Inconsistency of HAC standard errors in event studies with i.i.d. errors
Careless users of software with HAC adjustment should beware. In particular, in event study models with pulse dummies included as explanatory variables, the study shows that, in the presence of i.i.d. errors, HAC-adjusted t -statistics on the event dummies have divergent distributions and can, to the point of probability one, lead to spuriously identified significant events. Thus, point and click users of software with HAC adjustment would best be served by pre-testing the nature of the error term in regression models that contain pulse dummies before applying the adjustment.
Year of publication: |
2005
|
---|---|
Authors: | Fomby, Thomas B. ; Murfin, Justin R. |
Published in: |
Applied Financial Economics Letters. - Taylor and Francis Journals, ISSN 1744-6546. - Vol. 1.2005, 4, p. 239-242
|
Publisher: |
Taylor and Francis Journals |
Saved in:
freely available
Saved in favorites
Similar items by person
-
Inconsistency of HAC standard errors in event studies with i.i.d. errors
Fomby, Thomas B., (2005)
-
Equity price discovery with informed private debt
Addoum, Jawad M., (2020)
-
Financial Literacy and Food Security in Extremely Vulnerable Households
Millimet, Daniel L., (2015)
- More ...