Sensitivity analysis consists of studying the effects of changes in variables on the outcomes of a mathematical model. Furthermore, a model may consist of numerous input variables and one or more output variables. By changing an input variable, and measuring how the outcomes are affected by that change, the analyst can gauge how sensitive the model is to the individual input variable.
Use sensitivity analysis in business decision-making. It is a way of measuring and quantifying uncertainty. The analyst can create a model based on the relationships between inputs and outputs. Once the model is set up, the analyst can tweak the inputs to see how the outputs are affected. The analyst can also alter the model to create hypothetical scenarios such as a best case scenario, a worst case scenario, and a most likely scenario.
For example, an analyst might use sensitivity analysis to measure a project’s net present value (NPV) for various expectations of costs, revenues, capital investment, macroeconomic factors, and other relevant variables.
First, the accuracy of the sensitivity analysis depends on the quality of the assumptions built into the model. If the model contains erroneous assumptions, then the output of the sensitivity analysis will be inaccurate. Second, sensitivity analysis may not account for interdependencies among input variables. Finally, the assumptions built into the model may be based on historical data. Therefore, it cannot necessarily be relied upon to predict future results. Also, subjectivity may taint the analysis.