🚀 go-pugleaf

In article <l9nikaINNf99@mizar.usc.edu> bomar@mizar.usc.edu (Jason Bomar) writes:
>I would like to know if anyone knows how statistically something is determined
>to be a trend over time.  Is this a question of regression, analysis of

	In the macroeconomics, time series, and econometrics literature,
exactly this issue of definition/determination of a trend has been debated
with some ferocity.  The keypoints are a little involved to treat fully
with any justice in this forum, but those interested might look at papers
such as:

Beveridge, S. and Nelson, C. Journal of Monetary Economics 1981
	A New Approach to Decomposition of Economic Time Series
	into Permanent and Transitory Components

Nelson, C. and C. Plosser, Journal of Monetary Economics 1982
	Trends and Random Walks in Macroeconomic Time Series

Watson, M., Journal of Monetary Economics 1986
	Univariate Detrending Methods with Stochastic Trends

Stock, J. and M. Watson, Journal of Economic Perspectives, 1988
	Variable Trends in Economic Time Series

Cochrane, J., Journal of Political Economy, 1988
	How big is the random walk component in GNP?

Quah, D., Econometrica, 1992, The Relative Importance of
	Permanent and Transitory Components

	Any one of these papers will also contain references to
many others.  The list above is woefully short, but should serve
as a useful start to follow up on... (my apologies to those who feel
I should have included their papers too as one of the key works... :-))


--
Danny Quah dquah@atlas.socsci.umn.edu
College of Liberal Arts   University of Minnesota, Minneapolis, West Bank