Intel: It’s Years Too Soon To Call It A Buy

Summary:

• Intel Corporation’s competitive position has been slowly degrading for many years, and its financial position is beginning to follow suit.
• In response, Intel has rolled out an ambitious roadmap to reverse its decline and reclaim its dominant position.
• I am skeptical of Intel’s odds of success, and I have some concerns regarding the company’s plans.
• Regardless, since we are in such an early stage of the revamp and the company could still rise or fall on these plans, I recommend holding for now.

Introduction and Background

Intel Corporation (NASDAQ:INTC) has a long and storied history as one of the pioneer semiconductor companies of the 20^th century. The company even created a timeline of events to chart the company’s story, with an entry for virtually every year from its inception in 1968 to the present.

It seems fairly accurate for the most part, but after the 1990s, Intel’s history starts to feel a bit… revisionist. The entry summarizing the 2000s reads as follows:

Intel entered the new millennium with its characteristic energy. The company continued to advance its personal computing offerings, this time with the Pentium 4, while continuing to diversify with new products for the wireless and e-commerce markets. With ongoing success in sales, the company launched a significant expansion of its global fabrication capabilities and upped its philanthropic work through two major initiatives: Teach to the Future and the Computer Clubhouse Network.

I’m sure Intel did all these things and more, but it omits one of the company’s biggest flubs of the decade: its sale of XScale. This critical misstep cost Intel greatly for several years. The reason? XScale was an ARM-based chip maker. ARM is a low power-consumption chip architecture that is favored by users of tablets and smartphones, technologies that were beginning to take off in the years after the sale. Had Intel kept its options open, and kept XScale instead of doubling down on its x86 architecture, it would have been able to leverage XScale to compete in the smartphone and tablet markets as these devices were adopted by consumers. Intel missed its chance to compete in the smartphone era at its inception, and by the time Intel was ready to try, it sealed its fate by resorting to its power-hungry x86 architecture for its smartphone chips. With Intel’s refusal, or inability, to use the low-power ARM architecture, its lack of prioritization of its tablet and smartphone chips, and with some interference by Qualcomm (QCOM), Intel’s attempts to break into the smartphone and tablet markets ended in failure.

To further illustrate the point, Intel describes the 2010s thusly:

Intel became a key player in the proliferation of “smart” devices that maintained wireless connections to cloud servers to perform tasks standalone devices can’t. The company’s technology powered the countless end-user devices found in households throughout the world, from kitchen appliances to unmanned aerial vehicles (drones), as well as the high-intensity servers that processed the data those devices generate and the networks that kept the connected devices connected…

Again, all well and good, but it omits Intel’s decade of underperformance in chip manufacturing. Intel suffered, and still suffers, from years-long delays in decreasing its transistor size on chips, often a marker of performance improvement. From 14 nanometers to 10 nanometers, and recently from 10 nm to 7 nm, Intel keeps failing to shrink its transistor sizes on time. A snowball effect should be noted here: failing to shrink transistor sizes on time results in delays of new types of chips, which results in Intel’s delayed entry into markets and devices which require or prefer such chips, which results in Intel’s outdated chips falling out of favor compared to competitors who do release new chips on time.

Speaking of which, aside from transistor shrinkage and chip releases, Intel’s competitors are not standing still. Its x86-based rival, Advanced Micro Devices, Inc. (AMD) has become a leader in the Field-Programmable Gate Array (FPGA) chip business with its acquisition of Xilinx last year. With this type of innovative, low-power x86 chip technology, AMD is set to become a top chip supplier to support the next wave of new technologies, including the data centers that will power them. Bloomberg already projects that AMD’s new class of x86 chips will do very well in this new industry. AMD will not be alone, however. Nvidia (NVDA), an ARM-based chip maker, is also poised to do well in the data center industry, being a leading provider of supporting technology like supercomputer accelerators and GPUs. With the data center market set to grow at a compound annual growth rate (CAGR) of up to 10% by some estimates, and with AMD and Nvidia set to profit handsomely from their data center focus, Intel’s fading data center prominence and fewer chip innovations are becoming a problem, especially as its CPU market share is eroded and its status as a leading chip leader declines.

Intel has seen the writing on the wall, and has proposed a bold plan to reverse course. Along with a highly speculative product roadmap that eschews names and measurements which would allow it to be directly compared to competitors, Intel has announced that it will invest $20 billion in building new chip foundries, or fabs, in Ohio, and spend another $20 billion on new Arizona fabs. While this is better than doing nothing and letting competitors take share in lucrative future markets, I have three main concerns that make me hesitant to recommend Intel as an investment.

Financials: Intel vs Its Competitors

Before detailing my main concerns for Intel’s future, I will discuss Intel’s financial situation, and why it troubles me today.

Intel’s revenues were ~$70 billion, ~$72 billion, ~$78 billion, ~$79 billion, and ~$63 billion for the consecutive years from 2018-2022. Its gross profit held steady at $42-43 billion from 2018-2021, but dropped to less than $27 billion in 2022. Cash flow from operations for the 2018-22 period was ~$29 billion, ~$33 billion, ~$36 billion, ~$29 billion, and ~$15 billion for each respective year. Capital expenditure for this period was ~$15 billion, ~$16 billion, ~$20 billion, ~$20 billion, and ~$25 billion. Lastly, Intel has ~$42 billion in debt, and ~$28 billion in cash.

At first glance, things seem relatively alright. Intel is undeniably profitable, and isn’t precipitously losing money, which is certainly good. Revenues are generally up, which is good. CapEx is rising, which isn’t necessarily bad, though it may be something to monitor. The large debt is concerning, and should be addressed when possible, but profit can be banked to pay it as needed if the company is careful. Excluding outlier year 2022, which saw some sharp declines, Intel seems to be a picture of sustained financial health.

That is, until you look at the picture in context.

AMD’s annual revenue from 2018-2022: ~$6.5 billion, ~$6.5 billion, ~$10 billion, ~$16 billion, and ~$24 billion.

AMD’s gross profit from 2018-2022: ~$2.5 billion, ~$3 billion, ~$4 billion, ~$8 billion, and ~$12 billion.

AMD’s operational cash flow from 2018-2021, and Q1-Q3 of 2022: ~$34 million, ~$500 million, ~$1 billion, ~$3.5 billion, and ~$3 billion.

AMD’s CapEx for 2018-2021, and Q1-Q3 of 2022: ~$160 million, ~$220 million, ~$290 million, ~$300 million, and ~$320 million.

AMD’s debt: ~$3 billion. AMD’s cash: ~$6 billion

These financial trends are similar for all of Intel’s frequently talked-about peers, i.e. Nvidia, AMD, and Taiwan Semiconductor Manufacturing Company or TSMC (TSM). All these firms and Intel saw ballooning CapEx in the past few years; however, only Intel’s competitors saw revenue, gross profit, and cash from operations steadily or exponentially increase as well. The competition also grew their cash to outweigh their debt, instead of the other way around.

Compared to its competitors, Intel barely squeaked by during the COVID era. It then began to tank last year. This level of underperformance appears to be somewhat unique to Intel, since it goes against the trend among its most notable semiconductor industry peers. It’s no wonder the company cut its dividend by nearly two thirds – it isn’t financially healthy enough to give cash away anymore. Such financial weakness screams “red flag” to me, and I would personally avoid INTC just based on the company’s present financial stagnation/deterioration. Better opportunities exist today.

Now to explain my concerns for Intel tomorrow.

My Three Concerns Regarding Intel’s future

Intel’s x86 Chip Innovations Are Stifled By Delays That Cost It Opportunities

AMD’s aggressive pursuit of the data center market is likely fueled largely by its acquisition of Xilinx, whose expertise in FPGA technology will allow AMD’s x86 chips to be competitive in performance and power consumption against ARM-based chip makers like Nvidia. Intel actually bought an FPGA company in 2015, and then released its own FPGA chip line in 2019. However, AMD has had more consistent (though admittedly marginal) improvements in transistor shrinkage compared to Intel over the past few years, and AMD’s advantage may continue to hold steady. This is a big unknown though, considering the constant jockeying for denser chips and smaller transistors between AMD/TSMC and Intel in the CPU, GPU, and FPGA markets. But considering Intel’s penchant for being late, I favor AMD to retain the lead for x86 architecture.

Even if Intel could produce new FPGA chip designs on time, Intel recently suffered an FPGA chip shortage that set back FPGA chip delivery, to the sure delight of AMD. Furthermore, if AMD is able to maintain its transistor shrinkage and density lead, then the performance benefits of its chips once FPGA programming is added to the mix will continue to outpace even Intel’s FPGA chips. This performance advantage, however small, translates to a competitive advantage when data centers pick chip makers to keep their machines running. On the massive scale that data centers operate at, the tiniest advantage in power consumption or performance per chip can add up quickly, and is a major deciding factor in which chip makers data center operators choose. While Intel is a dominant force in the data center industry today with ~70% market share, its market leadership is not set in stone, and is already being eroded by AMD. Furthermore, Intel’s CEO has already admitted that the company will continue to lose momentum in the data center space, and will cede market share to competitors like AMD for at least a year or so. Predictably, this is due to a delayed release of a new planned chip, Sapphire Rapids.

With Intel’s track record of delays, right up to the FPGA shortage and Sapphire Rapids postponement last year, there’s a good chance data center operators will seek other chip makers who can provide better performing chips on time. This sets a bad precedent for Intel, especially for deployment of chips in data centers and other technologies that will require FPGA or other innovative x86 chips. Observers in multiple chip-related industries may consequently mark Intel as unreliable and list it as a chip supplier of last resort, reducing the company’s partnerships and business opportunities. This reduction in business would make INTC a subpar x86 chip stock to invest in for the future.

Intel’s Fab Investments and Research Initiatives MUST Yield Substantial Results and Advantages

Despite Intel’s investments in new American fabs, and talk of fabs in Europe, some of Intel’s planned research labs were recently scrapped due to revenue shortfalls. While the research that would have been conducted in those labs is expected to still occur elsewhere, R&D is not where I would make cuts if I were investing in reclaiming my company’s market dominance.

The $40 billion fab investment in the US (plus however many billions that will be spent in Europe) is a good start, but if it is only spent on making chips that are worse than competitors’ in terms of hardware (transistor size, transistor density) and software (FPGA programming, architecture type), then it will be a waste of money. Even for Intel, $40 billion is a lot of cash to spend on fruitless endeavors. $40 billion spent over 3 years from 2022-25 (assuming the same timeline for the Ohio and Arizona projects) means an annual hit of over $13 billion per year until mid-decade, and spending on possible European fabs would increase the cost further. Considering Intel’s $8 billion in net income in 2022, this may mean more debt for the company to take on, which is concerning when current debt already outstrips its cash.

With such large investments that will drag down the financials, the ability of these future fabs and research endeavors to produce popular, outperforming chips becomes that much more important. If at least $40 billion is spent, and Intel’s products are still less appealing than its competitors’ at the end of it all, Intel will be punished severely with flagging consumer demand, and by a selloff from investors when they realize the company’s big revamp has failed to change its downward trajectory. Combined with Intel’s deteriorating financial condition, such an outcome would be devastating for the company.

US/China Tensions and China/Taiwan Conflict Could Make Things Very Bad For Intel

Many investors see Intel as a safe value play for stable, “Made in America” chip supply chains. In their view, Intel’s lack of exposure to Asia would mean that the company, and their investment in it, would be out of harm’s way should China invade Taiwan, home to Intel’s foundry rival and the world’s dominant chip foundry company, TSMC. But this Seeking Alpha contributor makes a good argument for why this may not be the case.

As he highlights in his article, the Chinese market actually generates a plurality of Intel’s revenue – about $21 billion in revenue in 2021, according to the contributor’s data. This means that as geopolitical tensions between the US and China worsen, Intel would get the short end of the stick should the company face pressure to restrict its chip operations in China. Such restriction becomes much more likely if China and the US end up in a proxy war over Taiwan – as we saw with the Russian invasion of Ukraine, the majority of known foreign businesses pulled out of Russia in response. Imagine the financial consequences of Intel having to pull out of China and lose nearly $21 billion at once. $21 billion is 50% more than Intel’s operational cash flow and nearly 3x Intel’s net income for the year 2022. Such a hit to revenue would surely cause Intel to buckle, or worse.

Intel’s revenue being plurality Chinese also means the rewards it will reap from America’s recently passed CHIPS Act are less valuable compared to TSMC’s share. The Act encourages US chip manufacturing and doles out rewards based on US chip investment. Both TSMC and Intel have each made fab commitments of about $40 billion in the US, though Intel’s other US investments into research labs and other elements of the chip value chain will increase its investment total some. Assuming it’s an extra $5 billion of investment from Intel all told, the total $45 billion from Intel vs $40 billion from TSMC should mean slightly more CHIPS Act incentives for Intel, perhaps a few tens or hundreds of millions more. But Intel’s $21 billion of revenue from China compares to TSMC’s roughly $5.7 billion of China revenue based on total revenue of $57 billion (since China exposure is only about 10% of revenue for TSMC). This means that should both companies face China restrictions caused by geopolitics, Intel’s CHIPS Act gains will be diluted by up to $21 billion in losses; meanwhile, TSMC’s similar CHIPS Act gains will be diluted by at most $5.7 billion in losses. On balance, then, Intel’s CHIPS Act subsidies are more at risk than TSMC’s from China fallout. Additionally, TSMC is investing internationally to expand its operations in Japan, and is reportedly strongly considering expansion into Europe as well, almost certainly collecting subsidies from these two regions to add to its CHIPS Act funds.

So, both TSMC and Intel are diversifying their fab operations, and both companies are exposed to significant geopolitical risks in Asia, but Intel’s China exposure makes it less likely than TSMC to benefit from the subsidies in new American chip legislation. This means that the assumption of greater chip supply chain and revenue safety for Intel (instead of TSMC) is quite flawed. Current and prospective Intel shareholders may want to think twice about INTC if this is a major reason they are investing in it, as worsening US/China relations will likely mean a major hit in the future to Intel’s already-softening revenue and profits.

Intel’s Valuation

On most measures, especially for trailing 12-month ratios, INTC’s valuation looks quite attractive compared to its sector as a whole. Its Price/Earnings ratio is ~13; its Price/Sales ratio is ~1.5; its Price/Book ratio is 1; and its Price/Cash flow is ~6.5. This compares to valuations of ~18, ~2.7, 3, and 20 for the P/E, P/S, P/B, and P/Cash flow of the whole Information Technology sector. INTC’s ratios also compare favorably to its competitors, which all have ratios that are either mixed or overvalued compared to sector. TSM might be the closest to INTC on a P/E basis (sporting a ratio of ~13 as well), but its P/B and P/S are much higher than the sector ratios, let alone INTC’s. For different measures, it’s the same story for AMD and NVDA – similar to INTC on one metric, but very overvalued on at least one other. INTC’s valuation is the lowest among the pack overall, and therefore a bargain hunter’s dream chip stock.

To some of you reading, it may seem odd to put the valuation section so far away from the financials section, but this was deliberate. Many investors only consider Intel’s storied history and greatest achievements when thinking about the company. The foregoing analysis and breakdown of the company’s challenges and mistakes is meant to inject a dose of reality into the positive narrative many people have about Intel. Blue chip investors and value investors in particular should note that this reality about Intel is not captured by the valuation alone.

A well-worn quote by Warren Buffet, famous value investor, is that “it’s far better to buy a wonderful company at a fair price than a fair company at a wonderful price.” I introduce the valuation section here because on its face, INTC’s valuation is clearly quite wonderful. But the above sections should make clear why the company itself is indeed merely fair: Intel has dug itself into a deep hole over the past few decades, and it is fairly probable that it will fail to claw its way out of it, and also rise above its most noteworthy competitors, in just the next few years. As such, the company is appropriately valued below its competitors and its sector – it is simply less valuable due to the low-to-moderate likelihood that it will achieve its lofty goals.

To be fair, its revamp plans may yet succeed, so it might be too soon to argue that Intel is a value trap. On the other hand, if the shoe fits…

Risks to Thesis

As I have discussed, many things need to go well for Intel’s revamp to work out. However, that does not mean these things cannot happen. If Intel is able to put an end to its chip delays and become a leader in transistor shrinkage and density, program its x86 chips to be above and beyond its competitors’ software capabilities, successfully turn its investments in new fabs and R&D into well-oiled, competition-beating, mass chip-producing machines, and minimize the financial fallout of increasing geopolitical tensions, then the company will have successfully achieved the goals of its big revamp, and then some. Intel would consequently reign supreme in the chip world once again. Additionally, should this end up being the case, it would mean INTC is indeed undervalued and trading at a massive bargain, and should be purchased in the near term while shares are relatively cheap. While the stock is in a diminished position that matches its company’s circumstances, it would skyrocket if the company became king of the hill in the semiconductor industry, rising along with its company’s fortunes. Should all this occur, my thesis would be broken.

But again, many things need to go well for these risks to materialize and break the thesis, and I am fairly confident that at least some of these things may not come to pass.

Conclusion

It’s early days in Intel’s revamp of its operations – too early to say how the fab and research projects will go, and whether these and other initiatives by Intel can get the company firmly ahead of its competitors by mid-decade. But these competitors won’t stand still while Intel gets its act together, and their past and future operations stand to take greater market share from Intel should the company’s revamp not produce the necessary results. In addition to competitive pressure, geopolitical tensions also seem poised to impact Intel for the worse, and in the worst case could drastically damage Intel’s weak financial position via a significant revenue cut.

While I am personally skeptical of Intel’s chances of success, I recommend a wait-and-see approach for the next few years as it all plays out. This will give patient investors sufficient time to see if Intel seizes or fumbles its chance to reclaim its status as a dominant chip maker. Therefore, I would call INTC stock a hold.

---

Disclosure: I/we have no stock, option or similar derivative position in any of the companies mentioned, and no plans to initiate any such 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.

---