Investing.com — Here are the biggest analyst moves in the area of artificial intelligence (AI) for this week.
InvestingPro subscribers always get first dibs on market-moving AI analyst comments. Upgrade today!
Phillip Securities downgrades Nvidia stock
Phillip Securities cut its rating on NVIDIA Corporation (NASDAQ:) on Friday from Buy to Accumulate, citing recent price movements in the chipmaker’s stock. The firm raised its price target for Nvidia slightly, increasing it to $160 from $155.
“We downgrade [from] BUY to ACCUMULATE due to recent price movements, with a higher target price of US$160,” noted analyst Yik Ban Chong.
Phillip Securities highlights that Nvidia’s fiscal Q3 2025 results were in line with the firm’s expectations, with revenue exceeding Nvidia’s guidance by 8%. Profit after tax and minority interests (PATMI) grew an impressive 109% year-over-year.
In its note, the investment firm emphasized that roughly half of Nvidia’s data center sales come from hyperscalers, with the remainder driven by enterprises and sovereign clients. Production of the company’s Blackwell chips is expected to begin in Q4 2025, with projected revenues exceeding initial forecasts of “several billion dollars.”
Nvidia anticipates that the initial rollout of Blackwell will yield gross margins in the low-70% range, eventually improving to mid-70% as production scales up.
While the rating downgrade reflects short-term price movement, Phillip Securities has maintained its fiscal year 2025 revenue and PATMI estimates. The firm has also raised its forecasts for fiscal year 2026 revenue and PATMI by 5% and 7%, respectively.
These adjustments account for a stronger-than-expected ramp-up of Nvidia’s data accelerator platforms, including Hopper and Blackwell, alongside reduced corporate tax rates.
The firm also adjusted fiscal year 2026 margin assumptions, aligning with Nvidia’s guidance for lower margins due to the Blackwell product launch, while keeping its weighted average cost of capital (WACC) and growth rate assumptions unchanged.
Bernstein sees Apple hitting $290 in bull-case scenario
Bernstein analysts suggest that Apple (NASDAQ:) shares could reach $290 in their bull case scenario.
The firm regards Apple as “a quality compounder, with mid-single digit revenue growth, improving margins, disciplined capital return, and double-digit earnings per share (EPS) growth.”
“Given its negative cash conversion cycle, the stock is less expensive than it appears. Investors have fared well by maintaining AAPL as a core holding, and adding to positions on pullbacks,” added analysts led by Toni Sacconaghi.
Apple’s market position is highlighted by its ecosystem of over 2.3 billion devices and nearly one billion “unique, demographically attractive users.” Bernstein also identifies Apple as a key beneficiary of AI advancements in two significant ways.
First, the firm anticipates an accelerated replacement cycle for Apple products, potentially beginning in fiscal year 2026. Second, they see AI integration driving new revenue streams for Apple, particularly through the distribution of large language models (LLMs) and third-party applications.
“Encouragingly, given its position as a channel/platform, Apple’s capex has remained low. A key question is whether AI could structurally alter iPhone’s replacement cycle,” the analysts noted.
They also observed Apple’s distinct seasonal trading patterns, cautioning that while the iPhone 16 cycle might underperform, any drop in the stock price to $200 or below—particularly between February and April—would represent a buying opportunity.
Bernstein’s bull case envisions Apple achieving $9 in EPS by fiscal year 2026, which would value the stock at $290.
Nvidia earnings good for TSMC, BofA says
Nvidia’s strong third-quarter performance and outlook are bolstering the positive sentiment for Taiwan Semiconductor Manufacturing (NYSE:), according to Bank of America (NYSE:) analysts.
“The results underscore robust structural AI demand, with limited digestion periods as adoption accelerates,” said analysts led by Brad Lin.
They highlighted Nvidia’s steady one-year cadence for data center GPU advancements, which has proven advantageous for TSMC by driving average selling price (ASP) content growth.
“The continuous scaling of Al models strengthens TSMC’s leading-edge node demand and industry leadership. Additionally, we are pleased to learn NVIDIA’s robust GPM outlook into CY2025, which reflects the stronger value TSMC offers to clients, vs ASP hikes,” the analysts added.
AI demand continues to outpace supply, with Nvidia’s Hopper and Blackwell GPUs facing ongoing constraints. Bank of America notes that demand for Blackwell is expected to remain above supply into fiscal 2026, spurred by significant investments in AI—an encouraging signal for TSMC’s prospects.
Meanwhile, TSMC continues making efforts to address its Chip-on-Wafer-on-Substrate (CoWoS) bottlenecks by ramping up production. The company plans to expand its CoWoS capacity from 35-40,000 units per month in Q4 2024 to over 80,000 units by the end of 2025.
“As AI models grow in complexity and require greater computing power, we believe TSMC is well-positioned to meet these demands,” Lin and his team concluded.
SMCI too important for AI to be delisted: Lynx
Shares of Super Micro Computer Inc (NASDAQ:) soared sharply this week after the AI server maker unexpectedly introduced an independent audit firm and submitted a compliance plan to NASDAQ. The move aims to prevent the company’s delisting, though formal approval from NASDAQ and an extension for its 10-K filing are still pending.
An analyst at Lynx Equity Strategies, however, views NASDAQ’s approval as likely a “formality.”
In a recent note, Lynx analyst KC Rajkumar highlighted SMCI’s importance in the AI data center market, stating it is “just too important a player in the AI data center space for it to be allowed to delist and go fallow.” He emphasized that delisting, which would cut off access to capital, was a low-probability event, despite regulatory uncertainties.
The analyst also pointed out the stock’s substantial discount, setting a price target (PT) of $45. After SMCI’s 8-K filing, which significantly lowers the delisting risk, Rajkumar now predicts a “vicious short-squeeze” could follow, with shares potentially reaching his price target soon.
“SMCI has a leadership position in the rapidly expanding liquid-cooled GPU server data center market, a position it is unlikely to give up any time soon,” Rajkumar said.
Raymond James upgrades HPE to Strong Buy
Earlier in the week, Raymond James upgraded Hewlett Packard Enterprise Co (NYSE:) to Strong Buy, citing optimism about the company’s refined business model. The shift, which now distinctly separates AI platforms from traditional servers, led the investment bank to raise its sales estimates for fiscal year 2025.
HPE is expected to deliver in-line financial results when it reports on December 6, though the federal sector could pose risks. Despite this, Raymond (NS:) James forecasts growth to accelerate in the fiscal year 2025.
The firm’s analysts also expect the Juniper Networks (NYSE:) acquisition to close as planned, a move that could enhance HPE’s stock valuation multiple.
Raymond James projects robust growth in HPE’s AI server sales, increasing from $4.1 billion in fiscal year 2024 to $5.9 billion in FY25 and reaching $7.4 billion by FY26.
“AI sales are mostly coming from AI model training applications, and HPE cited traction with sovereign networks,” noted analysts led by Simon Leopold.
They highlighted that enterprise clients, currently in the experimental phase, accounted for a mid-teens percentage of the backlog.
“As enterprise adoption expands, we expect continued strong AI sales with improving margin. We include HPE within the context of an AI networking basket,” the analysts added.
