Analysts across the Street made notable revisions to ratings and price targets for a set of high-profile companies tied to the artificial intelligence investment cycle. The reactions ranged from calls for larger shareholder returns to warnings about the profitability of heavy infrastructure spending for AI. Below, we break down the main analyst actions, the arguments they advanced, and the financial metrics they highlighted.
Bank of America says shareholder returns could be the next re-rating trigger for Nvidia
Bank of America analysts led by Vivek Arya suggested that an increased commitment to returning cash to shareholders could act as a re-rating catalyst for Nvidia. The team argued that a more robust capital return program might broaden the chipmaker’s investor base and help close a valuation gap they consider unwarranted.
Despite Nvidia’s status as the largest company in the S&P 500 by market capitalization - roughly $5.08 trillion on the metrics cited - BofA notes the stock is trading at a substantial discount to its Magnificent Seven peers on a price-to-earnings basis. Their estimates place Nvidia at 26x and 19x for 2026 and 2027, respectively, versus a peer average of 49x and 41.5x for the same years.
The divergence appears even larger when viewed through a free cash flow lens. BofA estimates Nvidia will generate in excess of $400 billion of free cash flow across 2026 and 2027 combined - an amount the analysts say is roughly equivalent to the combined FCF of Apple and Microsoft - yet Nvidia trades at about a 30% lower market-cap-to-FCF multiple than those two companies.
One contributing factor, BofA argued, is Nvidia’s near-zero dividend yield of 0.02%, which effectively excludes the stock from many income-oriented portfolios. The analysts highlighted that equity income funds own Nvidia in only 16% of funds, whereas comparable tech peers are held by an average of 32% of such funds, with a range between 9% and 57%. The average dividend yield among the peer group is 0.89% versus Nvidia’s token 0.02% yield.
BofA suggested Nvidia could lift its yield toward a 0.5% to 1% range, which would align it more closely with Apple’s and Microsoft’s yields noted in the analysis. They estimated that achieving a 0.5% to 1% payout could cost the company between $26 billion and $51 billion, or roughly 15% to 30% of projected 2026 free cash flow.
The bank also pointed to Nvidia’s return-of-cash history as a relative weakness. Over the past three years, Nvidia returned just 47% of free cash flow through dividends and buybacks, materially below a peer average of roughly 80% and also beneath Nvidia’s own historical rate of about 82% over the 2013-2022 period.
Goldman sticks with Microsoft after results; sees potential turning point
Goldman Sachs reiterated a Buy rating on Microsoft and kept a $610 twelve-month price target following the company’s fiscal third-quarter report. Analyst Gabriela Borges framed the quarter as a potential inflection point for a stock that had been lagging peers.
Microsoft reported third-quarter revenue of $82.9 billion, up 18% year over year and approximately 2% ahead of Street estimates, along with non-GAAP earnings per share of $4.27, up 23% year over year and about 5% above consensus. Goldman emphasized the company’s guidance on Azure, which included a forecast of 39% to 40% constant-currency growth for the fiscal fourth quarter and implied modest acceleration into the first half of fiscal 2027.
Capital expenditures were a particular focal point. Microsoft projected calendar-year 2026 capex of $190 billion, which Goldman noted implies second-half spending of roughly $120 billion. That level represents a roughly 45% step-up compared with the Street’s prior $82 billion estimate for second-half capex.
Borges also pointed to operational signals tied to AI adoption. Microsoft disclosed 20 million Copilot seats and reported the fastest quarter-over-quarter net additions since the product’s launch. The firm called attention to internal silicon efficiency gains, citing a 67% improvement in GPU efficiency for transcription and up to a 260% efficiency gain for image generation workloads.
Goldman characterized Microsoft as well positioned in the AI ecosystem and described the quarter as a meaningful first step in reversing the stock’s multi-quarter underperformance, noting a “catalyst-rich path” into the second half.
JPMorgan downgrades Meta amid AI spending concerns and product visibility shortfalls
JPMorgan downgraded Meta Platforms from Overweight to Neutral and trimmed its price target from $825 to $725. Analyst Doug Anmuth acknowledged Meta’s strong recent top-line momentum but expressed concern about whether heavy AI-related infrastructure spending will deliver sufficient long-term returns.
While noting Meta’s revenue growth of 33% year over year, JPMorgan questioned the return profile of full-stack AI investment outside of advertising. The analyst wrote that while AI-driven ad stack optimizations, accelerating impression growth, and engagement gains have supported revenue, intensifying full-stack AI competition makes it more challenging for Meta to generate attractive returns on heavy AI capital spending beyond the core advertising business.
Anmuth pointed to Google and Amazon as competitors that may be better positioned because of deeper enterprise integrations and clearer multi-year return profiles on AI capital deployment. He also flagged that Meta had recently raised its 2026 capital expenditure guidance by $10 billion to a range of $125 billion to $145 billion.
JPMorgan projected that Meta’s capex could climb 42% to $202 billion in 2027, which the firm estimated would lead to negative free cash flow of $4 billion in 2026 and negative $24 billion in 2027 under their modeling. The analyst also outlined a cluster of near-term headwinds that could keep pressure on the shares, including tougher year-over-year comparisons, exposure to the Middle East conflict, the implementation of European regulations, and foreign exchange headwinds.
Summing up, Anmuth said the move to a Neutral rating reflects the need for greater clarity on agentic products and how Muse models will help drive incremental revenue beyond advertising.
D.A. Davidson starts Micron with a Street-high target, arguing AI is changing memory dynamics
D.A. Davidson initiated coverage of Micron Technology with a Buy rating and a $1,000 price target - the highest on Wall Street according to the firm. Analyst Gil Luria argued that the AI-driven shift in compute deployment is creating a longer-than-usual memory cycle, with demand generation and compute deployment forming a positive feedback loop that elevates pricing potential and demand ceilings.
Luria contrasted the current environment with prior memory cycles, which were characterized by a fixed demand ceiling. Historically, capacity additions eventually exceeded demand and margins compressed. By contrast, Luria contended that AI introduces incremental demand each time new compute infrastructure is deployed, unlocking use cases that did not exist before the underlying infrastructure was in place.
The analyst highlighted a structural change in commercial arrangements as well. Micron was the first memory supplier to announce a five-year supply deal in March, and Luria said Samsung and SK Hynix are reportedly engaged in similar discussions with hyperscale customers. Shifting away from one-year contracts, the analyst argued, improves visibility into demand and pricing stability across the memory industry.
On the product and cost side, D.A. Davidson cited Micron’s node leadership - four consecutive generations in DRAM and three in NAND - as a compounding cost advantage. The firm sees high-bandwidth memory as the central growth driver, noting that Micron’s HBM market share expanded from roughly 5% in 2024 to approximately 21% by the second quarter of 2025, overtaking Samsung to become the second-largest supplier.
While the analyst did not deny the existence of cycles, he argued the duration and magnitude of the current cycle are likely underpriced by the market. Combined with Micron’s node leadership and what the analyst describes as a long-duration earnings power story, D.A. Davidson sees significant upside to shares.
Mizuho splits coverage: upgrades CrowdStrike, trims Adobe
Mizuho delivered a split call that reflects differing assessments of companies' AI positioning. Gregg Moskowitz upgraded CrowdStrike to Outperform and lifted the price target to $520 from $490. The analyst highlighted robust platform demand, hyperscaler marketplace traction, and AI security initiatives as drivers that should allow CrowdStrike to capture AI-related security spend. Moskowitz also referenced Falcon Flex and flagged Project Glasswing as a potential future catalyst, calling CrowdStrike’s set of offerings arguably the strongest in the space.
Mizuho noted that CrowdStrike trades at roughly 14x calendar-year 2027 ARR, a valuation level Moskowitz described as more reasonable and supportive of a favorable risk/reward.
By contrast, Moskowitz trimmed Adobe’s price target to $270 from $315 and moved the rating to Neutral. The analyst cited intensifying competition in the prosumer and small-to-medium-business segments and said there is no clear catalyst for the stock. Moskowitz warned of a risk of margin erosion and now models Adobe’s organic revenue and ARR compound annual growth rate over the next two to three years as high-single-digits at best. He acknowledged having hesitated to downgrade earlier but concluded the risk/reward profile for Adobe is generally balanced from current levels.
What these moves mean for markets and investors
The week’s analyst activity highlights two recurrent themes: first, the trade-off between deploying capital into AI infrastructure and returning cash to shareholders; and second, the uneven way AI is altering industry dynamics, from memory vendors to security and software providers. Where some analysts are urging a reintroduction of shareholder-friendly policies to capture a broader investor base, others are focused on the long-term demand growth and structural changes that heavy AI investment can enable.
For income-oriented investors, the BofA note on Nvidia underscores how dividend strategy and buyback pacing can materially influence inclusion in certain fund universes and relative valuation metrics. For investors concerned about capital allocation efficiency, JPMorgan’s reassessment of Meta and D.A. Davidson’s bullish stance on Micron both emphasize the need to understand the return profile of substantial AI-related capital expenditure.
Quotes and explicit analyst views cited in this report
"Increased cash returns could signal sustainability, widen the shareholder base, and help narrow the valuation gap," BofA analysts led by Vivek Arya wrote regarding Nvidia.
Goldman’s Gabriela Borges wrote, "We are constructive on Microsoft’s positioning in the AI ecosystem and see a catalyst-rich path in 2H. We see this quarter as a meaningful first step in reversing the stock’s multi-quarter period of underperformance."
Doug Anmuth at JPMorgan observed, "While we’re encouraged by META’s +33% Y/Y revenue growth which has been supported by AI-driven ad stack optimizations, accelerating impression growth, & engagement gains, we believe full-stack AI competition is intensifying and Meta has a more challenging path to returns on heavy AI capex beyond advertising."
Gil Luria of D.A. Davidson wrote, "We believe artificial intelligence is creating a longer-than-usual memory cycle as compute deployment and demand generation exist in a positive feedback loop, creating a structurally higher ceiling for memory pricing and demand."
Bottom line
The recent flurry of analyst moves reflects the market’s struggle to price the rapid changes that AI-related investment is producing across technology sectors. Opinions diverge over whether capital should be redeployed to shareholders to close perceived valuation gaps, or reinvested to capture what some analysts view as a structurally higher demand trajectory. These divergent conclusions are translating into materially different ratings and price targets across major AI-exposed names.