Microsoft's (Mis)Adventures in Hardware

Caught flat-footed during the surge in popularity of smartphones and tablets,  Microsoft (MSFT) was forced to play catch-up as it entered the hardware business, trying to maintain the network effects and switching costs that had bound consumers and enterprises to its products. Due to a combination of unforced errors and bad circumstances, Microsoft's hardware strategies have historically been unprofitable and have represented departures from the firm's usual strategy of spurring license and services sales, despite the strategic importance of each product. We believe recent changes in Microsoft's overall strategy could lead to the company reducing or ending its direct investment in hardware, without ending its larger strategy of expanding the Microsoft platform and productivity ecosystem.

Hardware, a Necessary Evil
In 2013, Microsoft sought to remake itself into a devices and services company, embarking on designing, manufacturing, and selling smartphones and tablets in addition to its Xbox gaming console. With new CEO Satya Nadella taking the helm earlier this year, the company's strategy shifted slightly to "mobile first, cloud first," although the previous "devices and services" strategy was already being executed. The abrupt change in leadership has resulted in several hardware projects being set in motion, which may not have been approved by Nadella. Nevertheless, he is now in charge, and he is, for better or worse, responsible for the success or failure of these projects.

Late to the Smartphone Revolution, Unprepared When It Finally Showed Up
Microsoft's recent smartphone problems have been well-documented. Since the introduction of the iPhone in June 2007, the Windows phone OS' market share has declined from the low-double-digit range to low-single digits, all while smartphone sales have grown at a 41% CAGR (between 2007 and 2013). Seven years ago, feature phones still dominated the market, and therefore the early iPhone sales gains weren't terribly worrisome to Microsoft's management. But as the popularity and adoption of smartphones accelerated when Android joined the fray in late 2008, Microsoft found itself increasingly left out of the party.

Microsoft's response was to finally update its aging mobile OS, Windows Mobile, which was used from 2000 to 2010, to a new mobile OS, Windows Phone. Rushed to the market in 2009, Windows Mobile 6.5 had a rocky start as a stopgap OS and was largely incompatible with the slightly more refined Windows Phone 7, released just a year later in 2010. Windows Phone 8, released in late 2012, used a new architecture that, once again, created compatibility issues with prior versions (Windows Phone 7 and 7.5). These stutter steps stunted development of an app ecosystem and adoption by original equipment manufacturers (OEMs) and users, while giving iOS and Android time to solidify their market share gains.

Because of those difficult early years, and the proven popularity of the Android platform, OEMs were less likely to build and promote Windows Phone as an OS. They were even less likely to do so because Microsoft charged a licensing fee for Windows Phone, while the Android OS was free. With a small user base, limited app ecosystem, and weak OEM adoption, the network effects Microsoft historically enjoyed in the PC market were nonexistent and were even working against Windows Phone, as market share and adoption numbers continued to plummet.

A $9.5 Billion Shotgun Wedding Binds Nokia and Microsoft for the Foreseeable Future
To jump-start its moribund smartphone fortunes and support the platform, in 2011 Microsoft inked an agreement for  Nokia (NOK) to build Windows Phones in exchange for quarterly payments of roughly $250 million through 2016. Fast forward two years, and Nokia was building an estimated 90% of all the Windows Phones on the market. The Nokia phones had made modest gains in nonsubsidized markets, while outpacing a struggling BlackBerry in the U.S. These were signs of progress, but Nokia's Windows Phone business--and its handset business in general--was struggling. As the sole manufacturer of a vast majority of Windows Phones, Nokia knew it had the leverage to convince Microsoft to buy its handset unit, lest Microsoft be shut out of the smartphone market once again. Nokia's negotiating position was bolstered by the alleged threat that it would move its smartphone efforts to the more popular Android OS.

The original purchase price for the acquisition was $7.2 billion, but the final purchase price for the Nokia Devices and Services (NDS) business ballooned to $9.5 billion due to acquired cash, foreign exchange rate movements, and working capital and other adjustments. In return for its $9.5 billion, Microsoft received a business that has seen declining revenues, stagnant gross margins, and operating losses. The historical results of the Nokia Devices and Services business have been a 61% overall decline in revenue and approximately a EUR 3.8 billion drop in profitability from 2010 to 2013.

With the recent trajectory of NDS, this business is now a challenge for Microsoft and will be a drag on revenue growth and profitability for at least the near future. In Microsoft's fourth quarter in fiscal 2014, NDS' results contributed for just over two months, as the acquisition closed on April 25, 2014. Phone hardware revenue for that period was $1.985 billion, with a gross profit of $54 million and operating expenses of $750 million, resulting in operating losses of approximately $696 million. During that period, NDS sold 5.8 million Lumia smartphones and 30.3 million feature phones, declines from year-ago results.

Using Nokia's results for the first half of its fiscal year, we estimate that net sales for NDS for the six-month period would have been roughly $4.54 billion in revenue and an operating loss of $845 million with annualized revenue of $9.1 billion and an annualized operating loss of roughly $1.7 billion. It is important to note that the two most recent quarters are typically the seasonally weakest quarters for handset sales, and therefore it is possible that we are underestimating the annualized revenue and overestimating the annualized operating losses for the business. Nevertheless, this exercise gives us a rough estimate of the cost synergies that Microsoft needs to find to achieve its goal of operating break-even in fiscal 2016.

Using these estimates, we calculate that NDS' recent double-digit revenue declines (negative 18% in fiscal 2011, negative 35% in fiscal 2012, and negative 29% in fiscal 2013) continued into fiscal 2014, while operating losses grew significantly, from a EUR 359 loss in fiscal 2013 to an estimated annualized loss of EUR 1,272 million the following year. The $250 million quarterly payments from the 2011 agreement between Microsoft and Nokia were roughly offset by royalty payments from Nokia in recent quarters. With those quarterly payments, the purchase price for NDS, and the operating loss for the quarter, and excluding the cash from the acquisition, we estimate that Microsoft has invested over $11 billion in Nokia and Windows Phones since 2011, not including in-house research and development and marketing efforts and the offsetting OS license payments from Nokia.

Management has guided to NDS reaching operating break-even in fiscal 2016. In service of this goal, Microsoft announced a massive restructuring effort, with half of the 25,000 employees taken on from the NDS acquisition expected to be laid off. This move, combined with the closure of several manufacturing facilities and a trimming of the product lines, should yield over $1 billion in cost savings, which is slightly above our estimated annualized operating loss for the unit.

Microsoft Doesn't Have to Be a First-Party Phone Manufacturer (but Now It's Stuck)
What do you do with a problem like Nokia? Historically, Microsoft spurred hardware development and adoption of its software and services through marketing and other types of support payments to OEMs, similar to the arrangement it had with Nokia before the acquisition. We believe this is the firm's preferred strategy, as it keeps the firm from having to invest in and manage low-margin hardware businesses. Microsoft's stated rationale for entering the phone hardware business is to "set the standard for productivity experiences and stimulate more demand for the entire Windows ecosystem."

​

Strategically, we concur with the need to be competitive in the smartphone marketplace. As computing continues to move to mobile devices, and bring-your-own-device (BYOD) gains acceptance in the enterprise, Microsoft risks being shut out of an entire class of devices that could support Microsoft applications and services. So while it's important for Microsoft to be relevant in smartphones to maintain the Microsoft services and application ecosystem for current and prospective customers, it's unlikely that the firm would have made the choice to become a first party provider had it not been forced to do so by Nokia. We believe Microsoft would have preferred to continue making platform payments to support OEM adoption of the Windows Phone operating system without having to take on the added expense, complexity, and distraction of hardware design, manufacturing, and integration of a 25,000 person workforce. 

It's going to cost money, but recent strategic pivots should help Microsoft expand the Windows phone market. The company's first move was to make Windows OS free for devices with screens smaller than nine inches, which will lower the costs for OEMs, bringing them to parity with the Android OS. Additionally, Microsoft is making marketing and support payments to select OEMs such as  Samsung (005930), HTC, Huawei, and Sony to support the Windows Phone--a plan similar to the firm's original 2011 agreement with Nokia--although Windows Phone will not be the exclusive OS for any of these OEMs. The company has not disclosed the amount it will spend on these marketing and platform support payments, but we estimate they will range between $1.0 billion and 2.5 billion, in aggregate. We view this as a critical component of Microsoft's smartphone strategy, as the relative weakness and unprofitability of most Android handset OEMs (excluding Samsung) create an opportunity for Microsoft to find a foothold in its device lineups as the marketing and support payments improve the business' financial fortunes. We believe the strategy of NDS combined with support payments to OEMs has a good chance of improving Microsoft's smartphone OS market share over the next few years.

Xbox's Success Shows Possibilities of Microsoft's Long-Term Strategy
Microsoft's Xbox gaming console is one of the few examples of the company finding success in a hardware platform after over a decade of continual investment and product refinement. Initially conceived in 1998 and launched in late 2001, the Xbox platform has undergone three iterations and was the leading gaming console sold in its Xbox 360 version, with over 77 million total consoles sold by the end of 2013. Originally, the Xbox gaming platform was created to give the Windows platform a presence outside of the PC form factor to appeal to console gamers, which it did. Over time, the console gained additional functionality and services as it morphed into its current version, representing Microsoft's hopes of become a technology hub (PC, gaming, content) for the living room.

The Xbox has required a tremendous amount of investment over time to achieve profitability only in the past six fiscal years. The console was almost certainly unprofitable in fiscal 2014 as a result of the research, development, and marketing expenses associated with a new console, but reporting changes eliminated that level of detail for fiscal 2014. Nevertheless, with established game titles and a strong Xbox Live ecosystem, we believe the Xbox business will return to profitability in the near term.

The Xbox 360 was considered a sales success--outselling the PS3 early in its life cycle given its head start over the PS3, and finishing calendar 2013 with over 77 million total consoles sold. The Xbox Live gaming and social network was a more modest success, collecting 48 million user accounts by the end of 2013, while the rival PlayStation Network boasted 110 million global users.

Despite the relative success of the Xbox 360, the Xbox One appears to be stumbling out of the gate, compared with its main competitor, the PlayStation 4. A botched introduction resulting from digital rights management (DRM) concerns, a higher price, and (still) limited global availability translated into an early sales lead and higher global install base for the competing Sony console. Domestically, the competing numbers for the install base appear to be much closer, although Sony still holds an edge. We view Microsoft's gaming console as an established brand with a strong ecosystem of gamers, online gaming, and a growing library of game titles.

We Think Xbox Can Still Be the Hub in the Living Room
Microsoft has a strong and established brand name, but it has recently lagged its main competitor and posted only marginal profitability, prompting us to ask the question, why is Microsoft in the gaming console business? Xbox is meant to be Microsoft's connection to the consumer's living room, a goal the firm has pursued in the past (recall the 1997 purchase of WebTV, for example). As a connection to the living room, the Xbox is an attempt to reinforce Microsoft's ties to the consumer, increasing switching costs for customers by adding programs and services such as Skype and Bing. Xbox's goal of being the hub in the consumer's living room seems prescient given the recent precipitous decline in consumer PCs.

As consumers choose computing factors other than PCs and laptops, Microsoft's connection with the end user becomes even more tenuous due to the firm's currently limited presence in smartphones and tablets. However, the popularity and brand of the Xbox is a notable positive outlier in the consumer electronics marketplace for Microsoft. We believe the company can continue to build on the Microsoft ecosystem with further investment in its gaming console. Although the console is marginally profitable on a stand-alone basis, we believe the software, games, and cloud-based services such as Skype, Live, Xbox Music, and Bing can benefit the firm's existing network effects and switching costs, and consequently, the economic moat for the company. Microsoft's strategy with the Xbox is similar to its long-term approach with its smartphone business; the hardware sales are important, but the device is also an important conduit to offering different Microsoft services. However, the Xbox and its related services are geared toward the consumer side of the business, while the smartphone strategy targets both consumers and enterprises.

Surface Tablet: An Expensive Stopgap Measure That May Be Nearing the End of Its Usefulness
The introduction of  Apple's (AAPL) iPad in April 2010 was a key catalyst that led to the slide in PC sales as tablets and smartphones gained rapid consumer adoption. The popularity of the device that seamlessly integrates software and hardware, combined with the growing adoption in the enterprise due to bring-your-own-device policies, made it an immediate and significant threat to the Windows and Office ecosystem. When Android improved its app ecosystem and manufacturers (including Google) released more affordable devices, Android tablet sales exploded, eventually outselling iOS tablets in late 2012. In the consumer market, including tablets, laptops, PCs, and smartphones, the Windows OS has dropped from a better than 90% market share a decade ago, to less than a 25% share, according to several estimates. Part of this is due to the proliferation of alternative computing devices such as smartphones and tablets that didn't exist or weren't prominent a decade ago, but the other issue is that Microsoft does not have any meaningful market share in either device category.

The danger for Microsoft is that it loses relevance in the consumer market as PCs fade, which could lead to erosion of its current dominance in productivity software in the commercial market. With its lighter operating systems based on ARM architecture, Apple's iOS iPads targeted the premium end of the tablet market, while Android targeted the mass market and the affordable range of the tablet market. As tablet adoption accelerated, Microsoft products lagged, as its Windows Vista OS and Windows 7 OS and the available chips from Intel were ill-equipped for portable, power-hungry devices. Furthermore, being open-source made Android lower-cost relative to a Windows license for OEMs. This, combined with the mobile-ready specifications of ARM chips, made it a less expensive proposition for OEMs to develop and manufacture Android tablets versus Windows tablets.

​

In response, the release of a more tablet-friendly OS, Windows 8, combined with the introduction of the Surface tablet, was supposed to help reverse the fortunes of Windows tablets, given the initial dearth of OEM Windows tablets. Unfortunately, high prices, consumer confusion regarding the messaging and branding of Windows and Windows RT, and a limited app ecosystem all led to lackluster launches for the Surface RT and Surface Pro in October 2012 and February 2013, respectively. The second iteration of the Surface lineup came to market in the fall of 2013, with refreshed models and improved features, but the same high-price, high-value proposition swayed few buyers. The most recent iteration of the Surface tablet came in the spring and summer of 2014. The jury is still out on the recently released third generation of Surface tablets, although sales appear to be improving. Overall, the Surface has been a money loser for the firm, while also proving to be a source of friction with OEMs that eyed Microsoft's expansion into hardware manufacturing with suspicion.

Recent strategic moves should improve relationships with OEMs, while the anticipated introduction of Windows 9 next year should help with the app ecosystem. First, with the introduction of the third generation of the Surface tablet, CEO Satya Nadella made a crucial move by eliminating the smaller-size tablet that was scheduled to be introduced. In our view, this move was critically important, as it accomplished two key goals: (1) It kept the existing Surface team focused on enterprise-class tablets, and (2) it helped mollify the OEMs that were unhappy with the prospect of having to directly compete with Microsoft in tablet hardware. Second, the earlier announcement that Windows OS licenses would be free for devices with nine-inch screens and smaller should help spur development of Windows tablets by OEMs, further expanding the device ecosystem both on the consumer and lower end by OEMs, and on the enterprise and high end by OEMs and Microsoft. As further evidence that this new strategy may improve the firm's fortunes in the tablet and portable market, OEMs recently showcased more than forty different Windows tablets being brought to the market throughout 2014, a major improvement over the handful of Windows RT tablets from just a year ago.

Are These Ongoing Hardware Investments Needed? Unfortunately, Yes.
In the three hardware businesses outlined earlier, we believe Microsoft is attempting to or has attempted to build a presence in a market in which it has had little to no impact, previously. A case can be made that all of those businesses are critical for the firm's longer-term strategy and to help maintain its wide economic moat. However, given the lower-margin profiles of these businesses, it's important to ask whether Microsoft should be engaged in the development and manufacturing of these devices rather than playing its traditional role of nurturing the development of OEM devices through marketing and other incentive payments. In our opinion, with the new "mobile first, cloud first" strategy, the devices are conduits to Microsoft's cloud services.

Smartphones: Likely required. In the case of smartphones, we acknowledge that the circumstances surrounding the acquisition of NDS resulted in an imbalance of power that necessitated Microsoft's decision to purchase the business. With the overwhelming share of Windows Phones, Nokia had enough leverage to force Microsoft to increase its investment in the platform or risk losing essentially all Windows smartphone production. This, in turn, would have shut the firm out of a crucial and still-growing market in mobile computing devices. Despite the margin compression from the acquisition, we reiterate that Microsoft was likely in an untenable position with respect to its mobile strategy.

Xbox: Required. We have already estimated that the Xbox has cost Microsoft over $2 billion in operating losses from fiscal 2002 through fiscal 2013. This is another margin-dilutive business for the company, but we do see strategic rationale for continued investment. We view the Xbox console as another device to help tie consumers into Microsoft's cloud services. As the link to consumers and their living rooms, the Xbox can help the firm reinforce network effects and increase switching costs through services such as Xbox Live, Xbox Music, Skype, and Bing.

Surface: Probably not required. Though we understand Microsoft's strategic rationale for its investments in the handset and console gaming businesses, we view the Surface tablet business as a luxury that likely could be eliminated should the recent increase in OEM tablet production be sustained over time.

We estimate that the Surface tablet has generated approximately $2.9 billion in operating losses, including inventory adjustments, in its short life. Given the critical condition that Windows tablets were in even just a year ago (with market share of less than 1%), combined with declining OEM adoption, we understand Microsoft's decision to create and manufacture its own line of tablet computers. However, the recent Windows license moves should help revitalize OEM tablet manufacturing and may make Microsoft's Surface business redundant. We estimate that the Surface business will continue to lose money over the next two to three years due to its high production costs given the current lack of scale.

On the gross profit line, we are quick to note that two months of NDS fiscal 2014 and the $900 million Surface RT inventory adjustment in fiscal 2013 had significant impacts on gross margins. We estimate that the three hardware investments decreased gross margins by 460 basis points and 850 basis points in fiscal 2013 and fiscal 2014, respectively.

We estimate that operating margin decreased by 450 basis points in fiscal 2013 and 1,040 basis points in fiscal 2014, a decrease of $1,248 million and $4,433 million in operating profits in fiscal 2013 and 2014, respectively. We estimate that fiscal 2014 operating profits would have increased by approximately $4 billion (an operating margin improvement of 1,000 basis points), with roughly a $0.42 increase in earnings per share. For business units where data was not available, we used the corporate average to estimate operating expenses.

With their drag on profitability, Microsoft's hardware investments have weighed heavily on our discounted cash-flow based valuation. Our current fair value estimate of $46 is 17.5 times fiscal 2014 GAAP EPS of $2.63. If we exclude the revenue and cost contributions of the hardware investments from our financial model, our fiscal 2015 operating margin estimate increases to 43% and our fair value estimate moves to $54. If we assume that Microsoft continues to invest in smartphones and tablets through marketing and development payments to OEMs to the tune of $1 billion per year over the next five years, our hypothetical fair value estimate moves from $54 to $52.