Quick Thoughts: IBM and Intel, Canaries in the Coal Mine
– IBM and Intel blamed weak Q3 sales and forward guidance on the obviously troubling macro environment, but a closer look shows evidence of the seismic paradigm shift underway in TMT
– IBM wants to hang its 5% YoY revenue decline on currency, but the 13% drop in hardware sales hints at serious weakness in the enterprise data center.
– Intel was more explicit on enterprise data center weakness, with a sharp mix shift toward cloud operators driving prices and margins on server chips lower, with more of the same forecast for 4Q.
– The shift to the cloud will accelerate, driven by the dramatic cost and performance advantages public cloud data center architecture – a painful long term downdraft for traditional IT vendors.
Tech bellwethers IBM and Intel kicked off 3Q 2012 earnings season reporting dismal top line sales and offering downbeat guidance for 4Q. At the surface, management was quick to hang the weakness on the global economy and, in IBM’s case, currency headwinds, but a deeper dive into the numbers suggests a more pernicious threat. I have written about the risk to traditional IT hardware and software vendors (here, here and here) as global scale cloud-based data center operators with dramatic cost and performance advantages compete to host applications previously run within private data centers. While both IBM and Intel participate in the emerging cloud market, their bread is thickly buttered on the traditional IT side portending more pain than gain as the transition to the cloud plays out.
IBM is an old hand at TMT paradigm shifts. Founded in 1911, it went from making commercial scales and industrial timers, to focusing on “tabulators” which were the precursor to the computers that became the company’s primary business. IBM led a major paradigm shift in the 1950’s that saw computers move from cumbersome custom built systems to standardized products that were adopted for a wide range of commercial applications. This era was further fueled by IBM’s 1964 launch of its System/360 mainframe computing solution, which became the bulwark of commercial computing for the next two decades.
IBM played a key role in the next TMT sea change as well, commandeering the nascent personal computing market with its own 1982 PC launch, legitimizing the distribution of computing power to the enterprise desktop, enabling a real consumer PC market, and setting in play a paradigm conducive to the rise of networking, and eventually, the commercial internet. The journey to the new paradigm was painful, culminating an $8.1B loss for FY 1992, what was then the largest corporate loss in US history. Nonetheless, IBM was able to bridge the generational gap by shedding poorly performing businesses and instead focusing on higher margin software and services. Not all hardware was gone, with IBM keeping its still high margin mainframe business and investing heavily in R&D.
Now IBM faces yet another tectonic plate shift in the tech landscape. Arguably, IBM has been more prescient than most, selling off its PC device business to Lenovo in 2005, long before the decline of the category was apparent. Since that divestiture, IBM’s mantra has been margin expansion, offsetting tepid sales growth with aggressive “productivity initiatives”. This quarter is an extreme example, with sales down -5% YoY while net income rose 5% on a 90bp margin improvement. However, based on today’s 10% stock price drop in a generally strong tech market, IBM investors are focusing on revenue prospects. Management hung 3 percentage points of its sales decline on currency fluctuations, but the 13.1% fall in hardware revenues belies that simple explanation. The hardware unit is almost entirely tied to spending on enterprise data centers, and IBM weakness is an indicator of a real pullback on IT capex budgets. While some of this is a function of the economy, it is also a likely indicator of a change in spending priorities and an increasing conviction that the future of enterprise IT is in the public cloud. Of course, IBM is moving to position itself for that future with its own cloud hosting and Software-as-a-Service (SaaS) offerings. While the company did not break out the performance of these future friendly businesses explicitly, it did call out that the revenues of its Business Analytics software unit, which contains SaaS offerings like COGNOS, was up double digits for the quarter. IBM may see its way through to thrive in the cloud future, but quarters like this suggest that the journey may be as painful as the last time.
On to Intel – I’ve been writing PC obituaries for the better part of three years (link, link, link) and Intel’s 3Q numbers do little to drum up optimism. Intel’s PC group sales, which represent 64% of the total, were down 8.3% YoY. Gartner, which has had to revise its PC device forecasts downward in each of the last 8 quarters, will likely have to go to the well again after this disappointment. The “Other Intel” category, which includes the nascent mobile business, was just 8.7% of total sales and was down an even more disappointing 14% YoY, underscoring the challenge as surging sales of mobile devices squeeze the PC. Despite substantial investment and management focus, mobile device makers have largely resisted Intel’s serenade and have stuck with ARM-based processors. 4Q will bring 20 Intel powered Windows8 tablets from 6 OEMs, but a tiny piece of the rapidly growing tablet market will not make up for the decline and fall of the PC.
For the last few years, the good news story for Intel has been its Data Center group, comprising the chips it sells for servers, which is 19.7% of total sales and had been reliably delivering double digit growth. 1Q sales for the unit had been flat but attributed to an inventory correction, an explanation seemingly corroborated by 14.7% growth in 2Q. Against this story, 3Q’s 5.7% growth does not feel good, particularly given a 7% sequential drop in average selling prices. Management laid the ASP hit squarely on a big mix shift toward server chips sold directly to cloud data center operators – the Google’s of the world favor moderately priced chips for 2-way servers, while the enterprise market splurges on premium chips for 4-way servers. 4Q guidance suggests a 600bp further decline in margins chalked up to charges for excess capacity and an increase in inventory reserves for its new Haswell chips. While margin fluctuations are normal for Intel before commencing production of new chips, the implied margin recovery seems to hinge on robust spending on private enterprise data centers. I’m skeptical.
IBM and Intel are the canaries in the coal mine and their songs for 3Q are not sounding particularly healthy. If this is a sign that enterprises are scaling back internal data center investment in anticipation of a move to the cloud, the implications are stark for any IT vendor accustomed to earning premium prices for value-added turn-key systems and legacy software. Watch the likes of HP, Dell, Cisco, EMC, NetApp, Oracle, SAP and VMWare closely for further signs of the pending seismic event.
For our full research notes, please visit our published research site.