# Variance Equations is missing definition

here:

https://www.nrc.gov/docs/ML1208/ML12088A329.pdf

Campbell, Lo, Mackinlay: The Econometrics of Financial Markets

on page 159 i am looking at equation 4.4.9 in the last line,

= $I\sigma_{\epsilon_i}^2+X_i^*(X_{i}^{T}X_i)^{-1}(X_i^*)^T\sigma_{\epsilon_i^*}^2$

I would like to estimate $V_i$ and I am wondering what the terms $\sigma^2_{\epsilon_i}$ and $\sigma^2_{\epsilon_i^*}$ are?

The first term does not have a * and the second one does but neither term seems to be defined.

Also shouldn't the X matricies here: $(X_{i}^{T}X_i)$ have stars * because they are in the event window?

FYI - I already have the $X_i^*$ matrix.

Thank you.

• Aren't they defined in 4.4.5 and 4.4.7 ? – noob2 Mar 23 '17 at 21:56
• I do not think so. The first term does not have a hat or a start. It looks similar to 4.4.4 but without a hat. The second term is not defined at all. 4.4.5 has a hat and nothing in 4.4.9 has a hat. 4.4.7 has a hat and nothing in 4.4.9 has a hat. – user3022875 Mar 23 '17 at 22:13