Portfolio construction and regression

[Lutz Binder]

Hi,

So I have got monthly panel data of about 5000 firms over a period of about 10 years and need to construct equally weighted as well as value weighted top in class and bottom in class portfolios based on a sustainability measure. Afterwards I need to take the (equally weigthed or value weigthed) returns of those portfolios and regress the Fama &French factors against them. Since I am not very experience with stata I was wondering whether someone could help me with the commands to do so. Therefore I would need a command that select the returns of the top and bottom 25 % each month based on the sustainability variable followed by a mean command of those returns and then regressing them with three other varables. For the value weighted portfolios I would also need a command which would weighted each return by the fraction of this firms market value to the whole market value within that portfolio.


Any help is highly appreciated
Regards,

Lutz

[Friedrich Huebler]

You can make the question less abstract by sharing an excerpt from your data. See section 12 in the FAQ for useful advice.

[Lutz Binder]

 bysort date: egen portfolio_top= mean(RET), pctile( ESGSCORE), p(25)

by Data basically looks like this:


Date ID RETurns ESGSCORE MarketCAP

200501 A 0.25 1.44 3.0e11
200501 B 0.22 1.20 1.2e10
200502 A 0.23 1.43 2.9e11
200502 B 0.20 1.20 1.1e10


I have got panel data for about 4500 companies ( number of companies each year varies). These companies are scored according to ESG measures which is the ESGSCORE variable.
Now I want to construct portfolios EACH MONTH averaging the returns but selecting only the companies' returns given the ESGSCORE.. Hence in 200501 the Top and Bottom 25% of companies based on the ESGSCORE values.
I tried the code above which doesnt work (Invalid p) maybe somebody can help.

[Nick Cox]

That's fantasy syntax. For example, you are combining two egen functions in a way that is quite illegal.

You need to calculate the percentiles first, then means conditionally on those. See also http://www.statalist.org/forums/foru...nel-data-graph for a similar question.

 
bysort date : egen ESC25 = pctile(ESGSCORE], p(25)  
by date : egen lower25 = mean(RET) if ESGSCORE < ESC25

[Lutz Binder]

Thank you very much Nick Cox. The first part worked although now I get a type missmatch error for the second line

[Lutz Binder]

ok got it now. needed to destring my RET thank you