Dropping Observations by using if command

[Nihar Singh]

I have the daily return datasets for various companies from 1996 to 2021 (example of dataset attached below)-

input long co_code str8 co_stkdate float(bse_firm_level_return_percentage nse_firm_level_return_percentage)
11 "20081006" 0 0
11 "20081007" -9 -11
11 "20081008" -13 -12
11 "20081010" -11 -12
11 "20081013" 7 6
11 "20081014" -10 -10
11 "20081015" 9 10
11 "20081016" 14 13
11 "20081017" -20 -20
11 "20081020" -1 -1
11 "20081021" -3 0
11 "20081022" -5 -6
11 "20081023" -7 -6
11 "20081024" -13 -14
11 "20081027" -8 -10
11 "20081028" 19 20
11 "20081029" -9 -5
11 "20081031" 3 4
11 "20081103" 1 0
11 "20081104" 8 5
11 "20081105" -2 0
11 "20081106" 0 -1
11 "20081107" 14 15
11 "20081110" -3 -4

I want to drop the observations of the companies with zero returns on at least 75% of the days in the last 12 months. Kindly help me in this regard.

[Nick Cox]

I am having difficulty understanding what that means , but I can advise that your date variable is not fit for Stata purposes.

gen ddate = daily(co_stkdate, "YMD")
format ddate %td

[Nihar Singh]

Cox: I have used the following command for date:
tostring co_stkdate, replace force
list co_stkdate
gen date2 = date( co_stkdate, "YMD")
gen cal_year = year(date2)
gen month= month(date2)
gen year= cal_year if(month <= 3)
list year
replace year = (cal_year + 1) if(month >3)
list year
format date2 %td

Got this output after the following command-
input long co_code float(date2 cal_year month year bse_firm_level_return_percentage nse_firm_level_return_percentage)
11 17811 2008 10 2009 0 0
11 17812 2008 10 2009 -9 -11
11 17813 2008 10 2009 -13 -12
11 17815 2008 10 2009 -11 -12
11 17818 2008 10 2009 7 6
11 17819 2008 10 2009 -10 -10
11 17820 2008 10 2009 9 10
11 17821 2008 10 2009 14 13
11 17822 2008 10 2009 -20 -20
11 17825 2008 10 2009 -1 -1
11 17826 2008 10 2009 -3 0
11 17827 2008 10 2009 -5 -6
11 17828 2008 10 2009 -7 -6


What I mean to ask is that I want to exclude those stocks which had zero returns (calculated as the ratio between today's closing price and the last closing price) on at least 75% of the days in the last 12 months. Eg- a stock is traded for 300 days in a year and for at least 225 days (75% of 300) the returns were 0%. So, I want to drop this stock from my data.