The fact that appraisals can be so easily carried out on spreadsheets or by other computer models means that it is possible to look at outcomes based on endless arrangements of input parameter settings. This is fine, but can lead to the over-abundance of output information causing the truly sensitive parameters to be missed. Although simple in concept and with the limitation that only one parameter is changed at a time, the 'one at a time' sensitivity analysis reveals the sensitive parameters and indicates the materiality of changes in input parameters.
The growing tendency to rely on spreadsheet and similar calculations may mean that obvious business risks and sensitive parameters are overlooked. Basic arithmetical truths, such as the fact that the largest figures will be most sensitive, should not be forgotten. It is important not to rely solely on an arithmetical analysis to identify sensitive and thus potentially risky parameters. Sensitivity analysis should really only be seen as a back-up of an individual's understanding of the risks inherent in the business proposition.
The need for the most likely case
Since sensitivity to change in inputs is being reviewed, there must be a base position. For the analysis to reveal and quantify realistic sensitivities, there should be a 'most likely case' - neither optimistic nor pessimistic - base model of the project. The simple but effective approach of carrying out a 'one at a time' sensitivity analysis may then be adopted.
As the aim of the sensitivity analysis is to identify and thus quantify the parameters that are most sensitive to change (in an adverse direction), there is no point in carrying out the analysis on figures that are either pessimistic or optimistic. It is the most likely situation that is required. This demand for the most likely situation may also help in focusing on the amount of effort that goes into estimating.