Earlier this year I held a relatively small Intel (INTC) position at a cost basis of just over $24 that I decided to sell for a slight loss in hopes of re-initiating at a lower cost. Typically my investment time frames are pretty short, but a lower share price and good dividend yield make Intel a more attractive investment to me.
Typical Trading Range, Dividend Yield and Product Releases
I sold my Intel position prior to the last earnings release. When I looked at upcoming product releases and a softening PC market I could not see Intel surprising with a strong quarter and decided to take my losses early and exit.
INTC data by YCharts
Excluding the highest highs and the lowest lows, Intel has typically traded at ranges between roughly $18 to $28 over the past 10 years. At $24 a share in the midst of a softening PC market, Intel being in the upper end of that band was my motivation to sell. At $22 per share, Intel is now in the lower half of this range.
Recently fellow Seeking Alpha contributor Mr. Bill Mauer published an article regarding Intel's dividend yield compared to its peers. A dividend yield of ~4% on a stock that is now trading in the lower half of its typical trading range seems more of a value to me. There is an attractive dividend, as well as more potential for growth from a softer share price, making Intel a potential dividend/growth play.
Intel should be launching their Bay Trail mobile chips at IDF in September, meaning that products should start shipping prior to the next earnings call. At the beginning of next year we should start seeing Intel powered smart phones.
Intel's "Other IA Group" Financials
Intel's "Other IA Group" is where the accounting for mobile chips resides. You can see that this group has been taking substantial losses on lowering revenues as of late. However, based on the timing of upcoming product launches on their updated Silvermont architecture, in the coming quarter we could see these losses start to decline and revenues increase provided Intel can garner the design wins needed.
I am also long Advanced Micro Devices (AMD), and have often stated that I feel AMD's problem is more in getting the design wins than the actual silicon not being competitive enough; Intel has the marketing budget to land their products in the needed designs to be a potential revenue driver.
I had stated before that I felt that Intel would have a hard time competing on price (I, II), but since those articles have been published Intel's share price has come down, along with additional churning in the rumor mill regarding Bay Trail pricing. Note there is also a pricing list on Intel's website for Bay Trail desktop parts.
Intel will refresh their Pentium and Celeron processors using Bay Trail chips, and the pricing information suggests that these desktop parts will be roughly $80 at the midpoint, and this is likely before any bulk discounts are offered to OEMs.
Based on the rumors linked to above, and Mr. Krzanich's comments stating Bay Trail tablets will be as low as $150 it is possible to see that Bay Trail can conceivably be priced to compete with ARM based SoCs. During an interview with THG, when an Asus spokesperson was asked directly about any plans to introduce a Temash based tablet, they kind of skirted the question stating "supply chains" as a deciding factor, which suggests, at least to me, as Intel as a likely candidate.
A single Asus tablet will probably not do too much for Intel, but reading through user reviews of Windows 8 tablets, a common theme is poor OEM design choices and mediocre performance. Bay Trail supposedly offers twice the power of the previous generation Clover Trail Atoms it is replacing, so this should alleviate some of the performance concerns. Recently benchmarks have posted on Primate Lab's GeekBench Browser showing performance of the Bay Trail z3770 running in both Windows and Android platforms. I have set the comparisons up to show the differences between Bay Trail and Clover Trail based Atoms.
Given Intel's vertical business model as a chip design company that owns their own fabrication facilities, Intel has tighter control over the supply chain than the competition. Intel also has the ability to run both Android and Windows, which gives OEMs flexibility of OS choice within the same design which could minimize costs for the OEMs. If Intel is providing a compelling solution, it is likely that other OEMs will join Asus if in fact Asus will be selling Bay Trail powered tablets.
The Importance of Silvermont in Tablets and Smartphones
I still feel Intel will be more competitive in tablets than smartphones until we see a viable LTE solution. Intel purchased Infineon in 2011, and last week it was announced that Intel had purchased a unit of Fujitsu that has a multimode 4G RF transceiver, so it is not inconceivable that in the future Intel will have a competitive LTE solution and will be able to compete more seriously in the smart phone arena.
That being said, Intel has taken penalties in the form of write-downs from having their fabs running at lower capacities. A reader that has prior related experience reminded me on a previous article of the affect of having fabs running at a higher capacity mitigating some of these write downs. Even if the Bay Trail chips are sold at lower margins, the two fold affect of sorts in increasing fab utilization is still a positive. Smart phones are sold in higher volumes than tablets, so if Intel comes up with a competitive smart phone solution, shipping these SoCs will further drive fab utilization higher.
One thing that I have started realizing is that it seems the consumer is choosing to replace PCs with alternative computing devices. Right now Intel is not powering these devices in any major way. Regardless of whether or not fears Bay Trail chips will cannibalize Core CPU sales are warranted, the Bay Trail chip is still a necessity for Intel to compete in a market they're not currently as competitive in. To quote Mr. Tim Cook, "... Never fear cannibalization. If we don't do it, someone else will."
Keeping in mind we have not seen an overall major market correction so far this year, as Intel starts to head to the bottom of its trading range I am comfortable with adding to my long position (or re-initiating if I have sold). Keeping in mind that my investment time frames are typically short, I am also building cash to add to my position if I feel Intel is able to break out of its trading range.
The ultimate bull case for Intel is that they are shipping 14nm silicon while the competition is still trying to ramp 20nm. According to the conference call Intel should be in volume production of 14nm at the end of the year, leading to a performance and most likely a cost advantage over the competition. While I am not as bullish as some in this scenario, I do see the value in this logic. Provided Intel can get the needed design wins heading into the end of this year and throughout 2014, I see 2014 as a potential growth story for Intel.
If you look back at the history of the computer, there is a common theme of underestimating both the number of consumers that will want computers, as well as the required processing power of computers. For the average consumer, I feel that we are actually reaching the point where increased performance is not really the main driving factor for the average consumer; cost and form factor are becoming more important.
I feel this is why tablets are being blamed for eating into the PC market. At each node shrink, performance per watt is going up, which is why I feel eventually the myriad of these designs will start to converge -- and we are already seeing the beginnings of this convergence with 2-in-1 type devices (functions as PC or tablet). Soon tablets or even a smart phone will have the processing power to perform most typical functions at least a larger percentage of average consumers will need. If Intel does not build the silicon for these devices someone else will -- they cannot fear cannibalization. Enthusiasts, engineers, or other power users will always want more computational power, but for some users this extra power is a waste.
Intel will have stiff competition from Qualcomm (QCOM) in mobile; Qualcomm already has very competitive communications solutions along with aggressive, focused R&D spending. Qualcomm's bread is buttered in mobile, whereas Intel is more reliant on the overall health of the PC market. Qualcomm I feel will be Intel's biggest hurdle getting into mobile in a bigger way. For example, Anandtech notes that Bay Trail-T GPU performance will likely be on par with an iPad 4, which if true will most likely put GPU performance behind Qualcomm's Snapdragon 800. Intel's biggest advantage against Qualcomm, in my opinion, is being able to run Windows. If Microsoft (MSFT) is able to satisfy customers with Windows 8.1, Intel's chance of succeeding with Bay Trail should go up.
AMD is also pushing HSA computing. I look at this as a wild card of sorts, which if successful could make AMD much more competitive in both the data center and PC space, meaning Intel could possibly lose market share if HSA gains footing.
Intel has essentially owned both the data center market and the PC market in recent years, and has not traded above $30 a share since 2004. Design wins do not necessarily guarantee the share price will break out of the trading range.
Sometimes the hardest thing to do is nothing. My personal view is that as Intel approaches the upper end of the trading range the risk inherently goes up. Getting in toward the bottom of the trading range is a somewhat safer approach, although still not a guarantee as Intel does have competition and things can change very quickly in this sector.
In the near term, by waiting for better entries an investor may miss out on some of the upside, but I feel it would take a catalyst on the scale of major design wins to cause Intel to break out of this trading range. By being patient and waiting for either good entry points or a breakout, it may be possible to minimize some risk if you are initiating or adding to an INTC position.