Catalog

	Table of Contents
p. 33 of 74 BACK | NEXT I. Introduction II. Background III. The Economic Rationale of A Project IV. Macroeconomic and Sectoral Context V. An Integrated Approach To Economic Analysis VI. Identification and Quantification of Costs and Benefits VII. Valuation of Economic Costs and Benefits VIII. Large Projects, Linkages, and National Affordability IX. Least-Cost and Cost-Effective Analysis X. Investment Criteria: Economic Viability XI. Discount Rate XII. Uncertainty: Sensitivity and Risk Analysis XIII. Sustainability of Project Effects XIV. Distribution of Project Effects XV. Projects and Policies XVI. Appendices XVII. Others

Guidelines for the Economic Analysis of Projects

---

XII. Uncertainty : Sensitivity and Risk Analysis

148. The EIRR or ENPV for Bank projects is calculated using the most likely values of the variables incorporated in the cost and benefit streams. Future values are difficult to predict and there will always be some uncertainty about the project results. The effects of different values should be investigated. For directly productive projects, this means assessing the effect of possible changes on the ENPV or EIRR and, hence, on the project decision. For indirectly productive projects, this means assessing the effects of possible changes on a basic project parameter, such as the unit cost of service provision.

149. Sensitivity analysis is a simple technique to assess the effects of adverse changes on a project. It involves changing the value of one or more selected variables and calculating the resulting change in the NPV or IRR. The extent of change in the selected variable to test can be derived from postevaluation and other studies of similar projects. Changes in variables can be assessed one at a time to identify the key variables. Possible combinations can also be assessed. Sensitivity analysis should be applied to project items that are numerically large or for which there is considerable uncertainty. To facilitate mitigating action, variation should be applied separately to underlying variables, such as areas and yields in agricultural projects, and not just to aggregate values. The effects of variation in the basic parameters for shadow price analysis, the SERF or standard conversion factor, and the SWRF, should also be assessed.

150. The results of the sensitivity analysis should be summarized, where possible, in a sensitivity indicator and in a switching value. A sensitivity indicator compares the percentage change in a variable with the percentage change in a measure of project worth. The preferred measure is the ENPV. A switching value identifies the percentage change in a variable for the ENPV to become zero, the EIRR to fall to the cut-off rate, and the project decision to change. Where percentage changes in the variable cannot be measured, for example, for delays, simply the percentage change in the ENPV can be presented.

151. Where the project is shown to be sensitive to the value of a variable that is uncertain, mitigating actions should be considered. This can include project level actions, such as long-term supply contracts or pilot phases; sector level actions, such as price changes or technical assistance programs; or national level actions, such as changes in tax and incentive policies. Where there is exceptional uncertainty, the project may have to be redesigned or implemented first on a pilot basis.

152. Quantitative risk analysis associates a probability of occurring with different values of key variables. When such variables are varied simultaneously through a random selection of outcomes, a frequency distribution for the ENPV or EIRR can be produced showing the probability that the project is not acceptable. Decision makers will compare the scale of net benefits from different projects with their riskiness in selecting an individual project or a portfolio of projects. Quantitative risk analysis can be carried out for large and marginal projects, or projects where there is considerable uncertainty about a key variable, such as the world price for the project output.

153. Sensitivity and risk analysis can be used to assess the effects of changes in project variables that are quantified. The results can be presented together with recommendations on what actions to take or which variables to monitor during implementation and operation. However, many projects involve institutional and social risks that cannot be readily quantified. A statement of such risks and any mitigating actions should be included alongside the conclusions from the sensitivity and risk analysis (see Appendix 21).

Back XI. Discount Rate	Next XIII. Sustainability of Project Effects