|
|
|
|
|
|
|
The most appropriate methodologies vary with the nature of the problem under review, but any method chosen should |
|
|
|
|
|
|
|
|
help explore soundness of assumptions |
|
|
|
|
|
|
|
|
promote development and assessment of different options |
|
|
|
|
|
|
|
|
make explicit implications of each option |
|
|
|
|
|
|
|
|
highlight sensitivities of each element |
|
|
|
|
|
|
|
|
enable formulation of clear recommendations. |
|
|
|
|
|
|
|
|
Then, this robust analysis is used to underpin informed expert debate. |
|
|
|
|
|
|
|
|
There are a huge number of approaches to help give a better understanding of the relationship between the key variables. These include |
|
|
|
|
|
|
|
|
influence diagrams and decision trees |
|
|
|
|
|
|
|
|
Each of these techniques takes a structured approach to understanding risk and reward and leads to a consistent view of the likely value of each project to allow a quantitative assessment of its contribution to the total portfolio. It is fundamentally important to understand and manage the risk in the portfolio. |
|
|
|
|
|
|
|
|
Influence diagrams are a visual tool to help understanding the factors which affect the ultimate value of the project; they are a pictorial means of representing a decision tree. Initially, they are used by the project team to ensure a shared understanding of the key issues and how they relate to each other. A simple example is shown in Figure 4. |
|
|
|
|
|
|
|
|
Each of the important uncertainties, then, is quantified using expert judgment to evaluate the overall impact of risk on return. |
|
|
|
|
|
|
|
|
Having determined the various factors and influences as outlined in Section VI.A, decisions are made on which courses of action to pursue. Decision trees enable the modeling of risk situations. They are used to determine and display the results of an analysis of probabilities. Each decision may have several possible outcomes with different probabilities of occurrence. Note that the probabilities are multiplicative. Where a number of events are independent, the probability that they all occur is the product of their |
|
|
|
|
|