Reading: Coping with Measurement Error

The sensitivity analysis tests by how much the valuation changes when we change one variable at a time and keep all other variable at their expected mean value. Again, this analysis can be implemented fairly easily in Excel. All we have to do is change the cell containing the variable in question and note the valuation outcome.

In the context of our four potential value drivers, we could first change the expected costs of sales during the explicit forecast period and keep all other variables at their mean expected value. The following table summarizes the results:

Because the explicit forecast period is only 2 years, the specific estimate of the cost structure during that period does not appear to be a main value driver. Moving from the best case to the worst case reduces firm value by only 1'167. This corresponds to a sensitivity of 2% of the expected value [= 1'167/57'031].

We can proceed accordingly with the additional potential value drivers. The results are reported below:

The table shows that, based on our assumptions, the firm value exhibits the strongest sensitivity to measurement errors in the estimation of the equity beta. More specifically, the valuation range that results when beta moves from the worst case to the expected case corresponds toroughly 25% of the expected value. The sensitivities of the other variables are less pronounced (10% in the case of the credit spread and 8% in the case of the long-term EBIT margin).

The sensitivity analysis provides valuable complementary information to the scenario analysis. In particular, it helps us identify the relevant value drivers. Still the picture is not complete. The most important missing piece of information is how likely the various outcomes are. This question can best be addressed with a simulation.