Journal Issue
Share
Article

Euro Area Policies: Selected Issues

Author(s):
International Monetary Fund. European Dept.
Published Date:
July 2014
Share
  • ShareShare
Show Summary Details

External Rebalancing in the Euro Area: Developments and Policies1

Since the crisis, the euro area current account has moved from rough balance into a clear surplus. The rebalancing underlying this shift has been highly asymmetric, with some debtor economies (Cyprus, Greece, Ireland, Italy, Latvia, Portugal, and Spain) seeing large improvements in their current accounts (sometimes into surplus), while many creditor economies (Austria, Belgium, Finland, Germany, Luxembourg, and the Netherlands) have largely maintained their surpluses. In this context, the paper examines current account developments from the perspective of these economies’ financing needs by considering the evolution of saving and investment. Creditor economies appear to exhibit persistent over-saving and under-investment relative to model predictions. The under-saving and over-investment characterizing debtor economies prior to the crisis has improved, but there is still some way to go. Finally, scenario analyses suggest that the external rebalancing arising from proactive policy adjustments and structural reforms could also yield substantial growth dividends.

A. Introduction

1. The euro area has shifted into strong surplus since 2011. Prior to 2008, the euro area as a whole had a small current account surplus. But this surplus masked a growing divergence in current accounts across euro area members, with some moving into large deficits, like Greece and Spain, and others into large surplus, such as Germany and the Netherlands. With the crisis though, many debtor economies saw their current accounts begin to improve, narrowing the gap with other economies in the euro area. The external rebalancing though has been highly asymmetric, as creditor economies have seen little reduction in their surpluses. Together, these changes underpin the shift towards a strong euro area current account surplus.

2. In this note, we investigate the behavior of current accounts in the euro area through the lens of saving and investment, looking at the economy’s financing needs. To accomplish this, we examine the patterns of saving and investment in the euro area, both at the aggregate level and from a sectoral perspective. We also consider how net capital flows (the financing counterpart to the current account) and related net foreign asset positions have evolved since the start of the euro area. This then leads into a statistical analysis of saving and investment, following an approach to assess possible under- or over-saving and investment by economy, similar to the IMF’s external balance assessment (EBA) approach which focuses solely on the current account. Finally, to gain additional insight into the potential role of policies in narrowing the divergence of current accounts, we use the IMF’s EUROMOD simulation model to generate scenarios that contrast the external rebalancing that could occur by following IMF policy advice with rebalancing due to adverse shocks and delays in adjustment.

B. Financing Needs and the Current Account

Geography of the Euro Area Current Account

3. The euro area current account saw a strong shift towards surplus from 2011 (Figure 1). This largely reflects improvements in the current accounts of members with current account deficits, but a component is also due to further strengthening of surpluses in Germany and other creditor economies (Chen, Milesi-Ferretti, and Tressel, 2013; Tressel and others, 2014).2

Figure 1.Regional Composition of the Euro Area Current Account

(percent of GDP)

Source: IMF World Economic Outlook database and IMF staff calculations.

Note: Black line indicates the sum of the respective components (approximately the overall current account). The group of other EA creditor economies includes: Austria, Belgium, Finland, Luxembourg, Malta, and the Netherlands. The group of EA debtor, IMF program economies includes: Cyprus, Greece, Ireland, Latvia, and Portugal. The group of other EA debtor economies includes: Estonia, France, Slovak Republic, and Slovenia.

4. From the inauguration of the euro area in 1999 to 2012, two key groups had persistent current account deficits: the group of IMF program economies3 (Cyprus, Greece, Ireland, Latvia, and Portugal), and the aggregate of large stressed economies (Italy and Spain). Their negative contribution to the euro area current account grew over time, peaking in 2008. Since then, these current account deficits have shrunk, particularly for the group of IMF program economies, which experienced a “sudden stop” of private inflows with the global financial crisis (Merler and Pisani-Ferry, 2012; Catão and Milesi-Ferretti, 2013). In 2013, these economy groups moved into current account surplus. Although both large and stressed economies, Italy and Spain do differ significantly in the magnitude of their net foreign liabilities, with Italy at around 30 percent of GDP while Spain is at over 90 percent of GDP. The aggregate of other debtor economies (which is largely France by economic mass, but also includes Estonia, Slovakia, and Slovenia) remains in a small deficit.

5. At the same time, the group of creditor economies (Austria, Belgium, Finland, Germany, Luxembourg, Malta, and the Netherlands) has seen its surplus increase since 2009. Germany’s current account moved into surplus in 2002, peaking in 2013 (as a contributor to the euro area current account), and staying in surplus ever since (see European Commission, 2012 for assessments of current account surpluses in the European Union). The group of other creditor economies has been in surplus since the start of the euro. The recent strengthening may be partly due to the approximately 15 percent depreciation of the euro area real effective exchange rate versus its trading partners observed from 2009 to 2012, coincident with the aftermath of the global financial crisis and the onset of the European sovereign debt crisis.

Sectoral Composition of Current Accounts in the Euro Area

6. At the euro area level, the post-crisis shift towards surplus has been largely driven by corporate and household behavior and more recently, by a decline in government deficits (Figure 2).4 Household saving has actually declined slightly, while corporate saving has risen since 2009. Automatic stabilizers resulted in government dissaving in the wake of the crisis, but this has recently vanished, leaving government saving at basically zero, but below the levels prior to the crisis. After a long run-up before the crisis, both household and corporate investment fell, and precipitously in the case of corporate investment. Government investment rose slightly with the crisis, reflecting some stimulus, but has since receded to slightly below pre-crisis levels. Overall investment is about 4 percentage points of GDP below where it was prior to the crisis, and it is this depressed investment that is the primary contributor to the current account surplus and its recent improvement. In fact, as seen in the lower panel of Figure 2, the recent strength of the euro area current account compared to pre-crisis is mostly due to a large shift of the corporate sectoral current account info surplus from deficit. This looks like a classic investment driven boom bust, accompanied by a rise in saving as corporate attempt to deleverage.

Figure 2.The Sectoral Composition of the Euro Area Current Account

(percent of GDP)

Source: Eurostat Balance of Payments database and IMF staff calculations.

Note: Black line or dot indicates the sum of the sectoral components by variable. The sectoral current account is defined to be the difference between saving and (fixed) investment for the respective sector. The total current account is the sum of the sectoral current accounts. Euro area saving and investment rates peaked in 2007.

7. Zooming into different country/groupings, we see:

  • For Germany, the household sector is the largest contributor to the current account surplus, with a secondary role played by the corporate sector (Figure 3, left column). The component of the surplus due to the household sectoral current account (household saving minus fixed, residential investment) roughly doubled in the early part of the 2000s, going from about 2 and a half percentage points of GDP to just over 5 percentage points in 2007-2008. Since then, it has been relatively stable. Apart from a brief, small dip negative in 2008, the corporate sectoral component has also contributed positively to the current account, becoming a net lender since the early 2000s.

  • The German current account strengthening reflects both rising saving rates and falling investment rates, across the household and corporate sectors. The German investment rate has fallen from about 23 to about 17 percent of GDP since the early 2000s, with the larger component of this fall due to declining corporate investment. Household investment has been relatively stable as a share since 2005; after falling between 1999 and 2005 Government investment’s contribution has been remarkably stable since 1999, showing only small perturbations. At the same time, household and corporate savings have risen, although in the case of household saving, this rise has been more marginal.1

  • Unlike Germany, both saving and investment rates have fallen in the other euro area creditor economies, while still maintaining an overall current account surplus. The fall in saving is largely due to a gradual ecline in saving by the household sector over the past 15 years. Corporate saving has actually risen slightly over the same period. After earlier saving, the government began to dissave in 2009, but this dissaving has shrunk over time, such that the contribution of government saving has been negligible over the past couple years. The decline in investment has not been as large as the saving fall, but remains important; if investment had not fallen simultaneously, the aggregate of these economies would have shifted into current account deficit. On a sectoral basis, both household and corporate investment have fallen, although the contribution of corporate investment to the overall fall is more important.

  • In other creditor economies, the current account surplus is mostly due to the corporate sector, with a smaller contribution by the household sector (Figure 3, right column). Since 2009, the government deficit has been and remains a substantial drag on the current account surplus in these economies, albeit not enough to offset the corporate and household sectoral surpluses.

  • The recent improvements in the current accounts of debtor economies mostly reflect a large fall in investment across sectors (Figure 4). There were large rises in household and corporate investment prior to the crisis, but these have now reversed, with the overall investment rate below the levels seen in the early 2000s. The investment rate has fallen from over 25 percent of GDP in 2007 to below 15 percent in 2012, for the aggregate of the IMF program economies. The fall in investment for the aggregate of Italy and Spain is slightly smaller, from around 25 percent of GDP in 2007 to around 18 percent percent (for Italy, it goes from 22 to 18 percent, while for Spain, it goes from 31 to 20 percent). Across most of these debtor economies, the hump-shaped investment patterns echo the dynamics of a classic boom-bust business cycle.

  • From 2009 onwards, almost the entire current account deficit in debtor economies is due to government deficits (Figures 4 and 5). Government deficits have been shrinking over time, reflecting the fiscal consolidation policies undertaken in these economies. For others in this group, the government is the largest contributor to the current account deficit, with the corporate sector deficit playing a secondary role.

  • Overall, the greatest contributors to improvements in euro area current account surpluses over recent years appear to be falls in investment (both corporate and household) and fiscal consolidation. As mentioned earlier, this bears the hallmarks of a classic investment driven boom-bust, with an aftermath of heavy indebtedness that leads to greater saving. The potentially worrying implication is that the current account improvements may vanish as the euro area regains its footing and recovers. There are some differences across economy groups, with the aggregate of other debtor economies showing little fall in investment2 Unlike the clear upturn and downturn visible in debtor economies’ investment, like Italy and Spain, Germany’s investment fall is more puzzling, since it appears to be on a persistent downward trend rather than a cycle over the past 15 years.

Figure 3.The Sectoral Composition of the Current Account in Creditor Economies

(percent of GDP)

Source: Eurostat Balance of Payments database and IMF staff calculations.

Note: Black line indicates the sum of the sectoral components by variable. The sectoral current account is defined to be the difference between saving and (fixed) investment for the respective sector. The total current account is the sum of the sectoral current accounts.

Figure 4.The Sectoral Composition of the Current Account in Debtor Economies

(percent of GDP)

Source: Eurostat Balance of Payments database and IMF staff calculations.

Note: Black line indicates the sum of the sectoral components by variable. The sectoral current account is defined to be the difference between saving and (fixed) investment for the respective sector. The total current account is the sum of the sectoral current accounts.

Figure 5.The Sectoral Composition of the Current Account in Other Debtor Economies

(percent of GDP)

Source: Eurostat Balance of Payments database and IMF staff calculations.

Note: Black line or dot indicates the sum of the sectoral components by variable. The sectoral current account is defined to be the difference between saving and (fixed) investment for the respective sector. The total current account is the sum of the sectoral current accounts. The current account for the group of other debtor economies was at a high in 1999.

8. Mirroring the current account surplus, the euro area has seen a move towards net capital outflows, dominated by other (bank) investment flows to the rest-of-the-world (Figure 6, upper left panel).3 The composition of net financing flows for the euro area has also changed, shifting away from a large foreign direct investment (FDI) outflow towards predominantly other (bank) investment and portfolio debt outflows. The FDI component has largely vanished in recent years, possibly reflecting a pull-back from equity investing (whether FDI or portfolio equity) outside of the euro area and a shift towards debt investing (whether bank-related or portfolio debt). The rise in net debt outflows makes sense in the context of the euro area sovereign debt crisis, when many investors shifted out of the euro area. Debt flows tend to be more volatile than FDI flows, so this change may presage more variability in net flows for the euro area (Bluedorn and others, 2013). Zooming in to the economy or economy group levels, we see:

  • Germany’s net financing abroad has been mostly via other (bank) investment (Figure 6, upper right panel). German banks remain heavily engaged in lending abroad, although as we shall see, there are signs that previous bank flows from Germany to other euro area economies has fallen. FDI elsewhere has played a smaller role. In 2012, all the financing components of Germany’s net capital inflows switched to become outflows, for the first time since the introduction of the euro.

  • Net capital outflows from the group of other creditor economies are also in the form of other (bank) investment over the past few years (Figure 6, middle left panel). Prior to the crisis, other instruments played a larger role contributing to net outflows (in particular FDI and portfolio debt), but these have lately switched to become net inflows. Portfolio equity investments from elsewhere have remained a source of net inflows, although at a lower level than seen in the mid-2000s.

  • For the group of IMF program economies, net financing prior to the crisis was overwhelmingly portfolio equity or other (bank) investment flows (Figure 6, middle right panel). These flows are still important financing sources, but they have been more than offset recently by large net outflows into portfolio debt. These debt flows could reflect safe haven flows into euro area creditor economies. Net official financing flows are reflected in other (bank) investment. These peaked in 2012, when a number of these economies were simultaneously under IMF programs. Prior to 2009, other (bank) investment flows were largely private.

  • Italy and Spain were most dependent on net inflows for portfolio debt and other (bank) investment prior to the crisis, but these inflows dried up in 2011 (Figure 6, lower left panel). In fact, the aggregate of Italy and Spain saw net outflows of portfolio debt in 2011 and 2012 (residents were increasing net foreign portfolio debt holdings more than non-residents were increasing net domestic portfolio debt holdings). Note that portfolio debt includes government bonds. In 2012, the sovereign debt crisis hit these two economies hard, with investors shifting away from their large government bond markets, although this trend has been reversed recently. Italy and Spain have positive net inflows of portfolio debt in 2013.

  • Other debtor economies remained largely reliant on net portfolio debt inflows for financing (Figure 6, lower right panel). Large net portfolio debt inflows, particularly post 2008, may reflect a desire to hold more core euro area government debt, with the rise in concerns about the viability of the euro. Prior to 2008, the largest component of net inflows was other (bank) investment, but these vanished with the crisis (chiefly due to France’s experience). FDI has been a small, but persistent, net outflow.

Figure 6.The Composition of Net Capital Inflows by Euro Area Economy Group

(percent of GDP)

Source: IMF Financial Flows Analytics database and IMF staff calculations.

Note: Black line indicates the sum of the respective components; overall net capital inflows.

9. Taken together, a picture of creditor economies supplying bank-related and other investment financing to the non-euro area, rest-of-the-world in recent years emerges. At the same time, debtor economies have dramatically reduced their financing via bank-related and other investment inflows (apart from the group of IMF program economies, although this is mostly net official financing). Hence, if creditor economies are increasing their holdings of other (bank) investment assets abroad, it is not vis-à-vis other euro area economies. Along with these developments, the volume of gross flows (inflows and outflows) dropped dramatically post-crisis and across the board. Moreover, they have remained low for the euro area, indicating that there may have been some loss of depth in euro area-related international financial markets.

A View from Net Foreign Asset Positions

10. Although the current account has moved into surplus, the euro area as a whole remains a net debtor to the rest-of-the-world (Figure 7, upper left panel). This is largely through its net portfolio equity and debt offerings to the rest-of-the-world. The euro area holds a large stock of FDI vis-à-vis the rest of the world, which offsets its large negative net foreign asset position (NFA) in portfolio assets to some extent. Zooming in to the economy or economy group levels, we see:

  • Germany’s net foreign assets are on an upward trend, at above 30 percent of GDP in 2012 (Figure 7, upper right panel). Its net foreign assets are overwhelmingly in the form of other (bank) investment assets, with a secondary role played by FDI assets. At the same time it remains a major net supplier of portfolio debt abroad. That being said, its rise in net foreign assets remains less than the accumulation of its current accounts over the same period. This indicates that there may have been some negative valuation effects that offset some of the surpluses.

  • The group of other creditor economies switched from net debtor to creditor in 2008 (Figure 7, middle left panel). Their net foreign asset position has continued to rise, with the group seeing their net foreign asset position as a share of own GDP at almost the same size as that of Germany. These net foreign assets mostly take the form of FDI and portfolio debt holdings, although the share of portfolio debt has shrunk in recent years while the share of other (bank) investments has risen. They are large net suppliers of portfolio equity elsewhere.

  • Net foreign asset positions in the debtor economies are all negative, but differ widely in their underlying instrument composition (Figure 7, left middle and lower panels). For the group of IMF program economies, their net debtor status is larger, exceeding 100 percent of GDP. As highlighted by Catão and Milesi-Ferretti (2013), such large net foreign liabilities have historically been associated with a greater chance of a financial crisis. This may partly reflect the greater vulnerability of such economies to a number of forms of mismatch of assets and liabilities, combined with a greater exposure to valuation effects (Obstfeld, 2012). Net claims against the group are in the form of portfolio equity and other (bank) investment, while the group holds a net positive foreign position in portfolio debt. By contrast, the aggregate of Italy and Spain has large net negative foreign positions in portfolio debt and other (bank) investment. In portfolio equity and FDI holdings elsewhere, they are net creditors. For the group of other debtor economies, their negative net foreign asset positions are also largely portfolio debt and other (bank) investment assets.

Figure 7.The Composition of Net Foreign Assets by Euro Area Economy Group

(percent of GDP)

Source: External Wealth of Nations, Mark II revised database and IMF staff calculations.

Note: Black line indicates the sum of the respective components; the overall net foreign asset position.

11. Combined with the information on net capital inflows, net foreign asset positions indicate that most of the other (bank) investment asset exposures held by euro area economies is still largely versus other euro area economies, but this has been changing. Euro area economies’ other (bank) investment exposures to extra-euro area economies have been rising, replacing some of the earlier intra-euro area exposures. Table 2 shows how realized returns on net foreign asset positions over the past 15 years may be one motivation for these developments. 4 On average, many creditor economies saw negative real return differences between their foreign assets and liabilities, acting as a drag on their net foreign asset positions and also suggesting possible gains from portfolio rebalancing, either by shifting away from foreign towards domestic assets, or by changing the composition of their foreign assets and liabilities, away from euro area debtor economies. At the same time, many debtor economies had large negative real return differences on average, reinforcing their large net foreign liability positions and making adjustment more of an uphill climb.

Table 1.Net Foreign Assets and Current Accounts in the Euro Area Economies

(percent of GDP)

Current Account, 1999-2013
Economy GroupEconomyNet Foreign Assets, 2013AverageStandard DeviationMinimumMaximum
Creditor EconomiesAustria31.91.8−1.74.9
Belgium441.82.8−2.07.9
Finland163.63.4−1.78.5
Germany484.33.3−1.77.5
Luxembourg1849.02.35.413.2
Malta25−4.34.3−12.12.4
Netherlands536.32.82.010.4
Debtor EconomiesCyprus*−86−6.34.1−15.6−1.5
Estonia−45−6.36.3−15.92.8
France−21−0.21.6−2.23.1
Greece*−119−8.14.2−14.90.7
Ireland*−105−0.63.3−5.66.6
Italy−30−1.01.4−3.51.0
Latvia*−65−7.68.5−22.68.7
Portugal*−119−8.43.5−12.60.5
Slovak Republic−65−4.83.5−8.52.4
Slovenia−38−0.93.0−5.46.5
Spain−98−4.83.0−10.00.7
Source: Eurostat International Investment Position database, IMF World Economic Outlook database, and IMF staff calculations.Note:
Table 2.Average Real Return Difference between Foreign Assets and Liabilities Euro Area Economies, 1999-2012

(percentage points)

Of Which:
Economy GroupEconomyReal Return DifferenceCapital Gains on AssetsYield on AssetsCapital Gains on LiabilitiesYield on Liabilities
Creditor EconomiesAustria−0.4−1.42.01.0−2.0
Belgium−0.6−1.91.21.5−1.4
Finland4.6−0.81.84.8−1.1
Germany−1.1−1.01.5−0.3−1.3
Luxembourg−0.1−0.30.90.3−1.0
Malta0.5−0.43.81.9−4.8
Netherlands−0.37.01.6−7.6−1.3
Debtor Economies, IMF ProgramCyprus1.4−1.81.04.0−1.9
Greece−3.3−2.20.2−0.2−1.0
Ireland−2.8−2.71.21.6−2.8
Latvia2.81.00.42.2−0.7
Portugal0.21.21.3−0.7−1.6
Other Debtor EconomiesEstonia−1.3−0.32.91.1−5.0
France0.1−1.20.90.9−0.5
Italy−0.3−2.51.22.4−1.5
Slovak Republic−3.32.41.8−1.3−6.1
Slovenia−0.4−0.40.11.6−1.6
Spain−2.5−4.31.41.8−1.4
Source: IMF Financial Flows Analytics database, External Wealth of Nations, Mark II reviseddatabase, Eurostat Balance of Payments database, and IMF staff calculations.Note: Averages over 1999–2012 are shown. Rates of return are in annualized percentage points. Returns on liabilities are shown with a negative sign, reflecting their contribution to the real return difference, defined as: rA -rL =rcap.A +rinc.A -(rcap.L +rinc.L), where A indicates the foreign asset side, L indicates the foreign liability side, cap denotes the capital gains/losses, and inc denotes the yield or per unit investment income achieved.

12. In the next section, we conduct a statistical analysis of saving and investment by euro area economies, assessing whether or not there is evidence of under- or over-saving or investment that has contributed to the large net foreign asset or liability positions observed in the euro area.

C. Model-based Saving and Investment Imbalances in the Euro Area

13. A cross-country regression analysis reveals how aggregate saving and investment deviations from model-based predictions for the euro area economies are linked to substantial and persistent current account imbalances in recent years. To assess the relative importance of these two channels, we separately investigated the behavior of national saving and investment in a panel of major EU economies at an annual frequency over the period 1986–2013. Building on common models of the determinants of long-term saving and investment, we estimated predicted (equilibrium) values for saving and investment and then quantified the deviations of observed saving and investment from the model-predicted levels for each economy in each year (see Box 1 for full details on the model specification and estimation approach).

14. The exercise to identify saving and investment imbalances described here differs from, but complements, the current account assessments in the IMF’s External Sector Report (ESR). The ESR incorporates regression-based analyses of current accounts, net foreign asset positions, and real exchange rates, with the judgment of IMF country teams in its assessments of current account and real exchange rate imbalances. By contrast, the saving and investment deviations presented here are solely model-based. Furthermore, by having separate regression models for saving and investment with slightly constructed as the difference between the model-based saving and investment deviations—may diverge from the ESR’s identified current account imbalance. But interestingly, the implied current account deviations from our analysis are strikingly consistent with the IMF’s External Balance Assessment (EBA) estimated current account imbalances.5

15. The quantitative analysis suggests that in recent years, creditor economies have had investment persistently below the model-based prediction. For instance, investment in Germany has been consistently lower than what the model suggests since 2002, by an average of 2.0 percent of GDP each year (Figure 8). The Netherlands has experienced a similar degree of underinvestment over the past ten years, at a yearly average of 1.8 percent of GDP. Moreover, recent outturns do not indicate any reduction of these deviations and indeed suggest a possible increase. For instance, in 2013, estimated investment was 3.0 percent of GDP in Germany and 4.2 percent in Netherlands higher than the actual outturn.

Figure 8.Model-Implied Investment and Saving Residuals and the EBA Assessment of Current Account Imbalances

(percent of GDP)

Source: IMF Staff calculations.

Note: See appendix table for underlying data sources and estimation sample.

16. A tendency for higher saving in creditor economies since the mid-2000s has also contributed to the large current account imbalance. As shown in Figure 8, the analysis suggests that Germany has had saving higher than its estimated value since 2004, as its saving rate has been persistently above the model estimate, by an average of 2.7 percentage points each year for the last ten years. Aggregate savings in the Netherlands has also been consistently higher than what the model predicts, by an average of 1.7 percentage points since 2005. Importantly, the deviation has become even more substantial since the onset of financial crisis, with the saving in these two countries standing at 2.5 and 4.6 percentage points of GDP above the model prediction in 2012, respectively. This is about twice as large as the typical imbalance in the sample, as measured by the standard error of the regression.

17. The analysis also indicates that the current accounts in debtor economies primarily reflected a persistent and substantial over-investment prior to the financial crisis. During 2003-2007, Spain ran an average positive investment deviation of 6.2 percent of GDP each year, much higher than the positive saving deviation, and thus generating a large negative current account deviation. Greece and Portugal saw a similar degree of positive investment deviation during the same period, at 2.0 percent and 1.5 percent of their GDP, respectively. On the other hand, a negative saving deviation also contributed significantly to current account deficits in Greece and Portugal since the mid-2000s, on average 5.4 percent and 1.5 percent of GDP since 2005, but less so in Spain and France.

18. Current account balance improvements in debtor economies since the crisis have largely been achieved through investment shedding, and to a lesser extent increases in saving, driven partly by much tighter external financing conditions. Aggregate investment has declined drastically in several stressed economies, leading to substantial reductions in the magnitude of positive investment deviations in Spain and Portugal, and even a substantial negative investment deviation in Greece since 2008. The negative saving deviations have also been corrected rapidly, in particular in Portugal and Greece. Much of the investment drop that has helped to reduce current account deficits can be attributed to the large and negative output gaps that now exist in many of these economies. As these gaps close over time and investment and consumption gradually rebound, there is the risk that these deficits will re-emerge, unless recovery is accompanied by structural reforms to raise potential output. Moreover, there is considerable uncertainty regarding estimates of potential output, which means that the assessment given here that over-investment has largely been corrected in debtor economies must be approached cautiously.

19. Since the financial crisis, adjustments in both saving and investment by debtor economies reflect substantial “internal rebalancing” via rapid domestic demand shedding, but corrections in creditor economies have yet to occur. As Figure 8 reveals, saving and investment deviations in creditor economies have widened since 2010, leading to a further strengthening of their current account surpluses. This generates further upward pressures on the common currency and downward price pressures across the euro area in an already very low inflation environment. This also reduces the pace of relative price adjustments and short-term effectiveness of structural reforms in improving competitiveness in debtor economies, slowing down the recovery for the euro area as a whole. Finally, the analysis suggests that it is in the interest of creditor economies to undertake policies to stimulate investment and reduce saving, in addition to being in the interest of the euro area as a whole.

Box 1.A Model-based Analysis of National Saving and Investment Behavior

We develop two linear regression models to separately examine the determination of national saving and investment, with the goals of quantifying deviations from their equilibria as well as the relative importance of these two channels in determining the observed current account deviations at the economy-level. The backbone of the analysis is a fixed effects linear regression model, based on the following equations:

where i indexes the cross-section and t indexes time (years), I is the ratio of gross (fixed) investment to GDP, and S is the ratio of the current account (CA) plus gross (fixed) investment to GDP. XI and XS denote different vectors of regressors that determine I and S respectively, with βI and βS the coefficients to be estimated. η and ε are mean zero error terms. The choice of regressors is guided by economic theory, leading to specifications similar to those of earlier empirical studies. In particular, regressors in the investment equation, XI are:
  • Financial conditions, as captured by the real short-term deposit rate and long-term government bond yields;

  • Risk and uncertainty measures, which include the risks associated with the institutional and political environment, as well as global capital market conditions. The former is proxied by Safer Institutional/Political Environment Index, and the latter by VIX/VXO index interacted with the degree of capital account openness of each country;

  • Cyclical conditions, as captured by the output gap in each country (relative to the world average) and the expected 5-year ahead real GDP growth rate from WEO projections;

  • Aggregate Tobin’s Q, measured by the median of individual firms’ readings in each country.

The specifications of the aggregate saving regression also follow the previous literature closely. In particular, regressors XS in the savings equation (2) are:

  • Non-policy fundamentals, including 1) GDP per capita (PPP GDP divided by the size of the working age population) relative to the U.S., Japan, and Germany, as a proxy measure of the distance of the country’s productivity to the frontier of highest productivity around the world; 2) Demographic factors, including the country’s aging speed, population growth, and dependency ratio; 3) Lagged Net Foreign Assets (NFA)/GDP ratio, as well as an interaction term which allows a different slope when NFA is below negative 60 percent of GDP;

  • Risk and uncertainty measures, which include Safer Institutional/Political Environment Index, and the VIX/VXO index interacted with the degree of capital account openness of each country;

  • Cyclical conditions, as captured by the output gap in each country (relative to the world average), the expected 5-year ahead real GDP growth rate from WEO projections, and the cyclical component of the commodity terms of trade, interacted with the trade openness of each country;

  • Policy-related factors, which include a country’s fiscal policy stance as measured by the cyclically adjusted fiscal balance (CAPB), which is instrumented with respect to a set of macroeconomic variables, and public health expenditure as percentage of nominal GDP, as well as the real short-term deposit rate.

Empirical estimation is conducted based on an annual sample of 19 major EU countries, for the period of 1986–2013. Different specifications, or a larger sample which also includes the non-EU OECD countries, generate similar results.

As displayed in Table 3 and 4, most coefficient estimates have the expected signs, and nearly all are statistically significant. The investment rate rises in response to an increase in Tobin’s Q, as well as a decline in real short-term interest rates. A safer institutional environment promotes investment, whereas heightened uncertainties on financial markets tend to discourage investment. Higher output gap or stronger growth prospect encourages investment, and the coefficients are statistically significant at the 1 percent level.

Table 3.Regression Results for Investment Model
Explanatory Variable(1)(2)(3)
Tobin’s Q0.006**0.006*0.002
(0.003)(0.003)(0.003)
Real short-term deposit rate−0.115**−0.098**−0.144***
(0.050)(0.043)(0.046)
Long-term government bond yield0.030
(0.044)
Safer Institutional Environment Index0.131***0.123***0.149***
(0.025)(0.022)(0.024)
VIX/VXO index−0.110***−0.109***−0.109***
(0.017)(0.017)(0.018)
Output gap0.415***0.418***
(0.050)(0.050)
Expected 5-year ahead growth2.018***2.029***1.973***
(0.157)(0.156)(0.168)
Number of observations437437437
Number of countries191919
R20.1890.1930.140
Source: IMFstaff calculations.Note: The dependent variable is the investment rate (investment to GDP ratio). The estimated regression model includes fixed effects for each country. Standard errors are in parentheses underneath the coefficient estimates. *, **, and *** denote statistical significance of the coefficient at the 10, 5, and 1 percent levels respectively. See appendix table for underlying data sources and sample details.
Table 4.Regression Results for Saving Model
Explanatory Variables(1)(2)(3)
GDP Per Capita0.093***0.097***0.081***
(0.018)(0.017)(0.017)
Real Short-Term Deposit Rate−0.001**−0.001**−0.001**
(0.000)(0.000)(0.000)
Output Gap0.007
(0.056)
Expected 5-year Ahead Growth0.317
(0.203)
Cyclically-Adjusted Primary Balance to GDP0.402***0.456***
(0.120)(0.114)
Public Health Expend. to GDP−1.808***−1.811***−1.928***
(0.185)(0.181)(0.181)
Net Foreign Assets to GDP0.062***0.061***0.066***
(0.006)(0.006)(0.006)
Interaction of NFA/GDP and Large−0.034**−0.028**−0.034**
Debtor Indicator(0.015)(0.014)(0.015)
Terms-of-Trade Gap−0.046*−0.051*
(0.029)(0.028)
Aging Speed−0.074***−0.073***−0.068***
(0.019)(0.019)(0.019)
Population Growth Rate0.311***0.311***0.351***
(0.075)(0.075)(0.075)
Dependency Ratio0.133**0.109**0.144***
(0.054)(0.051)(0.051)
Safer Institutional Environment Index−1.499***−1.648***−1.480***
(0.540)(0.536)(0.530)
VIX/VXO Index−0.289***−0.317***−0.265***
(0.072)(0.066)(0.065)
Number of observations490490490
Number of countries191919
R20.5570.5740.513
Source: IMF staff calculations.Note: The dependent variable is the saving rate (saving to GDP ratio). Saving is calculated as the difference between the current account and fixed investment. Large debtor indicator takes the value 1 if net foreign assets to GDP are less than -60 percent; otherwise, it is equal to zero. The estimated regression model includes fixed effects for each country. Standard errors are in parentheses underneath the coefficient estimates. *, **, and *** denote statistical significance of the coefficient at the 10, 5, and 1 percent levels respectively. See appendix table for underlying data sources and sample details.

The saving rate in a country tends to be associated with higher productivity and faster aging speed. Faster population growth and a higher dependency ratio each tend to reduce the saving rate, reflecting the higher consumption that characterizes such countries. Larger net foreign asset positions are associated with higher saving rates, but such correlations are lower in high foreign debt countries. A safer institutional environment tends to reduce saving, possibly by mitigating the precautionary motive. However, heightened financial risks appear to weaken the saving rate. Finally, an increase in the terms-of-trade will raise the saving rate, and better social security (as proxied by public health expenditures) seems to reduce saving, again, possibly by mitigating the precautionary motive.

D. Good and Bad Ways to Achieve External Rebalancing

20. From a normative perspective, the underlying drivers of external rebalancing matter, for both debtor and creditor economies. A general equilibrium framework is required to make a proper assessment of policies simultaneously implemented across debtor and creditor economies in the euro area. As discussed in the ESR, adjustments in the euro area call for further efforts by both debtor and creditor economies, with higher investment and lower saving in some creditor economies, complemented by product market and service reforms, and labor and product market reforms by many debtor economies. We use the IMF’s general equilibrium model EUROMOD to estimate the consequences of such a joint policy push.6 The model also allows for endogenous productivity growth and the international diffusion of productivity shocks across economies.

21. Under a “desired” rebalancing scenario with further progress on structural reforms and higher investment in large creditor economies, external imbalances narrow across the board. The scenario assumes substantial progress is made in reforming product and labor market institutions in all countries, accompanied by measures to stimulate investment in Germany (via both fiscally expansive public investment and private investment incentives), and a small fiscal expansion in the Netherlands, in line with current IMF policy advice.7 Under this scenario, growth dividends in Germany and other large euro area economies could be substantial, ranging from 2–6 percent of GDP cumulatively over the medium term, as compared to the WEO baseline forecast. In addition to the direct impact of these policies on output, public investments are assumed to have a catalytic role in the economy, by raising the return on private capital and thereby increasing private investment through an additional channel. The current account surpluses in Germany and Netherlands would narrow by between 1–2 percent of GDP on the back of higher investments and imports, with positive spillovers to the rest of the euro area (Table 5).

Table 5.A Desired Rebalancing Scenario1/
EconomyReal GDPInflation 2/Real InvestmentReal ExportsReal ImportsCurrent Account2/Real Competitiveness Index 3/
Germany6.40.915.95.05.8−1.9−1.7
France1.70.5−0.33.2−2.40.7−3.9
Italy2.10.50.82.5−2.20.6−3.1
Spain2.20.61.72.7−2.40.9−3.2
Greece2.10.94.72.81.60.0−0.6
Ireland2.81.02.83.62.3−1.2−1.7
Netherlands3.30.98.13.32.7−0.7−1.4
Portugal1.90.73.12.01.3−0.4−2.0
Euro area3.50.76.33.51.1−0.2−2.4
Memo:
United States0.10.00.20.60.90.01.2
United Kingdom0.50.10.51.22.10.10.8
China−0.10.0−0.31.00.60.10.7
Sources: EUROMOD, IMF Research Department and OECD.Note: This scenario assumes that all economies advance structural reforms to close 10-20 percent of their gaps relative to the best practice of the latest OECD structural reform indices on product and labor market institutions. Structural reforms are assumed to be persistent. The scenario further assumes that public and private investment in Germany each increase by 1 percent of GDP, and the Netherlands loosens their fiscal stance by 1/2 percentof GDP. These additional impulses are assumed to be spread over two years (2014 and 2015). The model also allows for positive productivity spillovers from Germany to the rest of the euro area.

22. For the euro area as a whole, the desired rebalancing scenario could also bring remarkable benefits. Real GDP is 3 ½ percent higher by 2019 compared to the baseline, implying that annual growth is higher by about ¾ percentage points on average. Inflation is also higher under this scenario, by about ½ percentage points higher in 2014 and rising to ¾ percentage points over the medium term. Higher investment growth means that the euro area current account surplus is expected to decline by just under ¼ percent of GDP by 2019 (Table 5). It implies that the reform dividends could be large with continued efforts to push structural reforms in both creditor and debtor economies across the euro area. Moreover, as described in the IMF’s 2013 Pilot External Sector Report (ESR), cross-economy spillovers from simultaneous implementation of these policies help boost their overall effectiveness. Germany’s rise in consumption and investment reduce its current account and increase growth, while structural reforms raise productivity growth in all economies.

23. But narrowing current account gaps could also arise from adverse shocks and policy delays. In the case of a “bad” external rebalancing scenario, structural reforms are stalled, debtor economies see a rise in their risk premia (due to an intensification of concerns about the pace of debt deleveraging), and creditor economies keep their fiscal stance unchanged relative to the baseline. Under this scenario, all economies would experience rising current account surpluses, but largely coming from falling demand and slower growth (Table 6). In particular, debtor economies see a persistent shift towards surplus from more depressed domestic demand, while creditor economies see their exports decline due to deteriorating international cost competitiveness, but more than offset by the compression of imports. These results stress the need to avoid policy mistakes and reform delays.

Table 6.A “Bad” Rebalancing Scenario 1/
EconomyReal GDPInflation 2/Real InvestmentReal ExportsReal ImportsCurrent Account2/Real Competitiveness Index 3/
Germany−3.7−2.2−14.2−4.2−4.72.12.7
France−3.1−2.0−10.8−3.7−5.00.91.8
Italy−4.3−1.5−16.0−2.8−5.41.71.0
Spain−4.3−1.4−13.2−4.3−3.40.91.7
Greece−3.9−1.1−13.6−2.8−3.00.70.4
Ireland−6.0−2.3−22.5−8.2−4.54.77.0
Netherlands−3.6−2.3−13.0−4.7−4.21.72.6
Portugal−4.2−1.2−13.3−3.3−4.01.00.9
Euro area−3.8−1.9−13.4−3.8−4.61.51.9
Memo:
United States0.00.00.3−0.6−0.4−0.1−0.7
United Kingdom0.3−0.11.8−1.6−1.4−0.8−1.2
China0.20.01.0−0.90.1−0.2−0.3
Sources: EUROMOD, IMF Research Department and OECD.Note: This scenario assumes that structural reforms are stalled in all countries. It further assumes that there is a reintensification of market concerns about the pace of deleveraging leading to a risk premium rise of 100-300 basis points in debtor economies by 2015. The fiscal stance is assumed to be unchanged in creditor economies.

E. Concluding Remarks

24. External rebalancing in the wake of the financial crisis has been highly asymmetric. Debtor economies in the euro area have seen substantial improvements in their current accounts, primarily through drastic declines in investment across sectors and reductions in government deficits. Creditor economies (such as Germany and the Netherlands) have mostly maintained or even further strengthened their surpluses, with declining household and corporate investment rates and rising corporate and government saving.

25. Mirroring the current account surplus, the euro area has seen a move towards net capital outflows, dominated by net other (bank) investment flows from creditor economies to the rest-of-the-world. Combined with evidence on net foreign asset positions, most of the other (bank) investment asset exposures held by euro area economies is still largely versus other euro area economies, but this has been changing, with a rise in other (bank) investment exposures extra-euro area.

26. A regression analysis of aggregate saving and investment by the euro area economies highlights how deviations from predicted valuesin saving and investment have contributed to current account imbalances in recent years. Relative to the model-estimated equilibrium levels, creditor economies (such as Germany and the Netherlands) have seen large and persistent negative investment and positive saving deviations, since the mid-2000s, generating large current account surpluses. On the other hand, debtor economies (like Greece, Portugal, and Spain) experienced positive investment deviations and, in some cases, negative saving deviations prior to the financial crisis, leading to their large and persistent current account deficits in the run-up to the crisis. Investment and saving deviations in debtor economies have been reduced in the wake of the crisis, primarily through rapid investment shedding, and to a lesser extent by increases in saving. However, corrections in the opposite directions by the creditor economies have not been realized. In fact, the positive saving and negative investment deviations in these economies have increased since 2010, leading to further strengthening of their current account surpluses.

27. Scenario analyses highlight that external rebalancing can arise positively, from structural reforms and improved policies, or negatively, from adverse shocks and policy delays. Under a “desired” rebalancing scenario, with positive investment policies in creditor economies and more structural reform progress in debtor ones, external imbalances in all economies narrowed. More importantly, potential growth and productivity improve across the board, reflecting more efficient product and labor markets. The growth dividends for euro area economies can be substantial, ranging from a ½ to ¾ percentage point rise in growth annually over a five year horizon, accompanied by a faster closing of output gaps and higher inflation by around ½ percentage point.

28. The overall message from the evidence and scenarios is that further progress on policies to narrow current accounts imbalances in the euro area is needed. Policies to stimulate investment by creditor economies and the passage of structural reforms by debtor economies would narrow external imbalances across the board, while improving potential growth and productivity. A lack of policy adjustments may lead to a sustained deflation, and lower growth and productivity, even though the external imbalances would also shrink. External rebalancing alone is not sufficient—appropriate adjustment is about undertaking policies to achieve both internal and external balance.

Table A1.Data Sources and Estimation Sample for the Saving and Investment Regressions
VariableData Source
Current Account to GDP ratioIMF World Economic Outlook
Savings to GDP ratio (current account minus fixed investment, both to GDP)
IMF World Economic Outlook
Fixed Investment to GDP ratioIMF World Economic Outlook
Tobin’s Q (median of individual publicly listed corporations’ Tobin’s Q by economy)
Bloomberg and IMF staff calculations
Real short-term deposit rate (ex post)IMF World Economic Outlook
Long-term government bond yieldIMF World Economic Outlook
Political Risk Services ICRG and IMF staff
Safer Institutional Environment Indexcalculation (IMF EBA)
VIX/VXO times (1- Capital Control Index), once-laggedBloomberg, Quinn (2011), and IMF staff calculations
Output gapIMF World Economic Outlook
Expected 5-year ahead growthIMF World Economic Outlook
Per capita GDPIMF Inter natio nal Financial Statistics
Aging speed (expected change in the old-age dependency ratio over next 20 years)
United Nations World Population Prospects
Population growthUnited Nations World Population Prospects
Old-age dependency ratio (relative to world average)
United Nations World Population Prospects
Net foreign assets to GDPIMF Inter natio nal Financial Statistics
Net foreign assets to GDP times indicator for less than -60 percent of GDP
IMF Inter natio nal Financial Statistics
Commodity terms-of-trade gap times exports plus imports to GDPIMF International Financial Statistics and External
Balance Assessment database
Cyclically-adjusted Primary Fiscal BalanceIMF World Economic Outlook
Public health expenditures to GDP (relative to world average), once-lagged
IMF Government Finance Statistics
Sample Economies (1986-2013)
AustriaItaly
BelgiumNetherlands
Czech RepublicNorway
DenmarkPoland
FinlandPortugal
FranceSpain
GermanySweden
GreeceSwitzerland
HungaryUnited Kingdom
Ireland
References

Prepared by John C. Bluedorn, Shengzu Wang, and Tao Wu (EUR).

Based on the Net Foreign Asset (NFA) positions as of 2013, each euro area economy is assigned to one of two bins: creditor or debtor. See Table 1 for the full listing of assignments. Throughout this note, the euro area aggregate is based on the euro area membership as of January 2014 (the EA-18).

The group of IMF program economies includes countries that have or recently had an IMF program.

Due to data constraints, only fixed investment is available at the sectoral level; this is what is shown in Figures 2 to 5. Hence, there may be a difference between the current account calculated from the underlying sectoral saving and investment data and the current account from the balance of payments or calculated using overall investment (which includes inventory investment).

Please see Chapter 1 of Germany’s 2014 Article IV Consultation Selected Issues Papers for further details on saving and investment trends in the different sectors of the German economy.

In fact, as seen in Figure 5, relative to 1999, investment is slightly higher in these economies.

Note that net capital inflows may not equal the current account from the balance of payments due to statistical discrepancies. Moreover, the underlying data for net capital inflows and its components are expressed in current USD, and then divided by nominal GDP in USD (aggregated for whichever group of economies is relevant). This means that exchange rate fluctuations within year may account for some of the differences between the underlying current account data (recorded in local currency terms) and net capital inflows.

Returns and return differences are calculated in the manner described in Lane and Milesi-Ferretti (2007) and Habib (2010).

Due to differences in methods and data, the results here cannot be identical to those from the EBA. However, as shown in Table 3 and 4, the qualitative and quantitative implications from these two studies are very similar.

The model is a multi-economy, general equilibrium, overlapping generation model combining both micro-founded and reduced-form formulations of various economic sectors with non-Ricardian agents. It is part of the IMF’s Flexible System of Global Models (FSGM).

More specifically, structural reforms, as proxied by the OECD’s indices of structural economic characteristics, are presumed to close 10 to 20 percent of the gaps between each economy’s structural characteristics (such as product and labor market institutions) and the OECD’s best practice. The German public investment increase is assumed to be 1 percent of GDP, spaced over 2014 and 2015 (for an overall fiscal expansion of slightly below ½ percent of GDP per year), accompanied by measures to increase private investment by about 1 percent of GDP, spaced over the same horizon. The fiscal stimulus in the Netherlands is ½ percent of GDP, similarly spaced over two years.

Other Resources Citing This Publication