Intel's (INTC 0.55%) stock recently tumbled after the chipmaker posted its first-quarter earnings. Its revenue fell 1% year over year to $19.7 billion, but still beat estimates by $1.75 billion.
Its non-GAAP net income declined 6% to $5.7 billion. Its non-GAAP earnings per share, buoyed by buybacks, dipped 1% to $1.39 and cleared expectations by $0.25. Those numbers exclude its NAND business, which will be sold to SK Hynix, and other one-time items.
Intel expects its revenue and non-GAAP earnings to decline 10% and 15% year over year, respectively, in the second quarter. Analysts had expected its revenue and earnings to fall 9% and 19%, respectively.
Image source: Getty Images.
For the full year, Intel expects its GAAP revenue to dip 1% to $77 billion, and for its non-GAAP revenue to drop 7% to $72.5 billion, which misses the consensus forecast of $72.8 billion. It expects its non-GAAP EPS to decline 13%, compared to gloomier forecasts for a 19% decline.
Intel's first-quarter growth cleared Wall Street's low bar, but its mixed guidance indicated its troubles would persist throughout the year. Intel's stock might look cheap at 13 times forward earnings with a decent forward yield of 2.4%, but it's still not worth buying for five simple reasons.
1. PC sales will decelerate after the pandemic ends
Intel's Client Computing Group (CCG) revenue rose 8% year over year to $10.6 billion during the quarter as it sold more CPUs for PCs. Its total PC unit volumes increased 38% year over year, with a 54% jump in notebook volumes offsetting a 4% decline in desktop volumes.
However, PC sales mainly rose because the pandemic caused more people to buy new systems for remote work, online education, and gaming. That tailwind should fade after the pandemic ends.
2. It could lose more market share to AMD
Intel's ongoing chip shortage, which started in 2018 and hasn't been resolved yet, caused many PC makers to buy more chips from its fabless rival AMD (AMD 3.29%), which outsources its chip production to Taiwan Semiconductor Manufacturing (TSM 3.16%) instead of manufacturing its chips internally.
As a result, Intel ceded the desktop and notebook CPU markets to AMD over the past three years.
|
Company (Category) |
Q2 2018 Share |
Q2 2021 Share |
|---|---|---|
|
Intel (Desktops) |
71.3% |
51.9% |
|
Intel (Notebooks) |
91.3% |
77.8% |
|
AMD (Desktops) |
28.7% |
48.1% |
|
AMD (Notebooks) |
8.7% |
22.1% |
Data source: PassMark Software.
Those declines could continue for the foreseeable future as the pandemic-related tailwinds dissipate and AMD moves further ahead in the "process race" to create smaller and more powerful chips with TSMC.
3. Existential threats in the data center market
Cloud capex soared across the world over the past year as data centers upgraded their hardware to cope with more online activities throughout the pandemic.
For example, the three largest hyper-cloud companies -- Amazon, Microsoft, and Alphabet's Google -- collectively boosted their capex 32% in 2020. That increased spending should be good news for Intel, which holds a near-monopoly in data center CPUs.
Yet Intel's data center group (DCG) revenue declined 20% year over year to $5.6 billion during the first quarter. It blamed that contraction on a "challenging" comparison to its 43% growth a year ago, which benefited from upgrades across the cloud, enterprise, government, and communication sectors.
But those challenging comparisons could persist throughout 2021 as cloud and data center customers prioritize purchases of other chips, such as NVIDIA's (NVDA 1.45%) high-end GPUs for crunching AI tasks, over Intel's flagship Xeon CPUs. NVIDIA's recent launch of its own data center CPU and its planned takeover of Arm could exacerbate that pressure.
4. Unclear plans for its own foundry
To resolve its ongoing production issues, Intel will outsource the production of some of its chips to TSMC. But it also plans to accept third-party chip orders at its new Intel Foundry Services division.
This strategy contradicts itself. Intel still needs TSMC to produce some of its new chips because of the limitations of its existing foundries, yet it's building new plants to manufacture chips for other fabless chipmakers.
In other words, Intel would still be temporarily dependent on TSMC, which remains ahead in the process race, while trying to compete against TSMC and Samsung for orders of less advanced chips. The foundry services business will also likely generate lower-margin revenue than its CCG and DCG businesses.
5. Gross margin pressure and reduced buybacks
Lastly, Intel's adjusted gross margin -- which dropped 6.1 percentage points year over year to 58.4% during the first quarter -- could remain under pressure as its PC sales decelerate, its data center sales decline, and it expands its lower-margin foundry services business.
Intel also plans to reduce its buybacks to conserve more cash for its expansion efforts. That's a necessary measure, but it will throttle its EPS growth as its gross and operating margins contract.
The bottom line
Intel isn't doomed yet, but it's falling behind better-run chipmakers like AMD, NVIDIA, and TSMC. Intel's new CEO Pat Gelsinger claims 2021 will be a "pivotal year" for the company -- but I'd like to see it actually pivot before I can consider it a worthy investment.
