2

I am using cross correlation to find the correlation between two time series data say X and Y. I have read this somewhere that :

" If X or Y contains auto-correlation or is not stationary about mean , the cross- correlation will not reflect the true relationship between X and Y.

What to do in this case?

Thanks in advance

Arushi
  • 1,027
  • 4
  • 14
  • 22
  • Maybe this post will be helpful to you http://stats.stackexchange.com/questions/68425/how-to-determine-correlation-between-stationary-and-non-stationary-time-series/68592#68592 – Alecos Papadopoulos Dec 23 '13 at 20:49

1 Answers1

2

For non-stationary variables we should always think in terms of cointegration

Hibernating
  • 3,723
  • 2
  • 21
  • 34
  • What do mean by this term ? can you explain it more? – Arushi Dec 24 '13 at 20:29
  • 1
    The term is defined and explained well on page 5 of the slides that I linked to. You asked what to do and I mentioned the area of econometrics that you may need to study. There is a lot of research and expository material available on this topic, which you can summarize for yourself before asking further, more concrete questions. – Hibernating Dec 25 '13 at 01:44