DP12946 Managerial Spillovers in Project Selection
Choosing a portfolio of projects to undertake is a fundamental managerial problem: from determining which units or divisions to establish within a firm, or which acquisitions or alliances to pursue, to which R&D, financial ventures, or marketing campaigns to greenlight. In this paper, we analyze the portfolio-selection problem given a budget constraint and featuring value spillover across projects. We distinguish between managerial spillover, due to the exploitation of common resources or real assets, and statistical spillover, when news about the
value of a project is informative about other projects. This distinction, largely overlooked in the literature, has tangible implications for managers. Statistical spillover is consistent with decentralized project assessment and undertaking provided there is informational integration—namely, as long as information flows freely across companies’ divisions. Managerial spillover requires that projects be undertaken within the same management structure and assessed in blocks: The combined savings from passing on two projects at once may outweigh their marginal contributions. We showcase these managerial implications in the cases of R&D with product development and of a company consisting of an HQ and two divisions.