Exercise Solution 3.19

1. False: A positive definite covariance matrix is never singular.
2. False: A positive semi-definite covariance matrix may be singular.
3. False: If second moments for a random vector don’t exist, the random vector won’t even have a covariance matrix.