How to create a for loop with a matrice of S&P 500 asset for computing correlations.

Hi, I have a dataset composed about 2 years prices of all the S&P 500 assets (523x504 MATRIX), imported as a matrix. What I'm trying to do is to compute all the correlations between a single asset of the index respect to the remaining 503 (we have a total of 504 asset), and then plotting toghether all the computed correlations.

I wanted to ask you for some suggestion about how to proceed with the creation of a for loop, and if in your opinion could be better to import the data in a table format for then develop the for loop.

For the moment, I have just computed the correlation coefficient running this code:

tmp = corrcoef(SP500price);
plot(tmp,'.')

But in that way I obtained just the correlations matrix, while I'm looking for the correlation of a each single asset respect to the remaining 503.

Thanks in advance!

Answer 1

[Brahmadev]

Hi Gianluca,

I understand that you would like to get the correlation between each column of “SP500price” matrix. Each element (i,j) of the Correlation matrix given by “corrcoef ” gives the correlation coefficients of each asset "i" (size:523x1) with a different asset "j" (size: 523x1).Using a for loop is also a possible solution, but you can achieve the same result without it.

To plot the correlations of first asset with all the other assets, you can use:

plot(tmp(1, :),'.');

Please take note that the diagonal elements of the correlation matrix "tmp" are 1 as they denote the correlation of an asset with itself.

For more information on the function 'corrcoef', refer to the following documentation:

https://www.mathworks.com/help/matlab/ref/corrcoef.html

Hope this helps in resolving the issue you were facing!

Best Regards,

Brahmadev