The three most common violations of classical regression theory that occur in cross-sectional factor models are (1) the errors in variables problem, (2) common variation in residuals such as heteroskedasticity and serial correlation, and (3) multicollinearity. There are statistical techniques that address the first two. The third issue is best dealt with by removing collinear variables from the regression, or by increasing the sample size.
EXHIBIT 12.14 Total Return Report (annualized)