Identifying technology horizons for strategic investment decisions

Technology investment models usually assume that a known number of technologies will appear in the future, though their appearance, cost, and revenue parameters may be uncertain. In practice, however, it is not always clear how many future technologies or periods need to be considered. We introduce the concept of a technology horizon, which we define as the minimum number of future technologies and future periods that need to be considered to guarantee that the optimal decision would not change even if additional technologies or periods worth of data were to be considered. We also allow the costs and revenue of existing technologies in our model to vary depending on the latest technology introduced. Further, we allow for implementation costs, which enables us to model situations where it may cost more to leapfrog technologies than if one were to buy intervening technologies. We then provide a fast algorithm to identify such technology horizons.