Fama-macbeth procedure vs OLS test of abnormal return
Posted: 05 March 2016 01:57 AM   [ Ignore ]
Newbie
Rank
Total Posts:  1
Joined  2016-03-05

Hi,

  Suppose there is a new trading strategy, let’s say long on stocks with high R&D and short stocks with low R&D.

  What is the difference between A) versus B)
A) running a fama-macbeth regression as in
http://www.wrds.us/index.php/repository/view/fama_mcbeth_regression_in_sas

B) an three factor regression on the fama-french three factors to test if there is a significant alpha, as in:
https://wrds-web.wharton.upenn.edu/wrds/research/applications/port/governance/
proc reg ;
  where &where;
  model ret_diff = mktrf smb hml umd;
quit ;

Profile
 
 
Posted: 05 March 2016 09:02 PM   [ Ignore ]   [ # 1 ]
Administrator
Avatar
RankRankRankRank
Total Posts:  901
Joined  2011-09-19

hi,

Fama McBeth is the ‘old school’ way of dealing with pooled data, where some regression is estimated for each year (and the ‘average’ effect would be the distribution of the coefficients). So, if somebody says they use Fama McBeth, you won’t know the regression itself (what the dependent and independent variables are).

A factor regression (B) with pooled data (several companies, several years) would ignore the issues that come with pooled data (correlated error terms). These days ‘clustering’ of standard errors seems to be the way to deal with pooled data.

Hope this helps,

Joost

 Signature 

To reply/post new questions: Please use the group WRDS/SAS on Google Groups! http://groups.google.com/d/forum/wrdssas

Profile