construct 10 equally weighted decile portfolios by ranking stocks on the past 3 year volatility

[MarthelF]

Hey Dear members, 

I'd like to construct with my Data sets in the Attachments 10 equally weighted decile Portfolios, which are ranked on the past 3 year volatility. what i've done yet is to rank and create a colomn on  my data named volatility like you can see on my SAS Code. But the data inside are again the data from my Excel that mean from LogReturn, so my Question now is how to calcul the volatility of each firm and rank all this result on the past 3 year volatility by generating 10 equally weighted Decile Portfolios. Thanks for your Help in Advance!

FILENAME REFFILE '/folders/myshortcuts/MyFolders/SaS_Stoxx300.xlsx';
PROC IMPORT DATAFILE=REFFILE
  DBMS=XLSX
  OUT=WORK.IMPORT
  replace;
  GETNAMES=yes;
  SHEET="LogReturn";
  RANGE='a3:zz99999999999'n;
RUN;

proc sort data=work.import(keep=_numeric_);
  by date;
run;

proc transpose 
      data=work.import  
      out=work.transposed(rename=(_label_=Firm Volatility1=Volatility) where=(not missing(Volatility)) drop=_name_)
      prefix=Volatility;
  by date;
run;

proc rank data=work.transposed groups=10 descending out=work.ranked;
  var Volatility;
  ranks Decile;
  by date;
run;

proc sort data=work.ranked;
  by date decile;
run;

proc print data=work.ranked;
  by date;
run;

[Ksharp]

How do you define 'equally weighted decile portfolios' ?

Do these portfolios have the same expect return (  mean(return) ) or the same variance ( var(return) )?

[MarthelF]

Hi Ksharp, to come back on your Question , i understand now what you mean, on my Excel data we have the LogReturn, the second  step should be to calculate the volatility of all this data, then use a Code to sort all the data into equally weighted decile Portfolio, that means on my example with 303 firms , after sorting we'll  have 10x30  and  each decile will have 30 Firms having data of volatility. i have added the Excel data in the Attachments to Show you how my data will look for after sorting. PS:  the data inside my tables are not correct just selected for me like example.  

[MarthelF]

Hi Ksharp,
It means that the excess return are quite fast the same after sorting but the data are sorted on variance( Volatility).thanks!