Manulife Investment's latest observations on the semiconductor industry Since 2026, semiconductors have emerged as one of the strongest-performing sectors in global equity markets, supported by both sustained demand and improving fundamentals. Recently, the Canadian asset management firm Manulife Investment published a research note stating that the expansion of artificial intelligence (AI) infrastructure is not only driving investment in high-performance computing chips but also boosting demand for network and power technologies essential for modern data center operations. Additionally, beyond AI, initial signs of stabilization and recovery are appearing in certain other industry segments, including industrial and automotive. Manulife Investment believes the supply chain is becoming more resilient and localized, supporting continued investment in semiconductor manufacturing and equipment. Opportunities exist in network connectivity, select computing power, power supply, and equipment segments, but risks such as weak consumer demand, cyclical volatility, and potential policy or geopolitical shifts also warrant attention. Since 2026, semiconductors have been a key driver of global equity market returns, outperforming the broader market. Notably, this rally is not confined to a single niche; AI computing, graphics processing units (GPUs), memory chips, analog and power semiconductors, as well as the equipment and manufacturing tools supporting chip production, have all advanced. Several factors are contributing to this momentum: First, the global build-out of AI infrastructure continues to accelerate. As more enterprises invest in computing power required for training and running AI models, demand is rising across multiple segments of the semiconductor ecosystem—not only for well-known "AI chip" makers but also for underlying connectivity technologies, power management, and manufacturing. Second, the memory industry has re-entered a more supportive cyclical phase, with a positive interplay between capacity expansion and pricing dynamics effectively improving market sentiment. Third, corporate earnings momentum has strengthened, with upward revisions to profit forecasts supporting stock prices. Fourth, the capital expenditure cycle in semiconductor equipment continues to improve, benefiting from both accelerated investment and emerging capital discipline. Furthermore, an important structural backdrop for investors is government policy support and the ongoing push for supply chain localization. As more countries prioritize domestic supply chain resilience for strategic industries, investment across the value chain is being encouraged. From an industry perspective, this trend is likely to extend the growth cycle for semiconductor equipment spending and capacity-related investments. 1. How the Semiconductor Cycle is Evolving Beyond AI While AI has been a prominent driver, it is not the only narrative. After several quarters of "digestion" in non-AI markets, many areas are stabilizing or improving. Specifically, industrial and automotive end markets appear to be in the early stages of recovery, though still below previous cycle peaks. In the industrial market, manufacturing Purchasing Managers' Index (PMI) data has improved in recent months, with major regions returning to expansion territory. Aerospace and defense spending also provides support, and long-term themes like robotics continue to offer positive catalysts. In the automotive market, reported inventory levels have retreated to near target ranges. This could allow market growth to align more closely with the underlying "semiconductor content per vehicle" growth trend. Broader restocking activity could provide additional upside. The consumer electronics market shows more mixed signals. Personal computers (PCs) and smartphones face significant challenges, partly due to rising memory prices, with forecasts suggesting double-digit year-on-year shipment declines for both categories in 2026. For investors, this reinforces a key point: semiconductors are not a single-cycle industry. Different end markets may exhibit divergent trends. 2. Supply Chain Resilience and Export Controls: Current Risk Environment is Relatively Benign While geopolitics remains a consideration for global tech supply chains, current assessments suggest the semiconductor supply chain remains robust, with risks from geopolitical disruptions considered manageable. The ongoing trend of deglobalization is expected to continue driving focus on domestic supply chain resilience. The implications of this trend are noteworthy: Building supply chain resilience is seen as beneficial for semiconductor equipment spending, as establishing and upgrading supply capabilities typically requires corresponding tools, manufacturing processes, and infrastructure upgrades. Regarding export controls, the current view is that risks have diminished compared to a year ago. While developments still require monitoring, export controls are not expected to significantly impact the overall semiconductor industry. 3. A Practical Framework for Understanding AI Infrastructure: Compute, Network, and Power To identify where opportunities may expand, AI infrastructure can be broken down into three building blocks: first, compute, the chips processing AI workloads; second, network, the "connective tissue" linking systems together; third, power management, the components and architectures supporting increasingly complex data center power needs. Within this analytical framework, the current focus is on the network domain. As AI systems grow more complex, major operators are placing greater emphasis on system resilience and connectivity. Notably, as the number of compute nodes increases, the growth rate of network connections may accelerate even faster, potentially creating a strong demand curve for network connectivity semiconductor solutions. In the compute domain, beyond GPUs, central processing units (CPUs) are also gaining attention, linked to expectations for "agentic" AI. The logic is that as AI evolves toward more autonomous, task-driven systems, CPU demand could rise significantly. Some believe agentic AI could substantially increase the CPU-to-GPU ratio over time. Power is also becoming increasingly critical. As data centers scale and technology iterates, power management challenges intensify. With the introduction of 800V architectures, semiconductor content is expected to rise notably, creating opportunities for analog chipmakers, power-related companies, and supporting supply chain segments. 4. Investment Opportunities and Risks to Watch Manulife Investment sees opportunities distributed across different geographies and segments, particularly in data center connectivity and capital equipment. The sustainability of global AI investment is the core investment thesis, supported by evolving return on investment (ROI) dynamics across the value chain. Simultaneously, opportunities are not limited to hardware. The market continues to focus on some software companies enabling AI, especially as positive factors in data management and security strengthen. This is because, in certain areas, the gap between corporate fundamentals and market narratives can be exceptionally wide, creating structural opportunities. However, risks persist. Weak consumer demand in the PC and smartphone segments is one near-term challenge. Another risk is that, despite improved market confidence, the semiconductor cycle could still turn quickly if assumptions about end demand, pricing, or capital expenditures change. Furthermore, while export control risks are seen as lower than a year ago, policies can still evolve, and no supply chain, however resilient, is entirely immune to shocks. 5. Summary In 2026, the semiconductor sector led global markets. AI infrastructure build-out, improving industry fundamentals, and structural shifts toward supply chain resilience and localization are shaping the industry's development. Concurrently, investment opportunities are expanding beyond leading AI chipmakers, extending to network, power, equipment, and select enabling companies supporting the core operations of AI data centers. Special Note: This material is for sharing and exchanging research views and does not constitute securities investment advice, investment analysis opinions, or promotional material for the company's fund products.
Comments