Intuitively, we would expect that an increase in the military preparations of potential enemies imply that the rival perceives an increase in the likelihood of future conflict. In this paper, we present a simple model that suggests that, surprisingly, the relationship is ambiguous. We find that (a) the specification of the social utility function; and (b) the rate of substitution between long and short lead-time preparations in the production of defense capability play a role in determining whether rivals respond to an increased future threat, by increasing or decreasing their long lead-time preparations.