McCallum (2010) presented evidence that the spurious regression problem can be solved by standard means. We show using finite-sample evidence that the spurious regression problem cannot always be fixed using standard autocorrelation correction procedures and remains, therefore, a not-so-spurious...