Announcement

Collapse
No announcement yet.
X
  • Filter
  • Time
  • Show
Clear All
new posts

  • Clustering standard errors at year or firm-level.

    Let's say I am running a regression where Y is stock price and X is natural disaster. Is it appropriate to cluster my standard errors both at the year and firm level?
    1. Firm-level because stock prices are auto-correlated.
    2. Year-level because all firms at certain years might have had a similar impact because of the natural disaster. But then, I'm not sure if this is related to year fixed effects?.. or standard errors?
    Thank you!

  • #2
    Jun:
    you seem to have serial correlation of the epsilon within and across panels.
    If you're dealing with a T>N panel dataset, you may want to take a look at -xtpcse-.
    Kind regards,
    Carlo
    (Stata 19.0)

    Comment

    Working...
    X