A proxy is a substitute for something. In formal voting procedures, I can give you my proxy, meaning that you substitute and act (vote) for me.
We use proxies all the time in business valuation. The Direct Market Data Method is based on the Principle of Substitution, which holds that the sale price multiples of guideline companies are indicative of the price at which a subject company will sell. We use the guidelines as proxies for our subject, with adjustments for various facts. We use Ibbotson rate of return data (based on public companies) as proxies for rates of return on private equity investments. We use control premium data as proxies for the value of control of a private company. We use all sorts of proxy data to estimate lack of marketability discounts (if we benchmark them, or even if we use Chris Mercer’s Quantitative Marketability Discount Model). You can probably find proxies in just about every aspect of valuation analysis.
The thing to remember about proxies in business appraisal is that they are not perfect. Relative to the above examples, comparable companies may be different from our subject or too few to create statistical confidence, Ibbotson data (particularly the size and industry premiums) are not perfect measurements of those considerations, and the control premium and marketability data have well-documented inaccuracies and dissimilarities when compared to private companies.
We use proxies because they are, from a practical point of view, the best available, although imperfect, measurements of things we cannot directly quantify. We hope or assume they are sufficiently reasonable substitutes, warts and all, to get the job done. If we had perfectly comparable, accurate data on all the things I mentioned in the preceding two paragraphs, then business appraisal would be like pricing heavily traded publicly traded stocks: you would look up the price on the Internet and be done.
Some authors have field days writing about the weaknesses of proxies. They are almost always absolutely correct in everything they state. Proxies are imperfect! But that begs the question: What do they propose as an alternative, why is that better, and can they prove it to the satisfaction of others?
Except for the (admittedly selfish and therefore minor) facts that it would wipe out my job, professional accomplishments, income, and blog, I am all for the development of techniques and procedures that would allow me not to use proxies. Until those come along, however, I will continue to use every available proxy that makes sense to me in a given situation, doing the best I can with the limited, imperfect information that they provide. That is the art of business valuation!