RPM - Robust Portfolio Modeling

RPM-methodology

Robust Portfolio Modeling (RPM) is a decision support methodology for analyzing multiple criteria project portfolio problems. Research and development of RPM is carried out at the Systems Analysis Laboratory at Helsinki University of Technology.

RPM framework extends the use of Preference Programming methods into portfolio problems, where a subset of available project proposals are to be funded with available resources in recognition of multiple criteria. The overall value of each project is modeled through an additive scoring model as a weighted sum of the criterion-specific values, whereby the relative importance of the criteria is captured through weight coefficients (see, e.g., Multiple Criteria Decision Analysis eLearning Site by the Systems Analysis Laboratory). The overall value of a project portfolio is the sum of its constituent projects' values. In RPM, the decision maker can

• Provide incomplete information about the relative importance of the criteria (e.g., a rank-ordering or sensitivity analysis)
• Provide incomplete information about the projects' criterion-specific values (e.g., interval parameters)
• Simultaneously account for preference statements from multiple stakeholders (group decision making)
• Account for project interactions (e.g., synergies)
• Account for various logical and strategic constraints (e.g., mutually exclusive projects or balanced allocation of resources)
• Analyze the portfolio problem regard to variable budget levels and interval-valued project costs.

Towards this end, RPM builds upon the computation of the set of non-dominated (also, efficient) portfolios in the light of the incomplete information. Furthermore, novel performance measures help to analyze the attractiveness and robustness of portfolios and individual project prosals.

The approach provides a systematic and transparent decision support framework that can have a valuable contribution to project portfolio selection especially when the number of proposals is high. In particular, the RPM approach provides:

• Simple and widely applicable scoring model for project evaluation
• Proactive analysis of parameter uncertainties - robustness
• Identification of projects that should incontestably be included in or excluded from the portfolio
• Help to focus further analysis and/or negotiation on border line projects
• Transparency w.r.t. individual projects through performance measures
• Tentative/partial conclusions at any stage of the selection process

RPM approach has been developed and applied in several research projects. A detailed descriptions of the approach and its use can be found in the papers, theses and presentations in the publications section. The approach is implemented in the RPM-Decisions decision support software, which also provides a variety of functions for graphical interactive decision support.