Refining the use of Monte Carlo techniques for risk analysis in project planning

Monte Carlo approaches to risk analysis in project appraisal are re-examined. Limitations with conventional methods are identified, and refinements suggested that deal with increasing uncertainty about variables throughout the project life, along with correlations and cycles among variables. These are illustrated with an example. Emphasis is placed on a practical approach that minimises demands on the appraiser's prior knowledge. Modelling the objective function of an investment appraisal autoregressively enables complete distributional mapping of project outcome, given specification by the appraiser of 'likely bounds' for the value of key parameters in the first and last periods of the project life plus an approximate correlation matrix.