An inquiry on currency areas within Europe

(1)

An inquiry on currency areas within Europe

Gido ter Heijne

10318453

Economics & Finance

UvA

International Money & Macro Economics

Supervisor: Nicoleta Ciurila

(2)

2

Index

1. Introduction p.3

2. Literature review p.5

2.1 Advantages and disadvantages of an OCA p.5

2.2 Criteria of an OCA p.6

3. Data analysis (correlations) p.9

3.1 The northern European currency area p.10

3.2 The southern European currency area p.12

3.3 The eastern European currency area p.13

3.4 The Baltic area p.14

3.5 Summary of currency areas p.15

4. Data analysis (standard deviations) p.15

4.1 GDP growth p.16

4.2 Inflation p.19

4.3 Debt/GDP p.23

4.4 National budget deficit p.27

4.5 Interest rate p.31

5. Conclusion p.34

6. Limitations and further research p.36

(3)

3

1. Introduction:

In 1961, Robert Mundell started a discussion about the criteria of an optimum currency area. Several economists believe that countries that are interconnected and similar from an economic perspective have the potential to create a common currency. This is because a monetary union has a higher probability to succeed if countries have converged.

In this thesis, the different views of economists such as Corden, Cuyvers, Horváth, Kenen, Mankiw, Mc Kinnon, Mundell, Robson and Tavlas will be discussed in order to try to find whether the Euro area is an optimum currency area or whether there are several optimum currency areas within Europe. These views contain similar as well as different criteria of which a few important and measurable criteria will be chosen and analysed. Moreover, these views contain advantages and disadvantages of optimum currency areas that will be mentioned briefly.

The optimum currency area theory originates from the debate about fixed and flexible exchange rates around 1940 of the previous century. The first fixed exchange rate regime was established in 1944 when the currencies of western countries got pegged to the dollar that again was pegged to gold. This was called the Bretton Woods system and it was the first step towards monetary integration and optimum currency areas. One definition of an optimum currency area as stated by Mongelli in 2008 is: ‘An optimum currency area is the optimal geographic area for one currency, or several currencies, of which the exchange rate is fixed. The fixed (or hard pegged) currency or currencies fluctuate collectively relative to other currencies. Borders of an optimum currency area are determined by the participating sovereign countries.`

In 1992, the Maastricht Treaty was signed by all members of the European Community. In this treaty, rules were created in order to promote similarity, this was done through the European Monetary Union. In order to do so, five criteria were imposed on all participating countries. These requirements to become part of a European common currency area were as follows: ‘First, inflation should not be more than 1.5 percentage points above the average rate of the three European Union member states with the lowest inflation over the previous year. In addition, a national budget deficit should not exceed 3 percent of GDP. Furthermore, the net debt to GDP ratio should not exceed 0.6. A country with a higher debt level can still become part of the Eurozone provided its debt level is falling steadily. Moreover, long-term interest rates should be less than two percentage points above the rate of the three EU

(4)

4 countries with the lowest inflation over the previous year. At last, the national currency is required to enter the Exchange Rate Mechanism II two years prior to entry.’ At first, these criteria were fulfilled, however, during the recent crisis, countries have started to diverge again because they did not obey these criteria any longer.

For this reason, the crisis of the last seven years which was more severe in Europe than for example in the U.S. shows that the European Monetary Union is still not an optimum currency area at its finest. The economic, social and political differences between countries such as Greece or Italy and on the other hand Germany or Denmark are still significant. Either political and fiscal integration should be improved, so more centralized fiscal and political power, or the Eurozone should be split in more than one currency area. The latter is the option that will be analysed in this paper, because whilst political and fiscal integration is improved, which can take years or even decades, it might be better to create several common currency areas for the time being.

This topic is of interest for the whole European Union or even the whole world as the EU takes account for almost twenty percent of the world’s GDP. Moreover, it is interesting because it is an actual topic that discusses recent problems and seeks solutions for the near future.

The research question is: ‘Would it be better from the point of view of the

convergence criteria to split the European Union, including the Eurozone, into more than one currency area? A literature review will survey the views of different scholars to create a solid background with criteria, advantages and disadvantages of optimum currency areas to start the data analysis. After the data analysis is done in section 3 (correlations) and section 4 (standard deviations), the results will be evaluated and a conclusion will be drawn in section 5. At last, in section 6, the limitations to the research of this paper will be stated and ideas for further research will be exposed. The research methodology will not be treated separately but will be included in the data section.

In the data analysis section, all European countries will be put into a group according to their geographical location. After that, the correlations of all European countries’ GDP growth rates and inflation rates will be calculated in order to see whether these countries within those geographical areas could fit together and join a common currency area. When this is done, for every preliminary currency area, standard deviations of GDP growth rates, inflation rates, debt/GDP ratios, real interest rates and national budget deficits will be

(5)

5 compared to the Eurozone as a whole. These standard deviations are expected to be lower than the standard deviations of the Eurozone as a whole which would mean that shocks would be more symmetric and therefore, there would be more criteria fulfilled which consequently means that the economic situation within these optimum currency areas would be improved.

It has been surprising that countries like Portugal and Greece which were expected to be better off within a southern currency area were actually better off alone. The south was expected to be more converged than it resulted to be whilst the eastern economies were just as similar to each other as the southern economies. Furthermore, the United Kingdom and Denmark did show high correlations with the rest of the north whilst having a different currency. Also, it seemed that the smaller the optimum currency area is, the better it performs during a crisis and vice versa. This needs to be researched further though as more statistical evidence is needed. Nevertheless, some results were expected, for example, Norway was very different from the northern currency area which can be explained by the fact that it does not have the Euro and Iceland did not show any significant correlations with any country due to its isolated geographic location and different currency.

2. Literature review:

2.1) Advantages and disadvantages of an OCA

Mundell (1961) thought about what the advantages and disadvantages could be from adopting a common currency. He concluded there were both advantages and disadvantages. He

believed there were benefits, for example, price transparency increased, trade volumes increased, currency liquidity increased and transaction costs decreased. Nevertheless, there were drawbacks as well of which the loss of flexibility within the exchange rate mechanism (Mundell, 1961) was the most important mainly because it makes monetary policy ineffective (Horváth, 2002).

One of the main advantages of an optimum currency area is the fact that the

uncertainty of the exchange rate fluctuations disappears (Tavlas, 2004). Moreover, because transaction costs disappear, the trade increases within the optimum currency area which improves the efficiency on micro economic level.

Nevertheless, when exports decrease in one member state of the optimum currency area, the currency cannot be devalued anymore in order to increase competitiveness and

(6)

6 make exports more attractive again because that will lead to a loss for other members of the optimum currency area. In this case, letting unemployment levels increase or decreasing wages and prices would be an imperfect but obvious substitute (Corden, 1972). He also argues that on the short term, this factor mobility brings about costs due to moving the production factors (housing for labour, etc.) which cannot outweigh the advantages on the short term. This is the reason that he stresses wage and price flexibility, because these variables can react to a negative shock on the short term.

Mundell’s paper in 1969 stressed the importance and the advantages of optimum currency areas again which he publically supported. He argued that members of optimum currency areas cannot suffer from severe problems individually as other members can compensate this loss. Again, he stressed the importance of factor mobility but he admitted that capital is way easier to transport than labour.

In his second article, which is of great importance for this thesis, he already stated that Europe was facing an historical monetary decision, namely he thought that a European common currency together with a European monetary board would be the best solution for the economic situation at that time. This is the start of the Euro, the first idea of a European common currency which started to develop into a real plan about twenty-five years later. Instead of the dollar, it was suggested that the pound should become the new European peg because the British economy including the pound were still very powerful at that time (Mundell, 1973).

Participation in an optimum currency area is a dynamic and evolutionary process for which every country should weigh the advantages and disadvantages individually. Not only the transaction costs disappear, also, speculative capital flows are limited. However, there are huge economic losses such as the loss of monetary autonomy and the decrease in fiscal liberty of national economies.

2.2) Criteria of an OCA

Robert Mundell in 1961 was the first one to lay out his view on the determination of the criteria of optimum currency areas. These criteria consisted of prerequisites and conditions of countries that were potential optimum currency area members. He based his theory on the economic relationship between the U.S. and Canada. A peculiar fact is that the west of the

(7)

7 U.S. had a better economic relationship with the west of Canada than with the east of the U.S. (Cuyvers e.a., 2012). Maybe it would be better for Canada and the U.S. to share a common currency. However, for a monetary union to work properly, Mundell stressed that a high degree of factor mobility especially labour is of crucial importance because it substitutes de exchange rate mechanism and it compensates for asymmetric shocks. To clarify this

difference, an example of two countries sharing a common currency will be used (initially in full internal and external equilibrium) to explain the importance of factor mobility. Imagine a shift in demand from the goods of country B to the goods of country A. Assume that wages and prices are sticky in the short run (Dornbusch, 1985). The shift of demand from B to A causes unemployment in B and inflationary pressure in A, accompanied with a current account surplus for A. In order to restore initial unemployment levels in B, the money supply in the common currency area has to increase, however, this increases inflationary pressure in country A even more. The reason why a high degree of factor mobility is important (labour in this case) is because if the wages in country A would rise, labour will shift immediately from country B to country A which again lowers the wages in country A and consequently decreases inflationary pressure. The high degree of factor mobility offsets the loss of sacrificing the flexible exchange rate mechanism and the effective monetary policy that comes with it (Mundell, 1961). This legal framework for full factor mobility already exists within the EU, so this requirement is already met.

In addition, the more open the economy of a country is, the more it benefits from creating a monetary union. The characteristics of an open economy are large trade volumes, large trade values, no trade barriers, etc. (Mc Kinnon, 1963). This already exists to a high degree within the EU, therefore this will not be discussed during the data analysis.

Some economists such as Kenen (1969) believe that Mundell and Mc Kinnon are right but that they do not stress the most important criteria. He argues that the degree of product diversification within a country and fiscal integration are the most important features that a country should have when it joins a monetary union. Simply because a country that produces many different products will also export many different products. This is important because when the demand of one export product decreases, this will not hurt the balance of payments because the country does not solely rely on this product as it has many other export products which compensate for the loss of the other product, therefore the probability of a current account deficit becomes significantly smaller. Product diversification therefore makes the

(8)

8 current account of a country more stable and therefore the need of a flexible exchange rate is also smaller.

On the other hand, the similarities in inflation rates should be a main requirement in order to form an OCA. Fleming (1971) came to the conclusion after conducting research that trade between two countries with similar and low inflation is more stable than countries which do not fulfil this requirement.

The theoretical framework of optimum currency area theories was still weak and incoherent in 1987 when the debate about whether Europe should create a common currency started (Robson, 1987). It was/is impossible to perfectly fulfil all optimum currency area criteria. Therefore, one should distinguish between more and less important criteria in order to complete an optimal monetary integration which was not done by the scholars mentioned earlier. A new question arose, which criteria must be met and which should be met at least to a certain degree in order to create an optimum currency area. Moreover, a few optimum currency criteria are not fully compatible, such as countries with an open economy are usually specialized in fewer products (Ricardo, 1817). This is an incompatibility of the criteria of having an open economy and product diversification. In addition, having an open economy and factor mobility do not always have to go hand in hand, this is called ‘inconclusiveness` (Tavlas 1994).

In 1997, it was proved that countries that traded more with each other showed higher correlated trade cycles which made it more beneficial to create an optimum currency area. This was also applicable to the European Monetary Union. The entrance of countries into the European Monetary Union would lead to more correlated trade figures which leads to less and less asymmetric shocks which then again leads to even more correlated trade (Frankel and Rose, 1997).

However, Mankiw (2000) also argued that fiscal integration is important because when adopting a fixed exchange rate regime, monetary policy is ineffective and therefore fiscal policy should be as similar as possible for the countries within an optimum currency area (so for example same tax laws) (Mongelli, 2002). This is still not the case for the EU, however, the degree of fiscal integration is hard to measure using data and therefore this requirement is only mentioned now but will be ignored during the data analysis. Furthermore, labour mobility might be slow within Europe, nevertheless, the high degree of capital mobility could

(9)

9 Summarising all these scholars, some criteria still seem incompatible or impossible to achieve, however, the main requirement in order to create an optimum currency area is similarity in the economic structure of these participants (definitely regarding the European monetary situation today) such that the consequences of negative external shocks for deficit countries can be solved by the surplus countries within the optimum currency area. For this reason, in this paper, data will be collected in order to analyse the economic similarity of countries by looking at the GDP growth rates, inflation, current accounts, debt/GDP and interest rates.

3. Data analysis (correlations):

In this part of the paper, tables of the correlations will be analysed in order to identify currency areas which can then be analysed further in section 4. From section 4 onwards, standard deviations of these different preliminary currency areas will be calculated and analysed in order to draw a conclusion on whether these currency areas are more integrated than the Eurozone itself. In the following analysis, I consider that a country is highly

correlated to another country if it has a correlation coefficient of at least 0.6 which would mean that, for example, if I consider that Belgium has six highly correlated countries with respect to GDP growth this means that there were six countries with a GDP growth correlation of higher than 0.6 with respect to Belgium. Moreover, I consider that an average correlation is high (green) if it is higher than 0.45, medium (orange) if it is in between 0.4-0.45 and low (red) if it is lower than 0.4. All the average correlations and the amount of highly correlated countries are calculated within the group, so for example, when Belgium has a high

correlation with Greece this does not count because they are not in the same geographical group. The same counts for averages, these are the average correlations that for example Belgium has with the rest of the group. On top of that, every country that has a question mark behind it is an uncertain member of that group, this means that it might be better for that country to join another geographical area or stay on its own.

(10)

10

3.1) Northern European currency area

Table 1.

All with respect to the geographical group Average correlation (GDP growth) No. Of highly correlated countries (GDP

growth) Average correlation (π)

No. Of highly correlated countries (π) Belgium 0,57 6 0,50 4 Netherlands 0,54 5 0,44 1 Luxembourg 0,37 0 0,52 4 Germany 0,51 3 0,30 0 Denmark 0,37 1 0,56 4 Finland 0,51 4 0,51 2 Norway 0,15 0 0,34 0 Sweden 0,48 0 0,54 5 UK 0,49 1 0,51 5 France? 0,58 6 0,57 7 Ireland 0,51 4 Austria? 0,44 1 0,49 1 Switzerland 0,53 2 0,42 0

First of all, Belgium and France should definitely be part of this northern European currency area as GDP growth rates as well as inflation rates are highly correlated with the rest of the group.

Luxembourg and Denmark are a doubt, they have a really low GDP growth correlation although the inflation correlation coefficients are positive. As mentioned in the literature review, it is not clear yet which criteria of optimum currency areas are more important, so for simplicity, all countries get the benefit of the doubt for now when they have the same amount of positive and negative results regarding the correlations. It is an interesting remark that other small economies such as Malta and Cyprus (which will be included in Table 2) also showed relatively low correlations with their neighbour countries, this is probably due to the fact that they are either geographically isolated, economically insignificant (little amount of trade) or both. Denmark is also not part of the Eurozone which might be another reason that these indicators are still not converged (Frankel and Rose, 1997). Germany and the

(11)

11 Finland and Sweden should be part of this currency area without a doubt as they seem to be well integrated with the north. Norway, however, shows very different results, namely not a single significant positive correlation with the rest of the group. One reason for this is that it is not part of the Eurozone, nevertheless, it is going to be excluded from the northern European currency area in this thesis.

Even having the pound as a currency, the United Kingdom shows a lot of similarities with the mainland of northern Europe. This is the reason that they will be put with the northern group. For Ireland, it is hard to say. Due to the omitting statistics of GDP growth rates, there is less statistical power which consequently makes the decision less accurate. Moreover, in the past years, Ireland has been associated a lot with the southern European countries and their statistics such as debt/GDP, current account deficits, etc. Therefore, there is not sufficient evidence that Ireland should be part of the northern European currency area.

Switzerland is not part of the Eurozone, nevertheless, it still looks reasonably

correlated with the rest of the northern group. Also, due to the reason stated by Frankel and Rose mentioned above, these correlations become even higher if it joins a northern European currency area. Austria could be regarded as a northern or as an eastern European country but the results are doubtful for a Eurozone member. If Austria is less similar to the eastern

European group, it should join the north. So, the northern European group will now include all countries except from Norway and Ireland.

(12)

12

3.2) Southern European currency area

Table 2.

First of all, Malta is completely different from the rest of southern Europe and even from the whole of Europe, so that is why it is excluded from any group from now on. Greece was expected to be highly correlated with at least Spain, Portugal and Italy, however it is only highly correlated with Italy and Cyprus regarding inflation and with Spain regarding GDP growth rates. Because Italy fits the group very well and does not correlate that much with Cyprus, Italy will remain in this group, Greece will be excluded as well as Cyprus.

Spain is by far the most highly correlated country with respect to the rest of the group so it should definitely be part of the southern European currency area. Portugal, on the other hand, should not be part of this currency area because the only high correlation within its group is with Spain. Due to the fact that Spain will be part of the southern European currency area, Portugal will be excluded and will be on its own.

Croatia could be regarded as part of either east or southern Europe. After this part of the data analysis it will be clear to which group it fits better. Croatia has a high average correlation of GDP growth with the rest of the group, however, not that many countries. Also, its inflation path is not similar to the rest of the south, this might be due to the fact that it is not part of the Euro. Therefore, it is still unclear whether it would be beneficial to include them with the south or not.

No. Of highly correlated countries (π) Greece 0,43 1 0,34 2 Spain 0,62 8 0,44 3 Croatia? 0,50 2 0,29 0 France? 0,54 3 0,36 1 Italy 0,58 5 0,42 3 Cyprus 0,42 1 0,32 1 Malta 0,26 0 0,24 0 Portugal 0,46 1 0,37 2 Bulgaria? 0,53 2 0,03 0 Hungary? 0,52 3 0,14 0

(13)

13 France could be within the southern European currency area, however, it fits in better with the northern European currency area and therefore it will not join the south. Bulgaria has three out of four negative values regarding the correlations, probably partly due to not having the Euro as a currency, but it should not join the south. It should either join the east, create an even smaller currency area or stay on its own. Hungary, however, should join the southern European currency as it has more positive than negative results whilst not having the Euro either, that means that there is still a high potential convergence with the rest of the currency area.

For now, the imaginary southern European currency area will consist of Italy, Spain, Croatia (for now) and Hungary. The currency areas that other countries should be part of will be determined after the eastern European currency area analysis.

3.3) Eastern European currency area Table 3.

No. Of highly correlated countries (π) Bulgaria? 0,54 3 0,17 0 Czech Republic 0,56 3 0,21 0 Croatia? 0,51 1 0,33 0 Hungary? 0,48 1 0,42 1 Austria? 0,37 0 0,10 0 Poland? 0,27 0 0,37 1 Romania 0,43 2 0,26 0 Slovenia 0,55 4 0,34 0 Slovakia 0,23 0

Firstly, Austria and Poland are insignificantly correlated with the rest of eastern Europe, for

this reason, they will be excluded from the eastern European currency area. They are also insignificantly correlated with each other which makes them better off alone. Perhaps, Austria would benefit from sharing a currency with Switzerland as these two countries show high correlations with each other, however, further research is needed to be able to conclude this. Hungary is more correlated with southern Europe than with eastern Europe, so they will join the south in this analysis.

(14)

14 Croatia and Hungary are more correlated with southern Europe than with the east. However, Croatia was also a not surely beneficial to the south, it does show some high correlations with southern countries such as Hungary and Italy. Croatia is not sufficiently correlated with the southern European currency area, however, the southern European currency area got reduced to Spain, Hungary and Italy and therefore Croatia can join this currency area. Hungary did show significant evidence of high correlations with the south and not with the east, so it will join the southern currency area as well. Slovenia has got high GDP growth rate correlations with the rest of eastern Europe, so it should join this currency area.

To summarise all this, the eastern European currency area will consist of Slovenia, Bulgaria and Czech Republic which also makes it possible for Romania to join this currency area as it shows some high correlations with these countries.

3.4) The Baltic area Table 4.

No. Of highly correlated countries (π) Estonia 0,46 0 0,52 0 Latvia 0,45 0 0,56 1 Lithuania 0,38 0 0,56 1 Poland? 0,27 0 0,49 0

First, Poland should definitely be excluded from any currency area as correlations are way too low in general. Estonia could be part of a Baltic currency area as average correlations of both GDP and inflation are sufficiently significant. With the departure of Poland, Latvia and

Lithuania also become significantly correlated with each other, however, because a standard deviation analysis on three countries is not reliable at all, the Baltic area will be added to the eastern European currency area to see whether this adds benefits to this area.

(15)

15

3.5) Summary of currency areas

To conclude, based on the correlations analysed above, in the rest of the study we will use the following split of countries in currency areas:

- The northern European currency area: Belgium, Luxembourg, the Netherlands,

Germany, Denmark, Finland, Sweden, UK, Austria, Switzerland and France.

- The southern European currency area: Spain, Italy, Croatia and Hungary.

- The eastern European currency area excluding the Baltic States: Slovenia, Bulgaria,

Czech Republic and Romania.

- The eastern European currency area including the Baltic States: Slovenia, Bulgaria,

Czech Republic, Romania, Latvia, Estonia and Lithuania.

The other countries that were analysed are determined to be better of having their own currency at least at this moment in time and they will not be discussed anymore because they will not be part of an optimum currency area within Europe for the time being which makes them irrelevant for this paper from now on.

4. Data analysis (standard deviations)

Now that the currency areas are formed, the standard deviations of the five indicators will be compared between the Eurozone and the newly created hypothetical currency areas over time. According to the optimum currency area theory, the more similar shocks are between countries, the more benefits an optimum currency area creates. In addition, similar shocks means a low standard deviation of these indicators. In this section, the standard deviations of these indicators for each group will therefore be compared to the Eurozone and will be

subject to a time analysis. This is a similar method to the one used by De Grauwe and Mongelli in 2005. If these standard deviations are in general significantly lower than the Eurozone standard deviations, the conclusion to the research question will be that it would be

economically beneficial to Europe to create more than one optimum currency area. Moreover, a comparison of the development of these indicators with the Maastricht criteria will also be made to draw a conclusion about its usefulness, its feasibility and whether it is realistic to believe that these indicators will remain subject to these criteria on the long term. Also, plotting all the groups in one graph would have been too unclear so they are separated from

(16)

16 each other. In addition, every time the words north, south and east are mentioned, this refers to the hypothetical currency area and not the whole geographical area.

4.1 GDP growth

There was no GDP growth rate criteria included in the Maastricht Treaty, however, it is a really important indicator to state whether economies are similar because many supply shocks and changes in these other indicators have an effect on GDP growth. It could be said to be the major or covering indicator of all indicators and it displays the health of the economy as a whole.

Graph 1.

First of all, Graph 1 shows that the standard deviation on GDP growth is on average lower for the northern group which means that there is more similarity in GDP growth rates between its members and their economies are therefore more interconnected. It is also observable at what point the crisis started, this is where both lines peeked, in 2008. This is in line with the expectations as shocks become more asymmetric during a crisis were some countries are affected by the burst way earlier than others. Furthermore, a few years after the start of the

0,00 0,50 1,00 1,50 2,00 2,50 3,00 3,50 1995Q 1 1996Q 1 1997Q 1 1998Q 1 1999Q 1 2000Q 1 2001Q 1 2002Q 1 2003Q 1 2004Q 1 2005Q 1 2006Q 1 2007Q 1 2008Q 1 2009Q 1 2010Q 1 2011Q 1 2012Q 1 2013Q 1 2014Q 1 St an da rd d ev ia tio n

Periods (in quarters starting from1995)

GDP growth

Eurozone North

(17)

17 crisis, it is observable that countries slowly started to converge again (with respect to GDP growth) whilst during the crisis they diverged.

Graph 2.

The ‘South’ line crosses the ‘Eurozone’ line a few times which indicates that the standard deviation has been higher for the South than for the Eurozone sometimes. It is also notable that the peak of the ‘South’ line was a few years before the peak of the ‘North’ line, namely in the fourth quarter of 2006 which is about two years before the north started to diverge (after the collapse of the Lehman Brothers in October 2008).

0,00 0,50 1,00 1,50 2,00 2,50 3,00 3,50 1995Q 1 1996Q 1 1997Q 1 1998Q 1 1999Q 1 2000Q 1 2001Q 1 2002Q 1 2003Q 1 2004Q 1 2005Q 1 2006Q 1 2007Q 1 2008Q 1 2009Q 1 2010Q 1 2011Q 1 2012Q 1 2013Q 1 2014Q 1 St an da rd d ev ia tio n

Periods (in quarters starting from 1995)

GDP growth

Eurozone South

(18)

18

Graph 3.

In the late years of the twentieth century, eastern Europe was still very different from the north and the south. After the turn of the century, the east became more similar to the Eurozone and its members and also converged within their own group. The most recent years are more important than the earlier years when analysing whether or not this currency area would be an improvement to the current currency area so this group performs well on the whole. It is remarkable that the standard deviation of the east did not peak that much during the crisis. 0,00 0,50 1,00 1,50 2,00 2,50 3,00 3,50 1995Q 1 1995Q 4 1996Q 3 1997Q 2 1998Q 1 1998Q 4 1999Q 3 2000Q 2 2001Q 1 2001Q 4 2002Q 3 2003Q 2 2004Q 1 2004Q 4 2005Q 3 2006Q 2 2007Q 1 2007Q 4 2008Q 3 2009Q 2 2010Q 1 2010Q 4 2011Q 3 2012Q 2 2013Q 1 2013Q 4 2014Q 3 St an da rd d ev ia tio n

GDP growth

Eurozone East

(19)

19

Graph 4.

Including the Baltic States to the eastern group, the standard deviations get much higher and even higher than the Eurozone on the whole. This means there is no reason to include the Baltic States with respect to GDP growth rate. One difference between Graph 3 and Graph 4 is that the crisis peak does exist in Graph 4 and it does not in Graph 3. The reason for this would therefore be that the Baltic States and the east diverged with respect to each other. This could be because the Baltic States do have the Euro and many members of the eastern currency area do not.

4.2 Inflation

In the Maastricht Treaty, inflation was one of the five criteria. The criteria was as follows: ‘Inflation should not be more than 1.5 percentage points above the average rate of the three European Union member states with the lowest inflation over the previous year.’ This criteria was definitely not met starting a few years after the criteria were set. During the crisis, there was even a period of deflation for most Eurozone countries. Inflation is also an indicator of economic growth just as GDP. For this reason, similar standard deviations of these indicators between areas resemble similar economies which could have a similar exchange

0,00 0,50 1,00 1,50 2,00 2,50 3,00 3,50 4,00 1995Q 1 1995Q 4 1996Q 3 1997Q 2 1998Q 1 1998Q 4 1999Q 3 2000Q 2 2001Q 1 2001Q 4 2002Q 3 2003Q 2 2004Q 1 2004Q 4 2005Q 3 2006Q 2 2007Q 1 2007Q 4 2008Q 3 2009Q 2 2010Q 1 2010Q 4 2011Q 3 2012Q 2 2013Q 1 2013Q 4 2014Q 3 St an da rd d ev ia tio n

GDP growth

Eurozone East + Baltic

(20)

20 rate or even a common currency. This data, European inflation rates, were not seasonally adjusted, so the average of the twelve months prior to the month in question was taken in order to remove seasonality and the consequent hyper fluctuation of the graphs.

Graph 5.

The inflation rates of the north were quite stable during the whole period (1996-2014), this is because price levels and wages have been similar for a long time. The Eurozone however does not show this. The peak right before the crisis resembles the difference in inflation rates between mainly the north, the south and the Baltic States at that time. Inflation rates were reasonably stable in the north and increased rapidly in the south mainly due to the housing bubble. In the crisis, the southern inflation rates changed from fairly high inflation (above two percent) to deflation in a short period of time. Graph 5 shows again that the north is better of alone as the standard deviation of inflation is much lower in general than for the Eurozone as a whole. It is remarkable that in the years after the crisis the inflation rates of the north and the Eurozone converged again, this is probably due to ECB policy with the goal of maintaining a stable and low inflation rate of not more than two percent.

0,000 0,050 0,100 0,150 0,200 0,250 0,300 0,350 1996M 02 1996M 12 1997M 10 1998M 08 1999M 06 2000M 04 2001M 02 2001M 12 2002M 10 2003M 08 2004M 06 2005M 04 2006M 02 2006M 12 2007M 10 2008M 08 2009M 06 2010M 04 2011M 02 2011M 12 2012M 10 2013M 08 St an dar d d ev iat ion

Periods (in months starting from 1996)

Inflation

Eurozone North

(21)

21

Graph 6.

Regarding inflation rates, the south does not significantly outperform the Eurozone as a whole. Especially before the Maastricht criteria were set, the southern inflation rates were significantly different from each other. As mentioned before, inflation rates converged for the north and the south probably due to ECB policy.

0,000 0,100 0,200 0,300 0,400 0,500 0,600 0,700 1996M 02 1996M 12 1997M 10 1998M 08 1999M 06 2000M 04 2001M 02 2001M 12 2002M 10 2003M 08 2004M 06 2005M 04 2006M 02 2006M 12 2007M 10 2008M 08 2009M 06 2010M 04 2011M 02 2011M 12 2012M 10 2013M 08 St an dar d d ev iat ion

Inflation

Eurozone South

(22)

22

Graph 7.

In this graph, it is clearly shown that the east had completely different inflation rates at first and started to converge around late 1998 when more and more eastern European countries wanted to join the European Monetary Union. In the recent years, the inflation in the east become lower and more stable. If this trend continuous it could be better to create an own currency area, but for now, the Eurozone performs better.

0,000 0,200 0,400 0,600 0,800 1,000 1,200 1,400 1,600 1,800 2,000 1996M 02 1996M 12 1997M 10 1998M 08 1999M 06 2000M 04 2001M 02 2001M 12 2002M 10 2003M 08 2004M 06 2005M 04 2006M 02 2006M 12 2007M 10 2008M 08 2009M 06 2010M 04 2011M 02 2011M 12 2012M 10 2013M 08 St an dar d d ev iat ion

Inflation

Eurozone East

(23)

23

Graph 8.

When the Baltic States are included, it is better, but still not sufficient. The Eurozone still has a lower average standard deviation and thus is a more optimal currency area according to this criteria. The fact that the east has such a high standard deviation in the last years of the twentieth century is because Romania and especially Bulgaria suffered from hyperinflation.

4.3 Debt/GDP

According to the Maastricht criteria, the debt/GDP ratio should not be higher than 0.6. This worked in the first years after the Maastricht Treaty, nevertheless, it has not been achieved during the crisis and for some countries such as Italy and Greece it has never succeeded. Since 2008, the average of the European Union has always been higher than 0.6. If this upper bound of 0.6 cannot be achieved, at least the standard deviations of the debt/GDP ratio should be small. 0,000 0,200 0,400 0,600 0,800 1,000 1,200 1,400 1,600 1,800 2,000 1996M 02 1996M 12 1997M 10 1998M 08 1999M 06 2000M 04 2001M 02 2001M 12 2002M 10 2003M 08 2004M 06 2005M 04 2006M 02 2006M 12 2007M 10 2008M 08 2009M 06 2010M 04 2011M 02 2011M 12 2012M 10 2013M 08 St an dar d d ev iat ion

Inflation

Eurozone East + Baltics

(24)

24

Graph 9.

It is clear that the standard deviations of debt/GDP of the north are lower than those of the Eurozone as a whole. Just as for inflation, the north has had a constant debt/GDP standard deviation throughout the years for which data was available. This is also logical because the northern economies have been similar for a longer time than the southern and eastern economies. During the sovereign debt crisis, the debt/GDP within the Eurozone started to diverge because for the southern countries it increased at a much quicker rate than for the northern countries. 0,00 5,00 10,00 15,00 20,00 25,00 30,00 35,00 40,00 2000Q 1 2000Q 4 2001Q 3 2002Q 2 2003Q 1 2003Q 4 2004Q 3 2005Q 2 2006Q 1 2006Q 4 2007Q 3 2008Q 2 2009Q 1 2009Q 4 2010Q 3 2011Q 2 2012Q 1 2012Q 4 2013Q 3 2014Q 2 St an dar d d ev iat ion

Periods (in quarters starting from 2000)

Debt/GDP

Eurozone North

(25)

25

Graph 10.

For the southern currency area, the graph looks a bit similar to the graph of the north except from the fact that the southern countries actually converged towards each other because they all increased their debt/GDP ratio rapidly two years after the crisis started. Although this is not a positive cause, the standard deviation decreased for that reason, so the advantages of creating a common currency area improved.

0,00 5,00 10,00 15,00 20,00 25,00 30,00 35,00 40,00 2000Q 1 2000Q 4 2001Q 3 2002Q 2 2003Q 1 2003Q 4 2004Q 3 2005Q 2 2006Q 1 2006Q 4 2007Q 3 2008Q 2 2009Q 1 2009Q 4 2010Q 3 2011Q 2 2012Q 1 2012Q 4 2013Q 3 2014Q 2 St an dar d d ev iat ion

Debt/GDP

Eurozone South

(26)

26

Graph 11.

The debt/GDP ratios of the east have converged in the years prior to the crisis and have diverged again when the crisis started. This is the opposite of the south were countries debt/GDP converged during the crisis probably due to the fact that they shared a common currency. The standard deviations for the east have been constantly low, even lower than the north and the south.

Graph 12. 0,00 5,00 10,00 15,00 20,00 25,00 30,00 35,00 40,00 2000Q 1 2000Q 4 2001Q 3 2002Q 2 2003Q 1 2003Q 4 2004Q 3 2005Q 2 2006Q 1 2006Q 4 2007Q 3 2008Q 2 2009Q 1 2009Q 4 2010Q 3 2011Q 2 2012Q 1 2012Q 4 2013Q 3 2014Q 2 St an dar d d ev iat ion

Debt/GDP

Eurozone East 0,00 5,00 10,00 15,00 20,00 25,00 30,00 35,00 40,00 2000Q 1 2000Q 4 2001Q 3 2002Q 2 2003Q 1 2003Q 4 2004Q 3 2005Q 2 2006Q 1 2006Q 4 2007Q 3 2008Q 2 2009Q 1 2009Q 4 2010Q 3 2011Q 2 2012Q 1 2012Q 4 2013Q 3 2014Q 2 St an dar d d ev iat ion

Debt/GDP

(27)

27 In the case of adding the Baltic States, more or less the same graph is plotted. The standard deviations are on average a little bit higher than the scenario were the Baltic States are excluded but this group still receives a point.

4.4 National budget deficit

A maximum national budget deficit of 3% was the criteria stated by the Maastricht Treaty. Nevertheless, even for countries such as Germany and the Netherlands this was not achieved for the majority of periods (quarters). So far, none of the Maastricht criteria were met on the long run, this can either be because the criteria are too difficult to meet or because the penalties on crossing these limits are too low. Anyway, the result is that countries cross the line and diverge from a common benchmark which is needed in order to benefit from a

common currency area. The national budget deficit just like the GDP growth and inflation, are indicators for which the Eurozone countries diverged rapidly when the crisis started. The reason for this is probably that different fiscal policies exist within the Eurozone. This also shows that fiscal integration has been lacking and is way behind monetary integration. Some countries within the Eurozone have anti cyclical fiscal policies which increases the national budget deficit during the crisis and others have cyclical fiscal policies which decreases the national budget deficit during the crisis. For Debt/GDP this is slightly different because it is build up over time, it is not an indicator for one specific year or small period of time but changes slowly, this is the reason that peaks are less sharp or inexistent.

(28)

28

Graph 13.

As mentioned above, the Eurozone countries diverged rapidly during the later years of the crisis (especially the north and the south around 2010-2011) and converged again after the rock-bottom was reached. The north converged during the crisis, this is again evidence of bipolarity within the Eurozone during the crisis. The north outperforms the Eurozone on this indicator. 0,00 1,00 2,00 3,00 4,00 5,00 6,00 7,00 8,00 2003Q 1 2003Q 3 2004Q 1 2004Q 3 2005Q 1 2005Q 3 2006Q 1 2006Q 3 2007Q 1 2007Q 3 2008Q 1 2008Q 3 2009Q 1 2009Q 3 2010Q 1 2010Q 3 2011Q 1 2011Q 3 2012Q 1 2012Q 3 2013Q 1 2013Q 3 2014Q 1 2014Q 3 St an dar d d ev iat ion

National budget deficit

Eurozone North

(29)

29

Graph 14.

The graph of the south looks less convincing than the north, especially for the early years (2003-2008), the years prior to the crisis. The south receives half a point because in the later years it outperforms the Eurozone, but the result does not look significant for now. The line of the south also fluctuates a lot which is not positive because it shows instability of national budget deficits across its members. For example, around 2006, Hungary had a budget deficit of around 9% whereas the budget deficits of Spain was around 2%, this explains the peak of the south around that time.

0,00 1,00 2,00 3,00 4,00 5,00 6,00 7,00 8,00 St an dar d d ev iat ion

National budget deficit

Eurozone South

(30)

30

Graph 15.

This graph looks better than the graph of the south, however, in the recent years, the standard deviation of the east has increased a lot. Due to the fact that later years are more important, especially for the short run, it is still doubtful.

Graph 16. 0,00 1,00 2,00 3,00 4,00 5,00 6,00 7,00 8,00 2003Q 1 2003Q 4 2004Q 3 2005Q 2 2006Q 1 2006Q 4 2007Q 3 2008Q 2 2009Q 1 2009Q 4 2010Q 3 2011Q 2 2012Q 1 2012Q 4 2013Q 3 2014Q 2 St an dar d d ev iat ion

National budget deficit

Eurozone East 0,00 1,00 2,00 3,00 4,00 5,00 6,00 7,00 8,00 2003Q 1 2003Q 4 2004Q 3 2005Q 2 2006Q 1 2006Q 4 2007Q 3 2008Q 2 2009Q 1 2009Q 4 2010Q 3 2011Q 2 2012Q 1 2012Q 4 2013Q 3 2014Q 2 St an dar d d ev iat ion

National budget deficit

(31)

31 When the Baltic States are included, the graph of the east looks similar in shape but much better. It only crosses the Eurozone line once and by very little, so it receives one point. The reason for this divergence of the east in the later years is due to a budget deficit of around 13% for Slovenia in the last year. This is more than four times as much as the maximum budget deficit should be according to the Maastricht Criteria.

4.5 Interest rate

The interest rate is also one of the Maastricht criteria, it states: ‘Long-term interest rates should be less than two percentage points above the rate of the three EU countries with the lowest inflation over the previous year.’ This criteria was also not met by most of the countries, before, during nor after the crisis. The difference between the interest rates of countries such as Luxembourg and Hungary or Romania for example in 2005 with 2% and 7% respectively was bigger than 2% for most periods. The Eurozone interest rates diverged, first during the credit crisis (because ‘cash was king’ so interest rates had to increase) and later even more during the sovereign debt crisis (due to increased default risk). These two crises are shown on the graphs by the two peaks on the Eurozone line. Some countries such as

Greece almost went bankrupt as they started to default on their debt. Due to this default risk, the interest rates had to rise substantially in order to prevent a capital outflow or even a run on the banks. Graph 17. 0 1 2 3 4 5 6 2001Q 1 2001Q 4 2002Q 3 2003Q 2 2004Q 1 2004Q 4 2005Q 3 2006Q 2 2007Q 1 2007Q 4 2008Q 3 2009Q 2 2010Q 1 2010Q 4 2011Q 3 2012Q 2 2013Q 1 2013Q 4 2014Q 3 St an da rd d ev ia tio ns

Long term interest rates

Eurozone North

(32)

32 It is clear that the north has really stable, low and similar interest rates, around 3%-4% and even 1%-2% in recent years. This is probably due to the increased difference in default risk between countries such as Germany or the Netherlands and Spain or Italy, so that the interest rate in Spain and Italy increased even further and the opposite happened to the northern countries. The line of the north does not cross the Eurozone line once.

Graph 18.

The members of ‘South’ had similar long term interest rates for the last years, but the south had a higher standard deviation for the years 2001-2008. This graph does seem to show that during a burst it is better to have a small currency area and during a boom it is better to have a big currency area.

0 1 2 3 4 5 6 2001Q 1 2001Q 4 2002Q 3 2003Q 2 2004Q 1 2004Q 4 2005Q 3 2006Q 2 2007Q 1 2007Q 4 2008Q 3 2009Q 2 2010Q 1 2010Q 4 2011Q 3 2012Q 2 2013Q 1 2013Q 4 2014Q 3 St an da rd d ev ia tio ns

Long term interest rates

Eurozone South

(33)

33 0 1 2 3 4 5 6 2001Q 1 2001Q 4 2002Q 3 2003Q 2 2004Q 1 2004Q 4 2005Q 3 2006Q 2 2007Q 1 2007Q 4 2008Q 3 2009Q 2 2010Q 1 2010Q 4 2011Q 3 2012Q 2 2013Q 1 2013Q 4 2014Q 3 St an da rd d ev ia tio ns

Long term interest rates

Graph 19.

According to this graph, the last years have been positive for the east interest rate wise. The pattern of the standard deviation for the east looks more random than the graphs of the north and the south. Only the peak of the credit crisis of 2008 is the same as for the south and the Eurozone. Graph 20 0 1 2 3 4 5 6 2001Q 1 2001Q 4 2002Q 3 2003Q 2 2004Q 1 2004Q 4 2005Q 3 2006Q 2 2007Q 1 2007Q 4 2008Q 3 2009Q 2 2010Q 1 2010Q 4 2011Q 3 2012Q 2 2013Q 1 2013Q 4 2014Q 3 St an da rd d ev ia tio ns

Long term interest rates

Eurozone East

(34)

34 Including the Baltic States makes the standard deviations increase in general. For example, for the credit crisis of 2008-2009, this currency area diverges even more than the Eurozone

interest rate wise, this might be because some countries have the Euro and some do not. If this area keeps up with the low standard deviations of the last few years it might be useful to reconsider it, but for now, this is not enough for this criteria to get any points.

5. Conclusion

To conclude, the research question is: ‘Would it be better from the point of view of the convergence criteria to split the European Union, including the Eurozone, into more than one currency area? The conclusion that can be drawn from the data analysis, is that the north is definitely the best integrated currency area of all the hypothetical currency areas that were created. For the south and the east it is still doubtful whether they are sufficiently integrated to create a currency area or not because not all criteria are met and these currency areas do not significantly outperform the Eurozone. The eastern currency area should definitely not include the Baltic States as this increases the standard deviation in general and this group did not even pass half of the criteria. To answer the research question, the north should definitely create a currency area as it would be more optimal than the Eurozone is now. The south and the east however are still more diverged and might need more fiscal integration and policy coordination before a currency area would be beneficial. Moreover, perhaps the Baltic States should create their own currency area just like Switzerland and Austria should because they are highly correlated for the indicators GDP growth and inflation, probably for the other indicators as well. It was not possible to see whether this is true by the use of this standard deviation analysis because a standard deviation of two or three countries will always be small and this outcome would have a negligible amount of statistical power. Countries that were expected to become part of a currency area such as Greece seemed to be too economically different to include them in a currency area, even compared to their neighbours. Another remark is that in general, the Eurozone performed better before the crisis and the

hypothetical smaller currency areas performed better during the crisis. Basically, being part of a big currency area increases risk, either growth rates area really high (in a boom) or really low (in a burst). The main reason for this is interdependence, the real high growth rates in a boom are partly explained by free riding and the real low growth rates are partly explained by contagious shocks and contagious decreases in currency value. Sharing a common currency can be dangerous when one member of a currency area does not perform well economically

(35)

35 because this decreases the value of its currency and puts inflationary pressure on the other economies for whom the currency should still be strong e.g. Germany. Productivity levels and other characteristics of the economies within a currency area should be similar, if not,

countries such as Greece become uncompetitive as their exports become way too expensive. For countries such as Germany, exports become way too cheap because the Euro is too cheap and not in line with German productivity levels. In addition, the Maastricht criteria which are still relevant (these five indicators except from GDP growth) were not met at all. Either countries should be fined for exceeding these limits or these criteria were not realistic at all and should be changed. It is hard for that many countries to obey the same rules whilst their economies are still so different. Out of these two options mentioned above, it seems that the criteria were not realistic for this currency area with this amount of significant differences between member economies, but for example would be better criteria for a northern

European currency area. It can be concluded that political leaders of the Eurozone countries have been way too optimistic about this monetary union as a whole, about labour mobility and about symmetry within Europe and forgot about the importance of political integration,

economic convergence and decreasing cultural gaps before a currency area becomes beneficial. Of course there are benefits such as the disappearance of transaction costs and exchange rate risk, but for the benefits to outweigh the social and economic costs, more economic, social and political convergence is needed than what now exists amongst the Eurozone countries at this moment, if not, countries might be better off alone.

(36)

36

6. Limitations and further research:

One difficulty for this paper was that one should actually distinguish between more and less important criteria in order to complete an optimal analysis, however, it is hard to say which criteria should weigh more than another. Furthermore, the analysis is purely economic, if the political aspect of optimum currency area theory could have been included not only in the literature review but also in the data section, the outcome would have been more realistic and the decision would have more statistical power. Moreover, Germany could be on its own regarding the inflation rates because they are even different to other northern European countries, nevertheless, it was assumed in this paper that Germany would be more similar with respect to the other indicators and fortunately this was found to be true. In addition, as stated before, only the criteria of symmetrical shocks was analysed as the other criteria were either met or hard to measure, such as the degree of capital mobility which is already high and existent and the degree of labour mobility which is hard to measure and lower. Future research might question why the smaller currency areas perform better during a crisis and vice versa. Furthermore, is there a way to quickly switch to smaller currency areas within a big currency area (such as the Eurozone) and hedge an economy from foreign contagion? This latter question is really far away from reality still so it might not be answered for a long time.

(37)

37

7. Bibliography

Corden, W. (1972). Monetary integration. Princeton, N.J.: International Finance Section, Princeton University

Cuyvers, L. (2012). Export promotion: A decision support model approach. Stellenbosch: Sun Press

Dornbusch, R. (1985). Exchange rates and prices. Cambridge, Mass. (1050 Massachusetts Avenue, Cambridge 02138): National Bureau of Economic Research

Fleming, J. (1971). On exchange rate unification. S.l.: International Monetary Fund, Research Department

Frankel, & Rose. (1997). Is EMU more justifiable ex post than ex ante? European economic review

Horvath, R., & Komarek, L. (2002). Optimum currency area theory: A framework for

discussion about monetary integration. Warwick (GB): University of Warwick, Department of

Economics

Kenen, P. (1969). International economics. Englewood Cliffs: Prentice Hall Mankiw, N. (2000). Macroeconomics (4th ed.). New York: Worth

Mc Kinnon. (1963). Optimum currency areas. The American Economic Review

Mongelli, F. (2002). 'New views on the optimum currency area theory: What is EMU telling us?’ Ricardo, D. (1817). On the principles of political economy and taxation

Robson, P. (1987). The economics of international integration (3rd rev. ed.). London: Allen & Unwin

Tavlas. (2004). On the road again: An essay on the optimal path to EMU for the new member states. Journal of Policy Modeling