... regression models with several explanatory variables for two reasons. First, they quantify the effects of collinearity, summarizing the impact of redundancies among the explanatory variables. Second, they save a lot of time. Consider the Fama-French variable High-Low. Is High-Low not statistically significant in this multiple regression because it is redundant (as is the case for Dow % Change), or is it simply unrelated to the response? To decide, look at the VIF. Because the VIF of High-Low is near 1, collinearity has little effect on this explanatory variable. High-Low is not related to returns on Sony stock once we’ve taken account of the other explanatory variables.
There’s no definitive rule for what constitutes a large VIF. We could have VIF > 10 ...
Get Statistics for Business: Decision Making and Analysis, 3rd Edition now with the O’Reilly learning platform.
O’Reilly members experience books, live events, courses curated by job role, and more from O’Reilly and nearly 200 top publishers.
