Using standard deviation values as independent variables

  • #1

Main Question or Discussion Point

Hey. I am planning on doing some research, where I predict a change based on different types of risk.
The question is simple. Can I use values of standard deviation as independent variables in a linear regression analysis (OLS)? The standard deviation values over time will be calculated in advance from time series, and then used as independent variables in the regression analysis.

Mons
 

Answers and Replies

  • #2
BWV
592
495
You have to be careful of overlapping time periods, ie the rh var is sd from period t:t+n and lh variable is data at time t, with each increment of t having n overlap - this will lead to autocorrelation and spurious t-values unless you use a robust method like GMM to correct for these
 

Related Threads on Using standard deviation values as independent variables

Replies
3
Views
816
Replies
9
Views
4K
  • Last Post
Replies
1
Views
641
  • Last Post
Replies
6
Views
3K
  • Last Post
Replies
5
Views
992
Replies
6
Views
399
  • Last Post
Replies
2
Views
3K
  • Last Post
Replies
6
Views
4K
  • Last Post
Replies
4
Views
4K
  • Last Post
Replies
18
Views
911
Top