Investing.com — Here are the biggest analyst moves in the area of artificial intelligence (AI) for this week.
JPMorgan upgrades Meta to Overweight on AI agent opportunity
JPMorgan has upgraded to Overweight from Neutral, raising its price target to $820 from $640, saying the company’s push into frontier AI models and AI agents opens a new avenue for growth beyond advertising.
Analyst Doug Anmuth said Meta’s superintelligence lab has essentially delivered on its goal of reaching the frontier within a year, with an accelerating cadence of model releases culminating in Muse Spark 1.3, which he said is competitive with Claude and GPT models. He expects a coming model, dubbed Watermelon, to unlock further opportunities across consumer products, engagement, advertising and internal efficiency.
The Muse agent has seen strong early traction, reaching as high as No. 3 in the U.S. app store on its second day. While monetization is not a near-term priority, Anmuth pointed to potential from commissions and subscriptions in a market he sized in the tens of trillions of dollars. He also noted “there’s still meaningful headroom in core advertising” from better targeting and AI content creation.
Anmuth projected Meta capex of $243 billion in 2027 and $284 billion in 2028, both above consensus. His $820 target is based on 23 times his 2028 EPS estimate of $35.44, which he said “could prove conservative.”
Accumulate SK Hynix’s US shares, JPMorgan says
JPMorgan has initiated coverage of SK Hynix’s American Depositary Receipts (ADRs) with an Overweight rating and a June 2027 price target of $245, betting that the AI-driven memory upcycle still has years to run.
“We expect the AI-driven memory upturn cycle to last for 5+ years and model SKH to print a +34% EPS CAGR over the next two years,” the analysts wrote, describing the cycle as still in its “mid-cycle” phase, with ten quarters of rising average selling prices (ASPs) against an expected trajectory of more than 20 quarters of positive ASP growth from Q1 2024 through Q4 2028.
The analysts highlighted that SK Hynix has locked in more than 50% of its capacity through long-term agreements and raised its shareholder return pool to over 50% of free cash flow, marking “a strong re-rate factor.” JPMorgan forecasts a total shareholder return yield of nearly 42% from 2026 through 2028.
trades at a premium of more than 30% to its local shares yet still reflects roughly a 20% discount to U.S. peer Micron, “even though SKHY is not behind on fundamentals,” the analysts said, pointing to comparable or superior memory margins, scale and HBM execution.
While differences in corporate structure and governance remain a structural factor behind the discount, JPMorgan sees “significant room for a re-rating as these structural gaps narrow.”
“We are positive on the memory valuation re-rate story and recommend investors accumulate,” they added.
JPMorgan names KLA top pick in chip equipment on foundry mix shift
In a separate note, JPMorgan has raised its wafer fabrication equipment (WFE) market forecasts and named KLA Corp. as its top pick among U.S. chip-equipment makers, citing relative underperformance and an upcoming mix shift toward foundry and logic spending.
Analyst Mio Shikanai lifted WFE market growth estimates to 31% for 2026 at $163 billion, 38% for 2027 at $225 billion and 17% for 2028 at $263 billion, a 28% compound annual growth rate (CAGR) over 2025 to 2028.
The upgrades were driven primarily by DRAM memory and TSMC, with supply expected to lag demand even in 2028 and TSMC’s leading-edge capacity staying above 100% utilization. Shikanai said cloud service providers “are accelerating their investments in response to further growth in AI-related demand,” with spending by the top four U.S. players seen growing at a 58% annual rate through 2028.
Shikanai said she sees “the most attractive risk/reward in given its relative underperformance YTD, underappreciated exposure to foundry/logic-led WFE spending in 2027, and potential for upward revisions to consensus revenue estimates.”
KLA shares are up about 50% this year but have lagged peers Lam Research and Applied Materials, up roughly 85% and 82% respectively, as those names benefited more directly from memory-related spending. Shikanai expects the mix to rotate toward foundry and logic in 2027, disproportionately benefiting KLA, with rising capex budgets at customers such as Intel offering additional upside to consensus.
Bernstein highlights “AI-enabled entry point” in stock
This week, Bernstein initiated coverage of Shopify at Outperform with a $160 price target, with analyst Mark Shmulik arguing that while the company faces risks from agentic commerce, AI-native alternatives and “vibe coding,” these same forces should ultimately expand rather than shrink the market Shopify serves.
Shmulik called AI disruption risks “storms in teacups,” pointing to strong adoption of Shopify’s product catalog by AI chatbots and merchants’ preference to keep transactions on Shopify’s payment rails for security and fraud management. He highlighted that agentic checkout hype “faded quickly” following a lackluster launch of ChatGPT Checkout.
Shopify has grown to roughly 14% of U.S. e-commerce market share, with more than half of its sellers now based outside the U.S. and 50% of incremental gross merchandise value (GMV) generated outside North America. International, physical, and B2B channels now account for more than 60% of total sales volume.
Bernstein modeled global GMV growing at a roughly 24% CAGR from 2025 to 2030, with revenue at approximately 25% and free cash flow at around 32%, built on the thesis that “FCF growth > revenue growth > GMV growth > market growth.” Shopify currently trades at 9 times 2027 revenue and 45 times 2027 FCF, below its historical range of 10-12 times forward revenue and roughly 70 times FCF.
“Ignore the noise, this is your AI-enabled entry point,” Shmulik wrote.
Morgan Stanley upgrades to Overweight, downgrades
Morgan Stanley has upgraded Synopsys to Overweight and turned more selective on European semiconductor stocks, citing widening valuation dispersion and signs that the memory cycle is nearing a late-cycle turn.
Analyst Lee Simpson said the bank remains constructive on the sector overall, “underpinned by strong AI demand and a broadening cycle recovery,” but added that with memory approaching an inflection, “we turn more selective.”
For Synopsys, Simpson highlighted an attractive entry point after a recent de-rating, greater confidence in synergies from its Ansys acquisition, a recovery in design intellectual property, and an underappreciated opportunity from physical AI. He kept his $500 price target, based on a forward price-to-earnings (P/E) of 30 to 35 times.
On the other side of the ledger, Morgan Stanley moved Infineon to Equal Weight from Overweight, saying that while the structural data center opportunity remains compelling, it sees limited near-term upside and no clear catalyst, cutting its price target to €65 from €81.