[Abraham Wolde-Tsadick]

Hi Salma,

Using Nick's code, now you can collapse the daily data to a yearly data, add 1 year to the collapsed year variable and merge as follows:

egen SD = sd(ret), by(permno yr)
collapse (mean) SD, by(permno yr)
replace yr = yr + 1
save SD.dta, replace
use masterdata.dta, clear
merge m:1 permno yr using SD.dta

Abraham

[Abraham Wolde-Tsadick]

Hi Salma,

Using Nick's code, now you can collapse the daily data to a yearly data, add 1 year to the collapsed year variable and merge as follows:

egen SD = sd(ret), by(permno yr)
collapse (mean) SD, by(permno yr)
replace yr = yr + 1
save SD.dta, replace
use masterdata.dta, clear
merge m:1 permno yr using SD.dta

Abraham

[Nick Cox]

Abraham: I think you understand much more of the context than I do. I have not been following these threads until I saw a post just a short while back.

If you're collapsing, then the first two lines in turn can be just one:

  collapse (sd) SD=return, by(permno yr)

[salma ktat]

help varlist

gives a good start.

As you're dealing with panel data, you really need to know how these operators can help you.

Another massive issue is that if you wish to reduce a daily dataset to a yearly dataset, doing it yourself with egen is much less satisfactory than using collapse.

I looked for help varlist as you recommended and then used the lagged function. It worked well.
It was a big help, thanks a lot Prof. Cox.

[Abraham Wolde-Tsadick]

Nick, thanks for the shortcut. That makes it even more efficient and easy to read.
I was just following the thread. Salma needs to make sure that the code gives what she want.

[salma ktat]

Nick, thanks for the shortcut. That makes it even more efficient and easy to read.
I was just following the thread. Salma needs to make sure that the code gives what she want.

Thanks Abraham for the suggestion but I don't like the collapse function as it converts the data in memory into a dataset of standard deviation, permno, year. Instead I used the following codes:
tsset id date, daily
egen SD= sd(returns), by(id year)
bys id year: keep if _n==_N

tsset id year, yearly
gen L_SD= L.SD

And that worked well.

[Nick Cox]

You can pass anything else you like through collapse.

P.S. collapse is a command, not a function. In Stata commands and functions are different.

[Attaullah Shah]

I did not pay enough attention to your query Salma!. My apology. Unlike the previous case where you wanted to find standard deviation over a 5 year rolling window, if you want to find standard deviation in a single year from daily observations for each ID, then things get simplified enough. The three stage code is no more needed, the following simple loop can do the job.

forval i=1986/2012{
bys ID: egen SD_`i'=sd(ret) if year==`i'
}

[Nick Cox]

You are missing the point made earlier that these SDs can be packed into a single variable.