In their present paper. Professors Cheng and Deets (hereafter C-D) attempt to derive a measure of instantaneous systematic risk for securities and portfolios which is consistent with the Sharpe-Lintner-Mossin capital asset pricing model when the true market horizon is infinitesimally short. In so doing, they assert that Jensen's resolution of the horizon problem for such a market horizon is incorrect. In the comments which follow, I shall attempt first to indicate explicitly the causes for the differences in the Jensen and C–D results, and second, to evaluate their relative merits.