Matching SDC with I/B/E/S
Posted: 13 May 2015 05:40 AM   [ Ignore ]
Newbie
Rank
Total Posts:  3
Joined  2015-05-12

Hello everyone!  I have used this forum many times for solving my little problems with WRDS ect, but this is my first post! I will try to make my problem as clear as possible and I hope some of you will have some good suggestions.

For my research, I want to look at the effect of Merger means of payment on analyst recommendations. Therefore I need the following information:

1. M&A mergers dates, means of payment, ect.
2. Firm control variables (size, ect)
3. One analyst recommendation before the M&A date per analyst, and then the recommendation issued after 90days the M&A announcement date.

Where the first two should be easy to get, I’m mainly having problems with obtaining number 3. I have downloaded the M&A transactions from SDC, and converted their 6digit Ncusip to 8 digit Cusips. When I enter these cusips into the IBES database,  I get 32 000 analyst recommendations, which each have their own date ect. therefore it is hard for me to match these analyst reports into the right time frame of the right M&A transaction.

I have read that it is more easy to access WRDS databases with SAS. unfortunately i’m not educated in SAS, but in Stata. I have downloaded the SAS University version, but the programs are not as similar as I have hoped.

My questions to you are:
- Does anyone have some good suggestions to structure and match these two databases
- Is it possible to retrieve analysts reports from IBES with sas, based on a timewindow and CUSIP so it is easier to match?

Kind regards!
Kiwietje

Profile
 
 
Posted: 13 May 2015 09:41 AM   [ Ignore ]   [ # 1 ]
Administrator
Avatar
RankRankRankRank
Total Posts:  901
Joined  2011-09-19

hi Kiwietje,

I suppose you have constructed a dataset with merger events, where each record is a merger with a certain date, and control variables like industry, size etc. You would need to ‘decide’ at which point in time you want to measure analyst properties, as they change over time. For example, you may want to measure #analysts three months before and three months after the merger announcement (just an example). Or, when you want to measure EPS forecast, you need to decide on which period (next quarter, current year, next year, etc).

You then construct variables that capture a date or period that you will use when matching with IBES (INTNX is a helpful function to create date variables). So, keep your dataset ‘wide’, i.e., all relevant variables for each merger in one row (add new variables as a column, don’t separate information for the event across multiple rows).

You can then use rsubmit to do the following:
- proc upload to upload the dataset with events (with ‘cusip’, relevant variables holding dates, etc)
- proc sql to match your freshly uploaded dataset with IBES
- proc download to download the result.

See the ‘tutorials’ on this website for how to use rsubmit.

In step 2, you need to know how IBES organizes their data. A quick and dirty match will give too many results (many doubles).

Assuming you use IBES detail, you need to realize:
- analysts forecasts multiple periods for a given firm (different quarters, different years)
- analysts forecast different things, even in the EPS datasets: EPS growth and EPS itself
- that analysts revise their forecasts, so a given forecast may appear on multiple dates

To get a handle on IBES detail, it is probably a good idea to download records for a single company for a few months and inspect what is there.

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
 
 
Posted: 14 May 2015 07:46 AM   [ Ignore ]   [ # 2 ]
Newbie
Rank
Total Posts:  3
Joined  2015-05-12

Hi Joost!

Thanks for your reaction! it is really helpful. My dataset until now is pretty ‘wide’ and i’m adding all variables in columns.  I think that the way you described is th easiest way to get to the data I need for my project.

It seems that I need to learn how to program SAS anyway if I want to get the data for my project.  I have tried to connect to WRDS with SAS and it says that i’m not licensed to connect to WRDS. Do any of you have experience with SAS university? is it because of my SAS version (the free one) ?

Kind Regards

Profile
 
 
Posted: 14 May 2015 01:15 PM   [ Ignore ]   [ # 3 ]
Administrator
Avatar
RankRankRankRank
Total Posts:  901
Joined  2011-09-19

hi Kiwietje,

Did you try to log on with something like this:

%let wrds wrds.wharton.upenn.edu 4016;options comamid TCP remote=WRDS;
signon username=_prompt_

The same username/password you use for the ‘normal’ website (wrds.wharton.upenn.edu) should work.

I have not used SAS University myself. I have heard of difficulty assigning directories as libraries, but I am not aware of issues logging in to WRDS.

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
 
 
Posted: 16 May 2015 05:06 AM   [ Ignore ]   [ # 4 ]
Newbie
Rank
Total Posts:  3
Joined  2015-05-12

Thanks!  I will give it a shot !

Else I can access SAS 9.3 from university computer I have heard so It should work!

Profile