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.

Markowitz [12] and Tobin [19] pioneered in the development of a portfolio selection model resting on the assumptions that the investor1. Chooses among alternative investment opportunities solely on the basis of expected return (E) and standard deviation of return 〈σ〉, and2. Prefers more expected return to less but will refuse to incur additional risk (measured by standard deviation) unless compensated by increased expected return.