Optimal timing of technology adoption

In a dynamic programming framework, this paper investigates the optimal timing of technology adoption by a competitive firm when technology choice is irreversible and the firm faces a stochastic innovation process with uncertainties about both the speed of the arrival and the degree of improvement of new technologies. A numerical example illustrates how the optimal timing decision is affected by changes in parameter values reflecting market conditions, the firm's initial technological attributes, and the characteristics of the stochastic innovation process. Some of these effects turn out to be in sharp contrast to common intuition. Contrasting the optimal decision rule derived here with the rule obtained under the net present value shows that the former implies a slower pace of adoption than implied by the latter. The optimal decision rule is generalized for the case of multiple technology switches and it is shown that for all the switching decisions except the Inst one, the optimal rule satisfies the net present value criterion. (C) 1998 Elsevier Science B.V. All rights reserved.