3065 Rosecrans Place #203
San Diego, CA 92110
Date – August 10, 2020
Created in 1968, Intel has been through everything. In addition to creating the ‘cubicle,’ Intel created the Tick Tock process, both through internal operations and through synergistic R&D. As recently as 2009, Intel was investigated for antitrust by the state of New York, though headlines today suggest that Intel is being crushed by competition. Indeed, there have been questions about the rollout of 7 nanometer chip manufacturing, which likely will cause some turbulence in marketing efforts and/or market share.
While this loss of market share is plausible, traditional chip manufacturing and installation is becoming a smaller percentage of the Intel business structure. Thus, proficient execution in capital expenditures in the past several years have allowed for Intel to guard its relatively high margins. Those high margins historically have been achieved through Intel’s scale and Moore’s Law – both seem to be continuing as discussed below. In addition to core businesses, Intel is moving into future high-demand, high-margin operations that give us confidence in future durable profits.
Intel’s major sources of revenue today are sixfold:
Data Center Group
Internet of Things Group
Non-Volatile Memory Solutions Group
Programmable Solutions Group
Client Computing Group
In a vacuum, these figures feed an impressively diversified business today, but our focus turns to where these businesses have been and where they will be in the future. For instance, the figures from the first half of 2020 are all higher in every revenue center versus that of 2018, with 2019 being a volatile year in the industry as a whole. The future of these revenue centers is unknown, though Intel seems to be emphasizing Platform solutions, as the high margins are still quite durable.
Intel’s high margins are achieved both from worldwide business relationships and high R&D/M&A/CapEx spending. Intel’s cash generation from platform businesses shape high barriers to entry that will likely only become stronger in the future. Because of this,awe view regulation risks still relevant in discussing Intel, even more than a decade after the New York State antitrust investigation. Surprisingly, the headlines about Intel failing to maintain market share dampen that argument.
At time of publication, market share is stable across the trifecta: manufacturing; PC chip design/installation; and data center chip design/installation. The largest risk would be the loss of Apple’s Mac contract. While the news is unwelcome, the reality is that Apple’s move in-house will take years to manifest and will represent likely less than 3% (~$2.5 Billion) of Intel’s annual sales. Nonetheless, losing Apple is symbolic, and other clients will explore taking design/installation in house, or buying chips from Apple instead of Intel. One thing is for sure – Apple will not risk putting out a degenerative product, meaning that the transition will be very measured and very capital intensive. While future profits will probably be damaged by Apple’s strategy, determining the magnitude or the timing of such damages is purely speculative.
We believe the delays of the 7nM chip manufacture likely affects Intel’s ability to compete. However, no loss of sales have either materialized or have been forecast. We believe that when manufacture of the 7nM chip is complete, any lost market share in the brief interim will be at least partially recovered. Of course, we see the industry is also growing.
In our view, the semiconductor industry may reach a TAM over $1 Trillion alone, growing more than 50% from today over the next 5 years. We believe Intel’s revenue will grow along with the industry, especially given the many ancillary focuses and the “Big Bets” operations to develop chips in burgeoning technologies. Thus, the larger TAM may support revenue growth in spite of growing competitive forces. Declining margins are probable in core platform operations As Intel continues to diversify.
We select ancillary focuses at Intel that are relatively new and impactful – 3D chip design/installation, 5G collaborations, edge processing, autonomous driving, artificial intelligence, and quantum computing. In this publication we cover autonomous driving, which was also set back by the chip delay.
Our view is that Mobileye is an extraordinary catalyst, even more so after the acquisition of Moovit. Mobileye’s camera-only surround sensing system is the market leader of the nascent autonomous driving industry. Mobileye will do nearly $1 Billion in sales this year, and has signed many international contracts beginning in 1-2 years. Mobility as a Service (MaaS) is mostly synonymous with robotaxi businesses, probably because it solves several challenges with current ride- hailing business models. With no human operators, there will be no questions around whether the operator is a company employee or a contract worker while concurrently eliminating the fee those human operators demand. Intel has planned the MaaS product to be sold as an end-to-end solution that ultimately has scalable margins, encouraging the ubiquitous “Land and Expand” model. We believe that Mobileye’s “EyeQ” chips will lead to larger enterprise sales of chips like enterprise AI and Edge Processing as corporations adopt the chips.
Additionally, mapping data is being licensed to third parties, creating simultaneously a new revenue stream and a virtuous cycle of data fed by international MaaS contracts. This strategy has been reinforced by the Moovit acquisition, which is a brand name trusted by hundreds of millions of people worldwide.
Risks in autonomous driving include information flow to the vehicles, meaning that edge computing and 5G will both be extremely important to development. If those challenges are overcome, new challenges arise such as those posed by Tesla Motors. Tesla vehicles may be built with self driving software pre-installed, and/or will be controlled by a cell phone app. Tesla will eventually have rather trendy, electric vehicles using its software and may license the self driving software to other vehicles through applications, should there be a mass upgrade to vehicles. These risks are impossible to determine, though it appears Mobileye will have full automation eventually through the 7nM EyeQ5 chipset. Full automation was originally scheduled to be achieved by the end of 2020, but the 7nM chip delay has set the date back to 2021.
Since 2015, Intel has invested in Quantum Computing, including using government subsidies/grants through the National Quantum Initiative Act of late 2018. Because of the large cash flow streams and diversified operations, Intel seems like a top choice to take the lead in quantum, or at least find a niche with its “Tangle Lake” chips, which are currently functional on a small scale. Although this operation is pre-revenue and will be for many years, breakthroughs and capital investments will begin to create yet another moat for Intel. The risk, of course, is that quantum computing is not achievable in a reasonable time horizon, so that cash flows may be tied up more than expected. We include quantum as a nice-to-have element that may not contribute to the current price target.
Intel Corp By the Numbers:
Revenue Growth over the last 5 years is averaging over 7%, with forecasted revenue from 2019 to 2020 growing around 10%. We expect long term growth to normalize in the mid-single digits, increasingly away from the semiconductor design/installation operations.
Based on these assumptions, and holding other items mostly stable, the return on investment for a buyer of Intel today would be above 10% and 11% from the perspectives of free cash flow and net income, respectively. In comparison, treasury and investment grade bonds yield negative real returns at the time of this publication, meaning that the high return on investment signals a strong buy. To provide color, AMD and NVDA’s free cash flow are 8% and 38% of estimated 2020 revenues, though they make up only 3% and 20% of Intel’s free cash flow, respectively. What this means is that while Free Cash Flow yields are respectable for AMD and NVDA, the figures are dwarfed by that of Intel.
Moreover, with a 2.8% current dividend yield and a coverage ratio of 4x, the dividend is both relatively very high and safe. This could serve as a catalyst on its own, as investors look to find current, tax-favorable income in the face of negative real rates.
Further still, Intel’s corporate debt (about 20% of capital structure) was issued at between 3% and 4% interest rates up to 20 years, though operations generate about 28% after all operating and capital expenses. This makes us suspect that Intel is doing a great job managing capital allocations, though the question remains how Intel will use the proceeds, considering there have been seven debt issues in the last 12 months while it is also receiving government grants. Perhaps a very large acquisition will be announced soon – we don’t believe buybacks will be appropriate for months if not years due to equity market volatility. If buybacks do come, we think that is mostly accretive to shareholder value nonetheless.
R&D/M&A will likely eclipse $30 Billion in 2020. Other competitors like AMD and NVDA will spend $1.7 Billion and $2.9 Billion in 2020, respectively. We expect that gap to narrow in future years, however, as AMD and NVDA are trending higher in available cash flows.
Margins on the topline are consistently high at ~60%, above AMD (~45%) and below NVDA (~65%). We believe the consistencies in these margins, especially given the narrative that AMD is stealing market share, suggest Intel is maintaining topline competitive advantages.
Thus, our bull case is based on these high gross margins, the Operating margins (~30%), the consistently growing CapEx, and the ancillary businesses that have yet to take off.
Many of the risks and issues will be elucidated during Intel’s upcoming ‘Architecture Day’ (Aug/Sept 2020 expected). Traditionally, summit meetings like an architecture day have been used to announce new products and describe strategies. We will look to get more color on the 5G collaborations with Analog Devices and VM Ware. We will see you there.
Our investors and our stockholder accounts are long Intel at the time of this publication. Investors should consider this report as only a single factor in their investment decision making process. Investors should consult an investment professional to ensure suitability before making any investment decisions.
Portfolio Manager Certification: All of the views expressed in this report accurately reflect the Portfolio Manager’s personal views about any and all of the subject securities or issuers and no part of any of the Portfolio Manager’s compensation was, is, or will be directly or indirectly related to the specific recommendations or views expressed by the research analyst(s) in this report.
All research reports made available to clients are simultaneously available on our website, www.eslerfi.com. Any data discrepancies in this report could be the result of different calculations and/or adjustments.
3065 Rosecrans Place #203
San Diego, CA 92110