An Empirical Comparison of Seasonal Adjustment Methods on Philippine Time Series Data
Author: Jose Ramon G. Albert, Ph.D.
Abstract:
A number of time series data generated by the Philippine Statistical System contain seasonal and other systematic components. For such data sets, it is desirable to estimate their respective underlying components and seasonally adjust these data. We consider applying the two foremost leading deseasonalization approaches called X12-RegArima and TRAMO-SEATS to some Philippine time series data using a freeware from EUROSTAT called DEMETRA, which interfaces these two seasonal adjustment methods. Here, we compare the empirical performances of these seasonal adjustment schemes in terms of forecasting performance and the revisions resulting from the availability of new data.
Keywords: