For our purposes, there are three interesting points: first, Feinseth is skeptical of net income because of the twin perils of earnings management and earnings guidance; second, Feinseth focuses on economic profit, charging management for the cost of capital employed in the business; and third, he has his own ideas for how Microsoft might use its cash.
Net income? Sure, his position makes sense to us. Net Income is an accounting measure and, while perhaps earnings are higher quality today than in years past (accounting for some portion of P/E ratio creep), they’re still too easily massaged, if not managed or downright fudged.
Economic profit? Ever since a consulting firm that shall remain nameless trademarked “EVA” for the clever turn of phrase “Economic Value Added,” the concept should have garnered much wider respect than it has. We have undertaken dozens of analyses covering hundreds of companies, and the analyst community (and by extension, investors) still seem to ignore the issue of invested capital, at least in terms of discussing it as a material part of their analyses. At the same time, top-line growth, even through mergers and acquisitions, still gets lots of often-undeserved attention.
By comparing Microsoft’s return on capital (~39%) with cost of borrowing (~x%) and return on cash (~x%), he reaches the conclusion that not only should the company buy back shares (see our earlier post on using of excess cash) but that Microsoft should probably even borrow money to buy back shares. This second prescription sounds funny if you read it in a vacuum, but in light of our capital structure discussion, what he’s suggesting could make perfect sense.
It’s always refreshing to hear sensible, consistent advice coming from an equity analyst.
[Update: of course, most people would today look at a past MSFT decision to borrow money at a low rate to buy back shares as a big mistake. That might be true, but we can’t just evaluate decisions in hindsight. What’s important, in some ways more than ending up with the right answer, is understanding the process at which a decision should be arrived. We’ve talked about decisiveness, and that quality isn’t in conflict with this goal. Analysis paralysis, which is typically driven by fear of making a mistake, is often as bad a problem as making a good decision that led to a poor outcome. Economists call this expected value. We naturally work to maximize expected value because that’s all we can actually know: the range of possible outcomes and their distribution. We can improve that knowledge, by making sure, for example, that we include ‘Black Swan’ events in our calculations, but we still can’t know what the actual result will be. Much like a quantum equation, Schrodinger’s mutual fund turns out to have exactly one rate of return for us, and it’s hopefully in the universe we predicted, but all of the other possibilities fall away. Quantum investing, the idea of dealing with probabilities rather than certainties, is just a fancy name for what good leaders call risk “management.”]