INTC: New CEO Choice, Brian Krzanich, Signals Future Direction (INTC: NEUTRAL)

Intel names Brian Krzanich, formerly COO, as its new CEO and with that move signals the firm’s commitment to capitalizing on its manufacturing dominance, not only to maintain its PC and server CPU advantages, as would be expected,  but also to push harder into mobility and to bring its nascent foundry business to a much larger set of customers.  Both of the latter require innovations in manufacturing technology and in production management of the sort that Krzanich has masterminded for the last several years.

We remain NEUTRAL on INTC shares in the face of near-term headwinds, but view the selection of Krzanich as positive step towards a future with stronger growth and sustainably high margins.

Mobility: Progress on the mobility front in the transformation of the Atom line of processors from mere single CPU solutions to more complex and power efficient SOCs depended on close coordination between the design and manufacturing teams, and Krzanich brings fundamental understanding to the table as Atom approaches a critical juncture in its battle to best the performance-power profile of ARM-based solutions.  Bay Trail, coming later in 2013, will finally allow fan-less tablet and hybrid designs, but further progress will still be needed to deliver better battery life and lower costs. And in smartphones, only flawless moves to the 22nm Silvermont and the 14nm Airmont can make Intel a serious contender. The design blueprints have been drafted for these, but keeping leakage down through manufacturing advances will be critical to hitting power efficiency targets.

Server: Intel owns the server CPU market, for now, and well knows the importance of delivering better performance and power efficiency with each new generation. But ARM proponents may overturn the old order and the shift from traditional enterprise servers to massively redundant Cloud installations is changing the sort of server in demand by the leading data center providers. And those changes appear to be triggering a move down to lower-cost traditional servers and to be opening the door to ARM- or Atom-based microservers.  While the Xeon roadmap is well drawn with this year’s upgrade to Ivy Bridge-based Romley and the 2014/15 move to Haswell, the pricing pressures sure to arrive from those pressures indicate that Intel manufacturing will need to deliver far more in unit cost savings to support the 60-62% gross margins enjoyed of late.

Foundry: The selection of Krzanich signals most directly, in our view, the importance of the foundry business to Intel’s future. Otellini fought the tough battles to maintain and even regain market share in PC CPUs at a time of global slowdown; he accepted the move to a more sensible pricing strategy which helped to bring gross margins into the low 60s, and he oversaw the takeover of Intel CPUs in servers as IT managers took advantage not just of Xeon’s performance, but also of the simplicity of uniformly equipped data centers. Krzanich now has the opportunity to expand Intel’s reach as a foundry player; Intel has a 1-2 generation lead over the established foundry players, but it now must become a company that can development multiple manufacturing process variants for diverse customers, learn to manage smaller production runs reliably and on time, and to build trust among potential customers. Costs of building large-footprint FPGA chips would see disproportionate savings from a move to the leading edge of manufacturing that Intel can provide. Altera’s and Microsemi’s decisions to go with Intel may very well put other logic suppliers at a disadvantage and trigger more business for Intel — but only if Intel can convince potential customers of its ability to deliver.  We believe Krzanich’s track record in manufacturing brings credibility to the fledgling foundry business.

Commitment to Capital Spending — In Our View, a Necessity:

Intel’s manufacturing capabilities are a critical component of its continued high gross margins and of its ability to break into new markets. Through investments in more advanced manufacturing, Intel has been able to reduce unit costs and, we believe, this has helped to make PCs more affordable and this, in turn, has helped to drive sales in emerging markets. Contrary to the concerns of some, we have not seen the market become flooded with cheap chips and have seen limited impact on Intel’s gross margins suffer even in this weak demand environment.

Intel has been on an aggressive capital spending program since the beginning of 2011 and it is expecting to maintain its capital budget at the high end of its normal range through 2013. While lower levels of capital spending were appropriate when the company hovered around $30B in revenues for a decade, it is now approaching $55B in revenues and has been increasing capacity to match that higher level of sales.

Such investments do indeed drive up the capital cost per wafer, but the company is also driving greater cost reductions as transistor density improves. These offsetting cost reductions prevent gross margins from being negatively impacted by the increased wafer costs.

Intel’s 2011 capital spending plan left many concerned that the company was risking over-capacity, excess CPU supply, and falling ASPs by adding so much capacity when PC demand appeared so uncertain. At the beginning of this expansion, Intel’s Andy Bryant went through a quantitative exercise to show (1) that the financial cost of lower-than-expected demand is very low (if, e.g., demand is half of the expected level), and (2) the risk in lost profits of failing to build this capacity significantly outweighs the risk in additional fixed costs if the assumed demand growth fails to materialize. In the end, Intel was able to sharply reduce capacity utilization when demand slowed and it accelerated the conversion to 32nm production. Through the cycle, Intel demonstrated its ability to respond quickly and limited the downside to gross margin. And in the current period of weakened demand, Intel has once again downshifted on current capacity and accelerated the move to next-generation process technology, where lower unit costs come to the fore.

• We believe current levels of capital spending are warranted due to the cost savings achieved by moving down production nodes. As it becomes more expensive to equip factories with the latest production technology, fewer companies will be able to make the required investments and we expect Intel’s advantage to expand.
• The company is also able to reuse between 80% and 90% of its equipment at the next node. This effectively lengthens the useable life of the equipment and spreads the investment costs over a greater number of units and years.
• Pushing harder to the next node improves Intel’s competitive position versus ARM-based processors in mobility and enhances its advantage over established foundries.
• Current facilities are being constructed for use through the conversion to 450mm wafers.

Utilization levels are now in the 55-65% range on the heels of weak global demand and the incursion of tablets, and while this remains a concern, Intel is building for future technology needs, not just capacity needs. Accelerating the move to 14nm is essential to Intel’s success in smartphones and as in foundry services.

Capital expenditure is driven by not only expected unit growth, but also by greater feature integration requirements. The capacity increase for the move from the 45nm to 32nm process node involved both of these as the GPU was brought onto the CPU and the chipset was produced at the leading edge of manufacturing for the first time. Much greater graphics capabilities at 22nm contributed to expansion at that node, but the weakening macro conditions prompted a pull back in those investments and  an acceleration to 14nm where shrink will help contain costs.

Capital intensity has historically been in the 10-20% range and jumped from 12% in 2010 to 20% in 2011. We expect capex will remain at the high end of this range over the next six quarters as Intel continues to push performance for mobility and as it builds its foundry capabilities.

Intel Investment Thesis (INTC: NEUTRAL): 

Ongoing macro weakness is contributing to a weak outlook through 1H13 and dampening prospects for 2013.  We are reiterating our NEUTRAL rating in the face of a lack of positive catalysts, but see room for longer-term positive potential on the appointment of Krzanich as CEO. Retirement of Otellini and naming of Krzanich as his replacement signals clear direction the company which emphasizes new opportunities in foundry and in mobile applications.

Longer‐term Catalysts:

• New Intel product cycle. Upgrade to Haswell for PCs, coming in mid‐2013, offers performance boost and power savings for OEMs which we expect to help spur consumer and corporate interest. But our checks show a cautious ramp of Haswell (to comprise only 20% of shipments by 4Q13), so benefits are likely to have larger quantitative impact in 2014.
• Emerging market needs. Penetration rate of PCs and servers in emerging economies remain well below that of the industrialized world, and as GDP rises, investments in IT infrastructure have been rising too. This year, with China’s resolution of its leadership line‐up, could be a bit better than last year. We expect minimal cannibalization by tablets in the corporate segment of emerging markets since productivity remains essential and connectivity remains limited.
• Virtualization and data center expansion. Look for further investments in servers and data centers to meet the growing data traffic spurred by the proliferation of mobile connected devices. 2013 brings additional wireless spending on infrastructure by the world’s carriers and this is likely to spur a new wave of data center investments. Intel expects ~15% server CPU growth and for this segment to become 20% of sales by 2016.
• Global economic recovery. Intel’s leading market share leaves it well positioned to benefit from household and corporate computing investments as economic recovery resumes (eventually) around the world.

Downside protection for INTC is based on its appealing dividend yield in this time of particularly low rates of return on cash.

Investment Risks:

Potential risks to our investment thesis, rating and estimates include, but are not limited to, global economic slowdown, failure of the company to maintain its assumed pace of innovation, an unexpected degree of price competition, customer decisions to switch to other suppliers, and the semiconductor and electronics supply chain’s disruptions from global political, physical, and economic shocks (earthquakes, hurricanes, floods, etc.).