Autocorrelation in Time Series Data : Cochrane-Orcutt

From sasCommunity
Jump to: navigation, search

Cochrane-Orcutt procedure is used to correct Autocorrelation in time series data. (Kutner, Nachtsheim and Neter (2009), Applied Linear Regression Models, chapter 12, page 492)

The AUTOREG procedure of SAS has the option of correction using Yule Walker Method.

This is what the help says:

Pioneering work in the field was done by Cochrane and Orcutt (1949). The Cochrane-Orcutt method refers to a more primitive version of the Yule-Walker method that drops the first observation. The Cochrane-Orcutt method is like the Yule-Walker method for first-order autoregression, except that the Yule-Walker method retains information from the first observation. The iterative Cochrane-Orcutt method is also in use.

The macro CochraneOrcutt estimates the autocorrelation, performs the Durbin Watson test on the corrected model.

I used YW test with a very small sample (n=16) and the rho estimate was lower using Yule Walker than using Cochrane-Orcutt. So I wrote the macro.File:CochraneOrcutt.sas

Please feel free to comment/use the macro. --Piyas Bandyopadhyay 21:12, 21 April 2010 (UTC)