VoxEU Column EU institutions EU policies Europe's nations and regions

Returns to German savings invested abroad

Through the Eurozone rescue mechanisms, Germany provided the periphery with hundreds of billions in debt at very low rates. There is a widely held notion that these savings would have been better used at home. This column challenges this notion, presenting evidence that Germany’s net asset position held up well, remaining much higher than domestic returns. The main reason is that Germany’s part in the rescue operations was actually much smaller than its claims towards the periphery.

It has become by now almost a cliché that Germany’s excess savings are being wasted abroad. But this is a popular misconception based on the divergence between the cumulated current account of Germany and its net international investment position (NIIP). A closer look at the data actually suggests that the net position is probably not measured correctly and that the observed returns on German investment abroad have remained above most domestic returns.

Advocates of ‘German rebalancing’ policies point to the persistently large current account surplus and the low domestic investment rate (European Commission 2014), while others stress allegedly poor returns on Germany’s foreign savings (see among others Klär et al 2013). The latter argue that German investors appear to have suffered large losses after the crisis started because a wide gap then opened between the current account surpluses and the measured changes in Germany’s net investment position, and thus also between the cumulated current account balance and the net position.

Indeed, in comparing the cumulated current account balance to the net investment position, one finds a substantial ‘gap’ – by the end of 2015, the cumulated current account surpluses of Germany were nearly €600 billion higher than the net international investment position. Germany is one of the few countries for which 2015 data are already available and one can clearly see the gap (Figure 1). 

A first point to note is that a similar gap can be observed for most EU countries. For example, the Spanish data show a net investment position that is over €200 billion worse than the cumulated current account figure. The gap is smaller in absolute terms than the one for Germany, but larger as a share of GDP. Nobody is arguing, however, that Spanish savers are wasting their money by investing abroad (or that foreign investors are making extraordinary returns in the country).

Figure 1. Rising gap between cumulated current account and net investment position, 1995-2015 (€ billion)

Germany

Spain

Note: Cumulated from 1995 until 2015. In the data for Germany, both measures are based on BPM6, while the NIIP for Spain also includes BPM5 data (the measures for Spain for the years with both observations do not differ greatly).
Source: Own calculations based on 2015 data from Bundesbank and Eurostat.

Relative to (known) assets, the gap amounts to only about 8%. Table 1 below shows that many other countries record a gap of a similar amount, if compared to the size of total assets. Both Spain and Italy exhibit a larger gap as a percentage of total assets than Germany, while France is barely below the German figure. None of these countries is dubbed a bad investor and there is no convincing reason why Germany should be singled out.

Table 1. ‘Gap’ in European comparison: The cumulated current account balance and the NIIP

  Gap in national currency,
billions, 2013
Gap in % of IIP assets
Germany 600 7.8%
Spain 220 14.4%
Italy 190 8.8%
France 455 7.0%
Netherlands 435 6.0%
Austria 45 4.8%
Portugal -20 -5.9%
United Kingdom -280 -2.7%

Note: Cumulated from 1995-2013. Based on BPM6, gaps filled with BPM5.
Source: Own calculations based on 2015 Eurostat data.

The narrative of the wasteful German investor thus hinges on the precision of the current account and net investment position data. Current account data are usually assumed to be rather reliable, but the net investment number is subject to very large measurement errors since the net position is calculated as the difference between two very large stocks (total foreign assets and liabilities), both of which are imperfectly measured.[1]

Moreover, detailed calculations of individual asset and liability positions show that valuation losses account only for a fraction of the gap (about one-third, according to Allianz 2014). This was partly due to falling interest rates, which led the value of assets in Germany to rise and hence German external liabilities went up. This was the case particularly for the German stock markets (increasing the measured value of FDI in Germany and foreign portfolio equity holdings). Moreover, as interest fell, the value of German long bonds rose. Given the large foreign holdings of German bonds and equity this led to substantial defaults and write-off losses, which accounted for about one-third of the gap (see Figure 2). These losses were incurred between 2007 and 2011, due to the Global Crisis and its aftermath. Since 2012 this positive has turned balanced and in 2014 even substantially positive, with €80 billion.

Figure 2. Effect composition in Germany’s international investment position (IIP), cumulated change, 2008-2014, billion euros

Source: Authors’ elaboration on Schipper (2015a).

Given these problems with the reported net investment position, we propose to look at an alternative measure of the investment performance – the investment income balance; that is, overall income balance after deducting remittances and other non-investment income. This can be thought of as the return on the net position. Since 2003, Germany’s investment income balance has moved from a persistently balanced position to a large surplus of around 2% of GDP, not far behind Japan. In Europe only Denmark, Sweden, and the Netherlands show higher values.

The main evidence against the thesis that Germany is wasting its savings abroad is thus that the (measured) returns on German investment abroad have held up rather well.

We find that the German net investment position had a yield (net investment income divided by net position) of around 8% in the last 10 years with a decline since 2011 to ‘only’ 4.5% (Figure 3a).  Over the entire period, the return on the net investment has been higher than that on German government bonds.

Foreign returns have also been higher than alternative investments in Germany. The interest on bank loans to German non-financial corporations has also consistently been lower than the return on the (net) foreign positions (taking into account that German banks also had a non-negligible loss rate on domestic loans during this period would strengthen this conclusion). The conclusion by DIW (2013) and others that Germany is losing its shirt abroad and would gain from more investment at home is thus not supported by the data.

Figure 3a. Interest yields in comparison      

Figure 3b. Return on asset and liability position

Source: Eurostat (2015).

The relatively high return on the net assets position is not only due to a superior performance of German investment abroad, but also to the very low return on foreign investment in Germany. The detailed data confirm that German investors managed to earn a relatively high positive return on their foreign assets, which exceeds that paid to foreign investors in Germany (Figure 3b).[2] This difference in returns has actually increased substantially and is now much higher than before the crisis.

One could thus conclude that German savers are actually doing better than their foreign peers. If the returns on assets and liabilities were the same, Germany's investment income surplus would be cut by one-half.

It should not be surprising that Germany pays less on its liabilities than on its assets, especially since the start of the EZ Crisis, when capital fled to Germany, accepting very low returns. Moreover, German government bonds enter into the foreign exchange reserves of many countries; and the yield on ‘Bunds’ is close to zero (actually negative for many maturities). The high return on German assets abroad is also due to a change in their composition (see Figure 4). The share of high-yield FDI has persistently risen over the past 10 years.

Figure 4. Investment income (receipts) by type (€ billion)

Source: Eurostat (2015).

More surprising is the finding that even the EZ periphery countries benefitted from a lower interest cost of their liabilities than they earn on their foreign assets. For example, Spain paid on average 2.1% on its liabilities in 2014, but earned 2.8% on its assets. The average cost of liabilities was actually lower than on longer term Spanish government bonds, most likely because of the very large refinancing of Spanish banks by the ECB.  

The obvious objection to the data on returns is that they could also be affected by measurement errors. But the data on returns on assets and liabilities separately look much more reasonable and would not be affected by a +/- 10 measurement error, which would change a return from 3.3 to 3.0%, but would not affect the sign.

Concluding remarks

All in all, the data thus contradict the widely held notion that Germany lost out by the Eurozone rescue mechanisms that provided the periphery with hundreds of billions of euro in debt at very low rates, and that German savings would have been better used at home. On the contrary, it seems that Germany's returns on its net asset position has held up well and remains much higher than domestic returns. 

The key reason for this is that Germany's part in the euro rescue operations was much smaller (at less than 30%) than its claims towards the periphery. Germany has basically been able to rope in the rest of the Eurozone to provide about 70% of the cheap financing that had to be provided for the euro periphery, although these other countries (Italy, France, etc.) held only a fraction of the claims against the periphery. Consequently German savers have fared quite well over the last decade in their foreign investments. This is what one would expect after all – with the EZ Crisis, risk premia rose throughout the periphery, but there were no large losses except in Greece, which accounts for only a fraction of German foreign investment. Moreover, rates fell in Germany. It was thus to be expected that foreign investment would become more attractive than domestic investment once the initial panic had abated.

References

Allianz (2014) “German investments abroad: A bad deal?”, Working paper 75, 11 August.

Deutsche Bundesbank (2014) “Discrepancy between changes in net foreign assets and the cumulated financial account: An unsuitable indicator of wealth losses”, Vol 66, No 5, Monthly Report for May.

Deutsche Bundesbank (2014a), “Methodological changes affecting Germany’s international investment position”, Monthly Report, October: 22-24.

DIW (2013) “Deutschland muss mehr in seine Zukunft investieren”, DIW Wochenbericht Nr 26.

European Commission (2014) “Macroeconomic imbalances Germany 2014”, Occasional Papers 174, March.

European Commission (2015) “Alert mechanism report 2016”, COM (2015) 691 final, 26 November 2015.

Frey, R, U Grosch and A Lipponer (2014) “Traps and pitfalls in quantifying German investors’ losses on external assets”, Wirtschaftsdienst 94.

International Monetary Fund (2009) “Balance of payments and international investment position manual”, International Monetary Fund, Washington, D.C.

Klär, E, F Lindner and K Šehovic (2013) “Investition in die Zukunft? Zur entwicklung des Deutschen auslandsvermögens”, Wirtschaftsdient, 93 Jahrgang, 2013, Heft 3.

Schipper, U (2015a) “Transaction and valuation effects on Germany's international investment position (IIP) – new statistical approaches and IIP trends”, Deutsche Bundesbank.

Schipper, U (2015b) “Transaction and valuation effects in the German IIP”, presentation at IFC Satellite meeting at the ISI 60th WSC, Rio de Janeiro, 24 July.

Endnotes

[1] The methodology for calculating the NIIP has recently been revised from the BPM5 to the BMP6 (IMF 2009). The application of the new approach shrinks the German NIIP by €350 billion (for 2013).  That a simple change in methodology can have such a large impact on the NIIP highlights the problems in the measurement of the NIIP.

[2] These can be estimated by the current account income debit and credit position as a percentage of the NIIP liability and asset position, respectively. Since 2004, the return on assets has surpassed the payment on liabilities.

2,625 Reads