Calculating annual returns that start few months after year end from monthly returns:Monthly-to-Annual

[Mike Kraft]

Hi
I have a simple question that might be basic for most users here:
I have monthly stock returns and want to calculate annual returns that start 4 months after the end of the year.
I used this code to calculate annual returns:
gen compounded_annual_ret= (1+f1.ret)* (1+f2.ret)*(1+f3.ret)*(1+f4.ret)*(1+f5.ret)*(1+f6. ret)*(1+f7.ret)*(1+f8.ret)*(1+f9.ret)*(1+f10.ret)* (1+f11.ret)*(1+f12.ret)-1
or simply
gen annual_ret= f1.ret+f2.ret+f3.ret+f4.ret+f5.ret+f6.ret+f7.ret+f 8.ret+f9.ret+f10.ret+f11.ret+f12.ret

If I want returns that start 4 months after end of fiscal year for my monthly data, does the following code make any sense?
gen 4month_compoundedannualret=f4.compounded_annual_re t
or
gen 4month_annualret=f4.annual_ret

My questions are:
Are the codes used above correct?
If the identifier is "permno", do I need to add "bysort permno" to the beginning of the annual returns codes above ?

Thanks in advance

[Mike Kraft]

Hi
Does anyone have some guidelines on how to proceed with these returns ?
Thanks

[Roberto Ferrer]

Mike,

Your question is not clear to me:

1. Is the _annual return_ the simple sum of a set of monthly returns?

2. If so, when you say "want to calculate annual returns that start 4 months after the end of the year.", do you mean the sum should start in may? If so, when does the sum end? In December of that year or in April of the next year?

3. What does "permno" identify? Different stocks?

4. Please post example data, representative of your working data set. An example of code producing such a thing:

clear
set more off

set obs 36

set seed 32896

gen dat = _n - 1
format %tm dat
gen year = year(dofm(dat))

gen return = runiform()

list, sepby(year)

You can also post the data directly, appropriately formatted.

5. Please follow recommendations given below if you want to increase your chances of getting a good answer, if any at all.