Running head: REAL OPTIONS APPROACH TO SBIR PROGRAM 1 A Real Options Approach to Project Selection and its Application to NASA’s Small Business Innovation Program

We consider an important class of R&D investments in public and private entities that require selection from a large number of proposals such that: (1) the potential monetary value of the proposals is not a first-order factor in the selection; and (2) the investments are staggered in multiple stages, with relatively modest early support and higher subsequent funding for a subset of the projects initially selected. In this paper, we present a real options method to evaluate a portfolio of proposals in each stage. The proposals are characterized by a numerical value for each of several factors relevant to the awarding entity. Our methodology is flexible enough to consider additional factors or funding constraints. Its limitation is computational, depending on the number of dimensions (or factors) under consideration. Yet, numerical methods permit expansion of the dimensionality of the analysis. We apply this methodology to a time series of data from the Small Business Innovation Research (SBIR) program of NASA. A key factor in the assessment is the technology readiness level (TRL), estimated by the selection team for each project upon acceptance and evaluated again at the end of each stage. Furthermore, we demonstrate the flexibility of our methodology by proposing different specifications and tradeoffs, like the possibility to favor “microfirms” (fewer than 10 employees), shown to be critical for job creation; or increasing the number of proposals funded in the first stage. These findings illustrate the use of this model in management of public and private technology portfolios.