Collinearity in Vector Autoregression and Impulse Response (Time Series)

138 Views Asked by At

I am sort of new to time series, and I am jumping right into modeling one time series in terms of past values of itself as well as other time series. Simply from visualizing the predictor time series, it appears that two of them are related in some way or have some similarities. I know the issue of collinearity is relevant in linear models, and I was wondering how that is dealt with in vector autoregression (either on a conceptual level or in r specifically)? How is impulse response affected by this collinearity, and is there anything to address it?

1

There are 1 best solutions below

0
On

You should take a look at Grange Cuasility for association of two time-series. There is a simple package in R to implement this.