why checking correlation in square series?

Why MATLAB product help (Econometric tools-Getting started-Time series modelling-GARCH models) calculates the correlation in both return series and the square of the return series. What is the reason for investigating the correlation in square series?

Answer 1

[Wayne King]

I'm guessing because correlation in the squared returns indicates correlation in the process variance. And that would indicate the possibility that the process is a good candidate for GARCH modeling.