Risk-aversion and the investment-uncertainty relationship: a reply

In my (Nakamura, 1999) paper, I tried to analyze the investment decision of a competitive firm with a constant-return-to-scale technology under uncertainty. The main contribution of this piece is that, with enough risk aversion, the investment-uncertainty relationship can be negative. Unfortunately, as Saltari and Ticchi (2003) correctly point out, my formulation of the maximization problem prevents the existence of a closed form solution. Essentially, the fault resides with my assumption that the output price follows a geometric Brownian motion. Their reformation of the problem, in which the price in each period is assumed to be an i.i.d. random variable, assures a closed form solution and shows that my main contribution holds. In addition, the investment function derived in my paper has a couple of strange properties: a very risk-averse firm behaves like a risk neutral one and a rise in a capital depreciation rate may increase investment. Appealing to non-expected utility preferences (Kreps and Porteus, 1978, 1979) implicitly, Saltari and Ticchi separate the effect of substitutability from that of risk-aversion to present an important and clear explanation for the results. In the sense that they throw a new light on the coherent intertemporal aspect of the investment decisions, their comment certainly constitutes a useful addition to the investment-uncertainty literature. It is hoped that the comment stimulates further research in this area.