Vuong test for comparing R-squared

Hello. I know one can use the Vuong test for comparing non-nested models with the same data generating process. I am wondering if the test can be used to compare the same model's R-square but over different time periods - in which case the data would be different.
If it can't be used for that, would anyone know of a reasonable method to compare the R-squared of the two regression outputs for different time periods. Basically I am working on a pre and post study and want to know if the model improves after a spefic year.
Regards, Deep.
I want to see if the relationship between stock prices and two accounting variables (earnings and book value) has changed over time. Ofen research just shows the R-squared change but not the statistical significance of the change.


Not a robit
Not sure there is a test for that. Are the series the same length? You could report the R^2's with their 95% CI.
Last edited:
The series are the same lengths but over different time periods. How would I get a confidence interval for comparing the two R-squareds? I have not seen anything like that before.