Q1. What is this? Q2. Do we have this in L2? Q3. An analyst is estimating whether company sales is related to three economic variables. The regression exhibits conditional heteroskedasticity, serial correlation, and multicollinearity. The analyst uses Hansens procedure to adjust for the standard errors. Which of the following is TRUE? The: A) regression will still exhibit heteroskedasticity and multicollinearity, but the serial correlation problem will be solved. B) regression will still exhibit multicollinearity, but the heteroskedasticity and serial correlation problems will be solved. C) regression will still exhibit serial correlation and multicollinearity, but the heteroskedasticity problem will be solved. D)heteroskedasticity, serial correlation, and multicollinearity problems will be solved.
B-- it’s in reading 12
B? Hansen method is used if serian corrleation and heteroskedasticiy is a problem. Its mentiond in the CFA book, hardly in the Schweser.
Just a note (in case anyone forgets) Hansen method should not be used if only heteroskedacity exists.
Thanks all for pointing this ‘lost Hansel (In a jungle of Time Series)’ to the appropriate sections. I just read the Hansen paragraph from CFAI and now the question sounds familiar to me. B is the correct choice, well done. I surely think that I am doomed for L2 already, since I have only done Schweser.
Please you correct me more than I can remember.
Niblita75 Wrote: ------------------------------------------------------- > Just a note (in case anyone forgets) Hansen method > should not be used if only heteroskedacity exists. You mean to say that it should be not be used if only serial correlation is a problem…surely? It is a test for serial correlation and not heteroskedasticity and it can only be used if there is autocorrelation and heteroskedasticity.
I’m trying to remember. (Read my post on the I quit thread because this is what I am talking about). If you just have autocorrelation you can use the hansen method, but it will also correct heteroskedacity. If you conclude that there is only heteroskedacity and not autocorrelation, you should not use hansen method.
Niblita, Hansen method has not even been mentioned under heteroskedasticity.
Niblita is right. Hansen method should be used if serial correlation is present. It stimultaneously corrects for heteroskedasticity. Robust is preferred to use if only heteroskedasticity is present.
CFAI text book 1 pg 303. I don’t know the page for Schweser.
That is what I’ve been saying. Forget it. We are looking at the same thing from two different angles.
I beleive both of you are saying the same thing. Niblita is saying you should not use Hansen if only herteroskedacity is present.
Schweser p. 199. Last paragraph
ruhi22 Wrote: ------------------------------------------------------- > > You mean to say that it should be not be used if > only serial correlation is a problem…surely? It > is a test for serial correlation and not > heteroskedasticity and it can only be used if > there is autocorrelation and heteroskedasticity. I just got confused by your wording. Hansen is not a test, but a measure for correcting autocorrelation. It can be used I think if only autocorrelation is existent. But I agree, two angles, same end result.
This is what I felt after reading that section… Hansen method is used to CORRECT for autocorrelation by adjusting the standard errors, but the good thing about it is that, It simultaneously corrects for the conditional heteroskedasticity too. So … 1 bullet, 2 dead!! Also there is a 2nd method to CORRECT for autocorrelation, by modifying the regression equation itself, but this would just eliminate serial correlation and not the conditional heteroskedasticity.
Yup, that’s why its preferred by CFAI, but appearantly, not by Schweser. hmmm…