On Prices of Privacy Coins and Bitcoin

Round 1

Reviewer 1 Report

EVALUATION

The authors compare privacy coins to Bitcoin from their price viewpoint, mainly from visual and descriptive data inspection. They argue that Bitcoin is potentially leading privacy coins from a price discovery point of view, and that correlation is typically high and positive. They additionally derive some conclusions based on such visual inspective analyses.

The paper focuses on an interesting topic. Writing can be improved. The literature review can be improved. The methodological part is very weak.  Overall, I found the paper could be very interesting, and I believe it would also be much appreciated to readers of your journal if some more rigorous literature and statistical analyses are included. I would therefore suggest to consider the paper for publication, conditional on the authors’ willingness to address the point-by-point comments below in the course of their revision.

• With regards to the literature related to the paper, the authors propose a poor literature review. I suggest them to integrate their literature review in the paper with the basics of Bitcoin economics and finance, and with other methodologies which have more recently been developed to measure price and volatility spillovers across cryptocurrencies, price discovery, dominance and market interconnectedness : the relatively new methodological framework by Diebold et al. (2012), and relative applications on different market spillover and dominance domains through VAR models or VECM models (see references 1,2,3,4,5,6,7). This is closely related literature to capture interconnectedness across cryptocurrency prices.
• The authors could better stress and further expand that the results obtained by their methodology could be the starting point for building up other applications. These measures should not only be beneficial to describe what has happened in the past or recent past, but they could also be useful in the construction of several investment management purposes. Indeed, I suggest the authors integrate their paper part on implications with future developments on the use of spillovers, dominance and networks for investment management through the application of techniques such as those of references from 8 to the end, which turn correlations, connectedness and spillover indexes into portfolio optimization and financial management algorithms taking into account for various risks sources.
•  
• "Some simple bitcoin economics." Journal of Monetary Economics 106 (2019): 16-26.
•  
• The empirical analysis of the paper is poor. There is basically no methodology to verify the authors’ claims. Already in the abstract, the authors say: “It seems that Bitcoin is leading privacy coins in their pricing, 7 and correlation is typically high and positive.” This statements should be empirically supported even by relatively easy statistical analyses. The authors talk about correlations, which are however not shown. The authors talk about Bitcoin leading privacy coins’ prices, but they do not verify it empirically. I suggest the authors to do two simple analyses:
• A) correlation (and partial correlation) analysis. You compute the correlation (and partial corr.) matrix among cryptocurrencies and present a table or a figure as an output, with their standard errors or p-values of the associated test for significant correlations and comment.
• B) causality analysis. You can use either very simple pairwise granger causality tests, or techniques such as information shares or spillovers, so to determine the lead-lag relationships among your set of cryptocurrencies. In this way, you will be able to say if there is a statistically dominant role of Bitcoin on privacy coins, and on which of them. You can also see whether there are some dominant privacy coins over the others.
• These two simple analyses are fundamental, can then constitute your methodology, which can be written in the paper in mathematical terms, and then its outputs can be commented in the results part, along with the already existing plots. The two analyses can be carried out through R, STATA, matlab or any other statistical software in a very easy way. Of course the authors can address the problem with other additional approaches, these two analyses, in my view, are just an example of minimal analysis for the paper to be accepted in JRFM.
• It would be interesting to add a plot of the volume of crypto against privacy coin exchanges over time. I know there is a table, but this is a static measure. Volumes should be presented dynamically over time, and, along with the analyses suggested before, the authors could find a relationship between the price dominant cryptocurrencies and their trade volumes, as researchers often seek in the financial literature on cryptocurrencies.
• The English of the paper should be revised entirely

To sum up, the paper is interesting and the topic is original. However, there are two major flaws: a weak literature review, and an as much weak empirical analysis. Therefore, I need to recommend a major revision of the paper. If the reviewers will follow the proposed comments in a good way, I would be very happy to reconsider my choice about the manuscript.

Author Response

The authors compare privacy coins to Bitcoin from their price viewpoint, mainly from visual and descriptive data inspection. They argue that Bitcoin is potentially leading privacy coins from a price discovery point of view, and that correlation is typically high and positive. They additionally derive some conclusions based on such visual inspective analyses.

The paper focuses on an interesting topic. Writing can be improved. The literature review can be improved. The methodological part is very weak.  Overall, I found the paper could be very interesting, and I believe it would also be much appreciated to readers of your journal if some more rigorous literature and statistical analyses are included. I would therefore suggest to consider the paper for publication, conditional on the authors’ willingness to address the point-by-point comments below in the course of their revision.

With regards to the literature related to the paper, the authors propose a poor literature review. I suggest them to integrate their literature review in the paper with the basics of Bitcoin economics and finance, and with other methodologies which have more recently been developed to measure price and volatility spillovers across cryptocurrencies, price discovery, dominance and market interconnectedness : the relatively new methodological framework by Diebold et al. (2012), and relative applications on different market spillover and dominance domains through VAR models or VECM models (see references 1,2,3,4,5,6,7). This is closely related literature to capture interconnectedness across cryptocurrency prices.

The authors could better stress and further expand that the results obtained by their methodology could be the starting point for building up other applications. These measures should not only be beneficial to describe what has happened in the past or recent past, but they could also be useful in the construction of several investment management purposes. Indeed, I suggest the authors integrate their paper part on implications with future developments on the use of spillovers, dominance and networks for investment management through the application of techniques such as those of references from 8 to the end, which turn correlations, connectedness and spillover indexes into portfolio optimization and financial management algorithms taking into account for various risks sources.

References

"Some simple bitcoin economics." Journal of Monetary Economics 106 (2019): 16-26.

"The inefficiency of Bitcoin." Economics Letters148 (2016): 80-82.

"Price discovery on Bitcoin markets." Digital Finance 1.1 (2019): 139-161.

”High frequency price change spillovers in bitcoin markets." Risks4 (2019): 111.

"Bitcoin: Economics, technology, and governance." Journal of economic Perspectives2 (2015): 213-38.

"Vector error correction models to measure connectedness of Bitcoin exchange markets." Applied Stochastic Models in Business and Industry1 (2020): 95-109.

"Libra or Librae? Basket based stablecoins to mitigate foreign exchange volatility spillovers." Finance Research Letters(2021): 102054.

"Neural network models for Bitcoin option pricing." Frontiers in Artificial Intelligence2 (2019): 5.

"Technical Analysis on the Bitcoin Market: Trading Opportunities or Investors’ Pitfall?." Risks 8.2 (2020): 44.

“Network Models to Enhance Automated Cryptocurrency Portfolio Management." Frontiers Artif. Intell. 3 (2020): 22.

Author’s response:

Shall give sincere thanks for these remarks, and you are right that literature review and contribution in that regard needed improvement for this proposed work. Have included all mentioned new references, together with two other references, to revised version of this article. Changes have taken place mostly in Sections 1, 2 and 3. You may find changes and additions of manuscript as these are yellow coloured. Changes do not only concern literature, but also research environment and methodology section.

Manuscript is also proof read for the second round, and numerous small changes have been implemented all over manuscript.

The empirical analysis of the paper is poor. There is basically no methodology to verify the authors’ claims. Already in the abstract, the authors say: “It seems that Bitcoin is leading privacy coins in their pricing, 7 and correlation is typically high and positive.” This statements should be empirically supported even by relatively easy statistical analyses. The authors talk about correlations, which are however not shown. The authors talk about Bitcoin leading privacy coins’ prices, but they do not verify it empirically. I suggest the authors to do two simple analyses:

1. A) correlation (and partial correlation) analysis. You compute the correlation (and partial corr.) matrix among cryptocurrencies and present a table or a figure as an output, with their standard errors or p-values of the associated test for significant correlations and comment.
2. B) causality analysis. You can use either very simple pairwise granger causality tests, or techniques such as information shares or spillovers, so to determine the lead-lag relationships among your set of cryptocurrencies. In this way, you will be able to say if there is a statistically dominant role of Bitcoin on privacy coins, and on which of them. You can also see whether there are some dominant privacy coins over the others.

These two simple analyses are fundamental, can then constitute your methodology, which can be written in the paper in mathematical terms, and then its outputs can be commented in the results part, along with the already existing plots. The two analyses can be carried out through R, STATA, matlab or any other statistical software in a very easy way. Of course the authors can address the problem with other additional approaches, these two analyses, in my view, are just an example of minimal analysis for the paper to be accepted in JRFM.

It would be interesting to add a plot of the volume of crypto against privacy coin exchanges over time. I know there is a table, but this is a static measure. Volumes should be presented dynamically over time, and, along with the analyses suggested before, the authors could find a relationship between the price dominant cryptocurrencies and their trade volumes, as researchers often seek in the financial literature on cryptocurrencies.

The English of the paper should be revised entirely

To sum up, the paper is interesting and the topic is original. However, there are two major flaws: a weak literature review, and an as much weak empirical analysis. Therefore, I need to recommend a major revision of the paper. If the reviewers will follow the proposed comments in a good way, I would be very happy to reconsider my choice about the manuscript.

Author’s response:

Reviewer is correct that initial version was missing correlation tables. These are now included in empirical part, and they are tied to text with modified sentences, but as well with completely new sentences. Think that these changes will make empirical analysis stronger.

Reviewer 2 Report

The article covers an interesting topic of private coins price changes in respect to bitcoin.

However the authors just visually analyses the price development of cryptocurrencies. In the article they report correlation results, but these are not numerically shown anywhere. Such analysis based only on visual observations and percentage changes is insufficient. To make the presented results relevant I would recommend to the authors to calculate at least cross-correlation coefficient in rolling window.

There is also extensive study on the cyrptocurrency market and cross-correlations between various cryptocurrencies - see "Multiscale characteristics of the emerging global cryptocurrency market", Physics Reports 901 (2021) https://0-doi-org.brum.beds.ac.uk/10.1016/j.physrep.2020.10.005 - which authors omit.

Minor issues: 
line 190 "Civic coin in early 2000"?
There are no axis labels in Figs 1-4
What is the data resolution - daily?

Author Response

The article covers an interesting topic of private coins price changes in respect to bitcoin.

However the authors just visually analyses the price development of cryptocurrencies. In the article they report correlation results, but these are not numerically shown anywhere. Such analysis based only on visual observations and percentage changes is insufficient. To make the presented results relevant I would recommend to the authors to calculate at least cross-correlation coefficient in rolling window.

Author’s response:

Reviewer is correct that initial version was missing correlation tables. These are now included in empirical part, and they are tied to text with modified sentences, but as well with completely new sentences. Think that these changes will make empirical analysis stronger.

There is also extensive study on the cyrptocurrency market and cross-correlations between various cryptocurrencies - see "Multiscale characteristics of the emerging global cryptocurrency market", Physics Reports 901 (2021) https://0-doi-org.brum.beds.ac.uk/10.1016/j.physrep.2020.10.005 - which authors omit.

Author’s response:

Thank you for this valuable addition for literature – have included this article to improved version of the manuscript, and it is properly cited in the main text.

Minor issues: 
line 190 "Civic coin in early 2000"?
There are no axis labels in Figs 1-4
What is the data resolution - daily?

Author’s response:

Have corrected Civic’s year to 2020. Axis labels of Figures 1-4 are now mentioned in figure texts (US dollars). Frequency of data used is also mentioned in Section 3, and it is daily closing price of each crypto.

Round 2

Reviewer 1 Report

Two major flaws of the paper still persist:

• The English
• The lack of a methodological part to support conclusions

For this reason, I suggest 2 very simple solutions, for the paper to reach a higher quality standard and be definitely accepted:

• Get the English revised (by a mother-tongue, or by some paper revising services)
• Show simple rolling correlations plots

Some notes on both the things:

• For the English, I am afraid to say that hasn’t improved and is not sufficient for publication, so getting it revised would be definitely good for publishing
• For the analysis, I see you use excel. Even with a simple excel analysis, you can show the rolling regression coefficients of some of your coins (say, the top 5 in terms of market capitalization) and that of Bitcoin. How you do that? You calculate correlations among prices, but only taking 100 observation. Then you roll the window and each day you have a new value of the correlation, which represent the “daily rolling correlation” with a 100-day window. By doing so, you will be able to evaluate the dynamic relationship between the prices of some major coins and Bitcoin, and not only in a static way as the correlation matrix you have put.

This latter analysis would give you at least a basis for the paper, which is not really there at the moment as it is only a descriptive analysis. The partial correlation you skipped in the previous review report would have helped with this.

I acknowledge the paper has also improved in several aspects. For example the literature part has considerably improved, and now the framework in which the authors operate is clearer than before.

When these two major points will be definitely recovered, the paper could be considered for acceptance.

Best regards,

MINOR POINTS

Research environment and methodology -> change the title, you basically do not have a methodology, so you can call it “related literature and analysis”

Author Response

Following changes have been completed to the third-round revised version of this manuscript proposal (major changes highlighted with yellow within main text):

1. Manuscript has been proof-read.
2. Title and associated text of Section 3 has been changed as proposed by reviewer.
3. Rolling correlations of 100 days have been added to empirical data analysis. This has led to additional paragraph in the manuscript.

Reviewer 2 Report

The authors satisfied all my concerns.

Author Response

Thank you for your time and input regarding the further development of this article. All of your earlier review points were valid, and improved manuscript significantly.

Round 3

Reviewer 1 Report

The authors have taken into account most of the comments