Question #2
Reading: Reading 34 Hedge Fund Strategies
PDF File: Reading 34 Hedge Fund Strategies.pdf
Page: 1
Status: Unattempted
Correct Answer: A
Question
A process which is helpful for creating linear conditional factor models that avoid multicollinearity is:
Answer Choices:
A. step-wise regression analysis
B. correlation residual factor analysis
C. mean-variance collinearity analysis
Explanation
Step-wise regression was run by Hasanhodzic and Lo in their 2007 paper and
demonstrated that removing the variables Bond and Commodity from their model
resulted in a higher R2 than prior analysis which included these variables. They concluded
there was multicollinearity and removing those factors improved the model.