Buy Intel: 5.9% Shareholder Yield And Plenty Of Growth
Summary
- Intel is moving away from being PC-centric to being data-centric. As a result, Intel is benefiting from trends in datacenter growth, internet of things, and autonomous vehicles.
- Over the past 3 years, Intel grew its revenue over 25%, EBIT over 57%, and EPS over 37%.
- Intel has also rewarded shareholders with a 2.4% dividend yield and a 3.5% buyback yield which sums up to a 5.9% shareholder yield.
Intel’s Business Transformation
Intel (NASDAQ:INTC) is in the fourth year of a business transformation. Intel’s industry thesis is very cogent: the world of computing is moving away from being PC-centric to being data-centric. Furthermore, selling to datacenters is more than just powering servers, it involves powering the internet of things.
That may sound like a bunch of buzz words, but in practical terms this shift has catalyzed the company to make investments in markets like memory, programmable solutions, and autonomous vehicles while continuing to invest in the core PC and server chip business.
Intel’s strategy appears to be paying off. From 2015, when Intel announced its business transformation, to the last 12 months, revenue has grown over 25%, EBIT has grown over 57%, and Non-GAAP Diluted EPS has grown over 37%.
Shareholders don't have to wait to benefit from Intel's investments, they are paid to wait. The stock offers a 2.4% dividend yield and a 3.5% buyback yield which sums up to a 5.9% shareholder yield. This is quite attractive for a blue chip company that is able to grow both its top and bottom lines at double digit rates.
Intel’s PC Business has Stabilized
Over the past 10 years, the PC market has been soft. The market for PCs has matured and the rise of mobile computing has hurt PC demand. In 2015, Intel’s decision to embark on its ambitious turnaround was largely catalyzed by a weak PC market. However, after many years of decline, the PC market has stabilized and is poised to grow in 2018.
In the most recent quarter, Intel posted PC y/y revenue growth of 16%. This is quite impressive considering that PC shipments were only up low-single-digits. In fact, Intel has been able to consistently grow its PC segment at a much faster rate than the PC market. This is because the PC chip is one of the most valuable components and benefits from steady increases in average selling price. Furthermore, Intel has continued innovating with new products. For example, Intel was critical in the launch of the new HP 2-in-1 laptop which doubles as a tablet and achieves a 19 hour battery life. The chart below shows how Intel’s PC Chip average selling prices have risen over time.
The PC business has not been a sexy growth business, but it is a cash cow for Intel. Last year, Intel’s PC business generated $34.0 billion in revenue which accounts for 54% of the company’s total sales. The PC business generated $12.9 billion in operating profit before tax last year which implies a 37.9% margin. This was the highest segment operating margins since 2011. After years of top-line weakness and margin pressure, Intel’s PC segment looks like a bright spot.
Intel’s Server Business Is Growing Despite Competition
Unlike PC, Intel’s server business has been a growth business. There is a long term bull market in server chips due to growth in cloud computing and storage. At Splunk’s (SPLK) 2018 analyst day, the company projected that by 2020, annual data production will be 44 times greater than it was in 2009. All this data production requires computer processing power. Intel is a fine way to play this trend as it has a greater than 90% market share in chips for servers.
However, in recent years growth has slowed in Intel’s server segment. There are a couple of reasons for this beyond a previously faster rate of market growth (which was also true).
First, Intel’s server growth was amplified in the decade earlier due to AMD (AMD) discontinuing its line of server chips. According to research from Citi, in 2006, AMD had a 25% market share in server chips. By 2012, AMD’s share dropped below 5%.
The second reason growth has slowed for Intel’s server business is Nvidia’s (NVDA) entrance into the market. I recently wrote an article expressing my bullish opinion on Nvidia. Nvidia makes GPUs which historically were exclusively used for processing graphical images. In recent years, it has been found that GPUs are particularly good in some areas of computing, particularly in artificial intelligence and machine learning. Artificial intelligence and machine learning are rapidly growing areas of interest and many data centers are investing in GPUs to help tackle these challenges.
Nvidia is quickly gaining market share in the data center chip business. Last quarter Nvidia’s data center segment grew by 58% year over year. Over the last 12 months, Nvidia’s datacenter revenue was $2,859 million. This compares to Intel which over the last 12 months generated $22,504 million in data center revenue. Last quarter Intel grew its server chip business at a 26% year over year rate.
Because Intel and Nvidia together control virtually the entire market for server chips, we can add up their server revenue and calculate a total addressable market of ~$25.4 billion today growing at a ~30% rate. This implies that Nvidia has achieved ~11% market share today from <1% share before 2016.
Nvidia’s growth in the server market is impressive and many analysts have hedged their bets on Intel as a result. However, Intel’s server business will still have an attractive growth profile because the overall market is growing fast enough to accommodate NVDA’s share gains while still providing Intel double digit growth. Furthermore, although AI represents one of the fastest areas of growth within computing, there is a limit to how much market share Nvidia can ultimately take because CPU chips are still the default choice for most computation problems.
It is worth noting that Intel’s recent quarterly growth rate of 26% in servers is a significant re-acceleration compared to recent years.
Last, I will mention that AMD has been trying to re-enter the market for server CPU chips. AMD is currently at 1% market share, but this is essentially from 0% last year. Some analysts think AMD can take as much as 30% share, but I am more skeptical. AMD doesn’t have a stellar track record of execution; in fact, a couple years ago analysts speculated on an AMD bankruptcy.
In reality, AMD will take market share, but it will do so the same way it has done so in PC chips: discounting. Commentary following the announcement that Dell would use AMD server chips indicated that Dell received a meaningful price discount relative to what Intel chips would have cost. This is a notable departure from Nvidia’s successful strategy of taking market share through delivering a significantly differentiated offering to server customers. In direct competition with Intel, I just don’t believe AMD will be able to keep up with Intel’s R&D or marketing budget.
Critics have noted that AMD’s EPYC server chip line has good performance, but the performance improvement is marginal and by no means awe-inspiring. AMD hurt its reputation 10 years ago when it discontinued its last server chip line. The company will need to overcome Intel’s superior brand and install base which will be no easy task. If AMD can achieve a 10% share within a couple of years, that would be a huge win for AMD but will not hurt Intel too much because of the continued overall growth in the market.
Intel’s Other Bets On The Future
Intel’s PC and Server chip businesses account for 85% of revenue and over 100% of profits (due to unprofitability of remaining segments). Intel has small businesses tied to internet of things (5% of total sales), flash memory (6% of total sales), and programmable semiconductors (3% of total sales). Many of these side bets were boosted through acquisitions.
In March 2017, Intel announced it would acquire Mobileye for $15.3 billion. Mobileye sells technology for autonomous vehicles including sensor fusion, mapping, front- and rear-facing camera technology, and a crowd-sourcing software for real time street maps. The Mobileye acquisition in addition to smaller investments, such as the stake in mapping company here, have provided a strong platform for computing in the automobile as well as autonomous machines more generally.
In September 2016, Intel agreed to acquire Movidius, a company which makes chips for drones and VR headsets, for $400 million. Movidius produces a system-on-a-chip product which is a turnkey semiconductor with pre-installed hardware equip to use algorithms like deep learning for navigation and depth processing.
In December 2015, Intel agreed to acquire Altera for $16.7 billion. Altera is a maker of programmable chips that can be functionally repurposed after they leave the factory, unlike traditional microprocessors. Companies such as Apple and Microsoft have had to create their own chips or painfully repurpose standard chips to meet certain functions. Programmable chips give this more sophisticated subset of clients more flexibility.
The Mobileye and Altera acquisitions are Intel’s 2 biggest deals ever and both occurred within the last 3 years. This signals that Intel has gotten more aggressive in diversifying itself away from the PC and server businesses.
Some analysts and critics have called Intel’s M&A binge destructive to shareholder value because the company is over-paying for unprofitable assets. However, Intel is making small bets (relative to its overall size) to ensure its relevance in a world that is moving towards mobility, internet of things, and flexible computing solutions.
Valuation and Shareholder Yield
Intel currently trades for 9.5x its enterprise value to next twelve months EBIT. A group of semiconductor peers including NVDA, QCOM, XLNX, AMD, and AVGO, trade for a median multiple of 21.6x EV/NTM EBIT. Intel trades at more than a 50% discount to this group! Intel is also cheaper than the S&P 500 which trades for 9.8x EV/NTM EBIT. This is despite the fact that Intel has historically grown its revenue and earnings at a higher rate than the S&P 500 and has a higher dividend yield, 2.4% vs. 1.8%. (Note: I used Capital IQ as a source for the trading multiples; NTM EBIT uses the median Wall St. analyst estimate, which tends to be fairly accurate when looking out over the next 1 year.)
The valuation story is attractive, but what gets me more excited is Intel's shareholder yield. Shareholder yield is a calculation where total capital returned to shareholders is expressed as a percentage of the per share equity value. According to my calculations below, Intel offers a 5.9% shareholder yield to investors today.
Furthermore, Intel has a strong history of increasing dividends per share and buybacks over time. Intel has increased its dividends per share from $0.45 in 2007 to $1.20 today. That is a dividend per share compound average growth rate of 9.6% over 11 years! On top of that, Intel recently announced a $15 billion increase to its share repurchase program. This is extremely shareholder-friendly capital allocation and Intel has the cash flow and balance sheet to sustain it.
Concluding Thoughts (Capital Allocation Matters!)
Many investors view Intel as a stodgy old-world semiconductor company primarily serving the mature personal computer market. While that statement on its face is not inaccurate because the PC chip business is still Intel’s largest source of revenue and cash flow, it fails to give the company credit for its ability to stay relevant and reward shareholders through excellent capital allocation.
Intel’s PC chip business is a huge cash cow that is no longer in a state of decline. The company has organically invested in and grown its server chip business. Through M&A, Intel has expanded into the fast-growing areas of ‘internet of things’, ‘autonomous vehicles’, and ‘machine learning’. Finally, the company has paid shareholders through significant dividends and share repurchases.
Reader note: this article is a continuation in my series covering the semiconductor industry. If you liked this analysis, please give me a follow for more analysis. I recently shared my bullish thesis on Nvidia here. I have near-term plans to write about AMD (AMD), Qualcomm (QCOM), and Broadcom (AVGO).
This article was written by
Analyst’s Disclosure: I/we have no positions in any stocks mentioned, and no plans to initiate any positions within the next 72 hours. I wrote this article myself, and it expresses my own opinions. I am not receiving compensation for it (other than from Seeking Alpha). I have no business relationship with any company whose stock is mentioned in this article.
Seeking Alpha's Disclosure: Past performance is no guarantee of future results. No recommendation or advice is being given as to whether any investment is suitable for a particular investor. Any views or opinions expressed above may not reflect those of Seeking Alpha as a whole. Seeking Alpha is not a licensed securities dealer, broker or US investment adviser or investment bank. Our analysts are third party authors that include both professional investors and individual investors who may not be licensed or certified by any institute or regulatory body.
Recommended For You
Comments (48)
I cashed out recently as ACTUAL yield has bottomed out indicating overvaluation.
I'll be happy to get back in if/when it's out of favor again to result in a decent dividend yield.
But we should be seeing a dividend raise soon for the real yield.
• FPGA. Intel (via its purchase of Altera) is gaining on the only other player on the market. www.fool.com/...
• Wireless. Qualcomm just admits it lost Iphone to Intel. Qualcomm Admits That It Just Lost the iPhone to Intel. www.fool.com/...
• AI. Intel sold $1 billion of AI chips this year. www.reuters.com/...
• 5G. Ericsson, Telstra and Intel achieve first 5G commercial network data call www.pcworld.idg.com.au/... or Intel partners with Nokia and Ericsson on 5G worldwide www.zdnet.com/...
• Intel is also breaking into discrete dedicated GPUs with an all star team of engineers. More here www.forbes.com/...
• Incidentally, Lisa Su knew exactly how much power the Rome chip pulled on the above demo, but she refused to disclose that information.
• The Rome chip has a gigantic and wasteful 14nm die in the middle that does zero processing but sucks power. It's only there to act as interconnect. Poor design. www.anandtech.com/...
• Intel's upcomign Cascade Lake beat AMD's 7nm Rome for the 2019 35 PFLOPS supercomputer: "We took a look at AMD Epyc, both Naples and certainly Rome but with the combination of price, schedules, and performance, we felt like Cascade Lake was the way to get the best value right now." www.nextplatform.com/...
I guess the author sees flaws in buying as well.
What happens if a severe downturn comes.... they spent way too much on bad acquisitions....where the ROA...zero?