In a post last week I talked about the power of disruptive technologies and how they can kill your margin of safety in an investment. This is especially important in valuing technology companies like Microsoft. Aside from the obvious technological obsolescence risks in this sector, there are three additional big risks that I think are important in valuing Microsoft:
1) Business model changes: Will a cloud computing world change Microsoft’s business model?
2) The growing irrelevance of the PC: Is the demise of the PC and the Windows O/S overrated?
3) Accounting shenanigans: Were Microsoft’s revenue deferral practices in the last 1990′s really “cookie jar” accounting or a political move by the SEC after the Enron blow up?
I’ll spend most of my time on the first two in this post. But before I discuss these, it’s worth repeating a quote an engineer friend of mine recently told me while we were debating technology investments:
“People overestimate the pace of change and underestimate the scale of change.”
1. Business Model Changes
Microsoft has 5 divisions and each will be affected differently by the big changes in computing that the Cloud will bring. These divisions are effectively:
- Windows (Windows and Windows Live division)
- Office (Business division)
- XBox and Windows Phone (Entertainment and Devices division)
- SQL Server (Server and Tools division)
- Bing (Online Services division)
Other than XBox, all of these divisions primarily produce software. In a traditional software licensing model, the company bares the risk of spending money up front to develop software but then every license it sells is effectively pure profit. So for a giant like Microsoft, this model offers juicy profit margins (operating margins > 40%) that only increase as the company sells more licenses. Currently Microsoft has over 450mln Windows licensees and 100mln Office licensees. This model is similar to that of my favorite company, Paychex, for whom every new payroll check processed at a given client is pure profit.
But in the new cloud computing (or software-as-a-service) model, the business economics change in two important ways. Customers will now pay per use rather than pay one lump sum every time they buy/upgrade a license. The positive change is that revenues become recurring every month, like a support contract. As a result, one can even argue that cloud computing makes customers stickier because once their data is on your system, they will keep going there (highly debatable though). More importantly though, I believe profit margins will drop with the cloud model. In a pay-per-use model, revenues and costs should roughly scale together (beyond a certain point). As more users login and use, for example, Office 365 on the cloud, this requires more computing power, more storage space, more bandwidth, more support personnel, etc… Therefore, each new sale is no longer pure profit. This change in business model is important to consider in valuing Microsoft.
2. Irrelevance of the PC
As for the death of the PC, it is not breaking news that investors think the iPad/tablets and smartphones are taking over PC’s and laptops as the computing devices of choice. As a result of this belief, Apple is a market darling and Microsoft shares have been flat for about a decade. This is in spite of the fact that during the last decade, Microsoft has tripled earnings and returned over half their current value in cash to shareholders via dividends and share buybacks.
While Microsoft has been dominant in PC operating systems and business software for the last two decades, the computing world is changing and this may change their Windows business model. It’s hard to speculate on how important a Windows O/S will be on desktop computers in a cloud computing world, the O/S may not matter as much. And it’s equally difficult to speculate on whether Windows Phone will make significant in-roads into the mobile phone market. Just like with PC’s, it’s the apps that determine the popularity of an O/S and the question then becomes are there enough developers to support a third major phone O/S in addition to Android and iPhone.
Keep in mind the quote above. Investors are likely over-discounting the demise of the PC. As of today, it is still the primary tool for productivity in businesses around the world and PC use in the emerging markets is only growing.
3. Valuation
Incorporating the above thoughts into my valuation, let’s see where I come out on Microsoft.
Today, their total market value (net of net cash position) is about $175 bln. This reflects a forward earnings yield of almost 15% (P/E of <7) for a company with a history of very consistent sales and earnings growth of 7-10% over the last 10 years. And because the software licensing model offers exceptional economics, they have produced incremental ROIC’s over 40% for their entire history. Simply extrapolating this past into the future conservatively gives a valuation of closer to $270 bln. But there has to be more as to why investing greats like John Rogers at Ariel Investments and Jeremy Grantham at GMO hold Microsoft as their largest position in their focused/quality funds.
Let’s take a look at each division separately, starting with Office since it is the largest in terms of sales and profits. I believe Office is the most valuable franchise that Microsoft has. There is really nothing on the market that can compete, especially in the enterprise market. Google is trying with Google Apps and OpenOffice is an open source alternative but I have tried both and the reviews confirm they are low-end competitors at best. So as business use of computers in the emerging world continues to grow, I expect Office sales will grow with it. Therefore, for Office/Business division, my valuation comes out to $100-130 bln with the expectation that if I’m right that the software-as-a-service model is less profitable that the value is on the lower end of this range.
Assuming a wide range of outcomes based on the dynamics discussed above surrounding PC’s and Windows, I get a wide range between $25-80 bln for the Windows division. For the sake of assessing our margin of safety though, let’s assume it’s value is zero. We know that investors are likely to over-discount the demise of the PC making this assumption absolutely wrong but adequately conservative.
The Server and Tools division is a bright spot that few ever comment on. The division, is, like Windows and Office, highly profitable, and also growing fast with the most recent quarterly numbers showing year-over-year sales growth of 10%. Here I definitely don’t understand whether the cloud will disrupt their existing business model or actually help it prosper even more. But like Office, I believe this division has a big competitive advantage and I see tools like SQL used by businesses everywhere. For this division, my valuation is conservatively around $45-65 bln.
Finally there is XBox and the Entertainment division (I’m counting Online Services as a Corporate expense because they are losing money for now). XBox has become the dominant gaming system and with XBox Live is transitioning to a full on online entertainment console. Another division with a strong brand and a lot of potential. But again, without assuming any speculative changes and market share gains, I get a valuation for this division of $10-25 bln.
When you look at each of the pieces, what you see is a group of market leading products. The market concerns appear to be primarily about the future of Windows and the future of the PC and this uncertainty is reflected in the extremely low stock price. Even the most conservative sum of the parts valuation, ignoring Windows, comes out to $155 bln (or almost $24/share after adding net cash back), which excludes the value of the Windows division. There is a solid margin of safety here.
I still need to dig into their accounting policies (specifically revenue recognition and use of deferred income accounts) to ensure I am not missing anything major. Earnings have been extremely smooth including very little decline in 2009, which doesn’t seem realistic.
So far, Microsoft is looking like a good buy at these levels.