Stocks tied to the artificial intelligence supply chain have come under renewed pressure this week, prompting divergent reactions from sell-side research desks. Some firms view the recent rout as an entry point for long-term AI exposure, while others are trimming ratings and cutting estimates in response to deal risk, softer demand signals and potential structural costs.
Market context and the Yardeni view
Yardeni Research characterized last Thursday’s broad market selloff as excessive and said it represents a buying opportunity for investors willing to look through near-term political and logistical challenges. The firm noted that markets rallied on Wednesday, driven by short covering, only to reverse course the following day when the S&P 500 tumbled 4.4% - its largest one-day drop since the April 2025 tariff shock - and the Nasdaq fell 5.5%, pushing it into correction territory.
Yardeni reiterated an earlier stance urging caution through the midterm elections but said it had expected subsequent buying opportunities. The research shop argued that recent voter opposition to new data centers - citing an exit poll showing that 76% of respondents opposed new facilities in their communities - and changes at the state level are more likely to slow the AI-driven build-out than halt it.
Texas, Yardeni noted, accounts for nearly 30% of planned U.S. data center additions in 2028 and recent state-level election results could affect approvals there. The firm referenced an estimate from Goldman Sachs that about 45% of planned U.S. data center capacity could face potential restrictions in some states. Even so, Yardeni called these developments "speed bumps" rather than an endpoint for the AI investment cycle, arguing that data centers will still be built, albeit more slowly, and that the market’s selling pressure has likely overshot fundamentals.
"That’s why we view Thursday’s sell off as a buying opportunity," the firm said.
Marvell downgraded amid concerns over all-stock Lumentum deal
Two major research shops moved to a more cautious stance on Marvell Technology this week following the chipmaker’s announcement of an all-stock acquisition of optoelectronics company Lumentum.
TD Cowen lowered its rating on Marvell to Hold from Buy and cut its price target to $175 from $350. Analyst Sean O’Loughlin said the deal’s structure and timing - not the quality of Lumentum’s assets - are the primary drivers of the downgrade. TD Cowen expects the transaction to dilute Marvell’s earnings per share by a mid-single-digit percentage in 2028, the first full year after the deal, and expressed concern that the acquisition could occur at or near the peak of an optical mini-cycle.
O’Loughlin acknowledged that Lumentum could become a "crown jewel" within Marvell’s portfolio, but highlighted a marked divergence between bottom-up consensus and TD Cowen’s own market sizing. He noted that bottom-up consensus implies roughly a $100 billion optical transceiver market in 2027, while TD Cowen’s forecast is materially lower at about $47 billion. If the cycle softens, Lumentum’s operating leverage could exacerbate EPS dilution risks.
Bank of America took a similar step, cutting Marvell to Neutral from Buy and trimming its price target to $225 from $400. Analyst Vivek Arya flagged the possibility of a prolonged overhang on Marvell’s shares related to the deal, as well as a steep drop in the Semiconductors ETF (SOX) over the past month. Arya acknowledged the strategic rationale for combining electrical-to-optical connectivity assets but highlighted potential headwinds: a lengthy regulatory review with closing anticipated in the first half of 2028, limited EPS accretion given that $1 billion in planned cost savings falls short of offsetting roughly 50% share dilution, and the prospect of customer conflicts.
"There was no mention of any large M&A in MRVL’s recent analyst day, so the news comes as a considerable surprise," Arya wrote, adding that the firm is monitoring the potential implications for Lumentum’s valuation once industry constraints ease.
Citi: AI financing worries and oil helped drive pre-election weakness - Europe favored
Citi’s note to clients argued that the weakness in the AI trade and moves in oil - rather than the election itself - were the primary drivers of the pre-midterm pullback. While the bank still expects data centers will be constructed to meet growing demand for computing power, it cautioned that the sites, architecture and pace of construction could evolve in the wake of stronger political scrutiny following recent election outcomes.
The bank said this shift is "far from ideal" for the picks-and-shovels segments of the market, including semiconductor suppliers. Citi’s seasonal analysis shows November-December returns for the S&P 500 historically in the 3% to 4% range, but the bank noted the current pullback carries a different complexion because of AI financing worries and energy moves.
Citi also highlighted its "Generals" index, which measures how many of the seven largest S&P 500 stocks trade below their 200-day moving averages, and said it is nearing a threshold that would signal additional caution. The firm expects continued rotation out of technology names into other areas for the near term.
Given these dynamics, Citi expressed a preference for European equities over U.S. stocks. The bank said the sizable decline in oil prices should benefit European markets to a greater degree, and that lower energy prices may encourage a more dovish European Central Bank stance - factors that could help Europe outperform the U.S. in the near term.
Citi also touched on geopolitical considerations, noting that a renewed escalation in the Iran conflict did not appear imminent. It cited the timing of naval deployments, noting a third U.S. aircraft carrier was not due in the Gulf until mid- to late November, although it did not predict a specific outcome.
Citi cuts Adobe to Sell, warns of leadership and margin risks
In a separate downgrade, Citi reduced Adobe’s rating to Sell and lowered its price target to $207 from $250. The bank cited a confluence of challenges: a leadership transition with incoming CEO Anil Chakravarthy scheduled to take over Dec. 1, ongoing vacancies in critical roles including the CFO and the head of Digital Media, growing competition in the Creative business and broader macro pressures.
Citi said it sees an increasing probability of what it called a "hard reset" within the company - a near-term pullback in the stock that could, in the bank’s view, create space to rebuild management, refine strategy, and set a more achievable baseline for future growth.
The firm expects Adobe to emphasize top-of-funnel growth through AI initiatives and the Firefly product rather than relying predominantly on price increases within Creative Cloud. Citi predicted that a shift toward monthly active user metrics could weigh on total annual recurring revenue (ARR) in fiscal 2027 and 2028, with the greatest impact concentrated in core Creative products. The bank also cited weak demand among small businesses and competition from web-native and AI-first tools as additional headwinds.
Quantitatively, Citi lowered its net new ARR forecasts for fiscal 2027 and fiscal 2028 by approximately $470 million and $500 million, respectively, to $1.5 billion and $1.35 billion. That adjustment takes Adobe’s projected total ARR growth to 5.3% for fiscal 2027 and 4.5% for fiscal 2028 - roughly 200 to 300 basis points below current Street estimates.
The firm also trimmed its fiscal 2027 non-GAAP gross margin forecast by 70 basis points to reflect unmonetized AI inference costs and flagged a sharper revenue slowdown in fiscal 2028 alongside modest margin degradation.
Multiple firms downgrade HubSpot after soft guidance
Wall Street’s response to HubSpot’s quarterly update leaned negative after the company offered a softer-than-expected outlook for the fourth quarter. HubSpot reported third-quarter revenue growth of 17% on a constant currency basis - a two-percentage-point beat versus expectations - but guided fourth-quarter growth to 12%, below consensus of about 13.8%.
Management highlighted heightened budget scrutiny, greater price sensitivity and elongated sales cycles, and said these trends are likely to persist into 2027. HubSpot added 5,300 net new customers in the quarter, short of the 5,600 consensus, and now expects to add fewer than 5,000 net new customers in the fourth quarter. Net revenue retention also slipped to 101% from 103% a year earlier.
Following the guidance, four research firms adjusted their views. BTIG moved HubSpot to Neutral from Buy and lowered its 2027 growth forecast to 11% from 14%, with analyst Nick Altmann saying the stock appears fairly valued at roughly 10 times estimated 2027 free cash flow. KeyBanc downgraded to Sector Weight, stressing that there is limited visibility for upside before formal 2027 guidance is released next February. Cantor placed HubSpot at Underweight with a $170 price target, citing what it described as European pricing changes that effectively amount to a price cut and underperforming AI tools relative to peers. Morgan Stanley cut its rating to Equal-weight from Overweight and lowered its price target to $229 from $287.
Analysts flagged that HubSpot no longer fits the profile of a premier SaaS asset deserving the benefit of the doubt without clearer evidence of stabilization in customer acquisition and retention metrics.
What this week’s moves mean for markets and investors
The week’s analyst activity highlights a market recalibrating expectations for the pace and economics of AI deployment. Political resistance to data-center siting, potential state-level approval slowdowns, deal-related dilution and the mechanics of monetizing AI features all figured in firm-level research calls.
Some firms are interpreting these developments as transitory constraints that create a tactical buying opportunity, while others are reacting by reducing exposure to firms where near-term earnings and cash-flow profiles appear most at risk. The divergence underscores an active debate on whether current price moves reflect a temporary repricing or a more structural change in the investment case for certain AI-adjacent names.
Investors weighing these developments will need to reconcile differing assessments: Yardeni’s view that data-center headwinds are mainly logistical slowdowns versus the more cautious take from several sell-side analysts who see tangible downside to near-term earnings and valuation drivers at specific companies.
Bottom line
Research desks have delivered mixed verdicts on the selloff in AI-exposed equities. Some see the pullback as an attractive point to add exposure, while others have downgraded individual companies on concerns ranging from transaction structure and dilution to weakening demand and rising costs tied to AI initiatives. The near-term path for these names appears to depend on how quickly political resistance, regulatory timelines and operational constraints reshape the pace of AI infrastructure investment and the monetization strategies for AI-enabled products.