The crux of the matter is to identify factors with highest cause-and-effect relationship between the metric used and the industry it is used for. For example, in SIC 36, “Electrical equipment and supply,” EBIT in 2017 was ranked first with a 0.55 coefficient of variation (COV); pre-tax was second with a 0.52 COV. The COV ranking, however, is susceptible to context of the real financial values that are fed into the calculation.
A corroborative and perhaps a stronger predictor for the ranking is the correlation factor R-square. Correlation is strictly on the 1.0 to –1.0 spectrum; R-square is helpful to normalize the cause-and-effect among instances in the sample that have with different orders of magnitude (e.g., one company’s assets is in the thousands, versus another’s in the tens of millions). The evidence of this magnitude variation within the sample is evident with a COV >1.0.
The Authors Respond
We appreciate your interest in my article co-authored with Nicholas Mastracchio. We understand you have a valid point, and using the R-square may provide corroborative evidence or even a stronger predictor. This would be an interesting empirical exercise. However, the purpose of our inquiry was to focus on the methods that have been established by experts such as James Hitchner (author of Financial Valuations Application and Models) and Shannon Pratt and Alina Niculita (authors of Valuing a Business The Analysis and Appraisal of Closely Held Companies) to identify the best indicators under the guideline method as prescribed, and also to explore whether Bennett Stewart’s Best-Practice EVA (2013) could be an improvement over the guideline method. Thus, the scope of our article was limited to such comparisons, and we did not deviate from the methods we were examining. This allows readers to be able to follow the discussion with a basis of what is currently in practice, as far as Hitchner is concerned, and with the challenge that Stewart proffered in terms of using EVA instead of current practice. We hope this addresses your question.
