(1) IN A FURTHER COMMENT on Wallingford's [4] valuation model Lee [3] criticizes Brennan's [1] and my [2] comment for an improper derivation of the certainty equivalent of the aggregate dividends. Lee's critique stems from a misinterpretation. To see this note that the market value of the aggregate dividend can be derived by two methods leading to different results: (a) Wallingford first derives the expected utility of the random dividend per share D(t)/ N(t) which leads to the certainty equivalent d(t)=E[D(t)]/N(t)-a'Var[D(t)]/N(t)2. As all shares have the same market price, the market value (a) of the aggregate random dividend D(t) must be N(t)d(t) according to Wallingford's model. Hence N(t)d(t) is the certainty equivalent (a) of D(t) based on the valuation of a single share so that an infinite number of shares is optimal. Therefore Brennan and I propose: (b) Derive the expected utility of D(t) which leads to the certainty equivalent (b) (= market value (b)) D(t) = E [D(t)] a'Var[D(t)]. Hence D(t) is independent of N(t) which seems reasonable. (2) In his final result Lee claims that for k = 1 even Wallingford's model renders the number of shares, N(s), irrelevant. This result is due to an algebraic error in Lee's formula (9): As according to his formula (8) F(s) is independent of k, the correct version of formula (9) is: