First, I would like to ask if there is any way to reduce serial correlation in my residuals. I am applying the DOLS methodology and by manipulating the leads and lags, I achieve the correlogram that is at the end of this post.
The second question is, whether that much correlation as the one I have is actually a problem. I have heard that small correlation can be neglected, however, I am not exactly sure how much correlation can be regarded as "small".
Lastly, if I do have correlation, can applying HAC (Newey-West) compensate for this?
- ac.png (15.76 KiB) Viewed 2350 times