Applied Materials, Inc. (AMAT) Stock Forecasts

Summary

The AI Trade Turns Selective In the months following the launch of ChatGPT in fall 2022, investors piled into Nvidia, the perceived chief enabler of the AI revolution. Within a year, they were willing to invest in any stock that looked like an AI participant. Real technological transformation is called creative destruction for a reason, however, and over the past year, the market has become more discerning about the long-term implications of this transformation. Investors are now trying to determine which companies will flourish on the way to building AI infrastructure and thrive after it is built – and which companies risk seeing their hard-won market share and enterprise value wiped away by all the proliferating new models. Winners and Losers in the Tech Sector In the Information Technology sector, the industry shorthand is that technology hardware companies – particularly semiconductors and semiconductor equipment – are the current and long-term beneficiaries of the AI revolution. Simultaneously, software companies are seen, however accurately or inaccurately, as near-term and particularly long-term losers. For the 2026 year to date as of mid-February, the Information Technology sector within the S&P 500 was down 5% overall. There is a great deal of in-sector industry disparity, however. On a year-to-date basis as of mid-February, application software is down 27%; systems software is down 17%; and internet services is down 6%. At the other end of the industry spectrum, semiconductor materials & equipment are up 35% year to date. Electronic components are up 21%, and electronic equipment instruments are up 19%. Communications equipment and EMS (electronic manufacturing services) are both up 6% year to date. Semiconductors are up just 1% year to date as former leaders in the mega-cap space (i.e., NVDA) take a pause. AI creates enormous potential for efficiency, improved processes, targeted messaging, and much more. As technology scales to the opportunity, rising performance and lower power consumption are reducing cost per token (i.e. per unit of processed data) and turning more and more bystanders into industry participants. As more and more enterprises embrace AI, the positive case for semiconductors becomes more compelling. AI begins with compute implementations comprised of ever-bigger GPU clusters. In order to scale, fine-tune, disseminate, secure, and generally manage AI, an entire web of infrastructure – most notably networking and data storage – must be created and perpetually refreshed and improved. Semiconductor makers and the semiconductor capital equipment industry are providing the technology that enables ever more efficient AI implementations, in turn widening the pool of potential and actual enterprise participants. In the GPU space, Nvidia’s nearest competitor is AMD. Other semiconductor companies are producing XPUs, which share the massively parallel processing power of GPUs while being able to address specific tasks (hence, the X variable). Giant cloud service providers such as Google Cloud, AWS, and Meta Platforms design TPUs (tensor processing units) and other GPU-like processors built in merchant fabs. Beyond GPU compute, AI data storage requires high bandwidth memory in high-capacity drives, while AI networking requires high-end chips up and down the network stack. World Semiconductor Trade Statistics (WSTS), an industry trade group, estimates that global semiconductor revenue including semiconductor capital equipment sales rose over 20% in 2025 to the $750 billion range. WSTS looks for mid-20% growth in 2026, which would bring all-in semiconductor revenue to the $1 trillion range for the current calendar year. A few years ago, McKinsey & Co. stunned investors with a report forecasting that global semiconductor sales would reach $1 trillion — by 2030. The claim seemed outlandish in its optimism at the time; it now seems outlandish from the opposite perspective. Gartner is forecasting that worldwide AI spending will total about $2.5 billion in 2026. Using that forecast and the WSTS estimate suggests that semiconductor spending will comprise about 40% of the total spend. A few years ago, when cloud computing was becoming ascendant, received industry wisdom was that tech hardware was dying. That hardly seems to be the case in the age of AI. The corollary of ‘hardware is dying’ is ‘software is thriving’ — also seemingly not the case in the age of AI. The iShares Expanded Tech-Software Sector ETF (IGV) was down 24.3% year to date as of 2/17/26. The ETF’s top holdings include Microsoft, Palantir, Oracle, Salesforce, Intuit, Adobe, CrowdStrike, and more. The selling extends back into 2025, with many of those stocks down 30% to 60% over the past six months Large language models (LLMs), and particularly multi-modal and frontier models, are perceived as being able to perform the tasks of leading software programs more cheaply. AI models can write code, rapidly create new programs, and update existing programs. Anthropic’s Claude Cowork model is seen as threatening application software firms, which make products for creating and managing spreadsheets for payrolls and other tasks. The investment community is divided as to whether the selling in software stocks is just getting started or has already created an attractive oversold opportunity. Top-tier technology companies typically are more than their existing products. These companies have long-term customer relationships; are often deeply integrated into their customers’ most valuable processes; and integrate within a security stack that is proprietary or managed with other vendors. Every IT implementation is only as strong as its weakest link. Enterprises will need to think long and hard before discarding the SaaS vendors that have so far accompanied their success. Sector Rotation Continues Beyond the inter-sector dynamics in Information Technology, sector rotation continues to favor defensive, interest-rate-sensitive, cyclical, and inflati