We study industries where prices are not limited to their allocative and distributive roles, but also serve as an investment into lower costs or higher demand. While our model focuses on learning-by-doing and the cost advantage that it implies, our conclusions also apply to industries driven by network externalities. Existing literature does not have a clear verdict on whether the investment role of prices benefits or hurts the overall welfare, as there are a number of economic forces at work, e.g. motivation to move down the learning curve faster could be offset by the ease of driving a weaker rival out of the market. We compute both market equilibrium and first-best solution. The resulting deadweight loss appears small, in the sense that eliminating the investment motive from pricing decisions leads to much worse outcomes. Further investigation into components of deadweight loss shows that while pricing distortions are the most important driver of the deadweight loss, these distortions can be fairly small. Entry-exit distortions that arise from duplicated set-up and fixed opportunity costs also contribute to the deadweight loss, but these distortions are partially offset by more beneficial industry structure, as the market equilibrium tends to result in more active firms than the first-best solution.