Serving the Quantitative Finance Community

 
User avatar
chartreuse
Topic Author
Posts: 1
Joined: April 2nd, 2008, 12:39 pm

Orthogonalized residuals

July 1st, 2014, 7:50 pm

Hi.How does one compute orthogonalized residuals? I'm looking to regress changes in spreads on some explanatory variables like GDP and use "orthogonalized residuals" from the regression as an input as a regressor to another regression. Thank you!
 
User avatar
chartreuse
Topic Author
Posts: 1
Joined: April 2nd, 2008, 12:39 pm

Orthogonalized residuals

July 1st, 2014, 7:54 pm

Sorry, let me clarify: use "orthogonalized residuals" from the regression as a regressor in another regression.
 
User avatar
bearish
Posts: 5906
Joined: February 3rd, 2011, 2:19 pm

Orthogonalized residuals

July 2nd, 2014, 9:28 am

There are probably several ways to do this, but I would run PCA on the residuals. That way, not only will the end result be orthogonal, but you have a natural ordering of them in terms of their "importance", i.e. the amount of variance they explain.