Effectiveness of Principal-Component-Based Mixed-Frequency Error Correction Model in Predicting Gross Domestic Product

Round 1

Reviewer 1 Report

The article “Effectiveness of Principal Component-Based Mixed-Frequency Error Correction Model in Predicting GDP” addresses a relevant issue, is well structured, the methodology is well described and the analyzes are adequate. Therefore, I understand that the article has the potential to be published. However, some minor revisions are necessary:

Add two lines to Table 1: one for the number of observations and another for the frequency (monthly, quarterly).

Add a note to Table 1 to specify the letters GDP, C, I, T.

In Figure 1 add the identification of the vertical axes.

Add a note to Figure 1 to specify the letters GDP, C, I, T.

Review the description of all equations to ensure that all symbols are identified. For example,  in equation 2.

Ensure that all symbols are identified. For example, ECM.

Author Response

Please see the attachment.

Author Response File: Author Response.docx

Reviewer 2 Report

mathematics-2601107

Effectiveness of Principal Component-Based Mixed-Frequency 2 Error Correction Model in Predicting GDP

The paper examines the forecasting ability of economic growth by using a mixed-frequency model to predict quarterly GDP. The authors refer to previous literature and confirm the previous studies, that the MIDAS model's effectiveness in predicting GDP is also evidenced by including consumption, investment, and trade.

1). The paper is a comprehensive empirical exercise on MIDAS and modifications from advanced econometric techniques. The paper is interesting, well-written, and the methodology is clearly exposed. The discussion of the results is based on the empirical findings. However, the originality of the paper doesn’t lie in the methodology part. Therefore, I would suggest the authors discuss their ideas relative to previous studies and findings more in depth. Also, it would be interesting if the authors would provide the audience with future work.

2). “The time lag of macroeconomic policy …”. Lag-length selection is an important issue in macro-econometric modelling, and different information criteria produce differing outcomes in terms of optimal lag selection. Hence, there is no a priori guide as to what the maximum length of the lag should be. I would suggest the authors the following reference:

Polyzos, E., Siriopoulos, C. (2023). Autoregressive Random Forests: Machine Learning and Lag Selection for Financial Research. Computational Economics. https://doi.org/10.1007/s10614-023-10429-9

Although relevant to this work, the choice of time lag length is not discussed by the authors, except in section 3.2.2, but without further justification.

3). The originality of the paper is not clear. Although the authors provide an excellent Introduction section, the contribution of their work is not obvious. For instance, very similar works such as the following are not presented and critically discussed in the paper:

Degiannakis, S. (2023). The D-model for GDP nowcasting. Swiss J Economics Statistics, 159, 7. https://doi.org/10.1186/s41937-023-00109-8

4). I would suggest the authors add a section discussing and criticizing previous literature. It would be interesting if the authors would present the different methods used by macro-econometricians to forecast GDP, and position accordingly their contribution to this strand of literature, which is dated since 1990 (see Evgenidis et al 2020). dHence, the originality of their paper and the contribution of their work would become clear and evident.

5). GDP forecasting is mainly based on yield curve. The authors do not consider this case. I would suggest the authors the following references amongst others:

(i). Ang, A, M. Piazzesi, and M. Weid (2006). What does the yield curve tell us about GDP growth? Journal of Econometrics,  https://web.stanford.edu/~piazzesi/APW.pdf

(ii). Evgenidis, A., S. Papadamou, and C. Siriopoulos (2020), The yield spread's ability to forecast economic activity: What have we learned after 30 years of studies? Journal of Business Research, 106, pp. 221-232, https://doi.org/10.1016/j.jbusres.2018.08.041.

 

Author Response

Please see the attachment.

Author Response File: Author Response.docx

Round 2

Reviewer 2 Report

Dear Authors,

Thank you for the revising version of your paper.