Morgan Stanley Narrows 2026 Tech Picks as AI Growth Enters Volatile Phase

Morgan Stanley has trimmed its technology buy list for the coming year, underscoring continued faith in artificial-intelligence semiconductors while flagging a bumpier growth path after two years of outsized market returns.

In a research note dated December 22, the bank told clients that AI chips remain the sector’s key engine, but it advised against assuming the uninterrupted spending trajectory that dominated 2023 – 2025. Analysts cited “uneven expansion, consolidation periods, and occasional slowdowns” as likely features of the next market phase, steering investors toward companies with visible earnings streams and defensible competitive positions.

Market backdrop shapes a more selective stance

The call follows a period of extraordinary equity gains. The S&P 500 delivered total returns of 26.3 percent in 2023, 25 percent in 2024, and more than 16 percent through mid-December 2025, translating into an approximately 86 percent cumulative advance since 2023. At the same time, the group of mega-cap technology names popularly known as the “Magnificent Seven” swelled to roughly one-third of the benchmark’s weight, with some estimates placing the share as high as 37 percent in October.

Morgan Stanley’s strategists argue that the next leg of performance will rely less on multiple expansion and more on consistent profit delivery. They project the S&P 500 to reach 7,800 by the end of 2026—about 14 percent above last Friday’s 6,834.50 close—driven primarily by an “earnings grind” rather than a repeat of valuation-driven rallies. The forecast implies another year of positive, though more moderate, returns versus the recent boom.

The bank’s tempered outlook is also informed by broader macro considerations. While demand for computing power continues to rise at what the report calls an “impeccable pace,” executives acknowledge that corporate technology budgets can tighten in response to higher funding costs or shifting economic conditions. Those crosscurrents, the note says, warrant a focus on companies that can translate AI demand into measurable revenue and cash-flow gains even during intermittent spending pauses.

Refined list of recommended names

With that backdrop, Morgan Stanley highlighted six areas it believes remain mispriced:

• AI processors: Nvidia, Broadcom
• Data-center connectivity: Astera Labs
• Memory: Micron Technology
• Equipment & manufacturing: Applied Materials, Taiwan Semiconductor Manufacturing Co.
• Analog chips: NXP Semiconductors, Analog Devices

The selections represent businesses that, in the bank’s view, are positioned to capture incremental AI-related spending without relying on aggressive valuation assumptions. Nvidia and Broadcom anchor the processor segment, reflecting their leadership in data-center graphics processing units and custom accelerators. Astera Labs appears on the list for its high-speed interconnect solutions, which the firm considers essential as cloud providers upgrade internal networks to accommodate larger model sizes.

On the memory side, Morgan Stanley expects a rebound in dynamic random-access memory pricing to bolster Micron’s margins. For equipment suppliers, Applied Materials and Taiwan Semiconductor are seen as beneficiaries of sustained process-technology investments, even if capital-expenditure growth moderates. NXP Semiconductors and Analog Devices round out the group, with the bank citing steady demand in automotive and industrial end-markets that are less sensitive to short-term AI spending cycles.

Imagem: financial planning 23

Volatility, not collapse, anticipated

Despite signaling caution, Morgan Stanley rejected what it labeled “delusional forecasts” that call for a swift end to the AI spending wave. Instead, analysts envisage a multi-year build-out marked by alternating surges and plateaus, similar to past infrastructure upgrades. The note compares the likely pattern to the aftermath of the early-2000s broadband boom, when select suppliers continued to grow even as headline sector numbers vacillated.

Historical evidence supports the idea that tech profits can expand without constant multiple inflation. According to data compiled by S&P Dow Jones Indices, earnings contributions from information-technology constituents have risen steadily over the past decade, outpacing the index’s overall revenue growth.

Against that backdrop, Morgan Stanley sees scope for orderly earnings expansion across its preferred names. The bank expects bottom-line growth to be led by AI-driven efficiency gains at hyperscale customers, which should translate into sustainable orders for semiconductor manufacturers and equipment vendors. Because the thesis rests on profitability rather than valuation resets, strategists describe the upside as “sizeable but grounded,” with fewer assumptions about perpetual double-digit revenue acceleration.

Key metrics to monitor in 2026

Looking ahead, the firm advises tracking capital-expenditure disclosures from major cloud providers, lead-time data from chip-equipment makers, and memory-pricing trends as early indicators of potential slowdowns or re-accelerations. It also highlights inventory levels across semiconductor supply chains as a gauge for demand health. While the research team remains constructive, it emphasizes that longer build cycles could translate into periods of consolidation for sector indices.

For investors seeking exposure, Morgan Stanley recommends maintaining positions in the highlighted chip leaders while avoiding less established names that depend on rapid multiple expansion. The report concludes that, although the AI theme is “entering a more disciplined phase,” the underlying demand for computing power is strong enough to support further gains—provided investors temper expectations for a straight-line advance.

Crédito da imagem: ANGELA WEISS/ Getty Images