Quote:
Originally Posted by fluorescenthippo
I am trying to find out the meaning of the Hodrick test and its standard errors. Google is hardly helping. I am not sure if this is the same as the Hodrick–Prescott filter test that Google keeps taking me to as my professor is terrible.
Any help or links to resources appreciated.
I tried to find this as well (just to help you) and was somewhat unsuccessful. From the little information I gathered, it seems like a cubic spline is fitted to time series data (and the data looks like differenced data--probably returns data).
When you fit a model like this to data (consider simple linear regression, multiple regression), you can estimate error for the coefficients in your model and also get error sum of squares. I'm not sure what standard errors you want, but the problem is very likely parallel to multiple regression and you are looking at some specific quantity.
If you could ask your professor what the null and alternative hypotheses are that would be of some help.