Comment on ``Density-matrix renormalization-group method for excited states''

In a Physical Review B paper Chandross and Hicks claim that an analysis of the density-density correlation function in the dimerised Hubbard model of polyacetylene indicates that the optical exciton is bound, and that a previous study by Boman and Bursill that concluded otherwise was incorrect due to numerical innacuracy. We show that the method used in our original paper was numerically sound and well established in the literature. We also show that, when the scaling with lattice size is analysed, the interpretation of the density-density correlation function adopted by Chandross and Hicks in fact implies that the optical exciton is unbound.