Monthly Stock Return Calculation lagged L operator

[Suzan Bekci]

Hi,

I have a simple question. I have an unbalanced panel dataset with monthly prices for each company from 2001 to 2022. I want to create a new variable calculating the monthly return for each company and each month however L lagged operator does not work. it creates a wholly blank column, consisting only of dots. Can you help me?

[Carlo Lazzaro]

Suzan:
welcome to this forum.
As you do not provide the details requested by FAQ as far as your data are concerned, I can only figure out that you're facing a situation similar to what below:

. set obs 1
Number of observations (_N) was 0, now 1.

. g id=1
. g time=1
. xtset id time
. g A=10
. g lag_A=L.A

. list

     +------------------------+
     |  A   time   id   lag_A |
     |------------------------|
  1. | 10      1    1       . |
     +------------------------+

.

[Suzan Bekci]

Hi Carlo,

Thank you I am glad that I joined this forum. But I solved my problem. instead of using l lag operator I used the code below;

sort ticker_numeric date
by ticker_numeric: gen lagged_price=price[_n-1]
gen monthly_return =(price - lagged_price) / lagged_price

Good day,
Suzan

[Nick Cox]

My guess is that you have monthly data but daily dates.

[Suzan Bekci]

My guess is that you have monthly data but daily dates.

yes Nick, the dates are in day/month/year format. Do you think it was the reason for lag operator not working exactly as i wanted?

[Nick Cox]

Yes indeed. You should convert daily dates to monthly dates using mofd() and then tsset using the monthly date variable.

[Suzan Bekci]

Yes indeed. You should convert daily dates to monthly dates using mofd() and then tsset using the monthly date variable.

Thank you sir.

[Suzan Bekci]

Hi, still working on the same project. this time I want to merge two stata dataset with a common column however the same column is string in one and float in the other one. how can i merge them? when ı try to convert from string or float , stata assigns numeric values, which i do not want. ı want to keep the same values in the common variable

[Carlo Lazzaro]

Suzan:
did you -destring- your -string-variable?

[Suzan Bekci]

Suzan:
did you -destring- your -string-variable?

yes i did. it said "tmonth: contains nonnumeric characters; no replace" the variable is in the format of "2003m1"

[Nick Cox]

I don't know how today's messages are compatible with earlier posts, but if (and only if) you are holding monthly dates in a string variable, then

gen mdate = monthly(tmonth, "YM")
format mdate %tm

[Suzan Bekci]

hi,

I want to rank the best and worst performing stocks over the last twelve months. For this, I used the below code but it gave syntax error. Can anyone help me?

egen initialrank = rank(-mom12) if date==min(tmonthnew)

Best,
Suzan

[Nick Cox]

min() as a Stata function requires two or more arguments which are compared with each other -- within observations if two or more variables are named. min() as a Stata function does not return a minimum over a variable.

What you want isn't clear to me without an example. It might be that you want to use the egen function min(), or a saved minimum from summarize.