2017 External Sector Report

Global current account imbalances were broadly unchanged in 2016, with minor shifts adding to the reconfiguration under way since 2013.


Global current account imbalances were broadly unchanged in 2016, with minor shifts adding to the reconfiguration under way since 2013.


1. The 2017 External Sector Report (ESR) documents the evolution of global external imbalances and provides an updated assessment of the external positions of 29 economies for 2016. This Overview Paper complements the country assessments detailed in the Individual Economy Assessments paper, providing a global view, identifying cross-country patterns and discussing policies that should be considered to address excess imbalances from a multilateral perspective. This year’s report provides a more detailed account of the process of arriving at the assessment of external positions and devotes special attention to the issue of persistent imbalances. The paper is organized as follows: Section II documents recent trends in external flow (i.e., current account) and stock imbalances (i.e., international investment positions) and exchange rates. Section III presents the normative assessment of external positions (Box 1) and Section IV discusses the outlook and policy recommendations. Finally, Section V focuses on large and persistent surpluses, exploring historical evidence on persistence and reversals, along with the role of corporations and households in driving saving-investment dynamics.

External Assessments: Objective and Concepts

Current account (CA) deficits and surpluses can be desirable from an individual-country and global perspective. A country’s ability to run CA deficits and surpluses at different times is key for absorbing country-specific shocks and facilitating a globally efficient capital allocation. Some countries may need to save through trade surpluses (e.g., due to an aging population); others may need to borrow via trade deficits (e.g., to import capital and foster growth). Similarly, countries facing temporary positive (negative) terms-of-trade changes may benefit from saving (borrowing) to smooth out those income shocks. Thus, deviating from strict external balance is often desirable both from individual-country and global standpoints. Yet, in some cases, deficits or surpluses can be excessive if they depart from the levels that are consistent with country fundamentals and desired policies. The ESR therefore distinguishes between CA imbalances and CA gaps:

  • CA imbalance refers to any CA position different from zero, i.e., surpluses or deficits, without implying any judgement, which in principle may be warranted, too big, or too small.

  • CA gap, or excess imbalance, is the difference between the actual CA (stripped of cyclical and temporary factors) and the level assessed by staff to be consistent with fundamentals and desirable medium-term policies (or “norm”). This staff-assessed gap reflects policy distortions vis-à-vis other economies identified in the External Balance Approach (EBA) models as well as other policy and structural distortions not captured by the model. A CA balance deemed to be “stronger” (“weaker”) than implied by fundamentals and desired medium-term policies corresponds to a positive (negative) gap. Eventual elimination of such gap is desirable, though there may be good reasons for a gradual adjustment. Excess surplus (deficit) is also used to refer to a positive (negative) gap, irrespective of the sign of the actual CA balance. Assessments also include a view on the real effective exchange rate (REER)—normally consistent with the assessed CA gap. A positive (negative) REER gap implies an overvalued (undervalued) exchange rate. REER gaps do not necessarily predict future exchange rates, and may occur in any economy, including those with floating exchange rates.

While external assessments focus on CAs and exchange rates, they take other indicators (e.g., financial account balances, international investment positions, competitiveness measures) into account. Assessments are multilaterally consistent, meaning that positive CA gaps in some economies must be matched by negative gaps in others.

Evolution of Global External Imbalances

This section documents recent trends in global current account imbalances and exchange rates, with a focus on the drivers of the reconfiguration observed since 2013 and the implications for International Investment Positions (IIP). This discussion is not normative—i.e., observed levels and shifts in external imbalances may not be undesirable per se if they reflect warranted effects of cyclical factors, country fundamentals, or desired policies. A normative view (i.e., of excess imbalances) is provided in Section III.

2016 Developments

2. Global current account imbalances were broadly unchanged in 2016, with only minor compositional shifts. Following a marked narrowing in the aftermath of the Global Financial Crisis (GFC), overall global imbalances remained unchanged in recent years, at about 1.9 percent of world GDP (Figure 1, left panel). The configuration of current accounts saw only minor shifts during 2016, with some narrowing of China’s surplus and of deficits of key emerging and developing economies (EMDEs)—Brazil, Indonesia, Mexico, South Africa—amid a slightly higher surplus for Japan and a higher deficit for the United States. Most currencies, with the notable exception of the yen, depreciated in nominal terms against the U.S. dollar (Figure 1, right panel). These nominal changes vis-à-vis the dollar implied important real depreciations for the United Kingdom (related to Brexit), China (capital outflow pressures), and a few EMDEs, notably Mexico (partly reflecting trade-policy risks) and South Africa (partly due to political developments). These real depreciations were accompanied by large real appreciations for Japan as well as for some EMDEs (Brazil, Indonesia), whose currencies strengthened on the back of improving outlooks and policies. The euro and the U.S. dollar were broadly unchanged in real terms during 2016.

Figure 1.
Figure 1.

Evolution of Global Current Account Balances and Exchange Rates, 2002-16

Citation: Policy Papers 2017, 045; 10.5089/9781498346597.007.A001

Sources: World Economic Outlook, International Financial Statistics, Global Statistics Database and IMF staff calculations.1/ Surplus AEs: Hong Kong SAR, Korea, Singapore, Sweden, Switzerland, Taiwan POC; AE Commodity Exporters: Australia, Canada, New Zealand; Deficit EMs: Brazil, India, Indonesia, Mexico, South Africa, Turkey; Oil Exporters: WEO definition plus Norway.2/ 2016 average relative to 2015 average.

The Reconfiguration of Imbalances Since 2013

3. The relatively small current account shifts during 2016 built on an earlier trend of increasing imbalances in AEs. The most noticeable development since the narrowing of imbalances in the years immediately following the GFC was the reconfiguration that started in 2013. The latter was characterized by a marked shrinking of surpluses (into small deficits) in oil-exporting economies and a narrowing of current account deficits in EMDEs, accompanied by growing imbalances in AEs (Figure 2, left panel). The current constellation of imbalances entails an increased concentration in AEs—with large and persistent surpluses in some countries (Germany, Japan, Korea, the Netherlands, Singapore, Switzerland, Sweden) and higher or unchanged deficits in the United States and United Kingdom (see also Table 1). Real exchange rates have, for the most part, supported these current account shifts (Figure 2, right panel).

Figure 2.
Figure 2.

Evolution of External Current Accounts and Real Effective Exchange Rates, 2013-16

Citation: Policy Papers 2017, 045; 10.5089/9781498346597.007.A001

Sources: World Economic Outlook. International Financial Statistics, Global Statistics Database and IMF staff calculations.1/Surplus AEs: Korea, Hong Kong SAR. Singapore. Sweden, Switzerland. Taiwan POC. AE Commodity Exporters: Australia. Canada. New Zealand; Deficit EMs: Brazil, India. Indonesia, Mexico. South Africa. Turkey; Oil Exporters: WEO definition plus Norway.2/ 2016 average relative to 2013 average. For groups, weighted averages using USS GDP as weights are reported.
Table 1.

Selected Economies: Current Account Balance, 2013-16 1/

Source: World Economic Outlook and Fund Staff calculations.

Sorted by size (in USD) of surplus and deficit in 2016.

2/ For India, data are presented on a fiscal year basis.

4. The reconfiguration of global imbalances since 2013 was driven by a confluence of factors, including the sharp drop in commodity prices, the uneven demand recovery in systemic economies, and related differences in policies (Figure 3). The fall in commodity prices—particularly sharp for oil—redistributed income away from commodity exporters and towards commodity importers, while differences in the cyclical positions among systemic economies supported stronger net import growth and currencies in the United States and the United Kingdom (with the effects of Brexit still to play out), especially relative to the euro area and Japan. Yet, the observed reconfiguration of current accounts suggests that policy responses to these shocks (and other idiosyncratic factors) also played a key role.

Figure 3.
Figure 3.

Contributions to Change in Current Account Balance, 2013-16

(in percent of group GDP) 1/

Citation: Policy Papers 2017, 045; 10.5089/9781498346597.007.A001

Sources: IMF International Financial Statistics, and IMF staff estimates.1/ 2013-16 change. Deficit and surplus countries are classified based on CA balance in 2013. Country groups as in Figure 1. See details of the decomposition in Technical Appendix I.
  • Shifting current accounts in commodity exporters were dominated by the direct price effects (income losses associated with lower terms of trade)1, although with some noticeable differences across countries. For the group of oil-exporting countries as a whole, the terms-of-trade income shock was particularly pronounced, and net export volumes offset only a small portion of those income losses—mostly due to the constraints played by rigid exchange rate arrangements, although fiscal consolidation supported some compression of domestic demand (e.g., in Saudi Arabia). This pattern of adjustment to changing terms of trade contrasted with those in other AE and EMDE commodity exporters, where expanding net export volumes, supported by weaker currencies, nearly fully offset (Australia, Canada, New Zealand) or more than offset (Brazil, Mexico, South Africa) the exogenous price changes. In some cases, tighter financial conditions and idiosyncratic shocks (Brazil, South Africa) contributed to adjusting trade volumes (Box 2).

  • Among commodity importers, the extent of spending of the terms-of-trade income gains reflected mainly differences in cyclical positions and policies. In the United States and the United Kingdom, terms-of-trade income gains were more than offset by trade volumes supported by stronger domestic demand and appreciating currencies. Similarly, trade volumes in China showed a large offset of the terms-of-trade income gains, although in this case reflecting policy stimulus. In contrast, expanding net export volumes added to the terms-of-trade gains in Japan, while in the euro area spending of the terms-of-trade income shock was very limited—in both cases amid weakening currencies as domestic demand recovered more slowly. Similar behavior was visible in other large surplus AEs (Korea, Sweden, and financial centers) where only a fraction of the terms-of-trade income gains was spent. Excepting in China, and in contrast to previous years (2010-13), fiscal policy had a limited role in driving imbalances since 2013, as consolidation was more gradual and evenly distributed among deficit and surplus economies.

External Adjustment in Large Deficit EMDEs1

The sharp narrowing of current account deficits in large EMDEs (Brazil, India, Indonesia, Mexico, South Africa and Turkey) since 2012-13 was driven by a combination of domestic and external factors.

For most of these countries (excepting India), rapidly narrowing external deficits were driven by sharp domestic demand slowdowns, reflecting in part domestic idiosyncratic factors. This was especially marked in Brazil and South Africa, where political uncertainties and governance problems weighed on investor sentiment during part of 2013-16.

The impact of such domestic developments was exacerbated by a general tightening of external financial conditions on prospects of U.S. monetary policy normalization, and in some cases (Brazil, Mexico), a decline in terms of trade. Meanwhile, improving terms of trade contributed to the current account strengthening in Indonesia, South Africa, Turkey and India.

Policies generally supported the narrowing of external deficits. Weaker real exchange rates in most cases helped buffer the negative impact of external shocks, while tighter fiscal policies contributed to the adjustment in a few cases, especially South Africa.


Current Account Balance and REER

Citation: Policy Papers 2017, 045; 10.5089/9781498346597.007.A001

Sources: WEO, International Financial Statistics and IMF staff calculations.1/ In percent of group GDP.2/ REER Index (2010=100). Weighted average using US$ GDP as weights.
1 Prepared by Jaebin Ahn.

5. The patterns of private capital flows and foreign exchange intervention also shifted markedly during this period (Figure 4).

Figure 4.
Figure 4.

Non-reserve Capital Flows and Reserve Accumulation in EMDEs, 2008-16

(percent of group GDP, 4-quarter moving average)

Citation: Policy Papers 2017, 045; 10.5089/9781498346597.007.A001

Sources: International Financial Statistics and IMF staff calculations.
  • Overall net non-reserve flows to EMDEs were dominated by China’s abrupt reversal (from net inflows to outflows), with the latter reflecting increased uncertainties regarding the country’s growth prospects and financial stability concerns amid a process of gradual opening of the capital account.2 In sharp contrast to previous years (see Box 3), China’s foreign exchange intervention took the form of reserve sales, leading to cumulative reserve losses of about 7 percent of GDP during 2014-16 but preventing an even greater weakening of the renminbi. Net non-reserve outflows were also sizable for Russia and Saudi Arabia (although they stabilized in 2016), partly reflecting weaker growth prospects from sharply lower oil prices and geopolitical tensions (Russia). To deal with terms-of-trade and confidence shocks, and prevent an even more rapid slowdown in demand, Russia and Saudi Arabia both sold sizable amounts of reserves.3

  • In other EMDEs, changes in non-reserve flows were less dramatic. Net inflows slowed, reflecting a combination of lower growth prospects, higher borrowing costs (especially for commodity exporters), and improved terms of trade for commodity importers (which reduced demand for external financing). In most countries, foreign exchange intervention was limited during 2013-16 as a whole (see also Figure 12).

  • Bouts of financial turbulence in global markets—in part due to developments in China—were reflected in safe-haven flows into financial centers, which accumulated foreign exchange reserves as they intervened to mitigate the impact of large flows. Central bank balance sheets in Switzerland and Hong Kong SAR expanded markedly, with cumulative foreign exchange purchases reaching about 30 and 16 percent of GDP, respectively, during 2014-16.

Reserve Accumulation and Global Imbalances: A Longer-Term Perspective1

A common perception is that global current account (CA) imbalances have been driven by mercantilist policies in surplus countries. At the center of the mercantilism debate is the role played by reserve accumulation and foreign exchange intervention (FXI), although the use of this policy instrument may also reflect other motives (e.g., the need for liquidity buffers, or a desire to mitigate the effects of global capital flow cycles). Taking a longer-term view, this box sheds light on the role FXI policies may have played in driving global imbalances since the early 2000s.

The box chart below provides a way to visualize the relation between CAs and reserve accumulation. Along the diagonal dotted line in each panel, net non-reserve capital inflows equal the CA deficit, so the change in official reserves is nil. Other things being equal, increased purchases of international reserves will raise net private capital inflows, depreciate the currency and strengthen the CA, moving the economy up and to the right. Increased reserve sales move the economy down and to the left.

Pre-GFC. In the years preceding the GFC, when global imbalances reached their peak, large external CA surpluses—primarily in China, Japan, oil exporting countries and other EM economies—were indeed associated with significant reserve accumulation (Box Figure, left panel).

2010-13. In the years immediately following the GFC, CA imbalances of economies with previously large surpluses narrowed markedly—except for financial centers—reflecting primarily a sharp slowdown in external demand as key advanced economies deleveraged (Box Figure, mid panel). Facing sustained capital inflows—in part due to accommodative monetary conditions in advanced economies—reserve accumulation continued in many of these economies, but at a significantly slower pace. Meanwhile, reserve accumulation increased sharply in financial centers, amid sustained CA surpluses and increased capital flows. Facing sizable capital inflows, some emerging economies gained reserves despite negative CAs.

2014-16. Since 2013, the configuration of CA imbalances and reserve accumulation has shifted further (Box Figure, right panel). CA surpluses have become more concentrated in systemic advanced economies (euro area and Japan)—without reserve accumulation—while large reserve decumulation in oil exporters (reflecting lower oil prices) and China has helped to keep global imbalances in check. With slowing capital inflows, financial centers have continued accumulating reserves although at a significantly slower pace.

Beyond the motives behind reserve accumulation, this longer-term view suggests that its role as a possible driver of CA imbalances has diminished significantly over time—and recently reversed in some economies, notably China. Whether this shift reflects changing exchange rate policies (i.e., policy reaction functions) or simply a change in the external environment remains to be seen. But the configuration of global imbalances today points to factors other than currency intervention policy as the main drivers.


Non-reserve capital flows, current accounts and reserve changes, 2004-16 1/

(percent of GDP)

Citation: Policy Papers 2017, 045; 10.5089/9781498346597.007.A001

Sources: WEO and IMF staff calculations.1/ Includes EBA countries plus Hong Kong SAR, Saudi Arabia and Singapore. Green (red) circles correspond to economies with significant accumulation (decumulation) of reserves. Others are marked in light blue. Capital flows calculated as current account balance minus change in reserves. Circles are proportional to the absolute value of CA balance, as share of world GDP (i.e., contribution to global imbalances). Values for financial centers are denoted in the label, as they fall outside of the graph's scales.
1 Prepared by Gustavo Adler.

Implications for Stock Imbalances

6. Sustained current account imbalances since the GFC have contributed to diverging stock positions. After slowing in the immediate aftermath of the GFC, stock imbalances resumed their widening trend in recent years (Figure 5) reflecting, on the creditor side, the accumulation of net foreign assets mainly by euro area surplus countries (Germany, the Netherlands) and other AEs (Hong Kong SAR, Korea, Singapore, Sweden). The growth in creditor positions was mirrored almost entirely by a widening of the U.S. net debtor position, although the latter reflected also large valuation changes.4 Flow imbalances played an important role in driving the global widening of stock positions—as most countries that were net creditors in 2010 have run current account surpluses since then, while net debtors have run deficits. Excepting oil exporters with large creditor positions—that recently shifted to running current account deficits—and some debtor euro area countries (Italy, Spain)—that shifted to running current account surpluses—the recent rotation of current account imbalances has not materially changed their contribution to stock dynamics.

Figure 5.
Figure 5.

International Investment Positions, 2002-16 1/

(percent of world GDP)

Citation: Policy Papers 2017, 045; 10.5089/9781498346597.007.A001

Sources: World Economic Outlook and IMF staff calculations.1/ Surplus AEs: Korea, Hong Kong SAR, Singapore, Sweden, Switzerland, Taiwan POC; AE Commodity Exporters: Australia, Canada, New Zealand; Deficit EMs: Brazil, India, Indonesia, Mexico, South Africa, Turkey; Oil Exporters: WEO definition plus Norway.

7. Cumulative current account imbalances, however, have been partly offset by valuation changes in many cases. While their dynamics going forward is uncertain, valuation changes on stock positions tended to play a NFA-stabilizing role since the GFC, displaying the opposite sign to current account balances for the most part (Figure 6). Countries with large and persistent current account surpluses, such as China, Germany, Japan, experienced valuation losses, containing the increase in their NFA position that would have occurred otherwise. Conversely, countries with persistent current account deficits, such as the Brazil, Canada, South Africa, or the United Kingdom, experienced valuation gains, mitigating the weakening of their NFA positions.5 An important exception to this pattern was the United States, with both current account deficits and valuation losses, mostly due to the appreciation of the U.S. dollar (which increased the value of U.S. foreign liabilities relative to assets). Valuation changes played a particularly important stabilizing role in some financial centers, although also raising questions about IIP and BOP measurement issues in these economies (see Technical Appendix II).

Figure 6.
Figure 6.

Flow imbalances and valuation changes, 2010-16 1/

(percent of GDP)

Citation: Policy Papers 2017, 045; 10.5089/9781498346597.007.A001

Sources: International Financial Statistics and IMF staff calculations.1/ Dot sizes are proportional to countries' US$ GDP.2/ Valuation changes are calculated as the difference between NFA changes and cumulative current account balances (see details in Technical Appendix II).

Normative Assessment of External Positions

This section assesses current external imbalances—and their recent shifts—from a normative standpoint (i.e., whether they are deemed excessive relative to medium-term fundamentals and desired policies—see Box 1). It also discusses the process for arriving at the external assessments, and presents estimates of excess imbalances for 2016, while highlighting the contributions from key policy distortions.

How External Assessments Are Conducted

8. The ESR assessments entail comparing actual current account balances (stripped of temporary factors) with those deemed consistent with medium-term fundamentals and desired policies (dubbed “current account norms”) for individual countries. To this end, and as in previous years, assessments of external positions were conducted for 28 systemic economies plus the euro area that account for more than 85 percent of global GDP. The ESR exercise combines numerical inputs from statistical cross-country models with country-specific judgements based on IMF country teams’ knowledge and insights of each economy. Judgement is applied carefully and transparently to derive a multilaterally consistent set of norms.

9. Key inputs for the external assessments are the numerical estimates from the IMF’s External Balance Assessment (EBA) models.6 The EBA models estimate the average historical relationship between the current account or real exchange rate (REER) and a set of country fundamentals and policy variables from a panel of 49 countries over 28 years (1986-2013). Fundamentals include variables known to drive aggregate saving and investment rates (and thus the external current accounts), like an economy’s income level, its medium-term growth potential, the perceived quality of its institutions, demographic characteristics (see Box 4), and its net foreign asset position, as well as other features such as whether the country is an oil exporter, financial center, or enjoys reserve currency status. Policy (or policy-related) variables include the fiscal stance, health spending (a proxy for the extent of social safety nets), accumulation of foreign currency reserves, an index of capital account openness, and credit as a share of GDP. The model also estimates the impact of changes in terms of trade and the output gap, thus allowing construction of a cyclically-adjusted measure of the current account balance. Beyond the current account and REER models, EBA also includes a separate exercise that focuses narrowly on the sustainability of external stock positions. That exercise informs assessments in cases where external stabilization is a dominant concern.

10. To convert the estimated relationships into current account (or REER) norms, policy variables are assessed at their medium-term desired levels. Desired policies are identified by IMF country teams and assessed within a multilaterally consistent framework. For example, if a country team assesses that a country’s fiscal stance is too tight relative to the medium-term desired level, the actual fiscal stance is replaced with the desired level of this policy variable for computing the norm. Moreover, and to ensure multilateral consistency, all variables, including the identified policy gaps (or difference between actual and desired policies) are evaluated relative to the weighted average of the sample.

11. Estimated current account norms vary substantially across countries (Figure 7). Estimated norms were generally positive (and large)—for 2016 as well as in previous years—in countries with higher income per capita, lower projected output growth, higher longevity, and higher share of working age population (e.g. Switzerland, the Netherlands, Germany, Japan, Italy). Estimated current account norms tended to be negative (and large) in poorer countries with, higher growth potential, and faster population growth (e.g., India, Brazil, Mexico). However, other characteristics were also at play. For example, the U.S. current account norm is lower compared to peers because of the “exorbitant privilege” of having a reserve currency (i.e., reserve currency status reflects relatively high global demand for assets denominated in that currency, which tends to strengthen the currency and thus support larger current account deficits), other things being equal. By contrast, for some EMDEs, perceived institutional weaknesses—which can affect negatively an economy’s ability to borrow—push the estimated norms up.

Figure 7.
Figure 7.

Estimated Current Account Norms and Main Components, 2016 1/

(percent of GDP)

Citation: Policy Papers 2017, 045; 10.5089/9781498346597.007.A001

Source: IMF Staff assessments.1/ Norms are sorted from highest to lowest. Excludes Hong Kong SAR, Saudi Arabia and Singapore as they are not part of the EBA sample.

Demographics and the EBA Current Account Norm1

How are demographics modeled in EBA? The EBA current account (CA) regression has three demographic variables: the old-age dependency ratio, population growth, and aging speed. They are expressed in deviations from the world average since—as is the case with other regressors—only the relative magnitude of each variable should affect the overall-saving investment balance. Consistent with the life-cycle model, a higher old-age dependency ratio, and therefore a higher share of consumers/borrowers relative to savers, should imply a lower CA. Population growth, if driven by birth rates, should also exert a negative impact by increasing the youth dependency ratio and therefore the share of non-savers in the population. That said, the impact could be more ambiguous if population growth is driven by growth in the working-age population (e.g., because of migration), as this may imply more saving, but also more investment to stabilize the capital-labor ratio. Meanwhile, aging speed, defined as the expected increase in the old-age dependency ratio 20 years forward, should exert a positive effect on saving and the CA, driven by higher life expectancy and the resulting need for more life-cycle saving. To capture the non-linearities of demographics, interaction terms are also included (these were added with the 2015 model refinements).2 Specifically:

  • The relative dependency ratio is interacted with the aging speed: with a higher aging speed, a given increase in the dependency ratio implies a higher survival probability of the young cohort, increasing the need for life-cycle saving, and offsetting the negative composition effect of higher dependency ratio.

  • Similarly, the relative aging speed is interacted with the dependency ratio. With a higher dependency ratio, a given increase in the aging speed implies also a higher survival probability of the younger cohort and a greater need for life-cycle saving.

How do demographic contributions vary across countries ? EBA estimates of the different demographic inputs (and their interaction terms) have the signs that theory would suggest and are mostly statistically significant. The demographic contributions to the norms vary significantly (Box Figure). Countries with relatively high aging speeds and already high dependency ratios (e.g. Germany, Italy, Spain, the Netherlands) tend to show larger demographic contributions, followed by countries with high aging speeds but somewhat lower dependency ratios (China, Korea, Singapore, Thailand). On the other end of the spectrum, in countries that are young and have low aging speeds (India, Indonesia, Malaysia, Mexico, Saudi Arabia, South Africa), the contributions of demographics tend to be negative. However, the contributions of demographics to CA norms are expected to evolve gradually over time as countries jointly age and advance at different pace with their demographic transitions.


ESR Countries: Old Age Dependency vs. Aging Speed, 2015

Citation: Policy Papers 2017, 045; 10.5089/9781498346597.007.A001

Source: UN Population Statistics and Fund staff calculations.

While the EBA provides for a multilaterally consistent and rich treatment of demographics, in some cases, country-specific circumstances are not always properly captured. For example, while the aging speed variable is generally a good proxy for longevity risk across working-age cohorts, in countries with current high old-age dependency but higher fertility rates over the past few decades (e.g. Sweden, Denmark) the aging speed variable may underestimate the need for lifecycle saving.

1 Prepared by Mai Dao.2 Further details are provided in Annex I of the 2015 External Sector Report.

12. Cyclically-adjusted current account balances and EBA-estimated norms are adjusted for country-specific temporary factors and structural features, respectively, that go beyond what the EBA models can capture. Adjustments can be applied to both the estimated underlying current account position—to account for temporary factors that are insufficiently captured by standard cyclical adjustment techniques—and/or the estimated norms—to account for country features not included in the EBA models, provided there is strong and clear justification. For 2016, these adjustments (see also Box 5 and the 2017 Individual Economy Assessments companion paper) were small on average. As a result, for most countries the staff-assessed external gaps—i.e., after introducing adjustments—were close to the EBA gaps, especially for the assessment of current accounts (Figure 8).7 Staff-assessed gaps are presented in ranges, to reflect the uncertainties inherent in this exercise, which are generally consistent with the model’s standard errors (see also Table 2).

Figure 8.
Figure 8.

Staff Assessed vs. EBA Estimated Current Account and REER Gaps 2016 1/

Citation: Policy Papers 2017, 045; 10.5089/9781498346597.007.A001

Source: IMF Staff assessments.1/ Sorted by the mid-point of the staff-assessed gap. Hong Kong, Saudi Arabia and Singapore do not have EBA estimates. For Saudi Arabia, the current account gap reflects a fiscal policy gap.2/ REER gap is defined as the average of index- and level-based regressions. For Saudi Arabia, the REER gap is not identified.
Table 2.

Summary of Staff-Assessed Current Account and REER Gaps, 2016

Sources: IMF Staff Assessments and IMF International Financial Statistics (IFS).

Country-Specific Adjustments to the EBA Model Current Account Gaps1

The EBA model current account (CA) gaps, defined as the difference between the cyclically-adjusted or underlying CA balance and the EBA-estimated CA norm, are multilaterally consistent by design. However, since the EBA models cannot capture all country-specific factors, the estimates are accompanied by staff judgement to arrive at staff-assessed CA gaps. This box documents staff adjustments introduced in the 2016 assessments (see Box Figure and the 2016 ESR individual economy assessments paper).

  • Adjustments to the underlying CA. These staff adjustments generally reflect additional cyclical/temporary factors that are not fully captured by EBA models, or statistical issues related to the CA measurement. For example, staff adjusted the cyclically-adjusted CA balance for larger-than-estimated impacts of terms-of-trade changes (Thailand, Malaysia, and India), continued delays in the political transition (Thailand), temporarily high energy imports (Japan), and temporarily low income flows (United Kingdom). In addition, adjustments were made for mismeasurement arising from merchanting and financial-center activities (Korea, Sweden, Switzerland). These adjustments are intended to reflect more accurately the underlying CA balance (see Technical Appendix II).

  • Adjustments to the CA norm. These relate to certain structural features of an economy (fundamentals, desired policies, or financing risks, as assessed by country teams) that are not properly captured by the EBA CA regression. For example, given external financing risks from high net or gross external liabilities, the staff norm can be guided by estimates from the external sustainability approach (Brazil, India, Spain), to stabilize or strengthen the net international investment position as a share of GDP over the medium term. Demographic adjustments have been made where the estimated aging speed understates saving due to the impact of more recent increases in fertility rates (Sweden), or where it overstates the saving rate as the model does not capture the recent increases in migration (Germany). For some counties, CA norms were adjusted to consider factors that are not in the EBA CA model, such as projected changes in structural characteristics (Japan) and the low relative productivity of the non-energy sector (Canada).

Multilateral consistency. While country-specific factors might be well justified, they can compromise the multilateral consistency of the assessed gaps. Thus, arriving at staff assessments requires ensuring both their evenhanded application across countries, and also that adjustments (outside EBA results) in some countries are offset by the others. For the 2016 assessments, the median staff adjustment was roughly 0.4 percent of GDP (compared to the median standard deviation of the EBA estimated CA norm of 1 percent of GDP), and the overall discrepancy of staff assessments was small (excess deficits exceeded excess surpluses by only 0.07 percent of global GDP).


2017 ESR - Staff Assessed CA Gaps and Adjustors

(in percent of GDP)

Citation: Policy Papers 2017, 045; 10.5089/9781498346597.007.A001

* Hong Kong SAR, Saudi Arabia, and Singapore are not in EBA sample.
1 Prepared by Ruo Chen and Ruy Lama.

2016 Excess Imbalances

13. Excess current account imbalances for 2016 show a similar configuration to that of previous years (Figure 8 and Table 5). External positions were deemed “substantially stronger” than justified by medium-term fundamentals and desirable policies (excess surpluses above 4 percent of GDP) in Germany, Singapore, and Thailand; “stronger” (excess surpluses in the range of 2-4 percent of GDP) in Malaysia, Korea, Sweden, and the Netherlands; and “moderately stronger” (excess surpluses in the range of 1-2 percent of GDP) in China, and Japan. On the other side of the spectrum, external positions were considered “substantially weaker” (excess deficit of more than 4 percent of GDP) in Saudi Arabia; “weaker” (excess deficits in the range of 2-4 percent of GDP) in Turkey, the United Kingdom, and some euro area countries (Belgium, France and Spain); and “moderately weaker” (excess deficits in the range of 1-2 percent of GDP) in Australia, Canada, Italy, South Africa, and the United States. Meanwhile, external positions were deemed ”broadly in line” with medium-term fundamentals and desirable policies in Brazil, Hong Kong SAR, India, Indonesia, Mexico, Poland, and Switzerland. The euro area as a whole was also assessed as “broadly in line,” as positive excess imbalances in some countries (mainly Germany and the Netherlands) were offset by negative excess imbalances elsewhere.

14. Staff REER assessments map closely to current account assessments, except in a few cases. Exchange rate assessments are, for the most part, based on staff’s views on the current account (mapped into exchange rates by using trade elasticities estimated separately). Yet, discrepancies between current account and real exchange rate staff assessments may arise in the context of sharp real exchange rate movements deemed transitory or yet to be fully reflected in the current account. There were three cases in 2016. While Japan’s current account remained moderately stronger than warranted by fundamentals, its real exchange rate was assessed to be broadly in line, reflecting the yen’s sharp appreciation during 2016. China’s real exchange rate remained broadly in line with fundamentals, despite a moderately stronger current account, reflecting a projected further narrowing of the current account surplus. Mexico’s real exchange rate was deemed moderately undervalued while its current account was considered broadly in line, on expectations that protectionist risks, which led to a sharp depreciation in 2016, will recede somewhat.8

15. While the factors accounting for the staff-assessed current account gaps vary, some common features can be identified. Staff-assessed gaps can be decomposed into “identified policy gaps” and “other” distortions. Identified gaps refer to the differences between actual and desired policies as included in the EBA models (Table 3) and reflect domestic policy gaps relative to those of the rest of world.9 It is worth noting that even in countries where the overall identified policy gap is small, individual policies might still be important if their contributions offset one another. Other gaps are intended to reflect distortions not explicitly modeled in EBA (partly due to data limitations) that generate suboptimal saving and investment decisions.

Table 3.

Summary of EBA and Staff-Assessed CA Gaps, 2016

(in percent of GDP)

Source: Fund staff estimates.

Refers to the mid-point of the CA Gap.

Breakdown between norm and other factors (namely temporary or measurement errors) is approximate in some cases.

For Spain, we report the CA level required to reduce NIIP by 5 percentage points of GDP annually. The NFA stabilizing CA is -1.6 percent of GDP.

Weighted average sum of staff-assesed CA gaps.

  • In many “excess deficit” countries (Belgium, France, Spain, Russia, the United Kingdom), fiscal stances that were weaker than desirable (i.e., larger cyclically-adjusted fiscal deficits than desirable over the medium term) contributed to the negative overall gap, with higher-than-desirable health spending playing a role in the United States (Figure 9, lower left quadrant).10 That said, other structural factors holding back saving and competitiveness were also important, especially in countries where the identified gaps had a limited role (Australia, Canada, Italy, South Africa, Turkey).

    Figure 9.
    Figure 9.

    Contribution of Fiscal Policy Gap to Staff-Assessed Current Account Gaps, 2016

    (percent of GDP)

    Citation: Policy Papers 2017, 045; 10.5089/9781498346597.007.A001

    Source: IMF staff estimates and assessments.1/ Policy gaps after multilateral consistency adjustment.

  • Similarly, in many “excess surplus” countries (Germany, Korea, Sweden, Thailand, the Netherlands), tighter-than-desirable fiscal stances boosted current account positions, with insufficient health spending also contributed in some cases (China, Korea). Policies to address other distortions—such as structural features that constrain investment or lead to excess precautionary saving—remain necessary in countries with overall positive gaps, particularly those that need also gradually to tighten fiscal and credit policies (Japan and China). In 2016, foreign exchange intervention did not play a significant role in explaining excess imbalances.

Shifting Excess Imbalances Since 2013

16. Since 2013, excess current account imbalances have rotated somewhat with increased concentration in AE (Figure 10).11 Noteworthy changes include the narrowing of excess current account surpluses in China and Malaysia, and the narrowing of excess deficits in key EMDEs (Brazil, India, Indonesia, Russia, South Africa, Turkey). However, in some large AEs, excess deficits have increased (France, Italy, the United Kingdom, the United States), and in others excess surpluses widened (Japan, Sweden, the Netherlands, Thailand, Singapore). Excess current account surpluses in Germany and Korea have remained large and broadly unchanged since 2013.

Figure 10.
Figure 10.

Global Excess Imbalances, 2013 vs. 2016 1/

(in percent of global GDP)

Citation: Policy Papers 2017, 045; 10.5089/9781498346597.007.A001

Sources: IMF staff estimates and assessments.1/ Other surplus: Hong Kong SAR, Korea, Malaysia, Singapore, Sweden, Switzerland, and Thailand; Debtor EA: Belgium, Italy, France, Spain; Deficit EMs: Brazil, India, Indonesia, Mexico, South Africa, Turkey; Oil Exporters: Canada, Russia, Saudi Arabia; Others: Australia, and Poland.

17. Overall, this rotation resulted in a small increase of excess global imbalances. Total excess imbalances amounted to about 0.7 percent of global GDP in 2016—roughly the equivalent of one-third of total global imbalances.12 The overall size of excess imbalances edged up relative to 2013, as narrowing excess surpluses and deficits mainly in EMDEs have been more than offset by larger excess imbalances in advanced economies and oil exporters. Besides being increasingly concentrated in AEs, another distinguishing feature has been the persistence of large excess surpluses (Figure 11), a feature that is further analyzed in Section V.

Figure 11.
Figure 11.

Evolution of ESR External Assessments, 2012-16 1/

Citation: Policy Papers 2017, 045; 10.5089/9781498346597.007.A001

1/ Refers to current account assessments. Grouping and ordering based on countries' prevailing assessment during 2012-16.

18. In many cases, policies undertaken since 2013 contributed to the reconfiguration of excess imbalances.

  • Tighter fiscal and credit policies supported the reduction of excess deficits in some EMDEs (Indonesia, South Africa, Turkey), while an easing of these policies (beyond what would be desirable over the medium term) contributed to the narrowing of excess surpluses in China. Meanwhile, fiscal consolidation supported the widening or persistence of excess surpluses in key AEs (Germany, Japan, Korea), but did little to reduce excess deficits in other AEs (Australia, Canada, the United Kingdom, the United States).

  • With a few exceptions, foreign exchange intervention was limited or consistent with external rebalancing (Figure 12). China and Malaysia, which in 2013 had stronger-than-warranted external positions, sold significant reserves to support their currencies, without unduly compromising the adequacy of their reserve positions. Similarly, others (Brazil, Indonesia) with weaker-than-warranted external positions in 2013 purchased reserves, preventing a widening of their excess deficits. Exceptions included, Russia and Turkey, which, despite having weaker-than-warranted external positions, sold reserves, further deteriorating reserve coverage (Turkey). More recently Thailand, accumulated reserves despite having excess surpluses and comfortable reserve buffers.13 Among financial centers, large reserve purchases in Hong Kong SAR and Switzerland prevented a weakening of their broadly-in-line external positions against a backdrop of large capital inflows; while substantial reserve sales helped to prevent a widening of Singapore’ excess surplus, amid large outflows.

Figure 12.
Figure 12.

Selected Economies: Foreign Exchange Intervention, Staff Gaps and Reserve Adequacy, 2013-16

Citation: Policy Papers 2017, 045; 10.5089/9781498346597.007.A001

Sources: IMF International Financial Statistics, International Reserves and Foreign Currency Liquidity Template, and IMF staff calculations.1/ Estimated FX intervention (FXI) includes spot and derivatives. Spot FXI is estimated as BOP net international reserves flows (or change in reserves) minus estimated income flows on reserves. Derivatives (non-spot interventions) include aggregate short and long positions in forwards and futures in foreign currencies vis-à-vis the domestic currency (including the forward leg of currency swaps), and financial instruments denominated in foreign currency but settled by other means (e.g., in domestic currency), as reported in the International Reserves and Foreign Currency Liquidity Template.2/ Measured as a share of Risk-weighted Reserve Adequacy Unadjusted Metric, except for China and India, for which the metric is adjusted for capital flow measures.

Developments Since 2016, Outlook and Policies

This section takes a forward-looking view, discussing implications of post-2016 developments for the outlook and risks from the configuration of excess imbalances. It also discusses desirable policies to address excess imbalances from a multilateral perspective, both for surplus and deficit countries.

Developments Since 2016 and Outlook

19. Since 2016, currency markets have remained fluid, although in most cases observed movements do not point to material shifts in excess imbalances.

  • Systemic currencies. While remaining fluid—partly reflecting changing expectations about policy stimulus in the United States—currency movements through end-May 2017 have been limited, and do not point to material shifts in overall assessments. The U.S. dollar and euro have been broadly unchanged in real terms relative to 2016, while other systemic currencies are all down somewhat (about 4 percent in the case of the sterling and yen, and 2¾ percent for the renminbi).

  • Other economies. Following sharp depreciations in earlier years, real exchange rates for key EMDEs (Brazil, India, Russia, and South Africa) posted significant appreciations through end-May, relative to 2016, in response to improved outlooks—linked to better reform prospects in the case of Brazil and some initial hope for reduced sanctions in the case of Russia. These recent appreciations could exacerbate excess external imbalances in these countries, although conditions remain volatile, as evidenced by the Brazilian real’s recent sell-off on political uncertainties. On the opposite end, the Turkish lira weakened by about 10 percent in real terms, relative to 2016, although accompanied by some widening of the current account deficit.

20. In the context of weak automatic adjustment mechanisms, global flow and stock imbalances are projected to widen over the medium term. Absent decisive policy actions, weak automatic adjustment mechanisms—in part due to rigid currency arrangements in both key surplus and deficit economies (including within the euro area)—will continue to constrain a timely correction of excess imbalances.

  • Flow imbalances: Staff project broadly unchanged global imbalances, with moderately higher U.S. current account deficits offset by improvements in the external positions of the United Kingdom and key oil exporters, supported by fiscal consolidation in the case of oil exporters.14 In key surplus economies (China, Germany, Japan, Korea, the Netherlands) surpluses are projected to persist, marginally narrowing because of structural reforms, the reversal of temporary factors, and an expected realignment of competitiveness.

  • Stock imbalances. The projected configuration of current accounts will lead to a further widening of stock imbalances (Figure 13). The U.S. net debtor position is expected to deteriorate further, exceeding-50 percent of GDP by 2021. The net creditor positions of Japan and Germany will continue increasing (reaching over 80 percent of GDP over the next five years), while China’s net creditor position is projected to remain broadly unchanged over the medium term. In the event countries succeed in narrowing their excess current account imbalances, the projected widening of stock imbalances would moderate somewhat.

Figure 13.
Figure 13.

Selected ESR Economies: Actual and Projected NIIP, 2006-21

(percent of GDP) 1/

Citation: Policy Papers 2017, 045; 10.5089/9781498346597.007.A001

Sources: World Economic Outlook and IMF staff calculations.1/ Dotted lines correspond to projected NIIP if external gaps are closed.2/ Other creditors: weighted average of key oil exporters and financial centers (Russia, Saudi Arabia, Hong Kong SAR, Singapore and Switzerland).

Trade Protectionism, Global Imbalances, and Growth1

This box analyzes the implications of trade protectionism on external imbalances using the IMF’s Global Integrated Monetary and Fiscal (GIMF) multi-country model.2 The model considers three hypothetical large regions: a deficit country, a surplus country, and the rest of the world (ROW).3 The deficit country initially imposes a non-tariff barrier for 2 years on imports from the surplus country (roughly equivalent to a 10 percent tariff). The surplus country is assumed to retaliate one year after by imposing reciprocal trade restrictions for two years. In the fourth year, all trade restrictions are lifted.

This exercise complements the analysis presented in the Fall 2016 WEO by considering two specific features: (i) transitory protectionist policies (to limit fully offsetting exchange rate effects); and (ii) trade restrictions that do not raise fiscal revenues (to underscore the distortionary impact of these policies). Implications of protectionist policies for deficit and surplus countries, global growth, and imbalances are described below.

  • Deficit country: The imposition of trade restrictions improves the trade balance (0.3 percent of GDP after 3 years) by compressing imports (down 4 percent relative to baseline). Lower imports reduce the demand for foreign currency, generating a real exchange appreciation (averaging 1 percent over 3 years) and a transitory reduction in exports. While this policy can reduce somewhat the trade deficit, this is achieved at the expense of output (GDP declines on average by about 0.1 percent during the first three years), as protectionist policies not only lower exports (due in part to the exchange rate effect) but also reduce consumption and investment, with the non-tariff barriers increasing the final price of goods.

  • Surplus country: The surplus country experiences a decline in its trade balance (0.25 percent of GDP over 3 years), led by a reduction in exports ([1.25 percent over 3 years), which is only partially offset by a reduction in imports resulting from the real exchange rate depreciation. The combination of protectionist policies in the deficit country and the retaliation generates an average reduction of output of 0.3 percent during the first three years. As trade restrictions are lifted, the negative effects on output and the trade balances quickly unwind.

  • Global impact: The simulations suggest that trade restrictions can narrow global imbalances although their impact is small, as real exchange rates partly offset the effect of trade barriers. Moreover, the world as whole is worse off, as global GDP experiences an average decline of 0.1 percent while protectionist policies are in place. Imbalances in the rest of the world are roughly unchanged, although they observe very small output gains from protectionist policies elsewhere, as exports rise to satisfy unmet demand.

Further simulations show that global GDP losses increase with the duration of protectionist policies, while the impact on global imbalances lessens as real exchange rates moves to fully offset the intended outcome. Beyond the channels captured by the model, protectionism would likely increase global uncertainty and financial volatility with material effects on global investment.

1 Prepared by Michal Andrle and Ruy Lama.2 See more details in Anderson and others (2013) and Kumhof and others (2010), and alternative scenarios in the Fall 2016 WEO.3 Simulations are presented as deviations from the baseline scenario. In the baseline, the current account balance is -2 percent of GDP in the deficit country, and 1 percent of GDP in the surplus country. Deficit and surplus countries each represent about 20 percent of world GDP.

21. The new configuration of imbalances poses distinct global risks, particularly over the medium term. While the rotation of imbalances toward advanced economies entails lower deficit-financing risks in the near term, an increased concentration of imbalances in a few systemic economies could (i) heighten the risk of protectionist responses, which, if implemented, would reduce global output (see Box 6); and (ii) raise the risk of disruptive corrections down the road, including due to diverging stock positions. Although policy-related risks are now more balanced, global flow and stock imbalances could widen further should U.S. fiscal policy prove to be more expansionary than under the baseline. A more accommodative U.S. fiscal policy could also lead to a faster pace of monetary policy normalization and sharper appreciation of the U.S. dollar, with possibly disruptive effects in EMDEs, especially those with large external financing needs. Similarly, insufficient reforms in surplus economies could also promote wider imbalances, including in China (where tighter fiscal and credit policies will be needed over the medium term).

Policies to Address Global Excess Imbalances

22. Sustained global excess imbalances, amid smaller output gaps, point to the need for recalibrating the policy mix in excess deficit and surplus economies alike (Figure 14).

Figure 14.
Figure 14.

ESR Economies: External and Internal Imbalances, 2016

(percent of GDP)

Citation: Policy Papers 2017, 045; 10.5089/9781498346597.007.A001

Sources: World Economic Outlook, and IMF staff calculations.1/ Bubble sizes are proportional to the absolute value of the assessed excess external imbalance, in percent of world GDP (i.e., contribution to global excess imbalances).
  • Excess surplus countries with fiscal space and negative output gaps (Korea, Thailand) should use fiscal policy to support demand, while reducing reliance on monetary policy, to narrow both domestic and external imbalances. In China, fiscal and credit policies should support domestic rebalancing (from investment toward consumption) and address financial sector vulnerabilities. Where fiscal space is limited (Japan) amid continued economic slack, supportive fiscal policy could play a role provided it is matched with a credible medium-term consolidation plan. In countries without independent monetary policy internal appreciation is needed. In those countries with fiscal space (Germany, the Netherlands) but with relatively healthy cyclical positions, tax and spending policies could be used to promote private investment and boost public investment, labor force participation, and consumption without necessarily inducing long-lived overheating pressures. Where output gaps are closed and both fiscal and monetary policy are available, a tighter monetary and easier fiscal orientation would be desirable. Countries with adequate foreign exchange reserves levels (Thailand) should limit their reserve intervention to deal with disorderly market conditions and allow the exchange rate to move flexibly so that their currencies align better with fundamentals.

  • Excess deficit countries with slack should generally avoid easing fiscal policy, and instead rely on accommodative monetary policy to address domestic and external imbalances. In countries where near-term fiscal easing is envisaged (Canada), a timebound consolidation plan should be adopted to reverse the projected strengthening of the currency and widening of the current account deficit. Where output is near potential (United States) fiscal consolidation should proceed, along with gradual monetary policy normalization. Meanwhile countries with high inflation pass-through and above-target inflation (Turkey) may need to tighten monetary policy while opportunistically build foreign exchange reserves. This will not only reduce vulnerabilities to capital flow volatility, but support saving and reduce excess imbalances over the medium term. In countries without independent monetary policy an internal devaluation is needed (France, Italy, Spain, Saudi Arabia), supported through fiscal consolidation and growth-friendly tax and spending policies (e.g., lower payroll taxes and job training programs) to complement productivity-enhancing labor market reforms.

23. Reducing large and persistent excess global imbalances will require increased focus on structural reform policies. Reforms should be tailored to country-specific circumstances (see Table 5) and complemented by supportive macroeconomic policies (see above), especially reforms that may have negative short-run growth effects (Box 7).

Table 4.

Selected ESR Countries: Current Account Regression Policy Gap Contributions, 2016

(in percent of GDP)

Source: IMF staff estimates.

Total contribution after adjusting for multilateral consistency.

Total domestic contribution is equivalent to coefficient*(P-P*)

Table 5.

2016 Individual Country Assessments: Summary of Policy Recommendations

Source: 2016 Individual External Assessments.

This non-exhaustive list focuses on key recommendations for closing external imbalances.