How the euro synchronised EZ cycles

Ayako Saiki, Sunghyun Henry Kim 02 February 2014

a

A

Prior to the introduction of the euro, the topic of whether the Eurozone fulfils the conditions for an optimum currency area was highly debated (e.g. Bayoumi and Eichengreen 1992).

In their seminal study, Frankel and Rose (1996) argue that the conditions for an optimum currency area – specifically business-cycle synchronisation – can be satisfied ex post because business-cycle correlation can be endogenous. Based on their empirical analysis of 21 industrial countries, they conclude that a currency union would generate stronger trade linkages – defined as trade intensity among member countries – which would in turn generate higher business-cycle correlations among them.

Now, 15 years after its inception, has the Euro increased intra-regional trade intensity and business-cycle correlation as predicted? In our recent working paper (Saiki and Kim 2014), we attempt to answer these questions by investigating trade patterns, business-cycle correlations, and their connecting factors in the Eurozone. Our finding is that, while the euro has had some impact on business-cycle synchronisation in the Eurozone, it is not through the channel that Frankel and Rose (1996) predicted. To enrich our analysis, we compare our results for the Eurozone with those for east Asia during the same period (1980–2007).1 East Asia’s economies have rapidly integrated during this period without belonging to a currency union or fixed exchange-rate regime. Therefore, a comparison with the Eurozone should shed some light on the influence of a currency union from an additional angle.

Stylised facts

The stylised facts are as follows:

  • Business cycles: The increase in business-cycle correlation in the Eurozone has been quite gradual, and there is no indication that the introduction of the euro increased business-cycle synchronisation.2 On the other hand, since 1980, business-cycle correlation has rapidly increased in east Asia, and is now approaching that of the Eurozone (see Figure 1).

Figure 1. Business-cycle correlation in east Asia and the Eurozone

Source: Authors’ calculations based on the OECD database.
Notes: The lines denote the ten-year rolling window correlation between each individual country’s business cycle and the regional business cycle (the year on the x axis is the last year of the windows). The shaded area denotes the average of all correlations in the region. Correlations in this graph are calculated based on Hodrick–Prescott filtered logged real GDP in local currency terms.

  • Regional trade intensity: Various studies (e.g. Micco et al. 2003) find that there has been some positive impact of the euro on intra-Eurozone trade volumes of around 10–15%. However, when we measure trade intensity, intra-Eurozone trade after the introduction of the euro exhibits a different pattern. Figure 2 presents regional trade intensities measured by bilateral trade volume as a fraction of total trade (left panel) and GDP (right panel). As Figure 2 shows, regional trade intensity has been declining in the Eurozone since around 1998, while it has been increasing in east Asia. Our conjecture is that this is perhaps due to China’s remarkably rapid integration into the global economy, which increased intra-regional trade intensity in east Asia, and decreased it in the Eurozone.

Figure 2. Trade intensity in east Asia and the Eurozone

  • Trade patterns: We examine the degree of intra-industry trade (henceforth IIT) – this means trade within the same industry, measured by the Grubel–Lloyd Index – as well as the degree of vertical intra-industry trade (henceforth VIIT) – this means IIT in which goods are differentiated by quality – in the Eurozone and east Asia. The degree of VIIT has been increasing over time in east Asia, and it is currently higher in east Asia (0.8) than in the Eurozone (0.75), where VIIT is declining. This rising degree of VIIT in east Asia is consistent with the rapidly growing global supply chain in the region. The declining VIIT in the Eurozone may be an indication of supply-chain development between Eurozone and non-Eurozone countries, including China, although we need to look into this further.

Analysis and results

In order to test how these different trade dynamics affect business-cycle correlation, we regress the correlation of business cycles – as measured by Baxter–King filtered GDP, Hodrick–Prescott filtered GDP, and year-on-year growth – on

  • trade intensity;
  • the degree of IIT measured by the Grubel–Lloyd Index; and
  • the ratio of vertical IIT to total IIT;

along with dummy-variable controls for regional effects and the adoption of the euro. We look at business-cycle correlations for three sub-periods: 1980Q1–1989Q4, 1990Q1–1999Q4, and 2000Q1–2007Q4. So the data is panel data for three periods with approximately 55 country pairs. The results are presented in Table 1.

Table 1. Main results

Note: Numbers in parentheses are t-statistics. ***, **, and * denote statistical significance at the 1%, 5%, and 10% levels, respectively. Coefficients are derived from panel regressions with country fixed effects. The Eurozone dummy is set to 1 for countries in the Eurozone geographically, both before and after the adoption of the euro. The EMU dummy is set to 1 after the countries in the Eurozone adopted the euro.
The main take-away messages from these results can be summarised as follows.

  • First, unlike Frankel and Rose (1996), who show that trade intensity has an unambiguously positive and significant impact on business cycle correlation, our results indicates that trade intensity is significant for the Eurozone but not for east Asia.3 In fact, after controlling for (V)IIT, the coefficient on trade intensity becomes negative. This is consistent with the prediction from classical international trade theory that trade between countries with different factor endowments will facilitate product specialisation, which makes countries susceptible to asymmetric shocks that can dampen business-cycle synchronisation. Our sample east Asian countries have different endowment structures – the standard deviation of GDP per capita, spanning from Singapore to Vietnam, is almost three times that of the Eurozone.
  • Second, VIIT’s positive and significant impact on business-cycle correlation is more visible in east Asia. This result is robust to different measures of business cycles (Baxter–King or Hodrick–Prescott filtered GDP, or year-on-year growth) and sample periods (including the financial crisis period). This perhaps reflects the lack of progress in VIIT in the Eurozone during the sample period. When we pool the samples for Eurozone and east Asia, we observe that higher trade intensity increases business-cycle correlation. However, after controlling for IIT or VIIT measures, the effects of trade intensity become insignificant, indicating that it is IIT – especially the VIIT part – that drives business-cycle synchronisation, not trade intensity per se.
  • Third is the noteworthy result that the EMU dummy is positive and significant throughout the analysis, even after we control for the regional (Eurozone) dummy. This indicates that some other channels – most likely financial integration – are working to enhance business-cycle correlation in the Eurozone, offsetting the negative impact from declining trade intensity. However, it will require thorough investigation to determine the exact role of financial integration – a topic we leave for further research.

Our findings have policy implications for Eurozone enlargement. As new member states with lower income per capita join the Eurozone, developing an intra-industry network – or production-segmentation across countries – between old and new member states would help to amplify business-cycle correlation among them. In other words, actively developing a supply chain between old and new Eurozone members can increase the benefits of the currency union.

Disclaimer: This article reflects solely the opinion of the authors and does not necessarily reflect the views of De Nederlandsche Bank. The authors acknowledge the valuable comments from Jakob De Haan, Peter van Els, and Pierre Lafourcade.

References

Bayoumi, T and B Eichengreen (1992), “Shocking Aspects of European Monetary Unification”, NBER Working Paper 3949.

Frankel, J and A Rose (1996), “The Endogeneity of the Optimum Currency Area Criteria”, NBER Working Paper 5700.

Micco, A, E Stein, and G Ordoñez (2003), “The Currency Union Effect on Trade: Early Evidence from EMU”, Economic Policy, 18(37): 315–356.

Mink, M and J De Haan (2013), “Contagion During the Greek Sovereign Debt Crisis”, Journal of International Money and Finance, 34: 102–113.

Saiki, A and S H Kim (2014), “Business Cycle Synchronization and Vertical Trade Integration: A Case Study of the Eurozone and East Asia”, De Nederlandsche Bank Working Paper 407.


1 We use the following sample countries: China, Indonesia, Japan, Korea, Malaysia, Philippines, Singapore, Taiwan, Thailand, and Hong Kong for east Asia; and Austria, Belgium, Finland, France, Germany, Greece, Ireland, Italy, Netherlands, Portugal, and Spain for the Eurozone.

2 We use three measures of business cycles: Hodrick–Prescott filtered GDP series, Baxter–King filtered GDP series, and annual growth rates.

3 Note that their study focuses on 21 advanced countries, and does not include emerging economies.

a

A

Topics:  EU institutions International finance International trade

Tags:  eurozone, euro, trade, business cycles, East Asia, supply chains, optimum currency area

Associate Professor of Economics, Sungkyunkwan University

Economist, De Nederlandsche Bank