FDfilter
Posted: Tue Mar 06, 2012 10:52 am
Hi.
I am a new learner of frequency-domain analysis. I found that the FDfilter in eviews add-in is easy to use. However i am confused about several questions. I have read the document written by Corbea and Quliaris. They said in their paper that they developed this filter from Phillips (1991). My first question is , what is the advantage of FDfilter than Phillips(`1991)? And another silly question is, If i run a regression which, for example, of income on consumption,and i tried to find out the long-run relationship. i firstly filtered the variables( income and consumption) by FDfilter and then i used low-frequency components(what i got in the first step) of them to run regression. Then i say that the coefficient i got is the relationship in long-run. Is my method correct? Thank you!
I am a new learner of frequency-domain analysis. I found that the FDfilter in eviews add-in is easy to use. However i am confused about several questions. I have read the document written by Corbea and Quliaris. They said in their paper that they developed this filter from Phillips (1991). My first question is , what is the advantage of FDfilter than Phillips(`1991)? And another silly question is, If i run a regression which, for example, of income on consumption,and i tried to find out the long-run relationship. i firstly filtered the variables( income and consumption) by FDfilter and then i used low-frequency components(what i got in the first step) of them to run regression. Then i say that the coefficient i got is the relationship in long-run. Is my method correct? Thank you!