COST-VOLUME-PROFIT ANALYSIS UNDER UNCERTAINTY

Cost-volume-profit analysis is widely used as a means for enabling management to decide whether to make or buy, to continue or discontinue a particular product, to increase production of a product, to introduce new lines, and so on. In the past, an undesirable feature of C-V-P analysis was the assumption that demand and other quantities, such as price and variable and fixed expenses, were known with certainty. The first steps to introduce uncertainty into the analysis were taken by Bierman1 and Jaedicke and Robichek.2 The assumption in both works was that sales volume followed a normal distribution. Bierman retained the assumption that price and variable expenses were constant, although Jaedicke and Robichek extended the analysis to include price and variable expenses as random variables with normal distributions. More recently, Johnson and Simik3 introduced correlations between product demands in a multiproduct situation, while assuming constant prices and expenses. On the other hand, all three sets of authors cited above did not address the problem of the derivation of values for the expectations, variances, and covariances of the probability distributions of future sales demands. Jaedicke and Robichek briefly suggested that such values might be obtained from historic data, or that possibly values might be given subjectively. In the following sections of this paper, the reliability of using the obvious