From time to time, I disagree with some standard banking analytical practice or other. Ok, quite a few standard banking analytical practices. But there is one I fully agree with. That, in a comparable company analysis, the subject company should be excluded from averages.
Agreeing with an approach doesn’t mean I can’t find something to be curmudgeonly about however. And in this instance, too often people don’t know why they are removing the subject company from the average—it’s just “the thing that is done”. But the reason is actually pretty interesting (for some definition of “interesting”).
In a comps analysis, the subject company’s data is excluded from summary statistics because the pertinent piece of information—the answer to the question you’re actually asking—is the difference between the subject company and the average. Including the subject company in this calculation doesn’t provide any additional information, rather it dilutes the difference and slightly obscures the analysis.
Why do I care if people don’t understand why they are doing what they are doing? Because any opportunity to feel intellectually superior is an opportunity I will take? Not entirely. It’s because the rule of excluding the subject company cannot be universally applied, so you kind of need to know when it makes sense and when it results in an inaccurate analysis.
For example, I’ve seen it poorly applied in the estimation of the beta of a publicly listed company. An analyst derived equity betas for a subject and its comps, unlevered them, averaged the comps’ values, excluding the subject, and then relevered that average. This number was then used as the basis of a discounted cash flow valuation of the subject company. Which is to say, the analyst excluded the single most relevant piece of information (the subject’s observed beta) from the calculation!
If the purpose of your analysis is to estimate some not-directly-observable quantity relating to a subject company, as is the case in the beta example, then the subject company should be included in the list of similar companies used in the analysis. If you’re looking for a comparison, however—that is how is a company compared to its peers—you need to exclude the subject company from the average to throw the difference into starkest relief.
Or put more simply:
- To contextualize a number, exclude the subject company
- To estimate a number, include the subject company
Are there types of analysis that really get your goat? Or that you’d like further information on? Email me at email@example.com.
Pellucid is created specifically for investment banking pitchbooks and analysis. Request a demo.