Round 1

Reviewer 1 Report (New Reviewer)

This paper examines the relationship between the exchange rates and the oil price/OVX in five sub-Saharan countries. Using the quantile response model and Markov exchange model, the author(s) finds an asymmetric response of exchange rates on oil price/OVX: falling prices and rising OVXs are likely to generate more significant effects. The paper generally reads good, yet, I have the following concerns and suggestions, arranged from the most important to the least.

1. What is the big picture? After knowing the results that the author(s) found, how can it change the world better?

2. The particularity of countries. Why does the author(s) choose sub-Saharan countries? Data availability is understandable, but I wondered whether the author(s) have a non-country-specific result that differs from the research about other countries. In other words, does the author(s) have an aggregate effect on these five countries? If yes, what are these countries' differences compared to other parts of the world? These questions are essential because they may reveal why we have observed the correlations between exchange rates and price/OVX.

3. Robustness. Are the results sensitive to model selection? I appreciate the Markov exchange model, but the author may add a more robustness check (which is supposed to be easy). For example, changing the quantile bandwidth, adjusting controls, etc.

4. Control variables. I realize that in most regressions, there are very few control variables. The author(s) may consider adding some related variables as controls.

5. Causal relation. Is it possible to generalize the current shape of the asymmetric correlation-like model to a causal model?

6. Writing. There are several typos in the paper, especially missing the articles (a/the) and having too many hard-to-read sentences. I wondered if any author(s) are native English speakers. If not, I suggest proofreading by a native one. For the author(s) convenience, here is a proofread example of the abstract.

The theory demonstrates that oil price and oil volatility (OVX) are significant determinants of economic activity; however, studies seldom consider both variables in the oil-exchange rate nexus and ignore the distributional heterogeneity of the exchange rate. We investigate their joint effect and employ quantile regression and Markov switching models to address this. We differentiate between positive/negative shocks and control for the effect of the global financial crisis in 2008 and the Covid-19 Pandemic in 2020. We observe that OVX shocks significantly impact the exchange rate for all countries, whereas oil price shocks only affect the exchange rate of oil-importing countries. Rising (falling) OVX causes the local currency to depreciate (appreciate). The impact of rising or falling OVX is the same for oil-importing and oil-exporting countries, whereas the impact of rising and falling oil price varies. The impact of oil price and OVX on the exchange rate is affected by market conditions. The exchange rate responds to oil price and OVX mostly at lower quantiles (bearish markets) for all countries, which reveals investors' sensitivity. In contrast, a weak to no significant response is observed at the higher quantiles (bullish market). Our results are robust in model selection (Markov switching models).

7. Table 6 (along with other similar tables) is somehow hard to read. It may be better to replace betas with the variable name. Also, I would suggest putting some simplified results in the main part but the full result in the appendix.

8. Page 24, Line 691, the subsection is "4.3.1.2"; should it be "4.3.2.2"? I realize the numbering is not continuous, and it is confusing.

9. The notes for Table 2 still show "table 1."

Author Response

REVIEWER 1, economies, #1915072

The impact of oil price and oil volatility index (OVX) on the exchange rate in Sub-Saharan Africa: Evidence from oil importing/exporting countries

Answers to reviewers are highlighted in yellow. Second round changes in the actual text/submission are also highlighted in yellow. Older changes in the actual text/submission are highlighted in different colors.

Comments and Suggestions for Authors

This paper examines the relationship between the exchange rates and the oil price/OVX in five sub-Saharan countries. Using the quantile response model and Markov exchange model, the author(s) finds an asymmetric response of exchange rates on oil price/OVX: falling prices and rising OVXs are likely to generate more significant effects. The paper generally reads good, yet, I have the following concerns and suggestions, arranged from the most important to the least.

1. What is the big picture? After knowing the results that the author(s) found, how can it change the world better?

AUTHORS’ REPLY:

After knowing the results, we have indicated what the implications are to the countries concerned in section 5.1. we have also renamed this section: The ‘big picture’ in the exchange rate-oil nexus and implications for the countries concerned.

For example, the countries sampled rely heavily on the energy sector and thus a major policy concern is the impact the changes in the oil markets can have on the exchange rate. This is also because fluctuations in the exchange rate can affect the growth of their economy. Our results also show that OVX is the dominant factor in the oil-exchange rate nexus.

We have also updated this section by commenting on the current European energy crisis and the effect on exchange rates. See below (with current references from the press):

To abstract from the African countries under examination and capture the big picture we offer a brief account of the current situation in Europe and the ongoing energy crisis (oil, natural gas, electricity).  The findings of this study are not relevant only for the countries under examination here. Specifically, the euro and the GBP have dropped in value in relation to USD (the lowest in nearly 20 years), inflation is increasing at an alarming rate and industrial production is reducing. If energy prices (including oil price and oil price volatility) have such a pronounced effect on currencies that are used for international transactions/payments, then the situation will definitively be gloomier for the currencies examined here. A major lesson to be learned here is that policy makers must take action to dampen the effects of energy prices on currencies whether those are used for international transactions or are less well known (as it is the case with the currencies under examination here)

(https://www.euronews.com/my-europe/2022/08/23/europes-energy-crisis-haunts-the-euro-as-it-tumbles-to-20-year-low-against-the-dollar)

https://markets.businessinsider.com/news/currencies/dollar-dominance-euro-energy-crisis-parity-currencies-investing-analysis-citi-2022-8

2. The particularity of countries. Why does the author(s) choose sub-Saharan countries? Data availability is understandable, but I wondered whether the author(s) have a non-country-specific result that differs from the research about other countries. In other words, does the author(s) have an aggregate effect on these five countries? If yes, what are these countries' differences compared to other parts of the world? These questions are essential because they may reveal why we have observed the correlations between exchange rates and price/OVX.

AUTHORS’ REPLY:

The reasons for choosing these countries have been mentioned in section 1.0 paragraph 3 and 4)

I believe that the aggregate effect you are referring to in this study is the effect of the OVX on the exchange rates of all countries (both importers and exporters). A similar study (Breen and Hu, 2021) even though with a much smaller sample (Norway, Australia and Canada), shows that the OVX is a good predictor of future exchange rates. One can conclude that OVX affects exchange rates for both developing and developed countries regardless of importer/exporter status. Given the absence of studies for developed countries and developing (excluding this one), one cannot really draw any conclusions regarding the uniqueness (and any special characteristics) of our sample.in other words we do not think that there any characteristics in our sample/countries that make our chosen countries different form the rest of the world except the fact that they are developing countries. Effectively more work needs to be done to reach ‘critical mass’ so as to draw any conclusions about the effects of OVX on developed/developing countries and export/import status given the tiny sample of developed countries in Breen and Hu (2021). This could be another study altogether. The paragraph we added (direct quotation) in our paper follows below for your convenience:

‘Second, although previous studies demonstrate the role of oil shocks on exchange rate, this study to the best of our knowledge, is the first empirical one to document the joint impact of both oil price changes and OVX changes on exchange rate for developing countries (both importers and exporters). Breen and Hu, (2021) look at the effect of oil price and OVX on the exchange rates of developed countries: Canada, Norway and Australia.  OVX appears to be a good predictor of exchange rates for Canada, Norway and Australia regardless of importer/exporter status. Therefore, we contribute to the literature providing new insights on the oil-exchange rate nexus from the perspective of the OVX (and price) for developing countries employing a bigger sample that could possibly allow us to offer generalized results for developing countries’.

3. Robustness. Are the results sensitive to model selection? I appreciate the Markov exchange model, but the author may add a more robustness check (which is supposed to be easy). For example, changing the quantile bandwidth, adjusting controls, etc.

AUTHORS’ REPLY:

We are not exactly sure what you mean by changing the quantile bandwidth. Currently we use 9 deciles (). We then divide them equally into 3 groups, namely lower quantile ((, median quantile (and higher quantile. Do you suggest using 5 ().? If we do this, then we would be able to look only at the lowest 20%, middle 40%-60% and highest 20%. In this case, the results would be more dramatic/accentuated especially at the extremes, the lowest 20%, and the highest 20%. Perhaps you mean something else. we are not exactly sure. now as far ‘controls’ is concerned, this question is answered below.

4. Control variables. I realize that in most regressions, there are very few control variables. The author(s) may consider adding some related variables as controls.

AUTHORS’ REPLY:

Thank you for the suggestion, we would have loved to add more variables as mentioned in Subsection 3.1 paragraph 3 but most of the economic variables are of lower frequency with a lot of missing data. To be more specific, the frequency of our current data is daily and this is how we are able to capture the effects of volatility and price changes on the exchange rate. Macro data is always available at a quarterly basis. For example, industrial production or money supply are available quarterly. If we include those variables and change the frequency, then we cannot use oil prices or OVX which is the point of the whole article. Volatility occurs at short term clusters. In a quarterly basis you will not be able to observe any volatility which negates the core concept of this article. The only variable that we could use potentially is the central bank base rate for each of the counties under investigation but again the central bank base rate changes only every few months if at all so econometrically speaking the base rate would be a ‘constant value’ adding nothing to the regression since it would be repeating itself. From an econometrics point of view a ‘non-varying’ variable would introduce a number of unwanted issues. There is no such thing as LIBOR available for the countries under examination. We subscribe to both Bloomberg and DataStream and all research to identify such variables was futile. We also checked at the websites of the central banks. Overall, the reason we did not include any control variables is not because we did not want to but a combination of non-availability and econometric ‘unsuitability’

5. Causal relation. Is it possible to generalize the current shape of the asymmetric correlation-like model to a causal model?

AUTHORS’ REPLY:

Even though we are not entirely sure what you mean, we believe that you are suggesting running a Granger causality test. We could do that for each country separately (pairwise Granger causality) or for all countries simultaneously (Dumitrescu-Hurlin approach) however this would be between i) foreign exchange rate and oil price and ii) foreign exchange rate and OVX. Standard pairwise Granger causality tests and the Dumitrescu-Hurlin approach allow only for the inclusion of a single explanatory variable, not two. We do not believe that this would anything to the whole paper since the whole idea is to test/understand how the two variables (oil price and OVX) interact together rather than separately. In addition, for Granger causality tests and in particular for the Dumitrescu-Hurlin approach the variables must be of the same order I(0) or I(1). We cannot mix variables of different orders. Unfortunately, this is not the case here. Even if this was possible it is unlikely that there would be any causality running from the exchange rate (exchange rates of SSA countries) to OVX. Perhaps the opposite could be the case. See references below:

Galariotis, E & Giouvris, E, International Review Of Financial Analysis (2015) On the stock market liquidity and the business cycle: a multi country approach, 38, 44-69, http://dx.doi.org/10.1016/j.irfa.2015.01.009,

Said, H. & Giouvris, E. Financial Markets and Portfolio Management. (2019) Oil, the Baltic Dry index, market (il)liquidity and business cycles: evidence from net oil-exporting/oil-importing countries.  33, 349–416. https://doi.org/10.1007/s11408-019-00337-0   

6. Writing. There are several typos in the paper, especially missing the articles (a/the) and having too many hard-to-read sentences. I wondered if any author(s) are native English speakers. If not, I suggest proofreading by a native one. For the author(s) convenience, here is a proofread example of the abstract.

AUTHORS’ REPLY:

Thanks for that, we have incorporated this in the submission

7. Table 6 (along with other similar tables) is somehow hard to read. It may be better to replace betas with the variable name. Also, I would suggest putting some simplified results in the main part but the full result in the appendix.

AUTHORS’ REPLY:

Done and thank you for the suggestion 

8. Page 24, Line 691, the subsection is "4.3.1.2"; should it be "4.3.2.2"? I realize the numbering is not continuous, and it is confusing.

AUTHORS’ REPLY:

numbers changed to read 4.3.2.2. thank you

9. The notes for Table 2 still show "table 1."

AUTHORS’ REPLY:

changed to table 2. Thank you

REVIEWER 2, economies, #1915072

The impact of oil price and oil volatility index (OVX) on the exchange rate in Sub-Saharan Africa: Evidence from oil importing/exporting countries

Answers to reviewers are highlighted in yellow. Second round changes in the actual text/submission are also highlighted in yellow. Older changes in the actual text/submission are highlighted in different colors.

Comments

Economies-1915072: The impact of oil price and oil volatility index (OVX) on the exchange rate in Sub-Saharan Africa: Evidence from oil importing/exporting countries

This paper demonstrates that oil price and oil volatility (OVX) are significant determinants of economic activity however. Several studies consider both variables in the oil-exchange rate nexus and ignore the distributional heterogeneity of the exchange rate. To address this, we investigate their joint effect and employ both the quantile regression and Markov switching models. We observe that OVX shocks have a major impact on exchange rate for all countries whereas, oil price shocks only affect the exchange rate of oil-importing countries. Rising (falling) OVX causes the local currency to depreciate (appreciate). The impact of rising or falling OVX is the same for oil-importing and oil-exporting countries whereas the impact of rising and falling oil price varies. The impact of oil price and OVX on exchange rate is affected by market conditions in that the exchange rate responds to oil price and OVX mostly at lower quantiles (bearish markets) for all countries, which reveals the sensitivity of investors, whereas a weak to no significant response is observed at the higher quantiles (bullish market). Our results are robust in model selection (Markov switching models).his study identify a perspective for tackling the issues of human capital investment and youth employment in Sub-Saharan Africa (SSA) with data spanning 1995-2017, 23 macro panel series from 40 SSA countries, making a total of 920 macro panel observations.

1. The abstract is very extensive and should be reduced. In the abstract, only the most important points of the study should be presented. I advise the authors to be more precise with the paper's subject.

AUTHORS’ REPLY:

The last 3 lines are not part of the abstract above, so I am not sure how this ended up in this review. Obviously, there is a mix up. The abstract submitted and included below is 200 words:

The Theory demonstrates that oil price and oil volatility (OVX) are significant determinants of economic activity; however, studies seldom consider both variables in the oil-exchange rate nexus and ignore the distributional heterogeneity of the exchange rate. We investigate their joint effect and employ both the quantile regression and Markov switching models to address this. We differentiate between positive/negative shocks and control for the effect of the global financial crisis in 2008 and the Covid-19 Pandemic in 2020.  We observe that OVX shocks significantly impact the exchange rate for all countries whereas, oil price shocks only affect the exchange rate of oil-importing countries. Rising (falling) OVX causes the local currency to depreciate (appreciate). The impact of rising or falling OVX is the same for oil-importing and oil-exporting countries whereas the impact of rising and falling oil price varies. The impact of oil price and OVX on exchange rate is affected by market conditions. The exchange rate responds to oil price and OVX mostly at lower quantiles (bearish markets) for all countries, which reveals investors sensitivity. In contrast, a weak to no significant response is observed at the higher quantiles (bullish market). Our results are robust in model selection (Markov switching models).

2. The manuscript must be streamlined; in the present state, it is confusing for the reader.

AUTHORS’ REPLY:

We have reduced a few sections, but it is not clear what you are suggesting. would streamlining involve removing the OLS regressions from the text and move those to the appendix? The reason we are suggesting removing the OLS regressions is because we want to show that using OLS in this particular study will not produce reliable results because of the changing nature of the distribution. Effectively the OLS regression is used to show that it is not appropriate and as a consequence we use QR and Markov switching models.

3. The authors should be better motivated. What is the contribution of the paper to the international macroeconomics?

AUTHORS’ REPLY:

The paper contributes to literature in the following ways: First, we employed the OVX as measure of uncertainty instead of the VIX which has been highlighted in section 1.0, paragraph 7.  We are not aware of any other studies that use both oil price and oil price volatility to explain exchange rates for developing countries. Second, most empirical studies, particularly in developing countries rely on linear models and do not take into account how the effects of oil price on exchange rate may vary throughout the exchange rate returns distribution which has been highlighted in section 1.0, paragraph 8 and section 2.2.4. Last but not least, the common trend in the literature is to examine effect of oil prices on exchange rate in developed countries. this study examines the effect of oil prices and oil price uncertainty on the exchange rate in developing countries as highlighted in section 1.0, paragraph 3 and 4.

The most important findings are:

1 We observe that OVX shocks significantly impact the exchange rate for all countries indistinguishably (both oil importers and exporters)

2 Oil price shocks only affect the exchange rate of oil-importing countries.

3 The impact of oil price and OVX on exchange rate is affected by market conditions. The exchange rate responds to oil price and OVX mostly at lower quantiles (bearish markets) for all countries, which reveals investors sensitivity.

We have also connected the current energy crisis in Europe to the effect that energy prices (oil, natural gas, electricity) currently have on international exchange rates, namely the EURO and the GBP against USD. See section 5.1. For your convenience we have also copied and pasted the last paragraph of section 5.1:

To abstract from the African countries under examination and capture the big picture we offer a brief account of the current situation in Europe and the ongoing energy crisis (oil, natural gas, electricity).  The findings of this study are not relevant only for the countries under examination here. Specifically, the euro and the GBP have dropped in value in relation to USD (the lowest in nearly 20 years), inflation is increasing at an alarming rate and industrial production is reducing. If energy prices (including oil price and oil price volatility) have such a pronounced effect on currencies that are used for international transactions/payments, then the situation will definitively be gloomier for the currencies examined here. A major lesson to be learned here is that policy makers must take action to dampen the effects of energy prices on currencies whether those are used for international transactions or are less well known (as it is the case with the currencies under examination here)

4. Which variables affect the exchange rate (Fraj et al., 2018; Conrad and Jagessar, 2018; Ebenezer et al., 2022; Konstantakopoulou, 2016)? Why is it interesting to investigate?

AUTHORS’ REPLY:

We have included the articles that you have suggested and indicated that those articles are mainly looking at the effect that exchanges rates can have on growth from a number of different perspectives (e.g governance, oil price, etc). We have included those articles at the very beginning of our introduction section in order to indicate clearly how those articles are ‘positioned’ in the literature and how this specific article is ‘positioned’ in the literature. This article is more specific in the sense that it looks into oil price shocks and specifically oil price volatility and their joint effect on exchange rates. It is interesting to investigate because currently we are observing the effect that energy prices in general have on international currencies. For example, the recent energy crisis in Europe had led to a decline in Euro and GBP in relation to USD. Of course, the price of oil is not the main culprit in this case (natural gas is) but generally speaking energy prices can have a serious effect as indicated here. the price of oil has been blamed in the past for major swings in exchange rates

In addition, this study provides an insight into the workings on the oil-exchange rate nexus for the countries considered here and allows policy makers to develop effective strategies to reduce oil price risk. this has been highlighted in section 1.0, paragraph 2 ,3 and 4. The policy implications are also highlighted in Paragraph 5.1

Also, why is it essential to examine the impact of oil price and oil volatility index on the exchange rate?

AUTHORS’ REPLY:

It has been highlighted in section 1.0, paragraph 3, 5 and 6.

5. The presentation of the paper is not clear. Generally, there are interesting points in the article which are not highlighted. The authors should be more explicit in writing the paper.

AUTHORS’ REPLY:

We believe that the most interesting point is the effect that OVX has on both oil importers and oil exporters especially in bearish markets. We have tried to connect this finding to the current energy crisis in Europe. We are highlighting this connection in section 5.1.

6. The following sentence is not correct; make the relevant correction. “Then, the bound testing approach (Pesearan et al., 2001) is used to investigate the degree of integration between the variables.”

AUTHORS’ REPLY:

Sentence corrected. Thank you

7. The empirical section follows the correct econometric methodology but is tedious to read.

AUTHORS’ REPLY:

Thank you for this comment. We have deleted a few sentences to simplify this section

References

Fraj, S.H., Hamdaoui, M., Maktouf, S., 2018. Governance and economic growth: The role of the exchange rate regime. International Economics, 156, 326-364.

Conrad, D., Jagessar, J., 2018. Real Exchange Rate Misalignment and Economic Growth: The Case of Trinidad and Tobago. Economies 6, 52. https://doi.org/10.3390/economies6040052.

Ebenezer,O., Ogujiuba, K., Maredza, A., 2022. Exchange Rate Volatility, Inflation and Economic Growth in Developing Countries: Panel Data Approach for SADC. Economies 10, no. 3: 67, 2-19. https://doi.org/10.3390/economies10030067

Konstantakopoulou, I., 2016. New evidence on the Export-led-growth hypothesis in the Southern Euro-zone countries (1960-2014). Economics Bulletin 36 (1), 429-434.

Author Response File: Author Response.pdf

Reviewer 2 Report (New Reviewer)

Comments

Economies-1915072: The impact of oil price and oil volatility index (OVX) on the exchange rate in Sub-Saharan Africa: Evidence from oil importing/exporting countries

This paper demonstrates that oil price and oil volatility (OVX) are significant determinants of economic activity however. Several studies consider both variables in the oil-exchange rate nexus and ignore the distributional heterogeneity of the exchange rate. To address this, we investigate their joint effect and employ both the quantile regression and Markov switching models. We observe that OVX shocks have a major impact on exchange rate for all countries whereas, oil price shocks only affect the exchange rate of oil-importing countries. Rising (falling) OVX causes the local currency to depreciate (appreciate). The impact of rising or falling OVX is the same for oil-importing and oil-exporting countries whereas the impact of rising and falling oil price varies. The impact of oil price and OVX on exchange rate is affected by market conditions in that the exchange rate responds to oil price and OVX mostly at lower quantiles (bearish markets) for all countries, which reveals the sensitivity of investors, whereas a weak to no significant response is observed at the higher quantiles (bullish market). Our results are robust in model selection (Markov switching models).his study identify a perspective for tackling the issues of human capital investment and youth employment in Sub-Saharan Africa (SSA) with data spanning 1995-2017, 23 macro panel series from 40 SSA countries, making a total of 920 macro panel observations.

1.       The abstract is very extensive and should be reduced. In the abstract, only the most important points of the study should be presented. I advise the authors to be more precise with the paper's subject.

2.       The manuscript must be streamlined; in the present state, it is confusing for the reader.

3.       The authors should be better motivated. What is the contribution of the paper to the international macroeconomics?

4.       Which variables affect the exchange rate (Fraj et al., 2018; Conrad and Jagessar, 2018; Ebenezer et al., 2022; Konstantakopoulou, 2016)? Why is it interesting to investigate? Also, why is it essential to examine the impact of oil price and oil volatility index on the exchange rate?

5.       The presentation of the paper is not clear. Generally, there are interesting points in the article which are not highlighted. The authors should be more explicit in writing the paper.

6.       The following sentence is not correct; make the relevant correction. “Then, the bound testing approach (Pesearan et al., 2001) is used to investigate the degree of integration between the variables.”

7.       The empirical section follows the correct econometric methodology but is tedious to read.

References

Fraj, S.H., Hamdaoui, M., Maktouf, S., 2018. Governance and economic growth: The role of the exchange rate regime. International Economics, 156, 326-364.

Conrad, D., Jagessar, J., 2018. Real Exchange Rate Misalignment and Economic Growth: The Case of Trinidad and Tobago. Economies 6, 52. https://doi.org/10.3390/economies6040052.

Ebenezer,O., Ogujiuba, K., Maredza, A., 2022. Exchange Rate Volatility, Inflation and Economic Growth in Developing Countries: Panel Data Approach for SADC. Economies 10, no. 3: 67, 2-19. https://doi.org/10.3390/economies10030067

Konstantakopoulou, I., 2016. New evidence on the Export-led-growth hypothesis in the Southern Euro-zone countries (1960-2014). Economics Bulletin 36 (1), 429-434.

Author Response

REVIEWER 1, economies, #1915072

The impact of oil price and oil volatility index (OVX) on the exchange rate in Sub-Saharan Africa: Evidence from oil importing/exporting countries

Answers to reviewers are highlighted in yellow. Second round changes in the actual text/submission are also highlighted in yellow. Older changes in the actual text/submission are highlighted in different colors.

Comments and Suggestions for Authors

This paper examines the relationship between the exchange rates and the oil price/OVX in five sub-Saharan countries. Using the quantile response model and Markov exchange model, the author(s) finds an asymmetric response of exchange rates on oil price/OVX: falling prices and rising OVXs are likely to generate more significant effects. The paper generally reads good, yet, I have the following concerns and suggestions, arranged from the most important to the least.

1. What is the big picture? After knowing the results that the author(s) found, how can it change the world better?

AUTHORS’ REPLY:

After knowing the results, we have indicated what the implications are to the countries concerned in section 5.1. we have also renamed this section: The ‘big picture’ in the exchange rate-oil nexus and implications for the countries concerned.

For example, the countries sampled rely heavily on the energy sector and thus a major policy concern is the impact the changes in the oil markets can have on the exchange rate. This is also because fluctuations in the exchange rate can affect the growth of their economy. Our results also show that OVX is the dominant factor in the oil-exchange rate nexus.

We have also updated this section by commenting on the current European energy crisis and the effect on exchange rates. See below (with current references from the press):

To abstract from the African countries under examination and capture the big picture we offer a brief account of the current situation in Europe and the ongoing energy crisis (oil, natural gas, electricity).  The findings of this study are not relevant only for the countries under examination here. Specifically, the euro and the GBP have dropped in value in relation to USD (the lowest in nearly 20 years), inflation is increasing at an alarming rate and industrial production is reducing. If energy prices (including oil price and oil price volatility) have such a pronounced effect on currencies that are used for international transactions/payments, then the situation will definitively be gloomier for the currencies examined here. A major lesson to be learned here is that policy makers must take action to dampen the effects of energy prices on currencies whether those are used for international transactions or are less well known (as it is the case with the currencies under examination here)

(https://www.euronews.com/my-europe/2022/08/23/europes-energy-crisis-haunts-the-euro-as-it-tumbles-to-20-year-low-against-the-dollar)

https://markets.businessinsider.com/news/currencies/dollar-dominance-euro-energy-crisis-parity-currencies-investing-analysis-citi-2022-8

2. The particularity of countries. Why does the author(s) choose sub-Saharan countries? Data availability is understandable, but I wondered whether the author(s) have a non-country-specific result that differs from the research about other countries. In other words, does the author(s) have an aggregate effect on these five countries? If yes, what are these countries' differences compared to other parts of the world? These questions are essential because they may reveal why we have observed the correlations between exchange rates and price/OVX.

AUTHORS’ REPLY:

The reasons for choosing these countries have been mentioned in section 1.0 paragraph 3 and 4)

I believe that the aggregate effect you are referring to in this study is the effect of the OVX on the exchange rates of all countries (both importers and exporters). A similar study (Breen and Hu, 2021) even though with a much smaller sample (Norway, Australia and Canada), shows that the OVX is a good predictor of future exchange rates. One can conclude that OVX affects exchange rates for both developing and developed countries regardless of importer/exporter status. Given the absence of studies for developed countries and developing (excluding this one), one cannot really draw any conclusions regarding the uniqueness (and any special characteristics) of our sample.in other words we do not think that there any characteristics in our sample/countries that make our chosen countries different form the rest of the world except the fact that they are developing countries. Effectively more work needs to be done to reach ‘critical mass’ so as to draw any conclusions about the effects of OVX on developed/developing countries and export/import status given the tiny sample of developed countries in Breen and Hu (2021). This could be another study altogether. The paragraph we added (direct quotation) in our paper follows below for your convenience:

‘Second, although previous studies demonstrate the role of oil shocks on exchange rate, this study to the best of our knowledge, is the first empirical one to document the joint impact of both oil price changes and OVX changes on exchange rate for developing countries (both importers and exporters). Breen and Hu, (2021) look at the effect of oil price and OVX on the exchange rates of developed countries: Canada, Norway and Australia.  OVX appears to be a good predictor of exchange rates for Canada, Norway and Australia regardless of importer/exporter status. Therefore, we contribute to the literature providing new insights on the oil-exchange rate nexus from the perspective of the OVX (and price) for developing countries employing a bigger sample that could possibly allow us to offer generalized results for developing countries’.

3. Robustness. Are the results sensitive to model selection? I appreciate the Markov exchange model, but the author may add a more robustness check (which is supposed to be easy). For example, changing the quantile bandwidth, adjusting controls, etc.

AUTHORS’ REPLY:

We are not exactly sure what you mean by changing the quantile bandwidth. Currently we use 9 deciles (). We then divide them equally into 3 groups, namely lower quantile ((, median quantile (and higher quantile. Do you suggest using 5 ().? If we do this, then we would be able to look only at the lowest 20%, middle 40%-60% and highest 20%. In this case, the results would be more dramatic/accentuated especially at the extremes, the lowest 20%, and the highest 20%. Perhaps you mean something else. we are not exactly sure. now as far ‘controls’ is concerned, this question is answered below.

4. Control variables. I realize that in most regressions, there are very few control variables. The author(s) may consider adding some related variables as controls.

AUTHORS’ REPLY:

Thank you for the suggestion, we would have loved to add more variables as mentioned in Subsection 3.1 paragraph 3 but most of the economic variables are of lower frequency with a lot of missing data. To be more specific, the frequency of our current data is daily and this is how we are able to capture the effects of volatility and price changes on the exchange rate. Macro data is always available at a quarterly basis. For example, industrial production or money supply are available quarterly. If we include those variables and change the frequency, then we cannot use oil prices or OVX which is the point of the whole article. Volatility occurs at short term clusters. In a quarterly basis you will not be able to observe any volatility which negates the core concept of this article. The only variable that we could use potentially is the central bank base rate for each of the counties under investigation but again the central bank base rate changes only every few months if at all so econometrically speaking the base rate would be a ‘constant value’ adding nothing to the regression since it would be repeating itself. From an econometrics point of view a ‘non-varying’ variable would introduce a number of unwanted issues. There is no such thing as LIBOR available for the countries under examination. We subscribe to both Bloomberg and DataStream and all research to identify such variables was futile. We also checked at the websites of the central banks. Overall, the reason we did not include any control variables is not because we did not want to but a combination of non-availability and econometric ‘unsuitability’

5. Causal relation. Is it possible to generalize the current shape of the asymmetric correlation-like model to a causal model?

AUTHORS’ REPLY:

Even though we are not entirely sure what you mean, we believe that you are suggesting running a Granger causality test. We could do that for each country separately (pairwise Granger causality) or for all countries simultaneously (Dumitrescu-Hurlin approach) however this would be between i) foreign exchange rate and oil price and ii) foreign exchange rate and OVX. Standard pairwise Granger causality tests and the Dumitrescu-Hurlin approach allow only for the inclusion of a single explanatory variable, not two. We do not believe that this would anything to the whole paper since the whole idea is to test/understand how the two variables (oil price and OVX) interact together rather than separately. In addition, for Granger causality tests and in particular for the Dumitrescu-Hurlin approach the variables must be of the same order I(0) or I(1). We cannot mix variables of different orders. Unfortunately, this is not the case here. Even if this was possible it is unlikely that there would be any causality running from the exchange rate (exchange rates of SSA countries) to OVX. Perhaps the opposite could be the case. See references below:

Galariotis, E & Giouvris, E, International Review Of Financial Analysis (2015) On the stock market liquidity and the business cycle: a multi country approach, 38, 44-69, http://dx.doi.org/10.1016/j.irfa.2015.01.009,

Said, H. & Giouvris, E. Financial Markets and Portfolio Management. (2019) Oil, the Baltic Dry index, market (il)liquidity and business cycles: evidence from net oil-exporting/oil-importing countries.  33, 349–416. https://doi.org/10.1007/s11408-019-00337-0   

6. Writing. There are several typos in the paper, especially missing the articles (a/the) and having too many hard-to-read sentences. I wondered if any author(s) are native English speakers. If not, I suggest proofreading by a native one. For the author(s) convenience, here is a proofread example of the abstract.

AUTHORS’ REPLY:

Thanks for that, we have incorporated this in the submission

7. Table 6 (along with other similar tables) is somehow hard to read. It may be better to replace betas with the variable name. Also, I would suggest putting some simplified results in the main part but the full result in the appendix.

AUTHORS’ REPLY:

Done and thank you for the suggestion 

8. Page 24, Line 691, the subsection is "4.3.1.2"; should it be "4.3.2.2"? I realize the numbering is not continuous, and it is confusing.

AUTHORS’ REPLY:

numbers changed to read 4.3.2.2. thank you

9. The notes for Table 2 still show "table 1."

AUTHORS’ REPLY:

changed to table 2. Thank you

REVIEWER 2, economies, #1915072

The impact of oil price and oil volatility index (OVX) on the exchange rate in Sub-Saharan Africa: Evidence from oil importing/exporting countries

Answers to reviewers are highlighted in yellow. Second round changes in the actual text/submission are also highlighted in yellow. Older changes in the actual text/submission are highlighted in different colors.

Comments

Economies-1915072: The impact of oil price and oil volatility index (OVX) on the exchange rate in Sub-Saharan Africa: Evidence from oil importing/exporting countries

This paper demonstrates that oil price and oil volatility (OVX) are significant determinants of economic activity however. Several studies consider both variables in the oil-exchange rate nexus and ignore the distributional heterogeneity of the exchange rate. To address this, we investigate their joint effect and employ both the quantile regression and Markov switching models. We observe that OVX shocks have a major impact on exchange rate for all countries whereas, oil price shocks only affect the exchange rate of oil-importing countries. Rising (falling) OVX causes the local currency to depreciate (appreciate). The impact of rising or falling OVX is the same for oil-importing and oil-exporting countries whereas the impact of rising and falling oil price varies. The impact of oil price and OVX on exchange rate is affected by market conditions in that the exchange rate responds to oil price and OVX mostly at lower quantiles (bearish markets) for all countries, which reveals the sensitivity of investors, whereas a weak to no significant response is observed at the higher quantiles (bullish market). Our results are robust in model selection (Markov switching models).his study identify a perspective for tackling the issues of human capital investment and youth employment in Sub-Saharan Africa (SSA) with data spanning 1995-2017, 23 macro panel series from 40 SSA countries, making a total of 920 macro panel observations.

1. The abstract is very extensive and should be reduced. In the abstract, only the most important points of the study should be presented. I advise the authors to be more precise with the paper's subject.

AUTHORS’ REPLY:

The last 3 lines are not part of the abstract above, so I am not sure how this ended up in this review. Obviously, there is a mix up. The abstract submitted and included below is 200 words:

The Theory demonstrates that oil price and oil volatility (OVX) are significant determinants of economic activity; however, studies seldom consider both variables in the oil-exchange rate nexus and ignore the distributional heterogeneity of the exchange rate. We investigate their joint effect and employ both the quantile regression and Markov switching models to address this. We differentiate between positive/negative shocks and control for the effect of the global financial crisis in 2008 and the Covid-19 Pandemic in 2020.  We observe that OVX shocks significantly impact the exchange rate for all countries whereas, oil price shocks only affect the exchange rate of oil-importing countries. Rising (falling) OVX causes the local currency to depreciate (appreciate). The impact of rising or falling OVX is the same for oil-importing and oil-exporting countries whereas the impact of rising and falling oil price varies. The impact of oil price and OVX on exchange rate is affected by market conditions. The exchange rate responds to oil price and OVX mostly at lower quantiles (bearish markets) for all countries, which reveals investors sensitivity. In contrast, a weak to no significant response is observed at the higher quantiles (bullish market). Our results are robust in model selection (Markov switching models).

2. The manuscript must be streamlined; in the present state, it is confusing for the reader.

AUTHORS’ REPLY:

We have reduced a few sections, but it is not clear what you are suggesting. would streamlining involve removing the OLS regressions from the text and move those to the appendix? The reason we are suggesting removing the OLS regressions is because we want to show that using OLS in this particular study will not produce reliable results because of the changing nature of the distribution. Effectively the OLS regression is used to show that it is not appropriate and as a consequence we use QR and Markov switching models.

3. The authors should be better motivated. What is the contribution of the paper to the international macroeconomics?

AUTHORS’ REPLY:

The paper contributes to literature in the following ways: First, we employed the OVX as measure of uncertainty instead of the VIX which has been highlighted in section 1.0, paragraph 7.  We are not aware of any other studies that use both oil price and oil price volatility to explain exchange rates for developing countries. Second, most empirical studies, particularly in developing countries rely on linear models and do not take into account how the effects of oil price on exchange rate may vary throughout the exchange rate returns distribution which has been highlighted in section 1.0, paragraph 8 and section 2.2.4. Last but not least, the common trend in the literature is to examine effect of oil prices on exchange rate in developed countries. this study examines the effect of oil prices and oil price uncertainty on the exchange rate in developing countries as highlighted in section 1.0, paragraph 3 and 4.

The most important findings are:

1 We observe that OVX shocks significantly impact the exchange rate for all countries indistinguishably (both oil importers and exporters)

2 Oil price shocks only affect the exchange rate of oil-importing countries.

3 The impact of oil price and OVX on exchange rate is affected by market conditions. The exchange rate responds to oil price and OVX mostly at lower quantiles (bearish markets) for all countries, which reveals investors sensitivity.

We have also connected the current energy crisis in Europe to the effect that energy prices (oil, natural gas, electricity) currently have on international exchange rates, namely the EURO and the GBP against USD. See section 5.1. For your convenience we have also copied and pasted the last paragraph of section 5.1:

To abstract from the African countries under examination and capture the big picture we offer a brief account of the current situation in Europe and the ongoing energy crisis (oil, natural gas, electricity).  The findings of this study are not relevant only for the countries under examination here. Specifically, the euro and the GBP have dropped in value in relation to USD (the lowest in nearly 20 years), inflation is increasing at an alarming rate and industrial production is reducing. If energy prices (including oil price and oil price volatility) have such a pronounced effect on currencies that are used for international transactions/payments, then the situation will definitively be gloomier for the currencies examined here. A major lesson to be learned here is that policy makers must take action to dampen the effects of energy prices on currencies whether those are used for international transactions or are less well known (as it is the case with the currencies under examination here)

4. Which variables affect the exchange rate (Fraj et al., 2018; Conrad and Jagessar, 2018; Ebenezer et al., 2022; Konstantakopoulou, 2016)? Why is it interesting to investigate?

AUTHORS’ REPLY:

We have included the articles that you have suggested and indicated that those articles are mainly looking at the effect that exchanges rates can have on growth from a number of different perspectives (e.g governance, oil price, etc). We have included those articles at the very beginning of our introduction section in order to indicate clearly how those articles are ‘positioned’ in the literature and how this specific article is ‘positioned’ in the literature. This article is more specific in the sense that it looks into oil price shocks and specifically oil price volatility and their joint effect on exchange rates. It is interesting to investigate because currently we are observing the effect that energy prices in general have on international currencies. For example, the recent energy crisis in Europe had led to a decline in Euro and GBP in relation to USD. Of course, the price of oil is not the main culprit in this case (natural gas is) but generally speaking energy prices can have a serious effect as indicated here. the price of oil has been blamed in the past for major swings in exchange rates

In addition, this study provides an insight into the workings on the oil-exchange rate nexus for the countries considered here and allows policy makers to develop effective strategies to reduce oil price risk. this has been highlighted in section 1.0, paragraph 2 ,3 and 4. The policy implications are also highlighted in Paragraph 5.1

Also, why is it essential to examine the impact of oil price and oil volatility index on the exchange rate?

AUTHORS’ REPLY:

It has been highlighted in section 1.0, paragraph 3, 5 and 6.

5. The presentation of the paper is not clear. Generally, there are interesting points in the article which are not highlighted. The authors should be more explicit in writing the paper.

AUTHORS’ REPLY:

We believe that the most interesting point is the effect that OVX has on both oil importers and oil exporters especially in bearish markets. We have tried to connect this finding to the current energy crisis in Europe. We are highlighting this connection in section 5.1.

6. The following sentence is not correct; make the relevant correction. “Then, the bound testing approach (Pesearan et al., 2001) is used to investigate the degree of integration between the variables.”

AUTHORS’ REPLY:

Sentence corrected. Thank you

7. The empirical section follows the correct econometric methodology but is tedious to read.

AUTHORS’ REPLY:

Thank you for this comment. We have deleted a few sentences to simplify this section

References

Fraj, S.H., Hamdaoui, M., Maktouf, S., 2018. Governance and economic growth: The role of the exchange rate regime. International Economics, 156, 326-364.

Conrad, D., Jagessar, J., 2018. Real Exchange Rate Misalignment and Economic Growth: The Case of Trinidad and Tobago. Economies 6, 52. https://doi.org/10.3390/economies6040052.

Ebenezer,O., Ogujiuba, K., Maredza, A., 2022. Exchange Rate Volatility, Inflation and Economic Growth in Developing Countries: Panel Data Approach for SADC. Economies 10, no. 3: 67, 2-19. https://doi.org/10.3390/economies10030067

Konstantakopoulou, I., 2016. New evidence on the Export-led-growth hypothesis in the Southern Euro-zone countries (1960-2014). Economics Bulletin 36 (1), 429-434.

Author Response File: Author Response.pdf

Round 2

Reviewer 1 Report (New Reviewer)

I have no more questions. The author presented clear answers to my previous concern.

Reviewer 2 Report (New Reviewer)

The paper in this version is improved. But Proof corrections of this paper are necessary before publication.

This manuscript is a resubmission of an earlier submission. The following is a list of the peer review reports and author responses from that submission.

Round 1

Reviewer 1 Report

Thank you very much for having the opportunity to review the paper. The research is up to date and important. I have several remarks. Please find them below:

1. The literature review should be improved. Some valuable publications are missing such as Volkov and Yunh (2016), Malik and Umar (2019), Czech and Niftiyev (2021).
2. It would be crucial to include in the analysis data covering the COVID-19 pandemic period. Did the 2020 year implement changes in the analysed relationship? Why did the authors not apply data from 2020? A significant contribution to the study would be to include, apart from the year 2008, another time of instability and high uncertainty, which was undoubtedly the time of the new coronavirus pandemic.
3. The Methodology section lacks the aim of the study and research hypothesis.
4. Clear justification of the used models is not presented. Why did the authors decide to choose this model? Does the model fit the data? Are there other models that also might be applied for the research? Critical discussion of the selected method is missing. For example, the authors could use Markov-switching models, in which it can be assumed that the studied relationship varies over time and that the estimation of the slope parameter is related, among other things, to market volatility.
5. Oil prices and oil volatility measured by the OVX index seem to be correlated. So is there no problem with multicollinearity in the model?
6. Figures 1 are hardly readable, it should be corrected. Moreover, please clarify if the results are correct. On the graphs, the standard error band crosses the zero axis.
7. Additionally, I recommend the authors better organise the research results sections. There is no fully logical continuity in the presentation of the results.
8. The conclusion section should be corrected. The authors should provide also the limitation of the study and the plan for future research on the topic.

References:

Czech, K., & Niftiyev, I. (2021). The Impact of Oil Price Shocks on Oil-Dependent Countries’ Currencies: The Case of Azerbaijan and Kazakhstan. Journal of Risk and Financial Management, 14(9), 431.

Malik, F., & Umar, Z. (2019). Dynamic connectedness of oil price shocks and exchange rates. Energy Economics, 84, 104501.

Volkov, N. I., & Yuhn, K. H. (2016). Oil price shocks and exchange rate movements. Global Finance Journal, 31, 18-30.

Reviewer 2 Report

This study aims to examine the impact of oil price and oil volatility index (OVX) on the exchange rate in Sub-Saharan Africa. I think this paper needs major revision based on the following comments:

1. The introduction should state what to do, how to do and the final conclusions. It also should clearly give the contributions of this papers.
2. Authors should interpret empirical results rather than simply perform statistical analysis.
3. Authors should summary literature and find out the knowledge gaps, instead of listing literature. Some relevant papers recently published in this journal should be noted, such as Wang et al 2022 (doi: 10.1016/j.energy.2021.122501), Khan et al 2021 (10.1016/j.ocecoaman.2021.105955), Li et al 2021 (10.1080/1540496X.2021.2009339), Umar et al. 2021 (10.1016/j.energy.2021.120873)
4. This paper needs to be thoroughly checked for errors.