Using portfolio management and optimization techniques to allocate capital and select the best investment opportunities has been the subject of much discussion in recent papers and articles. However, there are still many practical problems associated with successful portfolio management within an organization. One of the biggest challenges is balancing the trade-off between modeling strategic-level issues versus capital allocation at the sub-strategic level. Two approaches are common. One is to take a high-level view and capture your portfolio at the business unit level. This approach enables clear strategic insights but is limited if business units are composed of several different types of assets contributing variously to the portfolio. A second approach is to start from the bottom up, with each project and sub-project captured in detail. This, however, is time-consuming and can produce confusing results, which are difficult to translate to meaningful capital allocation decisions.
This paper describes a third approach: projects were captured at a detail level appropriate to 1) the goals of the organization, 2) project size relative to the portfolio, 3) project range of uncertainty, and 4) the decision-making level for capital allocation. Also, similar projects were grouped together and a "typical" project chosen to represent the group – using surrogates to reflect the group population. The paper discusses the importance of developing an understanding of key portfolio drivers and adequately characterizing the range of diverse investment opportunities before proceeding with data gathering. This has enabled much greater clarity in answering the often-asked question "Did the portfolio like ‘it’?", and has also established thresholds for material portfolio impact for different sizes and types of projects.
A further benefit is identifying the critical economic and financial metrics that drive the company portfolio and feeding these back to the project originators so that they can focus on improving the accuracy of those metrics in the project economics.