Elvis Presley and Kurt Cobain; both were entertainers loved in their own time, both were musicians with unique talent, both had fatal drug addictions. Yet if one were to compare the talent, substance, and success of the two artists and attempt to make an objective decision about “who was better”, one would quickly find that the attempt is futile. Both musicians were successful in their own unique way, appealing to 2 widely different audiences. Honky Tonk != Grunge Metal, not on earth, not on any planet.
And yet this is exactly the types of comparisons we get with securities analysis. If you have had a chance to get on alpha.forcastix.com (now requires an invitation, sorry if you missed out when it was open!) you would have noticed that we attempt to point out peers, or companies that do similar stuff. What you surely found was that many of the “peers” really did not have anything to do with each other.
Take for example Google. Their business is relatively simple…advertising, software as a service. On Forcastix, as on many securities analysis sites, Google is in the Technology sector and the Computer Services industry. And that’s about as granular as it gets. Guess who else is in the Computer Services industry?
– RCM Technologies, an IT staffing company
– Novatell Wireless, wireless broadband provider
– Wipro, an IT staffing, computing hardware, home appliances, and baby care products company
Now really, how is an internet advertising company such as Google anything like an IT staffing/appliance/baby care conglomerate such as Wipro? For years analysts have been devising ways of properly classifying companies so that it reflects their actual business. A few of the attempts include the SICs, NAICs, GICS (S&Ps Brainchild), ISIC, OKVED, TRBC (the one we use), ICB, herp derp herp derp herp herp derp…
If you are still awake, I will give you one simple reason why every single one of these classification systems fails:
A company is defined, not by what it sells, but who are its buyers.
Therefore, companies who share a pool of buyers, who are buying the same or similar product and service, are competing. Companies who maybe have similar products and services, but whose buyers would never have to compare one from the other to determine the best choice, are not competing. It is that simple. And yet over and over again, groups of thinkers try to plug companies into simple models of categorization that do not work. Out of all the places I have searched to collect accurate classification data, only Revere does it right.
