How Can We Extract a Fundamental Trend from an Economic Time- Series?
AbstractThis paper attempts to extract a fundamental trend, which we call a " trend-cycle component," from an economic time-series. The "trend-cycle component" consists of a medium-term business cycle component and a long- term trend component. The objective is to eliminate the short-term irregular and seasonal variations that hide a fundamental trend in an economic time-series. We test five different time-series methods. Among them, the Henderson moving average (which is incorporated in an X-12- ARIMA seasonal adjustment program), the Band-Pass filter (which utilizes a Fourier transformation), and the DECOMP are found to be effective in extracting a "trend-cycle component" with a cyclical period longer than 1 .5 years. However, no method is found to be effective in extracting a " long-term trend component" with a cyclical period longer than that of a medium-term business cycle. Although the HP filter is somewhat successful , it still contains a component with a cyclical period of about three years that corresponds to a business cycle. These methods are useful for forecasting a wide variety of economic variables because they reveal a fundamental trend in the time series. In addition, statistical programs are available for easy application. They have, however, a few shortcomings. First, it is often difficult to provide a meaningful economic interpretation of the revealed characteristics of the "trend- cycle component." Second, the addition of new data can change the estimation results. In particular, an extracted component around the end of a sample period is likely to be revised with new data. Special caution is in order, therefore, in interpreting the estimation results and forecasting the time series when the data exhibit large variations. In this case, comparing the results of different methods provides a useful way to assess the reliability of an extracted "trend-cycle component."
Download InfoIf you experience problems downloading a file, check if you have the proper application to view it first. In case of further problems read the IDEAS help page. Note that these files are not on the IDEAS site. Please be patient as the files may be large.
Bibliographic InfoArticle provided by Institute for Monetary and Economic Studies, Bank of Japan in its journal Monetary and Economic Studies.
Volume (Year): 16 (1998)
Issue (Month): 2 (December)
Contact details of provider:
Postal: 2-1-1 Nihonbashi, Hongoku-cho, Chuo-ku, Tokyo 103
Web page: http://www.imes.boj.or.jp/
More information through EDIRC
Find related papers by JEL classification:
- E32 - Macroeconomics and Monetary Economics - - Prices, Business Fluctuations, and Cycles - - - Business Fluctuations; Cycles
- C22 - Mathematical and Quantitative Methods - - Single Equation Models; Single Variables - - - Time-Series Models; Dynamic Quantile Regressions; Dynamic Treatment Effect Models &bull Diffusion Processes
Please report citation or reference errors to , or , if you are the registered author of the cited work, log in to your RePEc Author Service profile, click on "citations" and make appropriate adjustments.:
- Michael F. Bryan & Stephen G. Cecchetti, 1994.
"Measuring Core Inflation,"
in: Monetary Policy, pages 195-219
National Bureau of Economic Research, Inc.
- John Moore & Nobuhiro Kiyotaki, .
1995-5, Edinburgh School of Economics, University of Edinburgh.
- Hirose, Yasuo & Kamada, Koichiro, 2003. "A New Technique for Simultaneous Estimation of Potential Output and the Phillips Curve," Monetary and Economic Studies, Institute for Monetary and Economic Studies, Bank of Japan, vol. 21(2), pages 93-112, August.
For technical questions regarding this item, or to correct its authors, title, abstract, bibliographic or download information, contact: (Kinken).
If references are entirely missing, you can add them using this form.