In the new environmental era, electricity markets change rapidly, requiring dynamic strategic responses from market agents and regulators. In this paper, we study the intertemporal dimension of the network planner decision process. In particular, we formulate a dynamic model of transmission investments and use this model to compare the behavior of a network planner when considering and when ignoring the dynamic nature of transmission investments. This paper's analysis focuses on the situation faced by some hubs with growing demand which meet their demand with power generated by distant sources subject to uncertainty. This is the case of a large city, which demand is satisfied by hydro or wind power supplied from sources located far away. For this kind of systems (typified as a simple two-node network that satisfies certain properties), we characterize the structure of a socially efficient transmission investment policy. It is shown that, in such a network, an optimal open-loop transmission investment policy has a multiperiod (s,S) structure. The results are illustrated with a numerical example.