A Methodology for R&D Portfolio Selection: Employing an Integrated DEA and BSC Model

We present a portfolio selection methodology that responds to three major goals that most R&D organizations are concerned with – effectiveness, efficiency, and balance. The methodology uses a strategy-oriented, multi-criterion, evaluation model that integrates concepts from data envelopment analysis (DEA) and balanced scorecard (BSC). It is based on two subsequent phases. At the first phase individual R&D projects are evaluated and project performance indexes are generated (i.e. project-efficiency, project-balance, and project-risk). By setting bounds to these indexes projects are screened and a subset of candidate R&D projects is generated. At the second phase potential portfolios are constructed using a branch-and-bound procedure that takes into account multiple resource constraints. Other possible restricting considerations, such as, the distribution of the length of the projects in the portfolio, may also be considered. The potential portfolios are then evaluated and an integrated portfolio performance index is computed. Based on this index the portfolios are ranked in order of decreasing rating, and a desirable portfolio is selected. Uncertainty is dealt with by adjusting output data to reflect expected values and by performing scenario analysis. The two-phase procedure reduces considerably the complexity involved in generating and evaluating all possible portfolios, and it considers both project and portfolio related factors.