This past week, we spent some time assessing Microsoft’s moat. Though its Windows operating system and its Office suite are cheap to reproduce, easy to transfer and store, and require only a modest sales force, Warren Buffett acknowledged that a paradigm shift in communications could quickly undermine Microsoft’s position. Today Microsoft’s moat may be wide and deep, but consumer preferences may change rapidly in the next decade.

If Buffett is right, and consumer preferences could change in communications more quickly than their preferences in other areas (e.g., carbonated beverages, razors), then the prospective investor should be more conservative in her valuation of Microsoft, or perhaps, demand a greater margin of safety.

One thing is clear; Microsoft’s margins show that they offer a set of products that grant significant pricing power. Even after lumping all of their marginal products together with their cash cows Windows and Office, Microsoft has averaged over 80% gross margins on sales for the last five years (2004-2008), and a 34% operating margin over that same span. These margin numbers even best those of eBay, one of our favorite wide moat companies.

Yet, even if Microsoft’s future is less certain than Coca Cola’s, what do we think the company is worth today? Following Seth Klarman’s recommendation, it makes good sense to value Microsoft by summing its book value with the net present value of its future cash flows. Many NPV analyses estimate a company’s earnings power over the course of the next two decades. But if we heed Buffett’s warning about a potential paradigm shift, perhaps we should dial back our estimates of future cash flows. Thus, in this valuation, we will only project Microsoft’s future cash flows over the next decade.

So let’s get to work. Using Morningstar’s data, we see that Microsoft generated 18.43 billion in free cash flow in 2008. Let’s conservatively project that Microsoft will grow its free cash flow at 5% over the next ten years (which may seem low, but which is rather close to their FCF growth over the past decade). Using these estimates, the net present value of the next decade of Microsoft’s cash flows, discounted at 15%, should be 115.6 billion. Add the 36.3 billion of Microsoft’s book value, and we get a total value of 151.9 billion. In other words, Microsoft should throw off 243.4 billion in cash over the next decade. But we wouldn’t pay that much solely to have it to trickle back to us over that decade; instead, we want a decent return on your capital outlay—say 15%. Thus, we should be willing to pay 151.9 billion today for 243.4 billion in cash outflows over the next decade. Per share of Microsoft, that amounts to a price of $17.08. With a 25% margin of safety, we should be willing to buy Microsoft today for $12.81—significantly below Friday’s closing price of $18.

Now, I can anticipate some of the objections to this meager valuation for what is today a market leader and cash cow. It is unlikely that Microsoft will have no earnings power after a decade; it is unlikely that a paradigm shift in communications will occur. It may be improbable that cash flows will only grow at 5%.

Leaving such objections aside for now, it is crucial that we highlight what looks to be the most important lesson in this analysis*. If we cannot predict what a company will look like in twenty years, our valuation of it and its future earnings should be far lower than we typically expect.* And here we see that for an investor like Warren Buffett, what looks today like a wide moat may not be sufficiently wide if we cannot reliably foresee its likely earnings in 2029.

Disclosure: I, or persons whose accounts I manage, own shares of Berkshire Hathaway and eBay at the time of this writing.

Hello,

I like your analysis. However, could you please let me know where you got your figures? You said that 5% is rather close to their FCF growth over the past 10 years, yet both Valueline and Gurufocus have Microsoft’s FCF growth over the past 10 years at close to 10%. Furthermore, FCF growth has been accelerating, with 18% growth per annum over the past 5 years. Even on a conservative basis, don’t you find that 5% is very low?

Also, have you done an FCFE valuation of the business ( 2 stage Damodaran Model)? I would be interested in hearing your figures for this. Using this model, with a terminal value of $80 billion, and your figures (5% growth), I get $25 per share. Even without any change in revenue, I get $17 per share. This would lead me to believe that a very conservative valuation for the company lies somewhere in this range, and closer to the $25, if I were to make an educated guess. What are your thoughts on this? Thanks!

Disclosure: I do not own any shares in MSFT

Thanks for your questions. I’ll give a more detailed response tomorrow. But two quick answers. First, I compute MSFT’s FCF growth from 1999-2008 to be 6.91%, which looks pretty close to gurufocus’ numbers (7.4%).

And second, I have done a two stage model for Microsoft, but here I am most interested in hypothetically positing Microsoft’s valuation as if a “paradigm shift” in communications would occur after ten years, rendering Windows and Office virtually obsolete. Under this hypothesis, Damodaran’s “second stage” of “stable earnings” never materializes, and so those earnings should not be added to our valuation.

My big point was that it is not sufficient for companies to have a wide moat today, but that they must also have a high likelihood of being in a dominant market position in twenty years. Buffett doesn’t invest in “technology” because he finds it difficult to judge that likelihood.

Thoughts?

Thank you for the quick response. You make some very good points. The gurufocus numbers are indeed different from the valueline numbers. I wonder why that is? Do you use an arithmetic or geographic mean?

What would you think of using an adjusted value model? One where:

adjusted value = Discounted cash flow value x (1- probability of distress) + Distressed sale value x (Probability of distress)

Obviously measuring the likelihood of distress (and the distressed value) is extremely difficult; just thought I’d throw it out there. Also, when you mention the “paradigm shift”, are you speaking about Google’s cloud computing? Has Microsoft been developing products in this area? What effect do you think this has on the analysis?

Finally, I must admit that my experience in terms of finding moats is quite limited. This is why I find your blog to be so interesting: I’m always looking to recognize and fill gaps in my investing acumen. Are there any books, articles, or resources you would recommend with regards to moats (aside from Buffet’s shareholder letters, and Munger’s speeches?

Thanks, I look forward to sharing more ideas,

Eric

My 6.91% figure is a geometric mean. I also have done geometric means for a variety of 5, 6, 7 year spans through the last decade. If I then use an arithmetic mean for these sub-spans, then I get a lower FCF growth figure, closer to 4%.

In terms of mathematical form, I think your adjusted value model is spot on. But, as you note, the probability is the crucial unknown. As an amateur, and thinking at this minute, I would estimate the likelihood of a paradigm shift in personal computing within ten years at 5%-20%. Thoughts?

Yes, “cloud computing” or open source software are two examples that quickly come to mind. I am not an expert on this though, and I don’t know what Microsoft has been working on in this area.

And finally, I do wholeheartedly agree that deep, real, and enduring moats are rare. And I’m still not sure what the proper criteria should be for defining them. Others point to Porter’s five forces and Dorsey’s The Little Book that Builds Wealth as important resources. I’ll get around to reviewing these eventually.

Thanks for your interest and questions. WM.

Pingback: Buffett, Technology, and Moats « Wide Moat Investing

Pingback: Microsoft–What’s It Worth? Re-examined « Wide Moat Investing

Pingback: David Einhorn and Return on Equity « Wide Moat Investing