What drives the German current account balance?

(1)

Universiteit van Amsterdam

Amsterdam School of Economics

What drives the German Current Account

Balance?

Master Thesis

Author: Laura Schreiber Student ID: 11825219 E-mail Address: laura.schreiber@student.uva.nl Supervisor: Dr. Naomi Leefmans Date: 15.07.2018 Programme: MSc Economics Track: International Economics and

Globalisation Second Reader: Dr. Dirk Veestraeten

Word Count: 13584

(2)

Statement of Originality

This document is written by Laura Michaela Schreiber who declares to take full responsibility for the contents of this document.

I declare that the text and the work presented in this document are original and that no sources other than those mentioned in the text and its references have been used in creating it.

The Faculty of Economics and Business is responsible solely for the supervision of completion of the work, not for the contents.

(3)

Abstract

Germany running the world’s largest current account surplus in 2016 has been a widely disputed issue by politicians and academics. Critics claim that this sizable imbalance is obstructing economic prosperity in the rest of the world and in particular, in the European Economic and Monetary Union. However, it is essential to understand the drivers of the German current account balance to suggest rebalancing measures. The results of a descriptive analysis of the macroeconomic environment in Germany between 1980 and 2016 and a subsequent panel data analysis including 24 OECD countries from 1995 to 2016 are in line with previous findings, that the rapid population aging has been susceptibly driving the German current account balance upwards, reflecting higher savings to sustain the German pension system in the future. Further, the regression outcome suggests that euro area membership significantly increases the current account balance if domestic inflation is below the euro area average. Moreover, controlling for high-technology exports as a share of overall exports, the results suggest that this measure of competitive advantage can account for at least 2 percentage points of the German current account balance to GDP between 2000 and 2016.

(4)

1 Introduction

In 2016 Germany ran the world’s largest current account surplus, amounting to nearly 300 billion US dollars which equals more than 8 % of its GDP. While many national politicians consider the latter a reflection of German producers’ high competitiveness, critics all over the political and academic spheres remark that this imbalance hampers economic growth in the rest of the world and threatens the effectiveness of a single monetary policy in the European Economic and Monetary Union (EMU). Therefore, the German government has been requested to take action in order to lower Germany’s current account balance. Yet, to give adequate policy recommendations, it is essential to understand the underlying drivers of the German current account balance. Thus, this paper discusses the question: What was driving the German current account balance between 1995 and 2016? While there are plenty of empirical studies tackling the determinants of current account imbalances, my research additionally examines the dynamic effects of euro area membership in form of the inflation differential of a member country to the euro area average as well as the effect of the share of high-technology products in overall exports, measuring product quality and the degree of competitive advantage in the export sector.

Source: The World Bank

Figure 1.1 shows that up until the 1970s, the German current account was close to balance as was the case for most other countries in the world. Increasing trade liberalisation and financial globalisation led to higher cross-border capital flows in the 1980s. Consequently, the German current account balance rose up to 4% of GDP during the decade. In the 1990s high public investment following the German reunification implied a negative shock to the German current account balance and entailed a decade of current account deficit. However, since the

-5 -4 -3 -2 -1 0 1 2 3 4 5 6 7 8 9 10 19 71 19 72 19 73 19 74 19 75 19 76 19 77 19 78 19 79 19 80 19 81 19 82 19 83 19 84 19 85 19 86 19 87 19 88 19 89 19 90 19 91 19 92 19 93 19 94 19 95 19 96 19 97 19 98 19 99 20 00 20 01 20 02 20 03 20 04 20 05 20 06 20 07 20 08 20 09 20 10 20 11 20 12 20 13 20 14 20 15 20 16 %

(6)

early 2000s the German current account balance has been on the rise again. Therefore, the decade following the reunification might be interpreted as an interruption of an upward trend that started in the early 1980s.

In what follows, chapter 2 defines the current account, what it reflects and why imbalances can be problematic. Subsequently, chapter 3 reviews relevant findings in the academic literature on the determinants of the current account balance. Chapter 4 presents a descriptive analysis of the macroeconomic environment in Germany from 1980 to 2016, plotting the evolution of key variables that affect savings and investment as well as variables influencing the trade balance. Due to restricted data availability for the examined variables for most countries before 1995 and to enhance the sample size, yearly panel data on 24 OECD countries from 1995 to 2016 is used in a subsequent panel data analysis in chapter 5. That is, the current account balance to GDP is regressed on a vector of determinants that are found relevant in the analysis of this paper. Consistent with previous research, the regression outcome suggests that the government budget balance, the real interest rate and the net foreign asset position increase the current account balance to GDP. In addition, including the high-technology share in exports to measure the effect of product quality and competitive advantage on the trade balance and current account balance, the results imply that high-technology exports explain at least 2 percentage points of the German current account balance to GDP between 2000 and 2016. On top of that, the impact of the common currency on imbalances in the euro area is measured by controlling for the inflation differential between member countries and the euro area average if the euro was used as the domestic currency in the respective time period. I find that being a member of the euro area and having an inflation rate of 1 percentage point above the euro area average significantly lowers the current account balance by at least 0.6 percentage points. Lastly, chapter 6 concludes my analysis.

2 The Current Account

To begin with, it is essential to define the current account within the framework of balance of payments accounting, see section 2.1. Further, section 2.2 derives theoretical determinants of global current account imbalances. Lastly, possible dangers of these imbalances are indicated in section 2.3.

2.1 Definition

Together with the capital and financial account, the current account makes up a country’s balance of payments (BoP). Under a floating exchange rate regime, the BoP is

(7)

balanced by definition such that the current account always equals the negative of the capital and financial account. The current account itself is defined as the difference between a country’s savings and investments. A country that runs a current account surplus is a net lender to the rest of the world and domestic saving is higher than domestic investment. On the other hand, if a country runs a current account deficit, it is a net debtor and investment exceeds saving.

Further, the current account balance consists of three components: 1. The trade balance (exports minus imports).

2. Net income from abroad (net earnings on foreign investment, net labour income from abroad and net property and entrepreneurial income from abroad).

3. Net current transfers (e.g. foreign aid, grants, donations, tax payments and remittances from abroad).

2.2 Current Account Imbalances

Generally, the existence of global current account imbalances increases the default risk in the global financial system due to the interconnectedness and interdependence of agents and institutions. For instance, the origins of the Global Financial Crisis 2008-09 can be traced back to increasing unsustainable global imbalances among other factors as e.g. the ongoing deregulation of financial markets (Obstfeld, 2017). Nevertheless, moderate and temporary imbalances are not necessarily an indicator of financial instability. Current account surpluses and deficits can be caused by business and financial cycles and allow countries to smooth consumption and investment over time. For instance, running a current account deficit enables especially developing countries to pursue investments which would not have been feasible without foreign capital inflows. On the other side, net savers can benefit from foreign investment opportunities. Motives for investing abroad can be the desire to improve one’s portfolio diversification, risk-aversion or -appetite that does not match the perceived risk in the domestic economy or the general scarceness of investment in a country. In this section, the determinants and dangers of current account imbalances are examined more closely.

2.2.1 Theoretical determinants of Current Account Imbalances

Since the current account is defined as the difference between savings and investment, it is largely determined by international capital flows and financial variables. Another approach to find current account determinants comprises the examination of trade in goods and services, as the trade balance usually makes up the major part of the current account balance. In the following, economic theories on both approaches are presented.

(8)

2.2.1.1 Financial Perspective

Initially, this section scrutinises current account balance determinants that directly influence savings and investment decisions of agents. It underlines which assumptions are made in economic theories that model these decisions and how these are restrictive or not sufficient for an application to reality.

In the neoclassical model as according to Obstfeld and Rogoff (1994), households and firms face dynamic optimisation problems with respect to savings and investment, subject to a no-Ponzi condition that requires foreign debt to be repaid. The model assumes rational agents with perfect foresight that prefer to smooth consumption inter-temporally. Within this framework current account imbalances are the result of the transition from autarky to an open economy with perfect capital mobility. Equilibrium is reached as international capital flows equalise the return on capital across countries. A country’s capital return in autarky is determined by its relative capital scarcity and productivity growth prospects. Thus, the model predicts a flow from capital-abundant to capital-scarce countries. Yet, the neoclassical model cannot explain major global capital flows from developing countries and emerging economies such as China to advanced economies as e.g. the US.

The asset market to exchange rate approach by Krugman and Obstfeld (2009) adds to the above model by taking into account liquidity and risk of different assets that are factored into the expected rates of return. In equilibrium, floating exchange rates clear the international asset market by equalising the expected returns on deposits when measured in the same currency. Under a fixed exchange rate regime, the endogenous variable reaching equilibrium concerns foreign currency reserves. Therefore, the model does not necessarily predict capital flows from rich to poor countries if asset liquidity and risk are not controlled for. However, risk and liquidity are not perfectly measurable and equally perceived by every agent when there is asymmetric information as it is often the case in reality. Additionally, it has to be taken into account that in the case of Germany, exchange rates are not solely determined by capital flows in and out of the country. That is, as Germany is member of a currency union, its exchange rates are affected by capital flows in and out of the whole euro area which consists of countries that differ in their rates of return on capital and attraction of foreign investment. Further, the above models have strong assumptions regarding perfect information symmetry and foresight as well as rational behaviour of agents. They can be supplemented by demographic variables and country-specific saving behaviour. For instance, relatively younger populations tend to borrow and invest more, middle-aged populations have a higher propensity to save for

(9)

retirement while a population with a high old age-dependency ratio will either consume or bequest their savings (Ciocyte and Rojas-Romagosa, 2015).

In summary, the theoretical approaches to current account determinants presented in this section attribute current account imbalances to international capital flows which are attracted by differences in rates of return as well heterogeneity in liquidity and risk across countries. Further, some restrictions of these models are strong assumptions, restricted applicability to countries that are part of a currency union as well as the omittance of country-specific saving behaviour and demographic variables.

2.2.1.2 Trade Perspective

After examining theoretical determinants of the current account balance that are connected to financial variables, the following section will focus on variables affecting competitiveness of domestic firms and industries that affect the current account balance of a country through the trade balance. Countries running current account surpluses usually exhibit trade surpluses, i.e. they export more than they import. But what makes a country an exporter of a product?

The most straight-forward explanation for international trade flows is the concept of absolute advantage (Smith, 1776). Accordingly, in a one factor model, a country will export a good if it can produce a higher quantity with the same input as other competitors, i.e. if its labour productivity is relatively higher. Notwithstanding, this implies that a country which has higher productivity levels in all sectors than its trading partners would be an exporter of all traded goods. This is not observed in reality, as especially labour-intensive goods are exported by developing countries which often display lower levels of labour productivity compared to advanced economies. The principle of comparative advantage introduced by Ricardo (1821) gives an explanation for this phenomenon. In his free trade model agents export a product that they have a comparative advantage in. A country has a comparative advantage in producing a good if its opportunity costs of producing that good are relatively low compared to its trading partners’ opportunity costs. Hence, even if a country has an absolute advantage in producing a good, it might import it in the case that its opportunity costs of not producing other goods instead are sufficiently high. Comparative advantages can arise through a country’s factor abundance, see Heckscher et al. (1991) who build on the Ricardian model with the Heckscher-Ohlin theorem: In a two-good model a country exports the good that for its production uses intensively the country’s abundant production factor and imports the other good. Yet, the latter implies that two countries will trade more bilaterally if they differ strongly in their factor abundance. In reality, however, we observe high trade activity between countries which are

(10)

similar in their fundamentals, e.g. Germany and the Netherlands. Another approach to trade flows and industry agglomeration that can account for these patterns is competitive advantage. Within the framework of the new trade theory (see e.g. Nielsen (1995)), competitive advantage arises from economies of scale over time which are based on the idea that the marginal cost of production is decreasing because of learning effects. This allows a large firm or industry to produce at lower costs than its domestic and foreign competitors by being the first-mover to expand production, to export to a foreign market and become a global monopolist. Economies of scale can arise externally, affecting a whole industry, or internally to the individual firm. In both cases, a degree of product differentiation1_{is required additionally such that in an}

imperfectly competitive market, a firm is able sell at a price above their marginal costs (Helpman and Krugman, 1985). Additionally, external economies of scale arise through location advantages, such as relatively low corporate tax, and reinforce themselves through the centripetal forces of the agglomeration theory (Neary, 2001). Accordingly, in a manner of circular causation, geographical concentration of production leads to a larger market as it attacks skilled workers. An increase in the workforce then also increases demand. In turn, a larger market allows for further exploitation of economies of scale and attracts more firms. Hence, if there are many industries in a country that benefit from competitive advantage due to location or first-mover advantages, it might have a relatively large number of global monopolists that drive domestic export activities, increase the trade balance and have a positive effect on the current account balance.

Moreover, trade flows are affected by the exchange rate. As according to Pilbeam (2006), the elasticity approach seeks to predict the effect that a depreciation of the nominal exchange rate has on the current account balance. An important and restrictive assumption of the model is that supply is perfectly elastic, i.e. changes in demand do not affect the producer price. Demand elasticities, on the other hand, can differ from country to country and determine to what extent a change in the exchange rate will affect the current account balance. Particularly, the elasticity approach is coined by the Marshall-Lerner condition. The condition states that, if the absolute value of the domestic import demand elasticity and the foreign export demand elasticity exceeds one, a depreciation of the exchange rate will improve the current account balance. There are two effects which play a role in this outcome: the price effect and the volume effect. In form of the price effect, a devaluation of the domestic currency makes imports more

1_{Product differentiation implies imperfect substitutability of products across producers.}

Through product differentiation producers gain market power which leads to an imperfectly competitive market where prices exceed the marginal costs of producers.

(11)

expensive for domestic consumers and exports cheaper in the foreign currency. Ceteris paribus, this would worsen the trade balance. Nevertheless, due to the volume effect, export and import volumes change, too. Since imports become more expensive in the domestic country, demand for imports will decrease while export volume will increase as domestic producers become more competitive in the foreign market. If the Marshall-Lerner condition holds, the volume effect outweighs the price effect. While there is empirical evidence that in the long run the Marshall-Lerner condition holds2_{, the International Monetary Fund (1984) finds that in the short}

run (over a 6-months horizon) price elasticities are lower and do not meet the condition. This pattern over time is called the J-curve and implies that a depreciation of the domestic currency temporary worsens the current account balance before it has a positive effect.

The above elasticity approach assumes the purchasing power parity to hold. However, nominal exchange rates do not always reflect the value of the currency in terms of purchasing power. A measure that can help examine if a currency’s real value is higher or lower than implied by market exchange rates is the real effective exchange rate (REER). The International Monetary Fund (2018) defines the REER index as the nominal effective exchange rate3_divided

by a price index of cost. A low REER could be a hint that the domestic currency is undervalued. If the REER index decreases, domestic prices become relatively lower in comparison to the rest of the world which makes domestic producers more competitive abroad and imports relatively more expensive in the domestic currency. Therefore, a decrease in the REER is expected to increase the trade balance and the current account balance. In turn, a high REER can imply that the domestic currency is overvalued A rise in the REER index reflects a relative increase in the domestic price level, national producers become less competitive and imports become cheaper for domestic consumers. Thus, an increase in the REER theoretically worsens the trade balance and the current account balance.

In summary, labour productivity, production cost as well as economies of scale and product differentiation in form of competitive advantage as well as the REER theoretically affect the competitiveness of domestic industries and individual firm, such that they are enabled to produce at lower cost than their competitors. The more competitive industries or firms in a country are to the rest of the world, the higher the country’s trade balance and current account balance is expected to be.

2_{See e.g. Gylfason (1987) who measures demand elasticity of exports and imports of 16 industrial and}

9 developing countries over the horizon of two to three years.

(12)

2.2.2 Dangers of Current Account Imbalances

Large and persistent current account to GDP imbalances can endanger the sustainability of a current account deficit and the ability of the borrower country to pay back the debt. This can deteriorate the economic situation of the debtor country, hamper growth and make its economy more prone to crises. While a high and persistent current account surplus does not have direct negative effects on the economy that is running it, a surplus reflects the existence of current account deficits in the rest of the world. If the deficit of one of the country’s trading partners becomes unstainable, agents might default on parts of their debt. This can cause a crisis which also affects the country running the current account surplus. Therefore, imbalances can increase systemic risk in the global economy.

Besides, current account imbalances in a currency union can create heterogeneity between countries that might threaten the effectiveness of a single monetary policy. Hence, the European Union created the macroeconomic imbalance procedure in 2011. Its indicator of macroeconomic imbalance is the three-year average of the current account in terms of GDP with the thresholds -4% (current account deficit) and 6% (current account surplus) which shows a higher tolerance margin for a current account surplus. However, Germany has been violating this criterion as its current account balance has been exceeding 6% of GDP since 2012. Thus, the country has been receiving in-depth reviews and monitoring by the European Commission.

3 Empirical Evidence Current Account Imbalances

While there is an abundance of literature on global current account imbalances, the following literature review will focus on the specific case of countries running current account surpluses. In particular, empirical evidence on the determinants of the German current account balance will be examined in the section 3.1. Finally, section 3.2 examines general empirical evidence on current account balance determinants.

3.1 The German Current Account Surplus and Policy Recommendations

Before looking at general current account determinants, this section examines the German current account balance in particular. The research presented in the following tackles the German current account balance during surplus periods in the 1980 and after 2000 as well as policy recommendations aimed at rebalancing of the German current account.

Initial publications tackling the German current account imbalance go back to the 1980s when a surplus emerged for the first time in the economy. For example, Neuthinger (1989) makes reservations about the growing imbalance and the fact that it is mostly considered a

(13)

positive sign of competitiveness and quality of the German export sector. He states that flexible exchange rates mostly offset relatively unit production cost advantages, since high demand for the domestic currency following increased demand for relatively cheaper domestic exports would lead to an appreciation of the domestic currency. Further, he notes that particularly the high savings rate of businesses and private households together with low domestic demand are major drivers of the German current account surplus. In order to rebalance, his recommendation consists of combining a medium-term fiscal stimulus to restore demand together with a revaluation of the Deutsche Mark.

In the 1990s the German reunification boosted public investment which entailed a current account deficit until 2001. In the early 2000s, current account growth picked up and reached a surplus of 6% of GDP just before the Global Financial Crisis in 2007. Thereafter, it remained at around 5% until it was on the rise again in 2011, reaching its highest to-GDP ratio of 8.5% in 2015. This emerging surplus sparked the academic debate on its drivers. At the same time the above-mentioned claim that the exchange rate offsets advantages in competitiveness of a country did no longer apply after the introduction of the Euro as a common currency. Also, the recommendation to combine fiscal expansion with a currency revaluation became unfeasible with respect to the latter part.

Investigating the determinants of the German current account surplus, Felbermayr et al. (2017) find in a descriptive analysis that the German current account surplus can be mainly attributed to the aging population which translates into a higher savings rate not only in the private sector but in the overall economy. They oppose the hypothesis that a wage increase is an essential factor, as they note that after 2007 the labour share of GDP has been increasing in Germany while the current account balance was on the rise, too. In order to reach a more balanced current account, they recommend a corporate income tax reform including loss offset, accelerated depreciation and R&D tax credits to boost domestic private investment.

Findings from Wörgötter and Coricelli (2012) investigating the drives of the German current account surplus suggest that the increasing productivity growth gap between the manufacturing and service sector has influenced the German current account in the 2000s. They note that this reflects productivity improvements in the manufacturing sector as well as a long-run weakness in services. Using an error correction model and OECD panel data over the period 1970 to 2007, they find that relative total factor productivity (TFP) of the manufacturing sector to the service sector significantly increases the current account. Yet, in a time series analysis for Germany their results are weaker and only suggestive. The latter is probably the result of

(14)

omitted variable bias and the insufficient sample size. Moreover, they find significant coefficients for the old and young age dependency ratio as well as highly significant coefficients for the working age population growth and the real interest rate.

3.2 Current Account Imbalances of other Countries

The following analysis is partially based on a literature survey by Ciocyte and Rojas-Romagosa (2015) who look at various empirical studies4_{on the determinants of the current}

account balance to GDP and complemented by an analysis of Han and Sin (2016) who particularly analyse the Korean current account surplus. All methodologies examined in this section are all based on the simple reduced form fixed effects regression of the current account determinants

𝐶𝐴

𝐺𝐷𝑃!" = 𝛼! × 𝐷! + 𝛽𝑋!"+ 𝜀!"

where CA is the current account balance in nominal terms, GDP is in nominal terms, 𝛼!

are country-specific intercepts, 𝐷_! are dummy variables that are 1 for country i and 0 otherwise, β is a vector of coefficients on the exogenous X variables, ε is the error term, and 𝑖 and 𝑡 are the country and time indices respectively. However, the studies reviewed by Ciocyte and Rojas-Romagosa (2015) and the regression analysis of Han and Sin (2016) differ slightly in their data sample and time horizon: That is, Chinn and Prasad (2000) examine a sample of 18 industrial and 71 developing countries from 1971 to 1995 using cross-section and panel data. Gruber and Kamin (2005) base their analysis on the same approach taking a panel of 61 countries in the period 1982 to 2003, focussing on the US current account deficit and the large Asian developing economies’ surpluses. Moreover, Higgins (1998) chooses time series data from 1950 to 1989 for 100 countries that are heterogeneous in their level of development using five-year-averages of the variables to avoid serial correlation and cross-sectional data from 1954 to 1992 using 13-year averages to ensure the measurement of cross-sectional instead of temporal variation in the data. Similarly, Barnes et al. (2010) use OECD panel data from the period 1969 to 2008 in a five-year period-average model to filter cyclical effects. In contrast, the International Monetary Fund (2013) purposely choses annual data instead of period-averages to find cyclical sources of current account volatility in the so-called External Balance Assessment (EBA). Within the period 1986 to 2010 they estimate the effect of current account drivers for a sample of 49

4_{Ciocyte and Rojas-Romagosa (2015) review empirical evidence of current account}

determinants found by Chinn and Prasad (2000), Gruber and Kamin (2005), Higgins (1998), Barnes et al. (2010), the International Monetary Fund (2013) as well as Gossé and Serranito (2014).

(15)

countries, consisting of mostly advanced and emerging economies that make up 90% of world GDP. Based on the latter methodology, Han and Sin (2016) use an OECD panel data set of 34 countries excluding Korea in the period 1980 to 2015 and predict the Korean current account balance based on the fitted model. Gossé and Serranito (2014) built on the simple reduced form regression of the current account determinants approach (see (1)) by additionally controlling for cointegration and define long-run equilibrium targets of the current account balance, using linear and asymmetric panel vector error correction models with a panel of 21 OECD countries over the period 1974 to 2009 and to test short and long run determinants of the current account balance. The specification is as follows:

∆ 1 𝐶𝐴

𝐺𝐷𝑃2_!" = 𝛼!× 𝐷! + 𝛽∆𝑋!"− 𝛾𝐸𝐶𝑇!"#$+ 𝜀!"

Additional to (1), this specification takes first differences of the dependent variable and the regressors. Further, the error correction term ECT measures the gap between the realised current account balance and the equilibrium value that the authors calculated while 𝛾 measures the speed of adjustment to the equilibrium value. Based on all of the above studies, the following variables in Table 3.1 are found to significantly affect the current account balance of a country in at least one study.

Firstly, according to evidence by the International Monetary Fund (2013), the productivity level affects the current account balance negatively, as higher productivity incentivises foreign capital inflows if free movement of capital is guaranteed. In their regression, an interaction term of productivity and capital account openness is found to be negative and significant. Although in the same study, expected GDP growth also results to have a significantly negative effect on the current account balance as it equally attracts investment from abroad, Ciocyte and Rojas-Romagosa (2015) note that expected GDP growth can improve the expected disposable income of households. This might increase or decrease the savings rate depending on whether households perceive it to be of temporary or permanent nature respectively (see also Chinn and Prasad (2000)). As for a stable institutional and political environment, there exists evidence that this factor ameliorates the investment climate in a country and therefore decreases the current account balance (see International Monetary Fund (2013) and Gruber and Kamin (2005)). On the contrary, they find that structural rigidities, particularly in the labour market, undermine domestic investment and thus, improve the current (2)

(16)

account balance; see e.g. Barnes et al. (2010) who estimate a significantly positive effect of the non-accelerating inflation rate of unemployment NAIRU5_{on the current account balance.}

Moreover, as already indicated in the previous chapter, demographic variables are found to be essential in the analysis of Ciocyte and Rojas-Romagosa (2015). As expected, there is evidence that the young age dependency ratio6_{has a significant negative effect on the current account}

since both of these age groups mostly consume and invest, see e.g. Chinn and Prasad (2000). Additionally, Higgins (1998) as well as Han and Sin (2016) find that a higher old age dependency ratio7_{represents a larger share of the population that deplete their savings and is}

therefore found to reduce the current account. Notwithstanding, Barnes et al. (2010) find that a higher expected old-age dependency ratio or aging increases rather than reduces the current

5_{Also known as the natural rate of unemployment, the NAIRU is connected to the concept of the Phillips}

curve that assumes a negative relationship between unemployment and inflation. The NAIRU is assumed to be the unemployment rate at which the inflation rate is constant. An unemployment rate below the NAIRU would accelerate inflation since the scarcity of available workers overheats labour markets and drives up wages and prices.

6_{Defined as the ratio of children under 15 years over the working age population (15-64).} 7_{Defined as the ratio of people older than 64 over the working age population.}

Table 3.1 Estimated Effects of Current Account Balance Determinants

Factor Estimated Effect

Productivity -

Expected GDP growth -

Institutional and political environment -

Structural rigidities +

Young age dependence ratio -

Old age dependency ratio -

Ageing (projected old-age dependency ratio) +

Population growth rate -

Lagged net foreign assets +

Real interest rate +

Fiscal balance +

Social security level -

REER -

Euro area membership + (core euro area country) / - (periphery euro area country)

(17)

account balance, as aging societies tend to save more in order to guarantee the sustainability of the pension system. Furthermore, the International Monetary Fund (2013) finds that the population growth rate deteriorates the current account balance as an increase in the labour force fosters foreign investment.

As for financial variables, the lagged net foreign asset (NFA) position is an essential determinant in the literature on current account drivers. It is usually found to have a significant positive effect on the current account balance (see Barnes et al. (2010); Chinn and Prasad (2000); Gruber and Kamin (2005); Han and Sin (2016); International Monetary Fund (2013)). The NFA position makes up the value of foreign assets held by domestic investors minus the value of the domestic assets owned by foreigners. As past investment usually yields a return, the NFA position increases future net income from abroad and hence, improves the current account balance with a time lag. Another factor that is commonly found to significantly improve the current account balance is the fiscal balance (Barnes et al., 2010; Chinn and Prasad, 2000; Gossé and Serranito, 2014; Gruber and Kamin, 2005; Han and Sin, 2016; International Monetary Fund, 2013). Essentially, previous studies suggest that an increase in public saving is not offset by a counteraction of private agents and will improve the current account balance. Moreover, Barnes et al. (2010) find a positive effect of long-term real interest rates on the current account balance as they increase the rate of return on savings and the cost of borrowing to realise investments. A variable that is found to deteriorate the current account balance is public expenditure on health or social security as it reduces the need of agents for precautionary savings (International Monetary Fund, 2013).

Furthermore, Gossé and Serranito (2014) present evidence that a temporary decrease in the REER, i.e. a depreciation of the domestic currency, improves the current account balance. They explain that a decrease in the REER has the same effect as a terms of trade improvement: as it raises the purchasing power of the currency, it increases real income of agents and, thus, savings. Also, Han and Sin (2016) find that including the REER as an explanatory variable of the Korean current account balance shows that the undervaluation of the Korean won has significantly been driving the country’s current account surplus as from 1999. Lastly, findings of Barnes et al. (2010) suggest that euro area membership contributed to the increase in current account imbalances in the currency union. While the effect of the euro core country dummy on the current account balance is positive, albeit insignificant, the coefficient of the periphery country dummy is both negative and significant. Additionally, in the specific case of the Netherlands Ciocyte and Rojas-Romagosa (2015) contribute unexplained residuals in the analyses to the large size of pension funds and amount of multinational companies based in the

(18)

country, noting that similar conditions might partly explain e.g. the Swiss current account surplus.

As seen in this section, variables affecting a country’s productivity, institutional, structural and political components, demographic variables, net foreign investment activity, interest rates, government finances and euro area membership are found to be significant determinants of the current account balance. Since my further analysis will focus on OECD countries which are highly developed for the most part and similar in structural and political environment, the further analysis omits these variables for the sake of simplicity. Moreover, I will include productivity, the net foreign asset position, the government budget balance and euro area membership in a following descriptive analysis of the macroeconomic environment in Germany from 1980 to 2016 and a subsequent panel data regression analysis including 24 OECD countries from 1995 to 2016.

4 Descriptive Analysis of the Macroeconomic Environment

Chapter 2 and chapter 3 alleged various causes for current account imbalances. The following section contains a descriptive analysis of the macroeconomic environment in Germany from 1980 to 2016 based on the aforementioned theoretically and empirically found drivers of the German current account balance. The analysis serves to understand the evolution of fundamental macroeconomic variables influencing the current account balance directly and indirectly before conducting a panel regression analysis including the German time series in chapter 5. Firstly, domestic savings and investment as well as their potential drivers are examined in section 4.1. Section 4.2 focusses on variables affecting the trade balance in particular.

4.1 Savings and Investment

In the following section, variables affecting saving and investment in Germany from 1980 to 2015 are examined. These variables comprise net lending of households, corporations and government, GDP growth in Germany relatively to world real GDP growth, the real interest rate, the number of multinational enterprises, net foreign assets, the age dependency ratios and the fertility rate in Germany.

As can be seen in Figure 4.1, the German savings rate shows an upward tendency between 1980 and 2016. From 22% in 1980 it increased to 25% at the end of the decade and stayed around that level until 2006. Just before the Global Financial Crisis 2008-09, savings spiked to 27% of GDP in 2007 and dropped by 4 percentage points in the following two years.

(19)

Since 2010 savings have been increasing again and reached their pre-crisis level in 2016. On the other hand, investment to GDP shows no clear trend. While investment made up 24% of Germany’s GDP in 1980, the ratio dropped to around 19% during the following decade. Nevertheless, the investment rate strongly increased due to large government expenditure related to the German reunification in 1990. The latter visibly drove investment above 25% to GDP during the 1990s. As from the turn of the millennium, the ratio continuously decreased until 2006. As well as the savings rate, the investment rate peaked in 2007 just before the Global Financial Crisis troughed in 2009. In contrast though, investment to GDP has remained below its pre-crisis level, fluctuating around 19% as observed in the 1980s.

Source: The World Bank and own calculations.

Source: OECD.

Note: Net lending implies that lending exceeds borrowing, while a negative net lending equals net borrowing.

0 5 10 15 20 25 30 198 0 198 1 198 2 198 3 198 4 198 5 198 6 198 7 198 8 198 9 199 0 199 1 199 2 199 3 199 4 199 5 199 6 199 7 199 8 199 9 200 0 200 1 200 2 200 3 200 4 200 5 200 6 200 7 200 8 200 9 201 0 201 1 201 2 201 3 201 4 201 5 201 6 % S I -12 -10 -8 -6 -4 -2 0 2 4 6 8 10 198 0 198 1 198 2 198 3 198 4 198 5 198 6 198 7 198 8 198 9 199 0 199 1 199 2 199 3 199 4 199 5 199 6 199 7 199 8 199 9 200 0 200 1 200 2 200 3 200 4 200 5 200 6 200 7 200 8 200 9 201 0 201 1 201 2 201 3 201 4 201 5 201 6 %

Total HH Corp Gov

Figure 4.1 German Gross Domestic Savings (S) and Investment (I) as a Percentage of GDP between 1980 and 2016

Figure 4.2 German Net Lending as a Percentage to GDP of Households (HH), Corporations (Corp) and the Government (Gov) between 1980s and 2016

(20)

Due to restricted data availability Figure 4.2 only depicts net lending in the period 1995 to 2016. However, it underlines that public expenditure was the main driver of the high investment rate in the 1990s. This is reflected in the immense net borrowing to GDP ratio of almost 10% in 1995. This budget deficit was mainly caused by the takeover of East German liabilities, transfers to the new federal states and special programs to foster infrastructure and business activity in the east. At the same time, net lending of corporations and households did not offset this increase in government expenditure as shown by the net borrowing for the total economy in the 1990s. In 2000 the government budget was temporarily balanced but troughed at -4% of GDP in 2003 after a tax reform implemented by former finance minister Eichel. The included corporate tax cuts were intended to boost economic activity but drastically depressed tax revenues and led to a government budget deficit. Although the government budget reached a small surplus again in 2007 and remained at balance in the following year, the Global Financial Crisis 2008-09 along a budget deficit. The underlying increase in government expenditure was caused by several bank bail-outs of e.g. Hypo Real Estate and Commerzbank as well as economic stimulus packages, e.g. the scrapping premium to boost sales in the automotive industry, under finance minister Steinbrück. Nevertheless, since 2011 the government budget balance continuously increased, turned into a surplus in 2014 and reached 1% in 2016 as a result of finance minister Schäuble’s “schwarze Null” break even policy. In contrast, net lending of households was less prone to fluctuations between 1995 and 2006. While the to-GDP ratio was close to 3% in 1995, it moderately increased to 3.5% in 2000. Until 2005 households’ net lending to GDP growth accelerated until the ratio reached 6% in 2005 and remained at that level for five years. One explanation for this continuously increased level of savings could be the economic downturn in Germany from 2000 to 2005 (see Figure 4.3) which increased the perceived risk in the economy and led households to increase their precautionary savings as well as to borrow and invest less. The Global Financial Crisis 2008-09 reinforced the latter. However, in 2011 households’ net lending to GDP dropped by a percentage point and remained around 5% until 2016. Moreover, corporate net lending to GDP decreased from 5% in 1995 to almost -6% in 2000 in consequence of increasing taxes and social security contributions that were required to sustainably finance the increased public debt after the German reunification. Yet, corporate net lending started recovering since the turn of the millenium and has not fallen below zero since 2001. The ratio peaked at around 4% in 2010, reflecting an overall lower global corporate leverage after the Global Financial Crisis. Nevertheless, since 2012 corporate net lending to GDP has been more or less constant at 2%.

(21)

Source: The World Bank.

Source: OECD, The World Bank and own calculations.

The next relevant variables concerning savings and investment concern German GDP growth, the real interest rate and the number of multinational enterprises as indication for the investment climate. Figure 4.3 indicates that between 1980 and 2006 German GDP growth was, as in most industrialised countries, below the average world growth rate except for a short period from 1989 to 1992 and in 2011. While the neoclassical model consequently predicts domestic capital outflows to countries with higher growth prospects, one has to consider the German financial markets have at the same time been characterised by low risk and high liquidity. This is partially reflected in the country’s continuous AAA credit rating by S&P since 1983 and the Bundesbank being one of the most autonomous central banks in the world (see e.g. Cukierman et al. (1992); Fernández de Lis Alonso (1996)). Further, Figure 4.4 depicts the

-8 -6 -4 -2 0 2 4 6 198 0 198 1 198 2 198 3 198 4 198 5 198 6 198 7 198 8 198 9 199 0 199 1 199 2 199 3 199 4 199 5 199 6 199 7 199 8 199 9 200 0 200 1 200 2 200 3 200 4 200 5 200 6 200 7 200 8 200 9 201 0 201 1 201 2 201 3 201 4 201 5 201 6 % Germany World -2 -1 0 1 2 3 4 5 6 7 198 0 198 1 198 2 198 3 198 4 198 5 198 6 198 7 198 8 198 9 199 0 199 1 199 2 199 3 199 4 199 5 199 6 199 7 199 8 199 9 200 0 200 1 200 2 200 3 200 4 200 5 200 6 200 7 200 8 200 9 201 0 201 1 201 2 201 3 201 4 201 5 201 6 % r

Figure 4.3 German and World Real GDP Growth between 1980 and 2016

Figure 4.4 German Real Interest Rate (10-year Government Bond Yield adjusted by the GDP Deflator) from 1980 to 2016

(22)

development of the real interest rate or the cost of borrowing adjusted for inflation. Between 1980 and 2000 the German real interest rate fluctuated heavily around 4%. Since the turn of the millennium it has been falling and even dropped below zero as from 2012. Although the empirical literature finds that the real interest rate is positively affecting the current account balance, this relationship does not become clear graphically in the case of Germany. Especially, while the current account balance is on the rise between 2002 and 2016 the real interest rate shows an inverse behaviour. The reason for that could be that the influence of other variables dominates, for instance, the effects of demographic variables which are discussed later in this section.

Source: OECD.

Lastly, Figure 4.5 provides estimates of the number of multinational enterprises in Germany practicing export activity between 1995 and 2016. Generally, the figure shows an upward trend of the variable, except for a trough in 2002. The latter may be attributed to the Dotcom Bubble in the early 2000s which caused a stock market crash that affected various internet-based companies. Yet, the number of MNEs has been increasing ever since and has almost doubled between 1995 and 2016. There are two effects that an increasing number of MNEs headquartered in a country can have on the current account. While on the one hand MNEs attract foreign investment and could therefore be expected to reduce the current account balance, MNEs also engage in export activities which on the other hand increases the trade balance and the current account balance. Graphically, one could presume that there is a positive relationship between the current account balance and the number of MNEs in a country. Still, correlation alone does not imply a causal effect of the number of MNES on the current account balance. 0 5000 10000 15000 20000 25000 30000 198 0 198 1 198 2 198 3 198 4 198 5 198 6 198 7 198 8 198 9 199 0 199 1 199 2 199 3 199 4 199 5 199 6 199 7 199 8 199 9 200 0 200 1 200 2 200 3 200 4 200 5 200 6 200 7 200 8 200 9 201 0 201 1 201 2 201 3 201 4 MNE

(23)

As explained in section 2.1, apart from the trade balance, net income from abroad and net current transfers form part of the current account balance. Further, the development of net foreign assets deserves a closer look as it drives future net income from abroad. Figure 4.6 illustrates the exponential growth of the German NFA position to GDP between the late 1990s up until the Global Financial Crisis 2008-09 that entailed a more moderate growth of the variable until 2016. As a consequence of the rapid NFA growth between 1998 and 2007, rising net earnings on foreign investment have driven net income from abroad since the early 2000s, see Figure 4.7. Between 1980 and 2016 German net income from abroad mostly had a positive effect on the current account balance except for the period between 1994 and 2003 when yields on foreign inward investment following the German reunification outweighed earnings from

0 10 20 30 40 50 60 70 198 0 198 1 198 2 198 3 198 4 198 5 198 6 198 7 198 8 198 9 199 0 199 1 199 2 199 3 199 4 199 5 199 6 199 7 199 8 199 9 200 0 200 1 200 2 200 3 200 4 200 5 200 6 200 7 200 8 200 9 201 0 201 1 201 2 201 3 201 4 201 5 201 6 NFA to GDP -3 -2 -1 0 1 2 3 198 0 198 1 198 2 198 3 198 4 198 5 198 6 198 7 198 8 198 9 199 0 199 1 199 2 199 3 199 4 199 5 199 6 199 7 199 8 199 9 200 0 200 1 200 2 200 3 200 4 200 5 200 6 200 7 200 8 200 9 201 0 201 1 201 2 201 3 201 4 201 5 201 6 NIA to GDP NCT to GDP NIA+NCT to GDP

Figure 4.6 German Net Foreign Assets (NFA) to GDP between 1980 and 2016

Figure 4.7 German Net Income from Abroad (NIA) to GDP and Net Current Transfers (NCT) to GDP between 1980 and 2016

(24)

abroad. Between 2004 and 2011 net income from abroad to GDP showed an upward trend. The decrease in the variable as from 2011 might be ascribed to the overall drop in interest rates in advanced economies after the Global Financial Crisis 2008-09 that depressed earnings on government bonds from these countries. Furthermore, between 1980 and 2016 net current transfers were continuously negative and thus decreased the German current account balance as the country provides foreign aid, grants and donations received by developing nations. Yet, as from 1995 the amount of net current transfers to GDP remained quite stable around -1.3% such that the sum of net income from abroad and net current transfers to GDP was visibly driven by net income from abroad and had an overall positive effect on the current account between 1983 and 1990, in 2006 and 2007 as well as from 2009 to 2016.

As seen before in plenty of empirical studies indicated in chapter 3, the young and old age dependency ratios are found to be negative and significant drivers of the current account balance. Comparing the years 1980 and 2016 the total dependency ratio is unchanged at 52% as seen in Figure 4.8. However, the shares of old and young age dependence have altered. In 1999 for the first time there were more people older than 64 than children under 15 in Germany. Since the turn of the millennium the wedge between the young and old age dependency ratio has constantly been increasing. While this development is partially caused by a decreasing young age dependency ratio, it is mainly driven by a rapid increase in the share of the older population. Related, Figure 4.9 depicts the tremendous fall in the German fertility rate during the late 1960s and early 1970s. While the fertility rate is an indicator of future population aging if it falls below the replacement rate of 2.1, it is also widely made available in public. A low or decreasing fertility rate might therefore influence agents to increase their savings for retirement

0 10 20 30 40 50 60 198 0 198 1 198 2 198 3 198 4 198 5 198 6 198 7 198 8 198 9 199 0 199 1 199 2 199 3 199 4 199 5 199 6 199 7 199 8 199 9 200 0 200 1 200 2 200 3 200 4 200 5 200 6 200 7 200 8 200 9 201 0 201 1 201 2 201 3 201 4 201 5 201 6 % TDR YDR ODR

Figure 4.8 German Total Dependency Ratio (TDR), Young Age Dependency Ratio (YDR) and Old Age Dependency Ratio (ODR) from 1980 to 2016

(25)

as they expect an increasing old age dependency ratio. Although since the mid 1970s the German fertility rate did not drop far below 1.5, it has not been meeting the replacement rate of 2.1 births per woman and has thus been causing an ever-increasing old age dependency ratio since 1985 as well as an overall aging population.

To sum up, both surplus periods of the German current account balance in the 1980s and between 2002 and 2016 are characterised by an increasing savings rate and a rather constant investment rate to GDP. The German reunification in the 1990s temporally boosted investment for a decade such that it exceeded savings which resulted in a current account deficit for Germany. Examining net lending in the public and private sectors of the German economy, the increasing net lending rate of the total economy was driven by moderately rising household net lending, strong growth of net lending in the corporate sector and ultimately, the improvement of the government budget balance since the Global Financial Crisis 2008-09. As for the variables determining investment, German GDP growth has been mostly below world GDP growth as for most other advanced economies. However, one also has to consider the strong independence of the Bundesbank and the high credit rating on German government bonds that signal high liquidity and low risk in the economy which attract investors looking for a safe haven. Furthermore, the declining interest rate in Germany since the early 2000s that accompanied a rising German current account balance does not match the theoretical concept that a decrease in the interest rate encourages investments and reduces savings. However, the latter suggests that other factors dominate the evolution of the German current account since the turn of the millennium. Additionally, the upward trend of the number of multinational enterprises since 1995 indicates that there could be a positive relationship between this variable

0 0.5 1 1.5 2 2.5 3 196 0 196 2 196 4 196 6 196 8 197 0 197 2 197 4 197 6 197 8 198 0 198 2 198 4 198 6 198 8 199 0 199 2 199 4 199 6 199 8 200 0 200 2 200 4 200 6 200 8 201 0 201 2 201 4 201 6 Fertility rate

(26)

and the current account balance as multinational enterprises largely engage in export activity and hence might drive up the trade balance and current account balance of the domestic economy rather than reducing the current account balance by attracting foreign investment. Moreover, the rapidly increasing NFA to GDP ratio since the late 1990s entailed a strong rise in net income from abroad since the early 2000s in Germany which is was an additional driver of the German current account surplus in the 2000s and 2010s. Lastly, the sudden drop in the German fertility rate around 1970 initiated a rapid population aging that may have increased the overall savings rate in the economy in order to maintain the pension system in the future.

4.2 Foreign Trade and Trade Balance

In the period 1986 to 1988 as well as in the year 1990 and between 2003 and 2008 Germany was the leading world exporter. That is why examining trade-related variables and especially the competitiveness of the export sector is crucial for the analysis of the German current account balance. In the following, variables affecting the German current account balance through the trade balance from 1980 to 2015 are scrutinised. That is, this section includes the evolution of high-technology exports, the change in unit labour cost, inflation and the REER.

As can be seen in Figure 4.10, German exports have continuously exceeded imports between 1980 and 2016. Yet, both variables appear to be cointegrated as they are subject to the cyclicality of international trade flows. Thus, both imports and exports to GDP have been on the rise from 1980 up until the Global Financial Crisis except for the reunification decade. After the crisis, exports have stagnated at around 40% of GDP, while the import share has declined

0 5 10 15 20 25 30 35 40 45 198 0 198 1 198 2 198 3 198 4 198 5 198 6 198 7 198 8 198 9 199 0 199 1 199 2 199 3 199 4 199 5 199 6 199 7 199 8 199 9 200 0 200 1 200 2 200 3 200 4 200 5 200 6 200 7 200 8 200 9 201 0 201 1 201 2 201 3 201 4 201 5 201 6 % X M TB

(27)

since 2011. This caused the German trade balance share of GDP to reach an all-time high of 8.18% in 2016.

A classification of exports that deserves special attention, are high-technology exports8

as they are related to the concept of competitive advantage as defined in chapter 2. Firstly, the R&D-intensity of the high-technology sector allows for strong learning effects that allow for decreasing marginal costs and dynamic economies of scale. Secondly, the sector is characterised by product differentiation which allows individual firms to gain market power or even become a global monopolist, see e.g. the automotive industry where cars are not perfect substitutes and there exists a large number of vehicle varieties. Another characteristic of high-technology exporters that fosters market concentration is the requirement of high R&D investments for the production and the existence of patents or special protection for producers. Finally, R&D intensive industries profit strongly from industry agglomeration and location advantages such as R&D subsidies, academic networks and the collaboration with leading universities that encourages the immigration of skilled workers and further fosters agglomeration. Therefore, the high-technology exports share of all exports of a country can be an indicator of the degree of competitive advantage of the export sector and the number of global monopolists headquartered in the country. Thus, a large high-technology export share might persistently increase domestic exports relatively to exports of other countries, create a positive trade balance and hence, increase the domestic current account. Figure 4.11 illustrates that the German share of high-technology exports has increased from 11.5% in 1988 to almost

8_{High-technology exports are R&D-intensive export goods that include e.g.}

pharmaceutical and sophisticated machinery.

0 2 4 6 8 10 12 14 16 18 20 198 0 198 1 198 2 198 3 198 4 198 5 198 6 198 7 198 8 198 9 199 0 199 1 199 2 199 3 199 4 199 5 199 6 199 7 199 8 199 9 200 0 200 1 200 2 200 3 200 4 200 5 200 6 200 7 200 8 200 9 201 0 201 1 201 2 201 3 201 4 201 5 201 6 % HTX

(28)

16% in 1995. In the late 1990s it stagnated and growth picked up in 2003. In 2005 the high-technology export share peaked at 18.6% and dropped below 16% in 2007. As from the following year, the variable was increasing again until 2015 when it dropped to 15.5%. However, between 1988 and 2016 there appears to be a visible upward tendency of the share of high-technology exports in Germany.

Source: OECD.

Further, competitiveness is often measured in the form of unit labour cost. This variable sets the value of a unit of output created in relation to the expenditure on wages it requires. An increase in unit labour cost is therefore expected to decrease exports, reduce the trade balance and worsen the current account. Figure 4.12 shows that between 1994 and 2003 unit labour cost growth in Germany fluctuated around zero. The reduction in unit labour cost between 2004 and 2007 is most likely connected to the labour market liberalisation program “Agenda 2010” including cuts in pensions and unemployment benefits among others. The aim of the reform was to reduce unemployment and to make German industries more competitive. Conceivably, the increasing trade balance and current account balance can at least partially be attributed to these reforms. During the Global Financial Crisis in 2009 unit labour cost increased abruptly as a consequence of the decline in output due to the global recession while labour market rigidities in advanced economies prevented wages from falling proportionally. After dropping to -1% in 2010, unit labour cost growth accelerated to almost 3% in 2012 which reflects the upwards pressure on wages caused by the decreasing German unemployment rate. Between 2013 and 2016 the variable remained at around 2%. Yet, at the same time, the German current account balance was on the rise, too. A reason for this could be that the level of German unit

-4 -2 0 2 4 6 8 198 0 198 1 198 2 198 3 198 4 198 5 198 6 198 7 198 8 198 9 199 0 199 1 199 2 199 3 199 4 199 5 199 6 199 7 199 8 199 9 200 0 200 1 200 2 200 3 200 4 200 5 200 6 200 7 200 8 200 9 201 0 201 1 201 2 201 3 201 4 201 5 201 6 % UCL

(29)

labour cost is still relatively low so it does not worsen the competitiveness of German competitors to the extent that exports decline.

Another important variable for my further analysis is the inflation differential between the German rate and the euro area average. Not only do high inflation differentials between member countries hamper the optimality of a currency union and the effectiveness of monetary policy. In particular, they also affect the purchasing power of one Euro in each member country differently and change the real exchange rates between member countries. It can be supposed that a negative inflation differential between the rate of country X and the euro area average increases overall competitiveness of country X and as this effect is not offset by an appreciation of external euro exchange rates, it can persistently increase the trade balance and the current account balance. Figure 4.13 shows that in 1980 inflation in Germany was at 6% which equalled

-2 0 2 4 6 8 10 12 14 198 0 198 1 198 2 198 3 198 4 198 5 198 6 198 7 198 8 198 9 199 0 199 1 199 2 199 3 199 4 199 5 199 6 199 7 199 8 199 9 200 0 200 1 200 2 200 3 200 4 200 5 200 6 200 7 200 8 200 9 201 0 201 1 201 2 201 3 201 4 201 5 201 6 %

Germany Euro area

80 85 90 95 100 105 110 115 120 198 0 198 1 198 2 198 3 198 4 198 5 198 6 198 7 198 8 198 9 199 0 199 1 199 2 199 3 199 4 199 5 199 6 199 7 199 8 199 9 200 0 200 1 200 2 200 3 200 4 200 5 200 6 200 7 200 8 200 9 201 0 201 1 201 2 201 3 201 4 201 5 201 6 REER

Figure 4.13 Percentage Change in the GDP Deflator from 1980 to 2016

(30)

half of the average rate of the countries that would later form the euro area. While there was a convergence in inflation rates during the 1990s that was driven by a decreasing average rate in the euro area, in 1988 inflation in Germany started to accelerate and exceeded the euro area average after the reunification in 1992 (3.6%) by almost one percentage point. From the following year on, the German inflation rate fell rapidly to -0.4% in 2000 while the euro area average remained around 3%. This gap did not narrow until the Global Financial Crisis in 2009. After the crisis the German inflation rate as well as the euro area average developed similarly. Yet, inflation in German has been higher than the average rate in the euro area since 2013 which is in line with the observed increase in unit labour cost.

Moreover, the REER is a measure of competitiveness. As indicated in section 2.2, a decrease in the REER increases competitiveness, which is expected to boosts exports and improve the trade balance as well as the current account. The opposite holds for an increase in the REER. Figure 4.14 depicts the development of the IMF REER index for Germany between 1980 and 2016. There is a remarkable maximum of the index that hints at an overvaluation of the Deutsche Mark in 1995. In the late 1990s the index fell strongly up until the turn of the millennium. Afterwards, the index accelerated again and remained around the same level between 2004 and 2009. After the crisis the German REER has shown a falling tendency until 2016. However, the graph does not evidently illustrate a strong relation between the German current account balance to GDP (see Figure 1.1).

In summary, both German exports and imports were on the rise between the early 1990 and 2016. However, exports increasingly exceeded imports which resulted in a growing trade surplus and current account surplus in Germany. As an indicator of competitive advantage of the German export sector and the number of global monopolists headquartered in Germany, the high-technology product share in all exported goods has shown an upward trend since 1995 and might be a positive determinant of the trade balance as well as the current account balance. Further, the fall in unit labour cost just before the Global Financial Crisis connected to labour market reforms in Germany may have increased the country’s trade balance through competitiveness gains of the exporting sector. Yet, the German current account balance did not worsen between 2011 and 2016 although domestic unit labour cost was on the rise. Further, the inflation differential between the German rate and the euro are average was negative for 7 years after the euro was introduced in 2002 and might also be a sign of relative competitiveness gains of German producers to producers in the rest of the euro area, increasing the trade balance and ultimately the current account balance of Germany. However, after the Global Financial Crisis, the German inflation rate was close to the euro average and even exceeded it between 2013 and

(31)

2016 while the German current account balance was increasing simultaneously. Lastly, a correlation between the German REER index and the current account balance does not become clear graphically.

5 Regression Analysis

Subsequent to the preceding graphical insights on the various possible determinants of the German current account balance, a regression analysis shall serve to further examine their validity. The first subsection will specify the applied methodology followed by a brief overview of peculiarities in the data in section 5.2 and lastly, by the results in section 5.3.

5.1 Methodology and Variables

Just as Chinn and Prasad, (2000), Gruber and Kamin (2005) and the International Monetary Fund (2017), the following reduced form fixed effects regression of current account determinants using annual data as described in chapter 3 will be used:

𝐶𝐴

𝐺𝐷𝑃!" = 𝛼!× 𝐷! + 𝛽𝑋!"+ 𝜀!"

Vector X contains several variables that are found to be significant determinants of the current account balance in the previous literature, such as the government budget balance as a percentage of GDP, the real interest rate (10 year government bond interest rate minus GDP deflator), the first lag9_{of NFA in billion US dollars, the young age dependency ratio, the old}

age dependency ratio, the fertility rate (births per woman) as a measure of projected aging, the percentage change in the REER and the percentage change in total factor productivity. Moreover, the effect of eurozone membership is not only included in form of dummy variables, but as an interaction term of a dummy variable that is 1 if country i has used the euro as a means of payment in year t times the euro area inflation differential of country i to the EU average in that year. This method allows for a dynamic measurement of the effect of the euro on the current account balance. The underlying assumption is that if country X is a euro area member, a domestic inflation rate below the euro area average implies a relatively slower growth of prices and will increase country X’s competitiveness relative to the other euro area members. This effect cannot be offset by a currency depreciation as between member countries nominal exchange rates are fixed and the external nominal exchange rate is determined by the overall

9_{The variable NFA enters the regression equation with its first lag as NFA affect the}

current account balance through net earnings on foreign investment which usually start to be paid out one year after the related investment.

What drives the German current account balance?

Universiteit van Amsterdam

Amsterdam School of Economics