Cross-correlation of returns matrix

Assume a returns table A(200 rows returns,30 cols stock names), trying to create a cross correlation matrix between each of the pairs of stocks.

Am I right trying to use xcorr function? Trying to do that in a for loop and also tried to xcorr with matrix name.

xcorr would take timeseries of A(,1) and create all cross correlations with rest of columns (in pairs), or am I wrong? I would expect the result would be D=cols+(cols-1)+..(cols-N), one cross-correlation for each unique pair. But thats not what I get from the below. Is my understanding wrong or my usage of function?

[r,lags]=xcorr(A)

or

for i=1:30
    for j=1:30
        D=xcorr(A(:,i),A(:,j))
    end
end

Answer 1

[Rik]

I'm not sure the rest of your code is doing what you want, but at the very least you need to index D so you don't overwrite it every iteration.