Article ID: | iaor1990359 |
Country: | Belgium |
Volume: | 29 |
Issue: | 2 |
Start Page Number: | 41 |
End Page Number: | 73 |
Publication Date: | Jun 1989 |
Journal: | Belgian Journal of Operations Research, Statistics and Computer Science |
Authors: | Raj Baldev, Siklos Pierre L. |
Keywords: | economics |
This paper restresses the importance of utilizing the spectrum of original and filtered univariate time series in assessing the impact of various filters that are commonly employed in time series and econometric analysis. The importance of linking the type and extent of filtering utilizes by researchers to the objective of the study is also stressed. The use of the log-change filter is illustrated to obtain a covariance stationary series for several annual macroeconomic time series.