Next Article in Journal
South African Government Bond Yields and the Specifications of Affine Term Structure Models
Next Article in Special Issue
Factors, Forecasts, and Simulations of Volatility in the Stock Market Using Machine Learning
Previous Article in Journal
The Influence of Board Diversity on Capital Structure Decisions: Examining Financial Risk Management Across Different Market Conditions in UK-Listed Firms
Previous Article in Special Issue
Who Is Leading in Communication Tone? Wavelet Analysis of the Fed and the ECB
 
 
Article
Peer-Review Record

Modelling Value-at-Risk and Expected Shortfall for a Small Capital Market: Do Fractionally Integrated Models and Regime Shifts Matter?

J. Risk Financial Manag. 2025, 18(4), 203; https://doi.org/10.3390/jrfm18040203
by Wafa Souffargi 1,* and Adel Boubaker 2
Reviewer 1: Anonymous
Reviewer 2: Anonymous
J. Risk Financial Manag. 2025, 18(4), 203; https://doi.org/10.3390/jrfm18040203
Submission received: 10 January 2025 / Revised: 4 March 2025 / Accepted: 5 March 2025 / Published: 9 April 2025
(This article belongs to the Special Issue Machine Learning-Based Risk Management in Finance and Insurance)

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

The paper examines and provide a comprehensive view into modeling and forecasting stock market volatility focusing mainly for the Tunisian index returns.  The introduction and literature review are appropriate and the review has the flow which helps in understanding the relevance for research.  The key difference in the paper is in examining the student’s t-distribution instead of the normal distribution helps in capturing excess kurtosis (fat tails) observed in financial data which indicates a benchmark improvement.  The study is found to be in line with the Aloui and Ben Hamida (2014) study.The final model offers better estimate as the ICSS-ARFIMA-FIAPARCH model with Student’s t-distribution plays significant in evaluation. 

A few key points which can help enhance the readability of the paper:

1. Why tunishian market ? and what is the significance of structural breaks?

2. Any specific movements in stock markets which makes the period of structural break significant?

3. Often studies with chow test is done to examine and confirm the presence of structural changes? if so, would this suffice?

4. State the underlying assumptions beyond the relevance of the test considered and their relevance with the structural breaks. 

5. why is the period of study important when compared to longer time frame?

6. The study seems to deviate far into econometric method and less into its relevance for tunishian markets. Can we state the research gap from the perspective of the financial markets and link objectives and conclusions to it.

 

The author/s can provide a much more comprehensive view in introduction on the need for the study. If more clarity is provided, it can help delve into econometric models better. If possible a comparative analysis.

Author Response

"Please see the attachement"

Author Response File: Author Response.docx

Reviewer 2 Report

Comments and Suggestions for Authors

The paper proposes an application of different financial time series models to the TUNINDEX 22 index returns data.

Although the paper presents a rich and well-argued introduction, it shows several problems in the presentation and lack in motivation.

First of all, the use of the ICSS for identify breaks in variance is not justified in any way. It is well known, since Mandelbrot (1963), that volatility clusters are a fundamental characteristic (Stylized facts) of the time series of returns, and volatility clusters ARE breaks in variance. In this framework the Engle's ARCH test is enough. Why use ICSS that was developed for independent data when observations of the financial series are not? Moreover, the authors also suppose that the analysed series presents Long Memory effects, that is they assume presence of autocorrelation, so a similar test is completely useless. Furthermore the authors do not clarify how k1 and k2 tests  overcome this problem.

The authors could test for the presence of LM effects, which they do not, in particular because looking at the figure 1, such effects do not seem to be present.

The rest of the paper is a simple application of well-known methodologies, and it is not clear what is the novelty in the paper contribution.

Minor point:

Pag4 lines 172-173: the sentence makes no statistical sense, reword

Pag. 4 line 176: from the sentence it seems that the proof of the asymptotic distribution of the ICSS was derived by the authors, while this result is present in the Tiao original paper.

Pag. 5 eqs (4), (5) and (6): the formulas of the tests k1 and k2 are missing, there are only those of their asymptotic distributions

Pag. 5 line 190: the references for the ARFIMA model in missing

Pag.7 line 2070: the V@R formula is missing

Pag. 11 table 2: 4 digital points are enough

Pag. 15 table 4: 4 digital points are enough

 

Pag. 17 table 6: 4 digital points are enough

Author Response

"Please see the attachement"

Author Response File: Author Response.docx

Round 2

Reviewer 2 Report

Comments and Suggestions for Authors

I remain of the opinion that the ICSS test is useless in this context.

However, the authors have provided adequate and detailed answers to my previous objections.
Although I believe that the paper does not contain any particular novelties, IMO it can be accepted for publication

Comments on the Quality of English Language

no comment

Author Response

"Please see the attachement"

Author Response File: Author Response.docx

Back to TopTop