“…Second, since the conditional variances and covariances likely are not constants, empirical estimates should allow for this possibility (for an example and related references, see McNew & Fackler, 1994). Nonetheless, such time-varying covariance estimation is costly and often does not result in greater hedging effectiveness relative to unconditional hedge ratios (Garcia, Roh, & Leuthold 1995;Myers, 1991). 4 Using estimated parameters implies that the estimated returns from the associated decisions can have a large variance around the true (but unknown) optimum (see Chalfant, Collender, & Subramanian, 1990, and references therein).…”