Simple correlation Question

Hi there. Here is my question.

Q) I have two time series of financial data. A = 1000 x 1 B = 1000 x 1

Now I want to find out the correlation of these at different frequencies.

So I want to test if A and all its lages is correlated with B and all its lags.

So I want to know for instance if the 5 minute price of A is correlated to the 2 minute price of B and etc etc....

I look forward to hearing a reply from you.

Answer 1

[Wayne King]

xcorr()

crosscorr()

Be careful with the word "frequency", there is a frequency domain correlation called coherence, but what you are after sounds like xcorr() or crosscorr()

For example:

    x = randn(1e3,1);
    y = randn(1e3,1);
    [xc,lags] = xcorr(y,x,'coeff');