Compute covariance and store in column

[George Papadopoulos]

My data are the following:
I have a dataset with 77000 rows and 3 columns.
I want to compute the covariance of dclose and dclose_lag for each identifier ticker2.
So I need 194 covariances.

How can I create a column consisting of the covariances?

I have already computed the variance for the first ticker2, but I can not create a loop.

That is the command i used to compute the covariance for the 1st ticker2:

correlate dclose dclose_lag if ticker2==1, covariance
gen float covname = r(cov_12)

[Nick Cox]

That may make sense to people using your kind of data, but not to me.

By column do you mean variable?

Here is an odd fact: of the 20 people who have posted most on Statalist since 2014, only 4 are identifiably economists and so far as I can see none of those specialises in this kind of data.

So to maximise attention you should give a better idea of what your data look like. Please do read the FAQ Advice especially #12 and come back with a data example making your structure much clearer.

[George Papadopoulos]

Dear Nick Cox,
thanks for your advice.Yes, by column I mean variable
dclose is the difference of Pt and Pt-1, where Pt is the price of a stock at time t
dclose_lag is the lag of dclose
ticker2 is the identifier for each stock
And i need a new variable that will give me the covariance of dclose and dclose_lag for each stock

[Nick Cox]

Sorry, but without a data example I am not much wiser. rangestat (SSC) can calculate covariances. If that's not an answer, then best of luck that someone else can decode your question.