Hi everyone,   I have a dataset like this (sample data). This dataset is organized by permno (stock identification), last date, and date. Variable "date" are the dates within the 2-year window from last date. For example, for the last date=31/7/2001, dates are ranging from 31/7/1999 to 31/7/2001. Symbol are stock symbols for each date   Permno Last date Date SYMBOL 12345 31/07/2001 31/07/1999 A  12345  31/07/2001 01/08/1999 A  12345  31/07/2001 02/08/1999 A  12345  31/07/2001 ………… A  12345  31/07/2001 15/02/2000 A  12345  31/07/2001 16/02/2000 A1  12345  31/07/2001 ……… A1  12345  31/07/2001 21/04/2001 A1  12345  31/07/2001 22/04/2001 A2  12345  31/07/2001 ………. A2  12345  31/07/2001 26/06/2001 A2  12345  31/07/2001 …… …… 12356 15/09/2009 ………. B   Now I would like a dataset that summarizes the begining dates and enddates of a stock symbol. The output should look like this. It shows that permno 12345 has a symbol = A from 31/7/1999 to 15/2/2000, and A1 from 16/2/2000 to 21/4/2001 and so on   Permno Last date Begdates Enddates SYMBOL 12345 31/07/2001 31/07/1999 15/02/2000 A  12345  31/07/2001 16/02/2000 21/04/2001 A1  12345  31/07/2001 22/04/2001 26/06/2001 A2 12356  15/09/2009 ……… ………… ………     Could anyone show me how to do it? Thank you very much ... View more

Hi everyone,   I have a data structured like this, where dates are trading days and month_year is a character variable created by this function: month_year=put(date,monyy7.);   Stocks date Month_year …………….. …………….. …………… ABC 01/04/2011 Apr2011 ABC 04/04/2011 Apr2011 ABC 05/04/2011 Apr2011 ABC 06/04/2011 Apr2011 ABC ……………………… Apr2011 ABC 28/04/2011 Apr2011 ABC 29/04/2011 Apr2011 ABC 02/05/2011 May2011 ABC 03/05/2011 May2011 ABC ……………………………… May2011 ABC 31/05/2011 May2011 XYZ 02/12/2010 Dec2010 XYZ 03/12/2010 Dec2010 XYZ 05/12/2010 Dec2010 XYZ 06/12/2010 Dec2010 ……….. ………….. Dec2010 XYZ 30/12/2010 Dec2010 XYZ 03/01/2011 Jan2011 XYZ 04/01/2011 Jan2011 XYZ 05/01/2011 Jan2011 XYZ 06/01/2011 Jan2011 ……… ……… ………….. XYZ 29/01/2011 Jan2011   Now I want to identify the last trading day each month for each stock. The output should look like this, where dates are last trading date of each month: Stocks Date Month_year ……….. ………. ………… ABC 28/04/2011 Apr2011 ABC 31/05/2011 May2011 ……….. ………. ……….. XYZ 30/12/2010 Dec2010 XYZ 29/01/2011 Jan2011   I use this code: data want; set have;  by stocks date month_year;      if last.month_year then output;   run;   However, this code produce the original data (with all the dates within month). That means SAS identify every observations in the month_year column as last.month_year. This is hard to understand because I convert the month_year to character. Could anyone show me how to fix it? Thank you for your time ... View more

Hi everyone,   I have a dataset with variable "meancost". Now I want to create new variable "lagmeancost" which equal to meancost in previous 4 quarters (4 lag). For example, the lagmeancost of GVKEY (firmid) 1001 in 2002Q1 = meancost in 2001Q1 (=11)   The problem is that I have to ignore the first 4 observations in each group to avoid the lagmeancost to be contaminated by data of other company.   The output data should look like this GVKEY Quarter Meancost lagmeancost 1001 2001Q1 11 . 1001 2001Q2 11 . 1001 2001Q3 11 . 1001 2001Q4 11 . 1001 2002Q1 12 11 1001 2002Q2 12 11 1001 2002Q3 12 11 1001 2002Q4 12 11 1002 2006Q1 65 . 1002 2006Q2 65 . 1002 2006Q3 65 . 1002 2006Q4 65 . 1002 2007Q1 70 65 1002 2007Q2 70 65 1002 2007Q3 70 65 1002 2007Q4 70 65   Any idea how to do that? Thank you very much ... View more

Hi everyone,   I have the following data with company name and TIC code. Company TIC A CAVR.1 B CAB.3 C 4330B   Now with TIC code that have character '.' , I would like to extract only characters before the '.' For example, TIC code CAVR.1 will be sub-stringed to CAVR   The output looks like this: Company TIC A CAVR B CAB C 4330B   Can anyone help me with this, please? Any idea would be much appreciated ... View more