could be nationalized, particularly debt associated with China’s SOEs. However, for this transition to proceed smoothly, greater transparency and development of the domestic banking system will be essential. In other emerging economies, the ability to stoke inflation is arguably greater than in developed economies, giving policymakers a tool to reduce their debt stock in real terms. In particular, the lack of hard currency debt presents EM policymakers with the possibility of using the foreign exchange channel to tackle excess debt. In general, we see excess EM debt as a likely drag on growth rather than an accident waiting to happen.
In our view, we are not on the precipice of another global credit crisis. Despite dire warnings to the contrary, we believe that—for the most part—DM policymakers have sufficiently inoculated their banking systems and households by allowing government balance sheets to take the strain. Meanwhile, emerging markets, notwithstanding the growth in their leverage, still have a broadly manageable aggregate level of debt. Where there are pockets of vulnerability in developed and emerging economies, they are largely isolated from the wider global financial system or are of sufficiently limited scale to render them a domestic rather than a global issue.
The buildup in debt levels across the system is the mirror image of the creation of a global savings glut—which, with aging populations and increasing regulation, is unlikely to disappear quickly. In plotting the trajectory of the credit cycle, it is crucial that we simultaneously consider that of the savings glut. While post-GFC policy intervention may have removed the most toxic elements of the credit overhang, we still face a drag on growth from excess capital—debt or savings, depending on which side of the ledger you sit—that is only servicing prior investment and consumption and not being put to work to fund new growth. In asset terms, this points to lower interest rates, flatter yield curves and reduced returns for riskier assets. Leverage probably won’t cause the next economic contraction, but it may take the next economic contraction to push policymakers into more direct efforts to reduce the stock of debt.