The chip stock bull market narrative is still “mainly rising”! The 2026 configuration theme focuses on the interconnection of AI chips, semiconductor devices, and optical devices
The Zhitong Finance App learned that the latest research report released by Wall Street financial giant Morgan Stanley shows that, driven by the unprecedented AI infrastructure boom and the strong pace of inventory removal of traditional analog chips/MCUs, the “long-term bull market logic” of chip stocks is still intact. Next year, chip stocks may still be one of the best-performing sectors in the US stock market. Nvidia (NVDA.US), Broadcom (AVGO.US), and Astera Labs (ALAB.US) are among the agency's preferred stocks in 2026.
At around the same time, a team of senior analysts from Bank of America Research (Bank of America Research), another Wall Street financial giant, said that the chip sector leading US stocks into this super bull market is expected to continue to rise in the bull market next year. They also emphasized that Nvidia and Broadcom are the chip stocks worth holding for a long time in 2026. The Bank of America said in its research report that the global AI arms race is still in the “early to middle stages”. Despite recent sharp downside fluctuations in popular chip stocks such as Nvidia and Broadcom, investors should continue to pay attention to industry leaders. For example, in addition to Nvidia and Broadcom, Bank of America is also optimistic about semiconductor device leaders such as KLAC.US (KLAC.US), Panlin Group (LRCX.US), and TER.US (TER.US).
According to the latest research reports from these two top Wall Street institutions, chip stocks, one of the core driving forces of the “US stock super bull market” since 2023, have recently regained global capital popularity after being drastically sold off, and are expected to continue to attract an influx of capital in 2026, so it is very likely that they will once again start a “mad cow” boom and become the core focus of the US stock market.
According to the 2026 semiconductor industry investment outlook recently released by these two major Wall Street giants, the three most popular investment themes can be described as the main investment lines in the chip sector that both sides are optimistic about, namely AI chip leaders, semiconductor equipment spending expansion, and leading players/leaders in the data center optical interconnection industry chain.
The main chip investment line proposed by Damo and Bank of America can be said to be closely related to the surge in demand for AI GPU/AI ASIC computing power clusters under the AI infrastructure frenzy since 2023. Demand for AI chips led by Nvidia and Broadcom continues to boom. At the same time, TSMC, Samsung, Micron, and SK Hynix continue to spend huge sums to buy advanced semiconductor equipment to expand production capacity. Optical connectivity is not a “supporting role in the AI narrative”, but rather a “performance delivery main line” in 2026-2027 where supply and demand are still tight and there are clear intergenerational upgrades. Whether it's Nvidia-led” The “InfiniBand+ Spectrum-X/Ethernet” high-performance network infrastructure+GPU cluster, or the Google-led “OCS (Optical Circuit Switching)” high-performance network infrastructure+TPU cluster are all inseparable from data center optical interconnection/optical module technology solutions.
The semiconductor boom cycle is far from over, and there is still plenty of room for the upside of chip stocks that have been rising like a rainbow
A team of analysts at Morgan Stanley wrote in a report to clients: “For three years in a row, the biggest controversy in the market was AI semiconductors. The index weight in the chip field was dominated by AI chip companies, and so far, the world's almost insatiable appetite for AI computing power is the most important variable.”
“We are skeptical about some five-year views — they make the strength of 2025 seem as simple as rounding up errors — in these views, industry players plan to spend several times their private market valuation on data center computing power infrastructure. The market's skepticism may be incorrect because the pricing trend and logic of market capitalization are constantly evolving, but even if the long-term outlook is in the right direction, we do expect a phased digestion cycle along the way, and we do not rule out that this positive cycle will affect the market pricing logic.”
“We see 2026 as the midpoint of an 8-10 year journey of upgrading traditional IT infrastructure to accommodate accelerated and AI workloads.” “Greater scrutiny of AI returns and cash flows from hyperscale cloud service providers may cause share price fluctuations, but this will be offset by updated/faster LLM developers and AI factories serving enterprise and sovereign customers. “We expect semiconductor sales to move towards the first $1 trillion in 2026, growing by about 30%, while fab equipment sales will grow by nearly double digits year over year.” Bank of America analysts said in the research report.
When first-tier Wall Street seller institutions such as Dama and Bank of America still list the chip sector (especially the AI chip industry chain/customized ASIC/semiconductor equipment chain) as the core position direction for 2026, it essentially reflects that the market's mainstream framework is still — AI capital expenditure and AI basic computing power requirements are likely to continue to provide an upward “narrative base” in 2026.
Compared to the slightly more cautious approach, Bank of America's statement is more direct: it still regards the AI competition as an “early/mid-term stage,” and it is expected that the AI semiconductor chip segment is still likely to maintain 50% + year-on-year growth drivers in 2026 (high data center utilization, tight supply, corporate adoption, LLM/cloud vendor/sovereign customer competition, etc.).
According to the latest semiconductor industry outlook data recently released by the World Semiconductor Trade Statistics Organization (WSTS), the global chip demand expansion trend is expected to continue to be strong in 2026, and MCU chips and analog chips, which have continued to weaken since the end of 2022, are also expected to enter a strong recovery curve.
WSTS expects that after a strong rebound in 2024, the global semiconductor market will grow by 22.5% in 2025, with a total value of 772.2 billion US dollars, which is higher than the forecast given by WSTS in spring; the total value of the semiconductor market in 2026 is expected to expand significantly to 9755 billion US dollars on the basis of strong growth in 2025, close to the market size target of 1 trillion US dollars in 2030 predicted by SEMI, which means it is expected to increase 26% year over year.
WSTS said that this strong growth trend for two consecutive years will be mainly due to continued strong momentum in the logic chip field dominated by AI GPUs and the storage sector dominated by HBM storage systems, DDR5 RDIMM, and enterprise data center NAND. Both fields are expected to achieve extremely strong double-digit growth, thanks to the continued strong expansion of demand in fields such as artificial intelligence inference systems and cloud computing infrastructure.
According to Wall Street giants Morgan Stanley, Citi, Loop Capital, and Wedbush, the global AI infrastructure investment wave with AI chip computing power hardware as the core is far from over and is only at the beginning. Driven by an unprecedented “AI inference computing power demand storm”, the scale of this round of AI infrastructure investment, which will continue until 2030, is expected to reach 3 trillion to 4 trillion US dollars.
The main line of chip stocks that both Dama and Bank of America are optimistic about: AI chips, semiconductor devices, and optical interconnections
Nvidia and Broadcom are Damo's two preferred stocks in the semiconductor market segment of AI chips, and are also Damo's preferred stocks in the overall chip stock market. Analysts from Damo explained, “At the time of writing this article, the market's enthusiasm for AI ASICs was continuing to heat up, and the market is betting that growth will be very strong, but as various ecological bottlenecks emerge, we still believe that the Nvidia AI GPU computing power cluster will still be the preferred solution for the highest ROI in the cloud, especially when the Nvidia Vera-Rubin architecture begins rolling out in the second half of 2026.” “Despite the limited leverage, we think the market has undervalued Nvidia's monopoly position.”
The agency also gave AMD (AMD.US) and Maywell Technology (i.e. Marvell, MRVL) a cautious bullish rating of “Equal-Weight” (Equal-Weight), believing that the two have “significant upside”, but there is also some uncertainty. Given its long-standing pessimistic stance on Intel's foundry business, Damo is more cautious about Intel (INTC.US). Astera Labs is the agency's most promising small-chip shareholder in the field of data center optical interconnection. It is particularly optimistic that the company will advance rack-level/scale-up optical interconnect ecosystem solutions through the acquisition of AIXscale, which emphasizes that the company is strongly entering the direction of optical interconnection/silicon photonics technology within data centers.
In the field of storage and semiconductor equipment, Morgan Stanley said that as the production capacity of AI chips and high-end memory chips continues to rise and fall short of market demand, it is likely to keep the supply of storage and logic wafers tight for a long time. As a result, Micron (MU) is the agency's preferred target in the memory chip market, while also giving SNDK.US (SNDK.US), an enterprise-class SSD leader, a “Increased” (Increased) rating.
In the field of semiconductor equipment and chip manufacturing, Applied Materials (AMAT.US) and TSM.US (TSM.US) are two of Morgan Stanley's most optimistic chip stocks. Damo and Bank of America said that as the global AI infrastructure construction process led by tech giants such as Microsoft, Google, and Meta becomes more intense, comprehensively helping to greatly accelerate the expansion of 3nm and below advanced manufacturing chip production capacity, the long-term bull market logic in the semiconductor equipment sector is still very strong. The various high-end manufacturing equipment introduced by semiconductor equipment manufacturers is critical to the AI training/inference related chip manufacturing capacity of chip manufacturing giants such as TSMC, Samsung, and Micron.
All the news about the production capacity of AI chips in advanced catalytic processes can be described as positive and positive for semiconductor devices. The agency said that the CPU/GPU package heterogeneity (based on NVLink high-speed interconnect and advanced chiplet packages such as CoWS/Emib/FoverOS) will greatly boost the structural requirements for EUV/high-NA lithographs, advanced packaging equipment, high-end etching equipment, customized deposition film equipment, and inspection and measurement, which is particularly beneficial to Terida, Fanlin, application materials, and Ke Lei.
According to Damo, the narrative basis for the analog chip market segment is roughly the same as the pricing logic of 2025 — that is, showing improvements, but at a slower pace. Therefore, the agency is optimistic about NXP Semiconductors (NXPI.US), believing that it is “the best combination of growth and value models in chip stocks”; while Adderall Semiconductors (ADI.US) is more expensive, it has “stronger growth potential in the analog chip field.”
In addition to Nvidia, Broadcom, and Fanlin Group, Bank of America favors KLAC.US (KLAC.US), Adderall Semiconductor (ADI.US), and CDNS.US (CDNS.US) in large-cap chip stocks; in small and medium stocks, it favors Credo Technology (CRDO.US), MKS (MKSI.US), Macom Technology Solutions (MTSI.US), and Advanced Energy Industries (AEIS.US).
In the field of data center optical interconnection, Bank of America stated that Lumentum (LITE.US) and Coherent (COHR.US) are regarded as absolute leaders in the field of optical interconnection. Google has embedded OCS (optical circuit switch) clusters into the Jupiter/AI data center network system on a large scale to support TPU AI systems and large-scale training/inference systems. Lumentum's OCS products such as R300/R64 are specifically aimed at “large-scale cloud computing scale + AI/ML data center networks”: using MEMS optical circuits to directly establish optical connections between endpoints, bypass intermediate power exchange and OEO conversion, focusing on high port numbers, low latency, and low power consumption. At the same time, Lumentum is also an important supplier of high-speed digital optical modules and optical interconnect chips such as 400G/800G. Officially, these products have been positioned as core devices that “provide expandable interconnection bandwidth for AI and hyperscale cloud data centers.”
The reason why Lumentum is one of the biggest winners of Google's AI explosion is mainly because it does exactly what it does is an indispensable lack of light interconnection in the “high-performance network base system” that is deeply bundled with Google's TPU AI computing power cluster — that is, OCS (optical circuit switch) + high-speed optical devices. Every time the number of TPUs is increased by one level, its shipments go up. Needless to say, the entire optical communication industry chain, including optical modules, is benefiting, because the optical interconnection AI model training/inference based on heavyweight computing power is essentially “weaving 100,000 class computing power chips into one machine”. The growth rate of network bandwidth and number of ports is no less than that of AI chips themselves. This is why optical modules have exploded in China's A-share market recently.