First posted on:
Money & Banking, 25 June 2018
“It is not speed that kills, it is the sudden stop.”
—Bankers’ adage, cited by Dornbusch, Goldfajn and Valdés.
After years of relative calm, in recent months several emerging economies have found that the cost of attracting foreign funding is going up. Faced with a halt of external financing, Argentina obtained a three-year financing deal worth $50 billion from the IMF, while funds also appear to be flowing out of Brazil, Turkey, and elsewhere. And, recent bond market turbulence in Italy suggests the possibility that political risks are triggering outflows there.
In this post, we explain balance-of-payments (BoP) crises – the sudden stops or capital flow reversals – that compel countries to restore their external balance between exports and imports,or, in the case of capital flight, shift to export surpluses. In addition to describing common features of BoP crises, and characterising sources of vulnerability that make them more likely, we examine one emerging-market example – the Asian crisis of 1997-98 – and one advanced-economy episode – the crisis of the euro area periphery from 2010 to 2012.
Before exploring the anatomy of crises, we start with the definition of the balance of payments itself. The BoP is an accounting identity stating that net cross-border flows of goods and services – the current account – must be matched by net flows of financial claims (including those arising from portfolio shifts and from direct investment)—the financial account. Simply put, if one country is importing more than it is exporting from another country, it must find a way to finance that difference. (For simplicity, we abstract here from the third BoP component, the official settlements balance; for details, see the appendix below).
To see the implications of this identity, start with the current account balance (NX). When exports – broadly defined to include not only goods and services sold to foreigners but also the income earned by residents on assets abroad – exceed imports, the current account is in surplus (NX>0). When exports fall short of imports, the current account is in deficit (NX<0).
The current account does not exist in isolation. It is the consequence of domestic consumption, investment, and saving behaviour. As we show in the appendix, there are two ways to think of the current account. The first is to consider the relationship between national income and domestic spending (the sum of consumption, investment,and government outlays). When income falls short of spending, the result is a current account deficit. Alternatively, we can think of the current account balance as equal to the gap between domestic savings (S) and investment (I). That is,
S–I=NW,
where domestic savings is the excess of national income over non-investment spending (which itself is composed of consumption plus government outlays).
The profound implication of this is that, when the current account is in surplus (deficit), savings must exceed (fall short of) investment. This means that current account balances are a macroeconomicphenomenon determined both by long-run factors like demographics that influence aggregate savings, and by cyclical factors that shift investment. If a country has a current account deficit that it wishes to bring closer to balance, then it must raise savings, lower investment, or both. Equivalently, it must lower domestic spending relative to national income.
Yet another way to think about the drivers of the current account is in terms of the financial account – the other side of the balance of payments. Consider a simple example where a country’s domestic saving is insufficient to meet its investment demand. Unable to find resources at home, the natural response is to seek funds abroad. This creates a capital inflow, which must be balanced by a current account deficit.
The inescapable consequence of the BoP accounting identity is that when a country can no longer finance a current account deficit, it must adjust instantly. Forced adjustment can precipitate a crisis requiring cuts in government spending (raising public savings) along with a recession that depresses both consumption and investment. In addition to the plunge in cross-border finance, firms anticipating a downturn will trim other investments. These large, rapid shifts typically are associated with bouts of high unemployment, excess capacity, reduced productivity, lost income, and business failures.
BoP crises occur precisely when the inward flow of capital needed to finance a current account deficit (or to offset gross capital outflows) abruptly halts. Such a sudden stop typically reflects foreign creditors’ doubts regarding the likelihood of full and timely repayment. When these concerns are mild, the result can be a simple rise in risk premia. But, if concerns become acute, risk premia can skyrocket, eventually halting financing altogether. Without access to foreign funds, a country must immediately restore the balance between exports and imports (or between national income and spending).
Even more disruptive than a sudden stop is a capital flow reversal – when sentiment turns so negative that inward capital flows suddenly become capital flight. Such reversals compel a shift to a current account surplus. The result is an even deeper recession.
Since the BoP is an identity, causality can run in either direction. This means that forces influencing the financial account can drive the current account, as well as the other way around. For example, should an emerging economy with a relatively high return on new capital liberalise cross-border fund flows, it may experience an inward rush of funds to finance an investment boom (I>S) that results in a large current account deficit (NX<0). However, if these investors lose confidence, then the capital flows can suddenly reverse, leading to a current account surplus and a deep downturn.
A range of factors makes a sudden stop or capital flow reversal more likely:
- First, free cross-border movement of capital makes it easy for funds to move in either direction.
- Second, when funding is short-term, foreign suppliers can reverse the flow by refusing to renew loans.
- Third, reliance on foreign-currency debt means that a domestic central bank can neither provide liquidity in the case of a bank run nor reduce the burden of the debt (as it can with domestic-currency debt) through a bout of unanticipated inflation. In both cases, this raises the risk of outright default.
- Fourth, credit inflows are more volatile than inflows of equity and direct investment (see Eichengreen, Gupta, and Masetti).
- Fifth, a country with a small tradables sector can support a smaller volume of external liabilities than can one that is more deeply integrated into the global economy. Put differently, when inward capital flows start to decline, economies typically must shift labour and other resources rapidly from the non-tradables sector (such as housing and government) to the tradables sector (exports; see Kehoe and Ruhl). A small export sector can make this adjustment difficult. Institutional rigidities—such as an inflexible labour market—add to this challenge.
- Other factors, like a rise in global risk perceptions or risk aversion (say, reflected by the VIX), can increase required risk premia, adding to the possibility of sudden stops.
- Finally, the combination of fixed exchange rates and free cross-border capital flows poses a classic vulnerability that we have previously discussed on our blog (see here and here).
Following on a long line of research regarding sudden stops in emerging economies (see, for example, here and here), Eichengreen and Gupta document several empirical regularities among countries experiencing sudden stops. First, sudden stops occur frequently: Eichengreen and Gupta identify 46 episodes in 34 emerging economies between 1991 and 2015 (see chart below). Second, these episodes tend to occur in waves, at least partly reflecting global conditions (such as the Asian financial crisis of 1997-98 and the Great Financial Crisis of 2007-2009). Third, sudden stops persist on average for about a year and are associated with a shift from capital inflows to outflows averaging about 3% of GDP. Finally, in addition to financial effects (including currency depreciation and a decline in asset prices), sudden stops have real economic impact: GDP falls on average by about 4% in the first year, with the associated current account adjustment largely reflecting a plunge of investment, not a rise in savings.
Number of countries experiencing a sudden stop (quarterly), 1991-2015
Source: Based on Appendix A in Eichengreen and Gupta (2016), Managing Sudden Stops.
With this as background, we now illustrate the impact of a sudden stop with two stark examples: the Asian crisis of 1997-98 and the more recent crisis in the euro area.
Asian financial crisis
In the emerging world, the Asian financial crisis of 1997-1998 presents a classic BoP crisis. As the next chart shows, the ASEAN-5 countries – Indonesia, Malaysia, the Philippines, Singapore, and Thailand – were running large current account deficits starting in the early 1990s. Rapid economic growth and high rates of investment in the region supported the confidence of foreign creditors. However, economists warned that the growth of Asia’s 'tigers' increasingly relied on rising capital and laborlabourinputs with diminishing returns, rather than a technologically driven expansion that is sustainable (see Young). During the pre-crisis years, a range of financial vulnerabilities accumulated (see, for example, Goldstein, The Asian Financial Crisis, Chapter 2):
- These economies operated with fixed exchange rates versus the US dollar, but (unlike China) they allowed relatively unimpeded cross-border flows of short-term funds, inconsistent with the open-economy trilemma. Furthermore, foreign currency reserves were generally small.
- In several countries, credit expansions that far outstripped GDP growth were financing investment booms, with a substantial portion of funds going to property investment, rather than boosting competitiveness. As a result, parts of the nonfinancial sector became highly leveraged, adding to the risks of a thinly capitalised financial system in several economies.
- To lower their interest burden, some banks, firms and governments borrowed in US dollars at short maturities, making them vulnerable both to currency depreciation and to funding (rollover) risk.
ASEAN-5: Current account balance and investment (percent of GDP), 1980-2018
Note: The ASEAN-5 includes Indonesia, Malaysia, the Philippines, Singapore, and Thailand. Source: IMF WEO database.
Against this background, a run on the Thai baht – leading to a devaluation in July 1997 – triggered financial contagion throughout the region. Motivated to reassess the risks of supplying funds to emerging market economies – especially short-term dollar credit – foreign lenders sharply raised the compensation they required. What followed was a series of speculative attacks that compelled a number of Asian central banks to give up their fixed exchange rate. In each case, the result was a plunge in the value of the currency resulting in a loss of competitiveness for those economies that maintained their currency pegs. Where dollar-denominated debt was high, sharp domestic currency depreciation increased the debt service burden, boosting non-performing loans, intensifying the credit squeeze, triggering waves of bankruptcies, and putting further stress on domestic banks. In the end, Hong Kong was the only economy in the region with an open financial account that proved able to preserve its fixed exchange rate – a tribute to the resilience of its financial system (and to the fortitude of central bank authorities who allowed overnight lending rates to rise temporarily over 200%).
The Asian crisis was far worse than a sudden stop: the collapse of investor confidence led to an abrupt reversal of net capital inflows, compelling a stunning turnaround of the current account deficits across the region. Between 1996 and 1998, the external balance of the ASEAN-5 shifted by almost 12 percentage points of GDP! To achieve this wrenching adjustment, per capita real GDP dropped by about 8%, while from 1996 to 1999 investment plunged by nearly 16 percentage points. (Similarly, in Korea, the current account shift exceeded 14 percentage points with a peak-to-trough GDP drop of 8%.) In contrast, shielded by its robust capital controls, China was largely immune to the turmoil.
Crisis of the euro area
It is common to view the euro area crisis of 2010-12 as both a sovereign debt crisis and a banking crisis. While these are important, one should not overlook the BoP crisis that featured the core countries of the euro area (including Germany, the Netherlands, Finland, and Luxembourg) on one side, and the periphery (Greece, Ireland, Italy, Portugal, and Spain) on the other. (For this BoP perspective, see Higgins and Klitgaard and Cecchetti, McCauley and McGuire).
At the start of the euro in 1999, the periphery of the euro area looked like a particularly attractive place to invest. Inflation had fallen dramatically, currency risk seemed to disappear, and capital per worker was relatively low. The resulting optimism of core-country investors led to a large expansion in the gross and net cross-border supply of credit – including short-term financing – with negligible compensation for default risk (let alone for a possible exit from the euro area). With funds cheap and plentiful, credit flowed into government finance (for example, in Greece and Portugal) and into real estate (in Ireland and Spain). This widened the periphery’s current account shortfall (see chart), without providing any real basis for future exports that would restore external balance.
Euro area periphery: Current account balance (percent of GDP), 2000-18
Note: The periphery includes Greece, Ireland, Italy, Portugal, and Spain. Source: IMF WEO database and authors’ calculations.
As the financial strains that began in 2007 wore on, in 2009 creditor confidence began to wane, boosting the risk premia required for financing external imbalances. At the peak of the euro area crisis – in 2011 and the first half of 2012 – creditors sought compensation not only for heightened risk of bank and sovereign default but also for the risk that a country would exit the euro, redenominate domestic debt into a new local currency, and encourage a sharp depreciation. These fears also drove depositors holding euros to shift their funds from peripheral-country banks into those located in the core. The result was sizable outflows of private capital.
As the chart above shows, this capital flow reversal triggered a large swing in the peripheral country current account balances: from a deficit of more than 4% of GDP in 2010 to a surplus of more than 1% in 2013. Unsurprisingly, this shift was associated with a plunge of investment and GDP, together with a surge in unemployment, non-performing loans and defaults.
Importantly, euro-area institutions – especially the Target2 real-time gross settlement system – prevented an even larger collapse from the sharp reversal of capital flows. To see this, we can look at how at the peak of the crisis the four peripheral countries financed their net capital flows. In the following chart, the blue bars are all below zero: these represent net private capital outflows from the peripheral economies. In the case of Spain and Italy, these outflows were equivalent to nearly 30% and 12% of GDP, respectively.
Euro area periphery countries: Net private capital inflows and their financing, 2011-1H 2012
Note: The black diamonds represent the net flows. Source: Based on Table 3 in Higgins and Klitgaard and authors’ calculations.
Countries in the euro area can finance net private capital outflows – on top of financing current account deficits – through two means. They can obtain official assistance from the IMF or from funding mechanisms sponsored by the US – like today's European Stability Mechanism (ESM)—as Greece and Portugal did (these are the grey bars). Alternatively, they can utilize the largely automatic Target2 mechanism (the red bars), which is operated by the ECB and the national central banks making up the Eurosystem.
In effect, Target2 provides an unconditional and unlimited interest-free loan from the national central banks of the core countries to the national central banks of the countries experiencing deposit flight, allowing them to provide credit to their commercial banks. That is, when investors flee a peripheral country bank, re-depositing funds in a commercial bank of a core country, this mechanism automatically replaces the peripheral bank’s funds with central bank credits. There is no limit to this process, so the amounts can be large. Indeed, Target2 financing exceeded all net private capital outflow from Italy and Spain at the peak of the crisis. Absent Target2, these countries would have been compelled to adjust their current accounts much more rapidly and to a far greater extent, triggering an even larger collapse of investment and GDP. It is questionable that the euro would have survived such a disaster.
Conclusion
Today, governments and international financial authorities understand a great deal more about sudden stops and capital flow reversals than they did two decades ago. They know that fixed exchange rates are generally incompatible with the free flow of capital across borders, so most countries have shifted to flexible or managed exchange rate regimes. Furthermore, to defend against the need for large, sudden adjustments and improve financial resilience, many countries also have built up foreign exchange reserves. And, the IMF has acknowledged the potential usefulness of some types of capital controls (or, euphemistically, ‘capital flow management policies’), especially where financial systems remain vulnerable and undeveloped.
Nonetheless, as much as we might hope, recent experience makes clear that sudden stops (or capital reversals) have not been relegated to the dustbin of history. They can occur both in emerging-market and advanced economies when confidence falters in the financial system, the government, or both. And, with the large increase in gross cross-border financial integration over recent decades, such a loss of confidence can trigger capital flight and a deep economic downturn even in places that have no history of current account shortfalls.
Authors’ note: We grateful to our friends and colleagues—Kim Ruhl and Venky Venkateswaran—for very helpful discussions.
Appendix: Savings, investment and the balance of payments
| Y |
Gross national income |
| C |
Consumption |
| I |
Investment |
| G |
Government purchases |
| NX or CA |
Net exports of goods and services (equals exports minus imports) or the current account balance |
| S |
National savings |
| FA |
Financial account (net capital flows; inflows are positive) |
| OB |
Official settlements balance (change of foreign reserves) |
The balance of payments (BoP) includes three components. The first – the current account balance (CA) – measures the net flow of goods and services across borders. The second component—the financial account balance (FA) – tracks the net flow of asset claims across borders. This includes purchases and sales of financial and other assets (such as equities, bonds, loans, and real estate). And the third component is the official settlements balance (OB), which represents the change in a country’s official foreign reserves (including foreign exchange and gold).
We can write the BoP identity as:
That is, excluding measurement error, the sum of the three BoP components must be zero.
In advanced economies today, official reserves positions generally do not shift much, so OB=0. This implies:
(2) CA = –FA.
That is, when a country’s current account is in deficit—imports exceed exports (CA<0)—the financial account is in surplus (FA>0): capital inflows exceed outflows. Put differently, foreigners accrue financial claims on a net-importing country in compensation for their net sales of goods and services. As a result, a persistent flow of current account surpluses (deficits) leads over time to a rising stock of net foreign assets (liabilities).
In economies that buy or sell foreign reserves to limit short-run currency-market fluctuations—as do many emerging economies—the financial account balance can differ (at least temporarily) from the current account balance.
It is important to keep in mind that since equation (2) arises from an identity, it says nothing about causality. While changes in the current account can trigger changes in the financial account, the reverse is not only possible, but quite common. For example, a country that liberalizes cross-border capital flows may attract large net capital inflows (FA>0) that then lead it to become a net importer of goods and services (CA<0). Similarly, a net-importing country that loses the confidence of foreign investors may face an abrupt halt to capital inflows (or even a reversal). Such a sudden stop of net financing from abroad would compel an immediate restoration of current account balance.
To help understand the factors driving the current account, we explore its relationship with national savings and investment. The national accounting identity states that national income or gross national product (Y) is the sum of consumption (C), investment (I), government spending (G), and net exports (NX):
(3) Y = C + I + G + NX.
Net exports – the gap between broadly-defined exports and imports of goods and services in the national accounts – equal the current account (CA) in the balance of payments. Thus, when imports exceed exports – that is, when NX is less than zero – the current account is in deficit. This also means that domestic spending (C + I + G) exceeds national income (Y).
We define national savings (S) as the gap between national income (Y) and non-investment domestic spending, which is the sum of consumption (C) and government spending (G):
Substituting the definition of savings (S) from (4) into (3) and subtracting investment from both sides, we see that the gap between savings and investment is equal to the current account balance:
(5) S – I = NX = CA
Equation (5) has profound implications. First, a nation has a current account deficit (surplus) if and only if investment exceeds (falls short of) national savings. This means that macroeconomic factors – like demographic influences on saving or cyclical swings in investment – determine the current account balance. Moreover, to improve its current account balance, a country must raise savings, lower investment, or both. Equivalently, it must lower domestic spending relative to national income.
Absent a change in the official settlements balance that would come from the sale of foreign exchange reserves, when a sudden stop occurs and the financial account shifts toward outflows, the current account balance must improve. In theory, this current account shift can occur through an increase of savings or a decline of investment. In practice, much of the adjustment usually occurs through a cyclical downturn (a recession) associated with a plunge of investment. In such a recession, national income falls, but the current account improves so long as domestic spending falls by more. Such a painful economic adjustment is a routine feature of BoP crises.
Importantly, sudden stops can occur even in the absence of current account imbalances. Because gross capital flows to and from a country can be volatile, a loss of confidence in the country can trigger both a plunge of inflows and a surge of outflows. Again, absent a swing in official settlements, these sudden net capital outflows must result in a sharp rise of the current account balance.
