Analysts from several large brokerages issued material coverage changes and guidance updates this week for companies they view as strategically positioned - or challenged - by the wave of AI-driven IT investment. Their decisions span the semiconductor equipment supply chain through to enterprise software, and reflect a mix of bullish convictions and cautionary warnings tied specifically to AI demand.
UBS: ASML reinstated as top Europe semiconductor pick and price target raised
UBS has brought ASML back as its preferred company in the European semiconductor equipment space, increasing the price target to 00 from 600 and raising earnings estimates for 2027 and 2028 to levels the bank says sit well above consensus. The bank framed the call in part as a response to the stock's underperformance this year versus peers.
UBS's valuation work showed ASML trading at a 6% premium to large-cap U.S. peers on a 12-month forward price-to-earnings basis, compared with a 10-year average premium of 84%. The analysts, led by Francois-Xavier Bouvignies, described ASML as offering "the most attractive risk/reward in the sector."
The bank set out three pillars underpinning its bullish stance. First, UBS disputed market concerns that ASML could become a supply bottleneck for advanced semiconductors, concluding those fears are overplayed. UBS estimated that ASML's 2027 capacity could support more than 50% year-on-year growth in leading-edge wafer output, compared with projected demand growth of roughly 25-30%.
UBS also acknowledged that productivity improvements across manufacturing steps could modestly reduce lithography intensity, but the analysts said they do not expect ASML to be a constraint over the next 12 to 18 months.
Second, UBS highlighted ASML's exposure to memory markets as an underappreciated positive. The bank noted ASML is "the most memory-exposed semi-cap name," forecasting roughly 30-35% of revenues tied to memory by 2026 versus 25-30% for U.S. peers. That exposure has already translated into stronger growth historically, with ASML posting a 23% memory revenue compound annual growth rate between 2020 and 2025 versus about 6% for peers, according to UBS. The bank expects this trend to persist as DRAM node shrinks increase lithography intensity through 2028.
Third, UBS maintained confidence in ASML's High NA EUV next-generation lithography technology despite some delays in adoption by a major customer. The bank estimated that High NA could deliver 20-40% cost savings on critical layers compared with alternative patterning methods, and anticipated adoption arriving within two to three years. On that basis UBS concluded ASML remains an attractive European semiconductor name and restored it as its sector top pick.
Evercore: Dell stands out among hardware and networking names ahead of results
Evercore ISI designated Dell as its top pick among hardware and networking companies heading into quarterly earnings for Dell, HP Inc., Hewlett Packard Enterprise and NetApp. The firm argued that demand for AI infrastructure and networking appears durable and that higher average selling prices should support revenue upside across the group.
In advance of April-quarter releases, Evercore said it views AI infrastructure and networking spending as among the most resilient categories of IT investment, citing higher hyperscaler and neocloud capital expenditures alongside continuing enterprise investment in campus networking and data center modernization.
Evercore identified Dell and HPE as companies likely to report upside and raise full-year guidance. For Dell specifically, the broker flagged potential incremental AI server revenue driven by neocloud and enterprise AI budgets. In HPE's case, Evercore pointed to networking upside tied to Aruba and Juniper as a key driver.
Evercore also pointed to positive near-term trends in the personal computer market. Citing IDC data, the broker noted approximately 3% year-on-year PC shipment growth, and observed that higher average selling prices should contribute an additional tailwind.
Bank of America: Salesforce coverage reinstated as Underperform on structural AI-driven reset
Bank of America returned to coverage of Salesforce with an Underperform rating and a $160 price target, arguing the company is undergoing a structural reset tied to AI rather than experiencing a transient cyclical slowdown. Analyst Tal Liani framed Salesforce's trajectory as a transition from a historically high-growth platform to a more mature cash-generating business, and modeled revenue growth near 10% annually going forward - a marked deceleration from the 18-28% growth rates recorded from fiscal 2020 through fiscal 2023.
Liani anchored the argument on three structural issues: weak net new customer additions, deteriorating upsell dynamics, and an unclear monetization path for Agentforce, Salesforce's AI product. He acknowledged Agentforce's headline metrics - 23,000 customers and $800 million in annual recurring revenue, growing 169% year-over-year - but said the product's penetration remains limited. Only about 9-10% of Salesforce's 200,000-plus customers have purchased paid Agentforce deals, and more than 60% of bookings for the product came from existing customers rather than new logos.
The analyst also warned of a structural tension between Agentforce's automation capabilities and Salesforce's traditional seat-based subscription model. As Agentforce automates tasks such as lead qualification and service case resolution, it could reduce the number of human users requiring Salesforce subscriptions, potentially pressuring the seat-based model that historically supported growth.
On competition, Liani cited expansion by several vendors into adjacent workflows and capabilities, including one company broadening into CRM-adjacent enterprise workflows, another deploying agent-based orchestration platforms above the application layer, and others competing in marketing and commerce. While those firms often target different subsegments, Liani said the net effect increases overlap and may exert pressure on growth and pricing power.
Liani did not predict widespread customer churn, noting that large enterprises remain entrenched on the platform. However he argued that entrenchment alone does not constitute a growth strategy, pointing out that roughly 90% of the Fortune 500 is already on Salesforce, which limits the addressable market for new customer wins.
Jefferies: ZoomInfo downgraded as AI disruption prompts business model pivot
Jefferies reduced its rating on ZoomInfo to Hold from Buy and cut the price target to $4 from $12, reflecting what it described as a challenging pivot away from subscription-based revenues toward a usage-based model amid weakening client demand and AI-driven disruption. Analyst Surinder Thind projected revenue declines of 4% in 2026 and a further 3% in 2027, reversing prior expectations for low-single-digit growth.
Jefferies said customers are increasingly building their own workflows around ZoomInfo's data rather than paying for software on a seat-license basis. The company lowered its 2026 revenue outlook to a midpoint of approximately $1.195 billion, below its prior guidance range of $1.247 billion to $1.267 billion and below consensus of $1.259 billion, according to the bank.
Management has outlined a three-part restructuring plan: move clients to consumption-based pricing with a hybrid rollout expected in the third quarter, shift to a largely product-led go-to-market approach to reduce reliance on more expensive sales-led processes, and reduce headcount by 20%, or roughly 600 roles.
Thind acknowledged the Jefferies downgrade was late relative to the stock's roughly 64% decline this year, but said there was no near-term catalyst visible to justify staying constructive. He recommended a cautious stance until there is greater clarity on the transition and its probability of success.
Morgan Stanley: Nokia's target raised and named top pick for AI data-center networking exposure
Morgan Stanley lifted its price objective for Nokia to 4 from 1 and reaffirmed the company as its top pick, arguing Nokia is well placed to benefit from increased data-center capital spending tied to AI and cloud expansion. The bank framed Nokia's shift from traditional mobile technology toward supplying data-center optical networking equipment as the central thesis for a potential re-rating.
Analysts led by Terence Tsui noted Nokia's AI and cloud revenues totaled 1.1 billion in 2025, which they described as well below peers but a potentially favorable starting point. Morgan Stanley argued that additional orders could materially move the revenue base and that the rate of change can be meaningful when beginning from a low level.
Morgan Stanley highlighted a recent company upgrade to guidance for its optical and IP networks unit - to 18-20% growth from a prior 10-12% range - and said its own forecasts sit above even that revised range, at 21% growth. The bank projected 2028 operating profit of .65 billion, versus the top end of company guidance at .2 billion.
The bank also emphasized Nokia's scarcity value in connectivity and networking within the AI story in Europe, noting relatively few Western suppliers operate directly in that segment. That positioning, Morgan Stanley said, has attracted investor attention.
What these moves collectively signal
Taken together, the coverage changes illustrate how analysts are parsing AI-related spending across multiple parts of the technology stack. They signal conviction that certain hardware and networking vendors can capture durable demand from hyperscalers, neoclouds and enterprise modernization projects, while also drawing attention to software vendors facing structural questions as AI alters product economics and customer behavior.
Investors weighing these analyst moves will see a mix of positive re-ratings based on capacity, product road maps and market positioning - and more cautious stances where AI appears to be reshaping monetization rather than only expanding addressable markets.
Bottom line
This week brought a string of analyst actions that reflect differing assessments of how AI investment will flow through semiconductors, servers and networking to enterprise software and data services. UBS, Evercore and Morgan Stanley expressed bullish views on ASML, Dell and Nokia respectively, while Bank of America and Jefferies raised caution on Salesforce and ZoomInfo as those companies confront structural shifts tied to AI adoption.