If Amazon.com (AMZN) isn’t the best company in this generation, then I think it has to be in the top few. By best, I mean the ability to create and build and entirely new business and to continue to build on it long after the initial concept (selling books over the internet; does anyone remember when that was what defined AMZN) has left its novelty in the dust. But great companies and great stocks don’t necessarily go hand in hand.

P/E – Dispensing With The Trite

I don’t care that Amazon has a triple-digit P/E. If the “drivers” of P/E are sufficient, a stock can be undervalued even with a quadruple digit P/E. And for the record, ideal P/E is 1 divided by the difference between the required rate of return (influenced by interest rates and risk) and the expected future earnings growth rate.

Actually, with AMZN’s still pursuing its spend-to-build approach to life, it’s hard to make much sense of anything relating to earnings except to say (i) it remains hard to credibly forecast growth because we still don’t have a good sense of what the business will look like in, say, 2025-30, and (ii) despite CEO Bezos’ revenue-oriented rhetoric, the company is no longer suffering from bottom-line deficiencies, as seen in Table 1.

Table 1

Ignore what Bezos says about his lack of obsession with margins, which are relevant but not in isolation. Profitability is about the combination of margin, turnover and leverage. In this overall regard, measured by the metrics in Table 1, while the company isn’t quit up to the S&P’s level of profitability, it has made considerable progress lately as it started to generate revenues and profits from its considerable recent investments.

While we can’t use P/E to value AMZN because we still can’t credibly project earnings growth, we have to do something because there’s no doubt AMZN is a solid, legitimate, and yes, even profitable, company.

Price/Sales

There’s no law that says we have to value AMZN, or anything for that matter, on the basis of earnings. The only perfect theoretical metrics for valuation are based on cash expected to pass directly into the hands of investors, which are dividends and eventual proceeds from selling; i.e., the iconic dividend discount model (DDM) which values stocks based on dividend divided by the difference between required return and expected future dividend growth rate. Earnings are used because for reasons that should be obvious to many, strict application of DDM is impractical in the real world. In fact, the P/E formulation I suggested above it just an algebraic reshuffling of DDM after substituting E for D, something we see as a practical market reality all the time.