HI!   I think I need different codes for 1)holding period return and 2)monthly, yearly return, or similar.   holding period return is tricky because I need to use days from a certain date to another certain date. In other words, I cannot simply make code someday+1's price and divide a previous day by the next day.   For exmaple, from July 12th, 2019, one-day HPR needs the next business day, July 15th, 2019. So... To calculate one-day holding period return, I may need a new code other than day+1.  I guess there should the code navigating adding or subtracting days by the order of rows on certain variables. (For example, if the interval is +- 1 business days, that would be July 11st and July 15th, 2019 as my CRSP daily data has the price row following business days) So far, I have looked for the code having date interval specified by the order of rows of interest. Please share your knowledge... Maybe my ability to search for the answer is so poor...   Another question is also related to the above question. If I need to know a month return from June 7th, 2019 to July 7th, 2019, I need to have stock prices on both dates. Well... 2019/07/07 is Sunday so I need next day, Monday. In that case, I need another code that automatically finds the next available date on my dataset.  I think this is quite an advanced SAS programming.     Maybe the gist of those coding is how to skillfully use the order of rows and I even wonder whether this is possible.    There is almost no way to express my thankfulness to you have helped me other than writing here. However, I would like to thank you again for your helpful replies very much!!!!   Jerry ... View more

Hello all, Importing Excel files shows corrupted files again and again..   I matched the bitness of my SAS and the Office to 64 bit and uninstall and reinstall SAS, the Office, and the SAS Documentation Server... but nothing changed... I changed .xlsx to .csv and all other formats available but... well... exhausted kind of.   I simply clicked the "Import Data" button over the menu bar. and followed processes... Not even sure what is the cause.   Absurdly, there is no error or warning sign on my log window at all. FYI, I attach the file(not only this file but other files does not properly is converted too with no error sign on the log window) so maybe one of you try to figure it out.   I asked my cohorts and my university's IT team and I know they have done their best. Sadly all did not work so far.   Please share with me any tips to move a bit forward!   Thank you!!!   ... View more

proc sql; create table Rstd as select cusip and year,std(dailyr) as with_in_std from rvol group by cusip, year; quit; I think this time I am even closer to the right answer than before. feels like I am learnings SAS more. Yes. I know I do not have to put and year after cusip but simply do not know how to have grouped std... I did both  select cusip,std(dailyr) as with_in_std and select year,std(dailyr) as with_in_std and it worked so good and saw groups std eight by cusip or year. BUT... I want grouping using both year and cusip simultaneously.   I appreciate your help very much!   ... View more

  proc sql; create table snp1 as select put(caldt,yymmn6.) as yymm, mean(SPINDX) as mean_price from snp group by calculated yymm ; quit; I think I am almost there.   My data set is monthly so... instead of having yymm6 I may need something to sum by year. I put some weird things like yyyyn6 and yes. it didn't work.   Could you tell me how can I get yearly average's' when using monthly data? I have months from 1964 to 2018 so.. there will be more than 50 yearly average!   Thank you!!!     ... View more

Hello!   I need to use one model used in previous papers but, of course, do not know the code for that, It is the regression of credit ratings on some indicator variables(like loss or income) and others(like sales). As there are more than 20 different credit rating scores and I need to use some indicator variables, I am lost how to make the code.   I find many about using categorical predictors but unsure about categorial regressees. Maybe using Logit/Probit thing...   e.g. : Credit Rating = b0+b1Sales+b2Loss indicator+b3dividend indicator...   If you already have worked on credit rating regression like this, please share your wisdome!   Thank you!!!   Jerry ... View more

Hello,   I merged two(monthly and daily) and made a merged daily one but the daily one is not like what I want... I used a merge code as below.   data CRSP_new; set CRSP1; year = YEAR(date); run; proc sql; create table crspcomp as select a.* , b.* from cstatcusip1 a inner join CRSP_new b on a.cusip8 = b.cusip and a.fyear= b.year; quit; and... it made this       What I rather wanted to have is no double dates on NamesDate row in chronological order and the same in I/B/E/S row too. As seen below, IBES is a monthly data set and it has median values that need to be matched by CUSIP and date.  My problem is... medians need to be changed when corresponding I/B/E/S dates change.         This seems kind of advanced than I usually deal with and cannot be done by the above code as this time needs chronological order + corresponding changes in median values. I referred to some related threads but failed to make it. and finally asking for your help again...   Please share some advice again    I sincerely appreciate all of your help!!   Jerry ... View more