Morgan Stanley’s (MS) downward revision of Applied Materials’ (AMAT) price target does not deny AI-driven demand for semiconductor equipment, but rather reflects a cautious judgment on the alignment between valuation premium and room for gross margin improvement. Earnings forecast upgrades and valuation multiple cuts proceed in parallel, reflecting the market’s search for a new balance between continued prosperity and prices that have already priced in expectations.
At a time when Meta Muse and the earlier-released GPT-6 Astra jointly reinforce market expectations for AI agent commercialization and AI computing power expansion, significantly pushing up the earnings trajectory of the semiconductor equipment sector, Morgan Stanley instead cut Applied Materials’ price target from $642 to $563, a reduction of $79, or approximately 12.3%. The institution simultaneously raised its 2027 revenue and earnings per share forecasts for Applied Materials and maintained its “Neutral/Hold” rating. The signal released by this combination is very clear: Morgan Stanley acknowledges earnings growth but believes the market has already priced in a considerable portion of the positive factors in advance, and the institution prefers semiconductor companies whose expectations have not yet been fully reflected in valuations.
The core of Morgan Stanley’s adjustment this time is cutting the valuation multiple applied to the stock from 26 times to 22 times, a reduction of approximately 15.4%. Even so, 22 times is still about 10% above Applied Materials’ cross-cycle average valuation of approximately 20 times since 2020, indicating that Morgan Stanley still recognizes a certain growth premium. The controversy centers on how large the premium should be. The institution expects Applied Materials’ gross margin to rise from 48.8% in 2025 to 52% in 2027, but remains cautious about whether it can significantly break through around 55% in the future. Analysts said that the valuations of semiconductor equipment companies are usually closely related to gross margins, and the room for margin improvement needs to match the valuation the market assigns.
From an engineering logic perspective, agents expand a single question-and-answer into planning, retrieval, tool invocation, code execution, and result verification. Multi-turn reasoning increases computing demand, background execution increases CPU load, and runtime state and data retention further consume memory and storage. Applied Materials management also pointed out that agent applications require a more CPU-intensive computing architecture and increase demand for DRAM and NAND. In chip manufacturing, Applied Materials’ high-end equipment covers atomic layer deposition, chemical vapor deposition, physical vapor deposition, rapid thermal processing, CMP, wafer etching, ion implantation, and other steps, and extends to advanced packaging equipment. The company previously disclosed that major customers provide rolling eight-quarter demand forecasts and plan to increase quarterly equipment system output to twice the current level by 2028. Morgan Stanley estimates that if quarterly equipment system shipments reach approximately $14 billion by mid-2028, this could correspond to approximately $68 billion in revenue and $31.70 in earnings per share.
Despite the price target cut, Applied Materials’ stock still closed up 0.36% at around $486.76 as of Monday’s close, showing a certain resilience. Morgan Stanley analysts said the market is willing to give Applied Materials trust amid remaining uncertainty, and the institution agrees with this judgment, but prefers companies whose expectations are not so fully reflected in valuation multiples. In Morgan Stanley’s view, the semiconductor equipment market has entered a “valuation downshift zone”—even as industry demand and company profits continue to grow, stock pricing may still come under pressure as the market lowers valuation multiples. Investment judgment therefore needs to simultaneously weigh order fulfillment, earnings upgrades, and premium expectations already reflected in prices.
