Forecasting Industrial Production Using Its Aggregated and Disaggregated Series or a Combination of Both: Evidence from One Emerging Market Economy
Round 1
Reviewer 1 Report
see attached file.
Comments for author File: Comments.pdf
Author Response
Please see the attachment.
Author Response File: Author Response.pdf
Reviewer 2 Report
See report
Comments for author File: Comments.pdf
Author Response
Please see the attachment.
Author Response File: Author Response.pdf
Round 2
Reviewer 1 Report
see attached review.
Comments for author File: Comments.pdf
Author Response
Please see the attachment.
Author Response File: Author Response.pdf
Reviewer 2 Report
The authors have implemented as many of the referee requirements as could reasonably be done in the 10 day period apparently suggested, and documented their changes in their reply. I remain astonished that ETS is 10 times better than any other method considered, with a tiny MSFE of 0.0024 even 12 periods ahead that is barely larger than the variance of the data. I have almost never seen an ex post forecasting exercise where even 10% improvements were achieved, never mind the 1000% claimed here.
The max-min is just 0.2 in Figure 1 and their reply suggests the Brazilian SD is twice the US whereas footnote 4 says 4 times, so one must be wrong. I am also surprised that the aggregating weights are fixed implying that the composition of industrial production has not changed in 20 years. That must play a key role in disaggregates forecasts helping, and would not apply to other economies I know.
I would rather use Delta_12y than an AR 13 given how seasonal the data shown in Figure 1 are: what is the standard deviation of Delta_12y? Please do NOT claim that differencing induces stationarity: it merely removes a unit root from the dynamics. Also Delta^12 should be Delta_12 just below the figure. A minor issue is the dates are too small to read; and on Figure 2 unless they want to print in color, use dashed and dotted lines to discriminate better in print.
The authors should save their pdf in a Word file (or use Scientific Word) to run a spell checker and use the information to correct a number of spelling mistakes in LaTex before a final recompile.
Author Response
Please see the attachment.
Author Response File: Author Response.pdf