Hello,
for my research I need fundamental data of US companies and want to match them to stock returns of CRSP. I have access to CRSP Compustat merged and want to know whether this is the right way to match the data:

1. Get the fundamental data from the CCM database (here I get gvkey and lpermno and lpermco as identifiers)
2. Get the stock data from the CRSP database (permno and permco as identifiers)
3. Match the data in stata via “permno”.

Thanks for the answer
orima

Sounds like a plan!

You can always ‘debug’ your code by inspecting observations with no matches, and multiple matches (and might as well look into successful matches) to make sure things are going as planned.

Hope this helps,

Joost

Hello,

I also have a question regarding the merging of ccm and crsp by permno in stata.

Earlier, I also have this question but till now I haven’t been able to merge them yet.

I follow exactly the steps listed by Orima.

1- download ccm
Since ccm has lpermno, which I think it is also permno in Crsp. Thus, in Stata, I also create permno which is lpermno.

2- download crsp

3- merge ccm and crsp by permno.

The results are that i lost almost half of the observations. But that is not a problem. However, the problem is that i think that merge is not correct.

For example:
in my ccm file, i have
permno         datadate
10001           31dec2003
10001           31dec2004
...
10001           31dec2015

in my crsp file, i have (for example)
permno         date (of individual day bid-ask spread)
10001           2jan2003
10001           3jan2003
10001           4jan2003
...             
10001         6dec2009

but when i merge them, the results is like this
permno         date             datadate
10001           2jan2003         31dec2003
10001           3jan2003         31dec2004
10001           4jan2003         31dec2005

And this is not true, because they bid-ask spread is still in the fiscal year from 1/1/2003 to 31/12/2003, but the datadate is mixed up.

I am so worried. I must have made a huge mistake here.

Could someone please help me with this question? Any suggestions/tips/hints are appreciated.

Kind regards,
An

hi An,

What was the Stata code you used for the match, and how would the results need to look like (what are you trying to get)?

Best Regards,

Joost