We are also maintaining our fair value estimates for leading communication service providers like Nokia (NOK) and Ericsson (ERIC). Huawei is the dominant provider of CSP equipment, software, and services. It is also one of the largest sellers of switches, routers, and wireless technology for enterprises, campuses, and data centers as well as an enterprise security vendor. We are assuming that Huawei’s addition to the list is more of a short-term leverage play rather than a long-term supply chain adjustment. However, we do believe U.S. companies not supplying into Huawei could create disruptions and potential doubts surrounding Huawei’s ability to provide products sustainably; in turn, we see opportunities for Huawei competitors in the CSP and networking spaces.
Near Term May Worsen for Chipmakers
Near-term chipmaker results may be worse than previously forecast because of even greater uncertainty about end-market demand and future production, likely leading to fewer chip orders. Further, any prior calls for a recovery in China in the second half of 2019 now seem premature, although a bounce-back would be likely if trade issues are resolved. Finally, in the near term, strength in 5G chip demand at companies like Xilinx, Skyworks Solutions (SWKS), Qorvo, and elsewhere may have been a bit of a mirage, as Huawei has reportedly been stockpiling chips in early 2019 in anticipation of such a ban. Nonetheless, we still see attractive margins of safety in stocks like Intel and Skyworks, while the pullback in Xilinx appears warranted compared with our $90 fair value estimate.
Huawei is the third-largest buyer of semiconductors in the world, according to Gartner, buying over $21 billion of chips in 2018 and lagging only Samsung and Apple. We struggle to see how Huawei can maintain its smartphone business if it were cut off from U.S.-based components, specifically radio frequency chips. According to Gartner, the top four RF chip vendors that make up 82% of global revenue (Skyworks, Qorvo, Qualcomm, Broadcom) are U.S.-based, and we struggle to recall a premium smartphone over the years that doesn’t include RF from at least one (if not several) of these U.S. vendors. Such RF design expertise, especially at a global scale, underpins our narrow moat rating on Skyworks in particular. We’re skeptical that non-U.S. RF suppliers can fill the technology gaps anytime soon and provide hundreds of millions of parts to Huawei per year, particularly on high-end phones using spectrum at higher-frequency bands.
Qorvo earned 13% of revenue from Huawei in fiscal 2019 and remains an important customer. Huawei was a 10% customer for Skyworks in fiscal 2017 but not fiscal 2018, so it’s likely a high-single-digit customer today. We would also be concerned about Austrian chipmaker AMS, which counts on Huawei for roughly 5% of total sales, based on our estimates. We believe AMS has design wins ramping with Huawei during the calendar year, but AMS has said that it has no plans to stop shipments to Huawei. If the Chinese company is unable to secure the other requisite components for its smartphones before burning through its inventory, AMS would certainly be affected. However, we believe AMS supplies illumination products into Huawei’s Chinese competitors, like Xiaomi and Oppo, which may help offset the customer substitution outside China in the event the ban extends beyond the next few months.
Xilinx benefited significantly from Huawei’s stocking up of chips before the ban. Our $90 fair value estimate is unchanged, as we had been anticipating lumpy and elevated 5G spending as well as the risk that escalating tensions between the U.S. and China could result in a ban of sales to Huawei. In fiscal 2019, communications revenue was 36% of sales for Xilinx. We assume roughly 20%-25% of revenue is derived from China, with about $300 million attributed to Huawei (also artificially inflated because of Huawei stocking up).
Qualcomm has multiple 10% customers (Apple, Samsung, Xiaomi), but Huawei doesn’t make the cut because of a combination of in-sourcing more of its chips and the fact that it’s not paying full royalties currently. There could be a risk here that Chinese original-equipment manufacturers don’t buy chips or pay royalties (revenue from China was 67% of last fiscal year’s revenue). We expect near-term pressure on Qualcomm’s financial results will be at the high end of those affected in the semiconductor space.
Approximately 50% of Broadcom’s revenue comes from sales to distributors, OEMs, or contract manufacturers in China, though the end customer may not be in China. Based on its customer concentration, we think Huawei is at least a 15% customer, which would imply about $3 billion in sales from Huawei. Broadcom reports in early June, and management will probably provide revised full-year guidance (it no longer provides quarterly guidance). We maintain our $300 fair value estimate, as we don’t think the Huawei ban will be long term. Broadcom competitor Marvell Technology (MRVL) also counts Huawei as a customer for networking switches and PHYs as well as chips for 5G base stations. We believe direct sales account for roughly 5% (or approximately $150 million in fiscal 2020 revenue based on our full-year outlook), and we are maintaining our $21 fair value estimate based on our view that this is a temporary headwind.
Intel derives about 27% of its revenue from China, though Huawei is likely below 10%. Intel probably sells an assortment of processors to Huawei, spanning the data center and networking segments. Out of our large-cap semiconductor coverage, we think Intel will be relatively less affected by a ban on Huawei, thanks to its scale, broader diversity, and lack of non-U.S. alternatives. Worth noting is Advanced Micro Devices’ (AMD) joint venture with various Chinese entities (public and private): Tianjin Haiguang Advanced Technology Investment, which licenses x86 technology to China. The initial processor was launched in 2018 and is a variant of AMD’s EPYC processor. Nonetheless, for many high-end and specialized workloads, we think Chinese customers will still require Intel Xeon processors, and thus we are maintaining our $65 fair value estimate for the wide-moat chip titan.
Also supplying products for devices is Corning (GLW), whose Gorilla Glass adorns the surfaces of several Huawei smartphones, laptops, and wearables. However, we believe Huawei-related sales account for a low-single-digit percentage of revenue at Corning. Other companies that could be affected with a similar severity are NXP Semiconductors (NXPI) (NFC chips for payments and transit), ON Semiconductor (ON) (sensors for flat-panel testing), and Cypress Semiconductor (CY) (Wi-Fi and Bluetooth modules). Under the assumption that this ban is temporary and does not extend beyond the next few months, we are maintaining our fair value estimates for all of these companies. However, interface designer Synaptics (SYNA) could be sorely affected if the Huawei ban is anything but short term, as it supplies touch and display solutions into several of Huawei’s devices. We estimate sales to Huawei account for more than 10% of Synaptics’ revenue.
In storage and memory, we believe both Seagate Technology (STX) and Western Digital (WDC) have counted Huawei as a customer, with Seagate supplying storage products for enterprise data centers and Western providing flash for a variety of applications. According to a report from Nikkei Asian Review, the latter has already suspended supplying Huawei. However, we believe Huawei accounts for just 2% of Western’s total sales. We estimate Seagate’s exposure is closer to 5%, or roughly $500 million, but we are maintaining our fair value estimates of $63 for Western Digital and $40 for Seagate based on our outlook for the ban.
The indirect effects of the Huawei issue could also weigh on U.S. tech companies in the near term while negotiations drag on, and perhaps longer if no deal is reached soon. First, Apple’s China business could continue to suffer–perhaps modestly and organically if Chinese consumers no longer favor buying U.S.-branded phones, but perhaps drastically if the Chinese government were to retaliate with some sort of action that might cripple Apple’s ability to manufacture iPhones in China, regardless of where they are sold across the world. Apple derived 20% of revenue from Greater China in fiscal 2018. The company has been negatively affected by weaker iPhone demand in China (for a host of reasons, including lower switching costs and Chinese nationalism). We had previously expected iPhone sales in China to be down nearly 30% this year, which could be lower if Apple tries to pass on the cost of tariffs to end customers (or margins could be lower if Apple maintains prices).
Second, chip demand paused in the fourth quarter of 2018 as customers were reluctant to build new manufacturing facilities or expand production in China. The latest bout of trade tensions around Huawei and the day-to-day tactics of the negotiations will probably lead to another bout of caution that could weigh on June quarterly results and perhaps the September forecasts for many chipmakers. We note that Analog Devices’ forecast for the July quarter fell short of consensus expectations due to the pause of sales into Huawei. A broad-based slowdown might be negative for other analog and microcontroller companies like Maxim Integrated Products (MXIM), Microchip Technology (MCHP), Texas Instruments (TXN), NXP, and others.
Third, regardless of the outcome of these negotiations, we expect more-aggressive chip investment and development in China, as there are some gaps in the tech supply chain there that the country simply can’t fill organically. Such ramifications would initially weigh on the memory space, though even this dynamic is a few years away from materially affecting the likes of Samsung or Micron Technology (MU). We don’t think domestic Chinese manufacturers will be able to encroach on TSMC’s (TSM) or Intel’s leadership anytime soon.
On the bright side, the Commerce Department ban on ZTE in May 2018 was crushing to the Chinese tech company, which we suspect helped lead to a relatively swift resolution; the ban was lifted in July 2018. Huawei’s placement on the Entity List might be crippling, too, perhaps leading to a similarly swift resolution. The ZTE setback was relatively immaterial to many of the chipmakers we cover, serving as no more than a blip to quarterly results last year. Huawei is certainly a larger company and a more important customer, but we still see potential for the U.S. ban being another blip for chipmakers rather than a roadblock.
Nokia and Ericsson Have the Most to Gain
Nokia and Ericsson have the largest potential gains in the mobile and fixed access networks, as telecoms are evaluating their current vendors for 5G rollouts. We believe that most communication service providers will continue to multisource vendors, but the two Nordic companies could end up being favored as primary vendors instead of Huawei. Other potential gainers for mobile infrastructure are the smaller players, Samsung and ZTE. Looking at the switches and routers used by the CSPs, Cisco (CSCO) and Juniper Networks (JNPR) could benefit from CSPs evaluating alternative options.
For switches, routers, wireless, and SD-WAN products sold outside of CSPs, Cisco, Hewlett Packard Enterprise (HPE), Juniper Networks, Arista Networks (ANET), and VMware (VMW) could benefit from Huawei supply chain issues. Any impact to Huawei’s enterprise storage and servers could be a boon for Dell Technologies (DELL), HPE, and NetApp (NTAP). Also, Huawei’s cybersecurity product sales could be affected, creating upside for companies like Palo Alto Networks (PANW), Cisco, Check Point Software Technologies (CHKP), and Fortinet (FTNT).
The overall CSP market spending is led by Huawei, with 21% overall market share, followed by Nokia (14%), Ericsson (14%), ZTE, and Cisco in 2018, according to Gartner (any subsequent market share data in this article is from Gartner and is as of 2018). Dissecting the CSP spending further, Huawei had 28% market share of mobile infrastructure, CSP routing and switching, and optical transport, collectively. Nokia had 17% and Ericsson had 13% share.
The mobile infrastructure market, which contains the equipment for mobile phone networks and radio systems, was split, with Huawei controlling 30%, Ericsson 25%, and Nokia 20%. As CSPs debate their potential vendors for 5G networks, we expect them to have increased interest in non-Huawei vendors now. Although 5G spending is not expected to start until late 2019 and ramp up in 2020, which gives Huawei an opportunity to recover from any supply chain disruptions, the CSPs are performing their important 5G radio access network, software, and core/edge network equipment evaluations beforehand. We do not believe that Nokia and Ericsson have the capacity to cover any Huawei product shortfalls overnight; however, the rolling nature of 5G trials and contracts should give the Nordics ample time to address capacity concerns if Huawei remains affected.
The market for CSP routers, switches, and optical transport was led by Huawei’s 25% share, followed by Cisco’s 18%, Nokia, Ciena (CIEN), and Juniper Networks. As the CSPs update the infrastructure for 5G capabilities, the increased speeds and data flowing through the core and edge networks will need to be upgraded as well. While we believe the radio access decision could be most critical for CSPs, in our coverage, we would expect Cisco’s, Nokia’s, and Juniper’s portfolios of switches and routers designed for the next generation of communication speeds to be demanded. If CSPs favor Ericsson for the radio access networks, Juniper would benefit through its close partnership. Huawei was also the leader for the fixed-access market. As CSPs look for a broadband access product that is not Huawei’s, we see Nokia’s portfolio as a leading alternative.
From a services perspective for CSPs, Ericsson’s market share is slightly ahead of that of Huawei and Nokia. In that segment, services related to designing, building, and running the networks are split fairly evenly among these three companies. In our view, Ericsson’s and Nokia’s market shares in services and support would increase as they win more contracts related to mobile and fixed-access infrastructure.
For the enterprise switches market, Huawei had 10% of the market but still lags Cisco as the dominant switching provider with 52% share. Cisco also has a similarly dominant position in campus switching and the data center. IT spending has been strong for switching equipment as companies create more data with more networking touch points, and Huawei disruptions could make the company strongly consider where to focus its resources among CSPs, enterprises, and consumer products. Additionally, the next wave of Ethernet speed, 400Gb, could affect Huawei’s ability to be considered for trials in 2019 in places like Europe, Japan, and South America, giving upside potential to Cisco, Arista, and Juniper. Huawei is also far behind Cisco in wireless local area network equipment, and any changes in competitive dynamics in the campus switching market could also weigh on the wireless segment. Huawei could also be affected as enterprises are deploying SD-WAN to connect their branches via Internet access to cloud-based applications and centralized data centers, which would be positive for market leaders VMware and Cisco, as their products could still be used across the entire networking ecosystem.
Any Huawei supply chain issues could also affect the storage and server markets. Huawei’s share in storage is much lower than Dell‘s 33%. Huawei has been growing at a nice clip, but this speed bump could affect sustainable growth and cause enterprises to favor one of the other mentioned providers. Enterprises continue to migrate to all-flash array storage, and Huawei could be hurt during one of few instances of rapid growth in the storage industry. For servers, Huawei trails Dell and HPE. As companies invest in server hardware, Huawei would be left out of consideration, positively affecting its competitors.
Finally, in enterprise security, Huawei is behind U.S.-based companies like Palo Alto, Cisco, and Fortinet, as well as Israel-based Check Point. However, Huawei’s 5% market share is next in line, and the company was the fastest-growing year-over-year option for enterprise security. With enterprises and governments putting a heightened focus on cybersecurity, Huawei’s rapid growth could be stymied as it looks to expand its global footprint, with the other players winning by being able to deliver products. We would expect the other main firewall vendors to be able to win new and upgrade opportunities if Huawei impacts remain, which would also give a boost to cybersecurity subscriptions and services offered.