The Financial Integration in the European Capital Market Using a Clustering Approach on Financial Data
Round 1
Reviewer 1 Report
In this paper, the authors propose to analyse the financial integration in Europe, with a clustering approach applied to financial data. The topic about financial integration is relevant, even more in a context like the actual one. Although, the paper needs to have several changes in order to be considered in conditions to be published. Some of the problems are the following:
1. The introduction has several statement which need to have some references. Just in the first paragraph the authors refer, at least, to: i) the effects of developed financial markets; ii) the need of an effective legislative framework; iii) the improvements in the transparency. At least these issues should be referenced.
2. Still in this context, and despite the benefits of developed financial markets, authors should clearly identify the benefits and risks of financial integration.
3. Figure 1 has some problems. It is difficult to be read (fonts are very small) and some statements are also meaningless. For example, authors state that "On the other hand, in Europe, the euro area capital market is larger than the capital market of countries outside the euro area.". This is natural considering the dimension of the Eurozone when compared with the non-Eurozone one.
4. Where does the authors found that the capital market union has emerged as an initiative in 2014? Do the authors know the history of the European Union?
5. The studies about financial integration could be made considering several different approaches and also have different definitions. Authors should reflect on this, identifying the different concepts about financial integration as well as the different approaches to study the phenomenon (see, for example, Adam, K., Jappelli T., Menichini A. M., Padula M., and M. Pagano (2002), Analyse, Compare, and Apply Alternative Indicators and Monitoring Methodologies to Measure the Evolution of
Capital Market Integration in the European Union, Report to the European Commission.
6. In a study like this, authors should make a literature review about the earlier evolution of the financial integration in the EU. It is possible to find dozens of works devoted to this topic Authors should concentrate their attention on studies about similar variables under study but they need to deepen their literature review.
7. Authors should also discuss their methodology and the others which are used for the same purpose. I have on important question: if the authors have access to time series, why they do not use them? Cluster analysis is not so powerful than other techniques applied to time series. Why those approaches are not used? And is cluster analysis robust to issues like non-stationarity, for example, a often common problem in time series? Did the authors control for this?
8. Authors refer, in the results, several times to the "performance" of financial sector. Where do they base that performance?
9. The conclusion does not fit with the graphs. And a study like this needs to have a relevant discussion on the findings. None discussion is made.
10. Another important question: why to use UK? Because it was made considering data before the Brexit And more important, if the objective is to assess financial integration in the EU, why Norway? Norway does not belong to the EU.
Author Response
Dear sir or madam,
thank you very much for very interesting and important points. We tried to do our best to improve and change our paper according to your recommendations. I believe that after editing, the article meets your requirements. Thank you for your time and excellent advice, if you have any further comments let us know and we will adjust the article accordingly.
1. The introduction has several statement which need to have some references. Just in the first paragraph the authors refer, at least, to: i) the effects of developed financial markets; ii) the need of an effective legislative framework; iii) the improvements in the transparency. At least these issues should be referenced.
We add correct citations there:
Tahir Akhtar: Market regimes and stock returns among emerging and developed financial markets
Ali C. Akyol: Do regulatory changes affect the underpricing of European IPOs.
2. Still in this context, and despite the benefits of developed financial markets, authors should clearly identify the benefits and risks of financial integration.
A paragraph dealing with this issue has been added to the Introduction chapter (see Agenor, 2001 or Schmukler, 2004)
3. Figure 1 has some problems. It is difficult to be read (fonts are very small) and some statements are also meaningless. For example, authors state that "On the other hand, in Europe, the euro area capital market is larger than the capital market of countries outside the euro area.". This is natural considering the dimension of the Eurozone when compared with the non-Eurozone one.
All the figures were completely changed to reach better clarity.
4. Where does the authors found that the capital market union has emerged as an initiative in 2014? Do the authors know the history of the European Union?
We forgot to write the source of this information, we have added author Robert Schuman to references. Thanks for your warning.
5. The studies about financial integration could be made considering several different approaches and also have different definitions. Authors should reflect on this, identifying the different concepts about financial integration as well as the different approaches to study the phenomenon (see, for example, Adam, K., Jappelli T., Menichini A. M., Padula M., and M. Pagano (2002), Analyse, Compare, and Apply Alternative Indicators and Monitoring Methodologies to Measure the Evolution of
Capital Market Integration in the European Union, Report to the European Commission.
We add paper Financial integration in the EU28 equity markets: Measures and drivers by M. Nardo which deals with it.
6. In a study like this, authors should make a literature review about the earlier evolution of the financial integration in the EU. It is possible to find dozens of works devoted to this topic Authors should concentrate their attention on studies about similar variables under study but they need to deepen their literature review.
We add to our paper a chapter that deals with financial integration.
7. Authors should also discuss their methodology and the others which are used for the same purpose. I have on important question: if the authors have access to time series, why they do not use them? Cluster analysis is not so powerful than other techniques applied to time series. Why those approaches are not used? And is cluster analysis robust to issues like non-stationarity, for example, a often common problem in time series? Did the authors control for this?
We described methodology a little bit more, also produced more clear graphs and 2 brand new graphical outputs that can help readers with understanding the methodology used.
It is a very good comment and discussion about time series. We wanted to produce some classification and the "classic" clustering we use allows us only to do it once per one year so maybe it looks static but because of that, we performed an analysis in 4 different years. We also mention it in conclusion as well: "In future analyzes, we plan to eliminate the shortcoming of this static clustering method and, instead of analyzing the values in a specific year, use the approach using dynamic time warping distance and then performing detailed and more dynamic cluster analysis.
9. The conclusion does not fit with the graphs. And a study like this needs to have a relevant discussion on the findings. None discussion is made.
We add policy recommendations to the conclusion chapter so this chapter was extended. Graphs now are clear and easy to read and interpret.
10. Another important question: why to use UK? Because it was made considering data before the Brexit And more important, if the objective is to assess financial integration in the EU, why Norway? Norway does not belong to the EU.
Through the Agreement on the European Economic Area, Norway is part of the European single market and also participates in some EU programs, but without voting rights. These are, in particular, the areas of business, environment, research, and education. The Capital Markets Union project can be another step towards successful bilateral cooperation.
Data from the United Kingdom were used, as the available data are from the period when the United Kingdom was part of the European Union.
Reviewer 2 Report
The paper investigates the financial integration in the European capital market. For this purpose, the author(s) use the cluster analysis. The paper is well structured and shows very interesting results. The paper has potential for publication and would like the authors to further develop the paper by considering the following comments.
1. Section Introduction. I suggest better explain the contribution and the objective of the research. Due to the relevance of the argument, I would like that the author(s) to deep the motivations in the introduction.
2. Section Data and methodology. I would like that the authors add the Table of data (for example summary statistics). I think is important to better explain the data used.
3. Section Conclusion. The results are very interesting. I recommend adding the policy implication of these findings.
Minor comment:
1. I think that some references are "wrong" (e.g. line 78). If the authors use Latex, I would suggest inserting the citations like this: "\citep{referee1, referee2, ...,}”
2. Line 118, citation missing (see "Acharya and Steffen")
I hope you will find my comments useful.
Author Response
Dear sir or madam,
thank you very much for very interesting and important points. We tried to do our best to improve and change our paper according to your recommendations. I believe that after editing, the article meets your requirements. Thank you for your time and excellent advice, if you have any further comments let us know and we will adjust the article accordingly.
1. Section Introduction. I suggest better explain the contribution and the objective of the research. Due to the relevance of the argument, I would like that the author(s) to deep the motivations in the introduction.
The introduction chapter was extended and new references were added.
2. Section Data and methodology. I would like that the authors add the Table of data (for example summary statistics). I think is important to better explain the data used.
As the data are time series, we did not consider computing summary statistics. Instead of that, we add more graphs and clearer visualizations of results.
3. Section Conclusion. The results are very interesting. I recommend adding the policy implication of these findings.
This is a very good point, thanks a lot for it. These policy recommendations were added to the chapter Conclusion.
Minor comments about citations:
We tried to edit citations according to your advice, is it ok now? Thanks for your precise review.
Reviewer 3 Report
Please see the attached pdf file.
Comments for author File: 
Comments.pdf
Author Response
Dear sir or madam,
thank you very much for very interesting and important points. We tried to do our best to improve and change our paper according to your recommendations. I believe that after editing, the article meets your requirements. Thank you for your time and excellent advice, if you have any further comments let us know and we will adjust the article accordingly.
1. I find the abstract to be a bit too lengthy. Only the last few sentences focus on what the authors actually do. So, I would suggest shortening the abstract, so that it does not sound like an introduction, and instead provide more description of the results and analysis.
According to the general rules, abstracts are typically 100–250 words and comprise one or two paragraphs. Our abstract has 196 words, so we do not feel it to be lengtly. These first sentences are important because it is like a short introduction to the whole topic, more details you can find in other parts of the paper. Sentences are complexly connected with describing our goal so we don´t think that abstract has to be changed. Thanks for your understanding.
2. Similarly, the introduction seems to be a bit lengthy and repetitive at times. For example, the description on lines 41-59 was largely said in the first two paragraphs of the intro. So, the first two paragraphs of the intro, as they do not seem to pertain so much to the analysis, could be shortened or made more to the point.
Chapter Introduction was edited. I hope now it meets your requirements.
3. I do not understand why the authors plot the market cap in China, US, in addition to EU in Figure 1. The study is on individual EU countries, so I feel like plotting the historical dynamics of the individual EU countries’ market cap would provide more information for the reader.
That is very good idea, thanks for giving us another view. From our point of view we wanted to descibe and visualize comparision EU versus China and USA. It could also be interesting to compare individual EU countries’ market cap.
4. I appreciate the detailed description of the clustering methods used, but I think it would be also helpful to provide a more big-picture/intuitive description of how the method clusters data into groups.
Thanks for the advice, we decided to visualize clustering in more details for households, I hope it can provide better intuitive description of this classification.
5. On lines 196-205 the authors list the financial indicators employed in the study. I have two suggestions here. First, I think it would be helpful to be more clear about which financial indicators are exactly used for the individual countries; stating, say, “unlisted shares” does not provide much information. Second, I think it would be also helpful if the authors could provide some summary statistics on the indicators across countries.
As the data are time series, we did not consider computing summary statistics. Instead of that, we add more graphs and clearer visualizations of results. From our point of view, it is not necessary to define terms such as "unlisted shares" for a reader oriented in the field of economics. All the indicators we analyzed were obtained from the Eurostat database and the names of these indicators are the same as in this database. According to the European System of Accounts (ESA2010) unlisted shares are equity securities not listed on an exchange. (see https://ec.europa.eu/eurostat/esa2010/chapter/view/5/ code: F.512)
6. The description of the results in section 4 could be more closely connected to the figures. Stating “Results can be seen in Figure 5” for example, is not really satisfactory. It was not clear to me how to see these results (such as that Norway is a special status based on Figure 4, plus what does that sentence even mean?) in the figure. How does the figure show this?
We have completely redesigned all visualizations and added new ones to make the interpretations and conclusions from the analysis clearer.
Through the Agreement on the European Economic Area, Norway is part of the European single market and also participates in some EU programs, but without voting rights. These are, in particular, the areas of business, environment, research, and education. The Capital Markets Union project can be another step towards successful bilateral cooperation.
Reviewer 4 Report
This is an interesting study and its quality is good, I think. However, I would like to see the following improvements in the manuscript before a making concrete decision:
1, The furthers should emphasize the motivation and the intended contribution of this research further.
2. The authors should avoid using multiple references. To this extent, no more than 3 references in a sentence are appropriate.
3. The authors should provide a footnote that they are willing to share their data set in Excel format with those who wish to replicate the results of this research.
4. The authors should identify the shortcomings of the selected cluster methodology.
5. The graphs are difficult to see, they should be presented in a better way.
6. The authors should provide some guidelines for future studies in this research area,
Author Response
Dear sir or madam,
thank you very much for very interesting and important points. We tried to do our best to improve and change our paper according to your recommendations. I believe that after editing, the article meets your requirements. Thank you for your time and excellent advice, if you have any further comments let us know and we will adjust the article accordingly.
1. The furthers should emphasize the motivation and the intended contribution of this research further.
The introduction was extended and also the conclusion is edited.
2. The authors should avoid using multiple references. To this extent, no more than 3 references in a sentence are appropriate.
Is this requirement mentioned somewhere in the instructions for the article? Standard scientific publications contain multiple references, but if this is a big problem, we can remake it.
3. The authors should provide a footnote that they are willing to share their data set in Excel format with those who wish to replicate the results of this research.
Done, thank you very much for this useful advice.
4. The authors should identify the shortcomings of the selected cluster methodology.
We add policy recommendations to the conclusion chapter and this chapter was extended. Graphs now are clear and easy to read and interpret.
5. The graphs are difficult to see, they should be presented in a better way.
We've completely changed the plots and added some new ones. Graphs now are clear and easy to read and interpret.
6.The authors should provide some guidelines for future studies in this research area.
We mention it in conclusion: "In future analyzes, we plan to eliminate the shortcoming of this static clustering method and, instead of analyzing the values in a specific year, use the approach using dynamic time warping distance and then performing detailed and more dynamic cluster analysis.
Round 2
Reviewer 1 Report
The authors didn't reflected in the text some of the changes, even if they tell that they do it. I will use here the numbering of my original review.
1. In the introduction remains the problem I had identified in some cases. It is true that the authors added one reference, but others are missing.
3. The statement "On the other hand, in Europe, the euro area capital market is larger than the capital market of countries outside the euro area." continues there. As I have noted in my review, This is natural considering the dimension of the Eurozone when compared with the non-Eurozone one.
4. It is incredible that the authors keep the statement about the emergence of the capital market union has an initiative in 2014... The authors introduce a reference in the end of the paragraph which does not match with my concern.
5. The authors disregarded completely my suggestion about the identification of the different ways to assess financial integration. They also referred a new paper, but it is not referenced through the text (althoug it is listed in the reference list).
6. Where is the section about financial integration in the EU?
9. The discussion is still scarce.
10. The answer is not correct. Despite Norway is a partner of EU, it is not a member. The authors identify the use of the UK because it was a member of the EU during the duration of the sample (and it is ok), so it does not justify the use of Norway.
Morevoer, the authors disregard also other basic suggestions of at least two reviewers. Why not to calculate summary statistics justifying it as working with time series?
In my opinion the work was not serious.
Author Response
Dear sir or madam,
thank you very much for your time and comments. We tried to do our best to meet your requirements. I will try to explain this with the example of all the points mentioned in the review. I hope you will be satisfied with our adjustments. If you have any additional questions and comments feel free to contact me.
1. In the introduction remains the problem I had identified in some cases. It is true that the authors added one reference, but others are missing.
We added references to the introduction (see Akhtar and Akyol). Benefits and risks of financial integration were also discussed in a more detailed way (see Agenor and Shmukler: lines 123 - 136).
3. The statement "On the other hand, in Europe, the euro area capital market is larger than the capital market of countries outside the euro area." continues there. As I have noted in my review, This is natural considering the dimension of the Eurozone when compared with the non-Eurozone one.
I don´t understand what is the problem with this sentence. We just wanted to point out that euro area countries have a higher market capitalization than non-euro area European countries.
4. It is incredible that the authors keep the statement about the emergence of the capital market union has an initiative in 2014... The authors introduce a reference in the end of the paragraph which does not match with my concern.
It is probably a misunderstanding, you asked if the authors know the history of the European Union, then you used slightly offensive wording "It is incredible that the authors keep the statement...". We deal with the capital market union, maybe you mean monetary union. We cited the paper called Capital-Markets Union: tentative shape and priorities (see https://www.robert-schuman.eu/en/doc/questions-d-europe/qe-335-en.pdf) where is mentioned: "The idea of market financing was first put forward by Jean-Claude Juncker in July 2014 in order to engage financial and political players with this issue...". If you still feel that this sentence is not properly worded, please write what you do not agree with.
5. The authors disregarded completely my suggestion about the identification of the different ways to assess financial integration. They also referred a new paper, but it is not referenced through the text (althoug it is listed in the reference list).
Yes, you are right, we did not incorporate this suggestion into our paper according to your previous review. But it is appropriate to add it to our paper, so we add there reference you recommended, thank you for this advice. (lines 261 - 264)
6. Where is the section about financial integration in the EU?
Please see lines 61 - 122 (Knudsen)
9. The discussion is still scarce.
We have extended the discussion by political implications and options for further analysis. (see lines 447- 457)
10. The answer is not correct. Despite Norway is a partner of EU, it is not a member. The authors identify the use of the UK because it was a member of the EU during the duration of the sample (and it is ok), so it does not justify the use of Norway.
We use Norway we used because as we mentioned in the paper, despite the fact that Norway is not a member of EU, it is a part of the European single market and we believe, that the capital markets union project can be another step towards successful bilateral cooperation. (see lines 301 - 308)
Morevoer, the authors disregard also other basic suggestions of at least two reviewers. Why not to calculate summary statistics justifying it as working with time series?
As we stated in the responses to the other reviews, it is not a good idea from a statistical point of view to calculate the arithmetic mean or standard deviation of separate time series. But it is a very good point and we consider computing the arithmetic mean for all countries in each year for all indicators. Thank you very much, it makes sense and it is good for comparison. (see Tables 1 - 4, page 8 of 20)
Reviewer 4 Report
The authors responded to all my recommendations apart from one. I would strongly suggest that they should avoid multiple references. To this end, no more than three references in a sentence are appropriate.
Author Response
Dear sir or madam,
thank you very much for your recommendations and positive review. There was just one sentence with more than three references, so we fix this problem. Thanks again, we appreciate it.
Round 3
Reviewer 1 Report
Dear authors,
Firstly I would like to recognize that over the different phases you made some improvements on the paper. Note that when the reviewers make some remarks, they are not necessarily offensive, but it is necessary to ensure that one paper works as a whole piece of work.
Regarding this second review round, you "cleaned" some of my concerns but I still have some comments which I believe that you could easily solve.
3. Your answer was "We just wanted to point out that euro area countries have a higher market capitalization than non-euro area European countries". this is true but it is meaningless because of the dimension of the euro area regarding to the non-euro area. It was the same if you identified that the market capitalization of the EU is higher than the one of the Euro area. Reconsider to eliminate just that statement.
4. I understood now what you want to say, although it is not clear in the paper. You refer to the CMU as the the program of the EC. But the way as it is written at least for me seemed to be confusing. Note that you have written, in lines 41 to 43 "The idea of Capital Market Union (CMU) has emerged as an initiative in 2014 with a view to create a stable economic environment and contributing to a higher degree of integration of the member countries of the European Union". But the capital market union comes from the single market. Please, make clear that it was the political/economic programa of the EC.
About the descriptive statistics, you refer that "As we stated in the responses to the other reviews, it is not a good idea from a statistical point of view to calculate the arithmetic mean or standard deviation of separate time series". I don't understand you is not good from a statistical point of view to identify descriptive statistics. In other words, it could make sense or not depending on the conditions of the time series themselves. But that is now ok for me.
Author Response
Dear sir or madam,
thanks a lot for the quick response. We appreciate your effort and comments and recommendations, all of them helped to reach better quality and clarity of our paper.
3. Your answer was "We just wanted to point out that euro area countries have a higher market capitalization than non-euro area European countries". this is true but it is meaningless because of the dimension of the euro area regarding to the non-euro area. It was the same if you identified that the market capitalization of the EU is higher than the one of the Euro area. Reconsider to eliminate just that statement.
The problematic sentence "On the other hand, in Europe, the euro area capital market is larger than the capital market of countries outside the euro area." was deleted.
4. I understood now what you want to say, although it is not clear in the paper. You refer to the CMU as the the program of the EC. But the way as it is written at least for me seemed to be confusing. Note that you have written, in lines 41 to 43 "The idea of Capital Market Union (CMU) has emerged as an initiative in 2014 with a view to create a stable economic environment and contributing to a higher degree of integration of the member countries of the European Union". But the capital market union comes from the single market. Please, make clear that it was the political/economic programa of the EC.
We changed the sentence:
"The idea of Capital Market Union (CMU) has emerged as an initiative in 2014 with a view to create a stable economic environment and contributing to a higher degree of integration of the member countries of the European Union."
to sentence:
"The idea of Capital Market Union (CMU) has emerged as the political-economic program of the EC in 2014 with a view to creating a stable economic environment and contributing to a higher degree of integration of the member countries of the European Union."
About the descriptive statistics, you refer that "As we stated in the responses to the other reviews, it is not a good idea from a statistical point of view to calculate the arithmetic mean or standard deviation of separate time series". I don't understand you is not good from a statistical point of view to identify descriptive statistics. In other words, it could make sense or not depending on the conditions of the time series themselves. But that is now ok for me.
Maybe I didn't make myself clear enough. I had in mind that it is not appropriate to compare the dynamics of the development of time series using averages or standard deviations and it is necessary to use the characteristics of the dynamics of time series such as the average absolute growth rate or even better would use the average growth coefficient for comparing time series expressed in different units.
