After a sharp selloff in technology stocks at the end of October, Wall Street analysts' outlook reports are beginning to show noteworthy shifts in judgment: pressure from AI capital expenditure and AI infrastructure financing still exists, but the recent steep declines in overall tech sector share prices driven by AI computing themes and broad semiconductor concepts, the pullback in relative valuations, and investor de-risking, combined with the still-strong long-term fundamental growth trajectories of most AI computing leaders, have made the potential returns on some tech assets more attractive.
Deutsche Bank has pivoted back to a bullish stance on US tech stock themes, focusing on adjusted prices; Morgan Stanley's chief equity strategist Mike Wilson continues to favor a rebound after risk release, focusing on the overpricing and absorption path of AI financing cost pressures, midterm elections, and liquidity constraints.
Another Wall Street financial giant, Goldman Sachs, bases its "seven repair signals" mainly on signals from the S&P 500, Nasdaq 100, US equity options, and hedge fund positioning and fund flows in the US market, focusing on the long-term bullish narrative for US tech stocks and AI-related shares.
A recent US equity market research report from Deutsche Bank shows that the institution has suddenly turned bullish on US tech stocks, completely reversing its cautious decision three weeks ago to downgrade its rating on the entire US tech sector to "Neutral." The bank's strategist team said the sharp selloff has created attractive buying opportunities, and a considerable portion of AI-related risks has already been reflected in prices.
As investors reassess whether AI-related infrastructure investment may fail to recoup its full costs and AI-driven economic growth expectations, the Nasdaq 100 index, known as the "global tech stock bellwether," saw its total market capitalization evaporate by more than $2 trillion last month.
The Deutsche Bank research report shows that over the past three weeks, tech stocks and mega-cap growth stocks underperformed the rest of the S&P 500 constituents by more than 16 percentage points: the former fell about 11%, while the rest of the index rose nearly 6%, bringing the sector close to the lower end of its historical trading range.
The "rest of the S&P 500 constituents" cited in the Deutsche Bank report refers to a comparison group composed of the remaining S&P 500 constituents after Deutsche Bank excludes the group of "tech stocks and mega-cap growth stocks" in its statistics, including stocks in finance, industrials, energy, healthcare, and other sectors. Therefore, this statement compares the performance of two groups of assets—"tech and mega-cap growth stocks" versus "the rest of the S&P 500"—rather than comparing with the entire broad S&P 500 index, and it should not be simply understood as excluding only the "Magnificent Seven."
Deutsche Bank Turns Bullish After Tech Selloff: The Market Is Reassessing "the Price of Growth"
The earnings of the "Magnificent Seven" US tech giants have reinforced investors' concerns about the pace at which massive AI spending will deliver returns, becoming the most core factor behind the recent sharp weakness in large tech stocks and even the core driver of the US bull market in recent years—namely AI computing and broad semiconductor stocks.
Deutsche Bank strategists said the relative performance of tech stocks has fallen to the bottom of its long-term range, a level that in recent years has marked attractive entry points. Deutsche Bank's proprietary statistics show that previously, when funds rotated back into tech stocks, the sector rose by an average of about 32%, outperforming the rest of the S&P 500 constituents by about 30 percentage points.
Although long-term earnings growth expectations for the US tech sector remain solid, overall equity positioning is still relatively light; Deutsche Bank's strategist team said that if profits at these AI-related framework companies remain resilient, investors still have ample room to increase exposure. Therefore, Deutsche Bank strategists upgraded US tech stocks from "Neutral" to "Overweight," its most optimistic long-term bullish rating.
The severity of this correction was first reflected in the rapid reversal of the early-week rebound. On October 29 and 30, the Nasdaq 100 fell 1.38% and 3.46%, respectively, for a cumulative two-day decline of about 4.79%; the Philadelphia Semiconductor Index fell 1.42% and 7.20%, respectively, for a cumulative two-day decline of about 8.52%. The repair gains previously formed by consecutive semiconductor counterattacks were almost entirely given back within two trading days.
The Nasdaq 100 fell about 6.65% in October, with tech growth stocks once again becoming the center of market risk release.
The historically significant shock was concentrated in individual stocks and trend structures: Intel plunged about 28.58% at the close on October 30, its largest single-day drop since 1974; according to Goldman Sachs trading desk statistics, the Philadelphia Semiconductor Index fell below its 200-day moving average, ending a streak of 355 consecutive trading days above that line. Intel's data center business growth of nearly 80% year-over-year still failed to offset the valuation reset caused by insufficient forward guidance, sufficient to show that the market's demands on AI beneficiary companies have risen from "whether AI computing demand is strong" to "whether they can deliver on schedule, maintain profitability, and realize optimistic cash returns."
Deutsche Bank saw an opportunity to reallocate to tech stocks precisely in this pessimistic pricing. The institution noted that tech and mega-cap growth stocks fell about 11% over the past three weeks, while the rest of the S&P 500 rose nearly 6%, with the former underperforming by more than 16 percentage points and having fallen back to the bottom of its long-term relative performance range.
For tech stocks, the most investment-significant change is that the price relationship between tech stocks and other sectors has adjusted markedly: the premium investors were previously willing to pay for AI growth has been rapidly compressed, while safe-haven rotation has pushed up the relative prices of other assets. Deutsche Bank therefore believes that a considerable portion of AI-related risks has been overpriced by the market.
This divergence also explains why tech stocks were hit hard while the broader market did not simultaneously fall into a full-blown bear market. According to AI versus non-AI stock portfolio statistics cited by Kobeissi, in October the two performed at roughly -13% and +8%, respectively; Goldman Sachs calculated that on Thursday, 406 S&P 500 constituents rose while the market-cap-weighted index closed lower. Funds are reducing highly concentrated, capital-intensive AI exposure while shifting toward other sources of earnings.
The core of Deutsche Bank's renewed bullishness is the judgment that this rotation has made the relative prices of tech stocks attractive; the historical experience it cites shows that previously, when funds rotated back to tech stocks, the sector rose by an average of about 32% and outperformed the rest of the S&P 500 constituents by about 30 percentage points—but it should be noted that these figures describe historical repair magnitudes, not this round's target gains.
Morgan Stanley's Wilson Bets on a Counterattack After Risk Release: 7,100 Is a Potential Entry Zone, 8,000 Is the Repair Direction
In a high-profile media interview on November 1 US Eastern Time, Wilson further clarified his long-term bullish judgment on US stocks, especially AI-related tech stocks: midterm elections, AI investment and financing controversies, and liquidity constraints may still constitute three types of events that could drive a brief market dip; but after these pressures are released in a concentrated manner, they may instead create opportunities for investors to rebuild equity risk exposure. He believes the S&P 500 could fall back to 7,100 points, about 9% below its recent record high, before accelerating a repair toward 8,000 points.
Based on the S&P 500's October 30 close of 7,580.07, 7,100 implies about 6.3% further downside, while 8,000 corresponds to about 5.5% upside; if the index first falls to 7,100 and then repairs to 8,000, the gain would be about 12.7%. This price path reveals his investment thinking: remaining bullish on medium-term returns while accepting that short-term risk release may offer better buying prices. However, his previous official investment strategy commentary also emphasized that lower-quality AI participants may still face widening credit spreads and further stock price declines.
On the midterm elections, Wilson places more emphasis on the dissipation of uncertainty rather than betting on a particular party winning. The statistics he cites show that since 1950, the S&P 500 has risen in both the six months and twelve months after every midterm election; but he also explicitly acknowledges that the period from the end of elections to the end of that year does not necessarily see gains, with 2018 being a counterexample. The political variables truly affecting AI investment now lie in data center approvals, community opposition, electricity price allocation, and regulatory policy: if election results significantly exceed scenarios already priced in by the market, project construction could still be impacted; if opposition during the campaign subsequently turns into negotiations with local governments seeking investment and jobs, construction resistance could also ease marginally.
On the AI financing pressure that the market increasingly worries about, Wilson's attitude is more nuanced than the market's panic narrative. He does not yet believe that equity issuance volume has become large enough to overwhelm funding demand, because corporate fundraising involves not only AI but also M&A and broader real investment. He judges that a large amount of financing has already been completed this year, and that this concentrated fundraising round may pause for a period over the next roughly 30 days. The positive implication of his latest judgment is that the new securities supply the market needs to absorb may decrease in stages, thereby easing pressure on capital competition; this is a different repair path from expecting the Fed to cut rates immediately.
He particularly emphasizes companies proactively adjusting between bond and equity financing: when equity issuance faces market resistance, companies can evaluate debt financing; when debt burdens become too heavy, they can replenish capital through equity. Rising credit spreads will also force projects with insufficient economics to slow expansion, forming the market's own constraint mechanism. However, this process will intensify corporate divergence. Under Wilson's framework, an environment with nominal economic growth of about 7%–8% and a 10-year Treasury yield of about 5% may still support high-quality corporate growth; for companies with borrowing costs of 8%–10% and weaker earnings and cash reserves, the same environment may be difficult to withstand. This is also why Wilson's long-standing bullish logic on US stocks has always included screening for balance sheet and earnings quality.
Undoubtedly, the strong cloud computing business growth reported last week by Google parent Alphabet and Microsoft, the two major cloud giants, is providing revenue-side validation for AI infrastructure investment. And from the series of positive growth developments among AI computing industry chain leaders not long ago, it can be seen that from memory carrying models and inference tasks to power supply and liquid cooling systems ensuring high-density rack operations, industry chain earnings reports are providing mutually corroborating demand signals, with memory makers' profits and long-term supply arrangements being especially representative.
AI application-side players including Google parent Alphabet as well as Microsoft, Meta, and Oracle, while bearing the pressure of rising AI financing costs, have also continued to actively provide positive signals about AI computing demand. Social advertising and AI application giant Meta's third-quarter revenue was $64.601 billion, up 26% year-over-year, with ad impressions up 10% and price per ad up 14%; daily active users across its apps reached 3.64 billion. Muse surpassed 10 million registered accounts in the US and Canada, about a quarter of which use it daily; enterprise database and cloud infrastructure giant Oracle explicitly stated that free cash flow will turn positive in the second half of fiscal 2028, exceed $30 billion in fiscal 2029, and revenue will be at least $225 billion in fiscal 2030. The former shows application growth coexisting with cash burn, while the latter provides a verifiable timetable for cash recovery after the peak of capital expenditure.
As Meta Muse and OpenAI's GPT-6 Astra fully ignite the wave of AI agents, "AI FOMO trading" (referring to investors buying or replenishing previously reduced positions out of fear of missing gains in AI-related assets) is expected to return. From the perspective of the underlying architecture of AI computing infrastructure, Muse and Astra are expected to further expand the scope of AI agent use; the massive expansion of AI inference workloads they bring will simultaneously increase demand for model inference, tool execution, and state management, while also deepening competitive pressure on some software companies.
JPMorgan strategist team's AI capital expenditure forecasts and outlook show that hyperscale cloud providers' capital expenditure will be about $950 billion in 2026, about $1.4 trillion in 2027, and about $3 trillion in 2030. However, the next stage of a more solid and strong upward trajectory for global AI computing themes and the semiconductor sector should be driven jointly by upward earnings estimate revisions, procurement payment realization, and credit spread improvement, especially in scarce capacity and scarce technology segments.
From a technical mechanism perspective, AI agents expand a single answer into repeated reasoning, retrieval, tool execution, and state preservation. GPUs and TPUs handle model computation, CPUs handle tools and task orchestration, HBM holds data for high-bandwidth access, server DRAM carries the runtime environment, and SSDs and HDDs serve high-performance access and high-capacity retention, respectively. At the same time, network infrastructure, high-speed optical interconnects in data centers, power supply, and cooling determine the actual utilization of expensive accelerators. Therefore, accelerating cloud revenue can support multiple hardware segments simultaneously; how much profit a specific company captures still depends on performance advantages, supply bottlenecks, and bargaining power.
NVIDIA, with a market capitalization of about $5 trillion, remains an important bullish pricing anchor on Wall Street amid this round of sharp market disagreement. As of the October 30 close, NVIDIA's stock price was $204.64 and its market capitalization was about $4.94 trillion; target prices of $300, $315, and $400 recently given by Morgan Stanley, Citi, and Bernstein correspond to potential upside of about 46.6%, 53.9%, and 95.5%, respectively, reflecting that Wall Street's most core sell-side institutions still maintain positive judgments on its earnings outlook.
Seven Repair Signals Emerge in the Stock Market: Rebound Capital Is Getting Ready, but Cash Returns Still Determine Who Leads
The seven types of repair signs compiled by senior derivatives strategist Brian Garrett from another Wall Street financial giant, Goldman Sachs, supplement the trading-level evidence in Deutsche Bank's and Wilson's judgments: index options beginning to price in panic, fund positioning shrinking markedly, downside volatility regaining dominance, non-AI stocks showing more resilience, defensive asset prices becoming more expensive, AI financing demand possibly lower than previously expected, and retail investors shifting from unconditional bottom-fishing to watching first.
The latest view array from this Goldman Sachs derivatives strategist collectively shows that the market has gradually abandoned its previous one-way pricing that "growth can cover everything." Among them, the most weighty are positioning and market breadth. Goldman Sachs statistics show that US fundamental long-short funds' net leverage has fallen to 47%, in the 7th percentile of the past year; in historical regressions, such positioning is usually accompanied by index drawdowns of more than 20%, while the current S&P 500 is only about 3% below its record high. To a large extent, this also means that some institutions have already carried out considerable risk reduction in advance; once earnings remain resilient and event shocks ease, the room to re-add exposure could become rebound momentum.
However, the latest data flow compiled by Goldman Sachs strategists shows that 73% of net fund buying that week came from macro-type fund products and was mainly reflected as short covering, indicating that current buying is more about reducing bearish exposure, and evidence of sustained active increases in tech stock holdings still needs to form further. The severe negative catalysts from heavy AI financing costs and free cash flow turning negative are precisely the key to whether such active allocation can continue strongly and long term.
Amazon's third-quarter cloud computing unit AWS revenue grew 39.3%, and management expects growth rates above 40% in both 2027 and 2028; at the same time, the group's free cash flow outflow over the past twelve months was about $42.4 billion, mainly affected by increased AI-related capital expenditure. Google parent Alphabet's third-quarter free cash flow was negative $9.065 billion, but over the past twelve months it remained positive at $19.747 billion. Demand growth and cash burn coexisting means investors are reassessing when construction spending occurs, when fee revenue is realized, and how much external capital is needed in between. The group's free cash flow turning negative itself does not prove that the cloud business is losing money or that long-term project returns are insufficient.
Credit pressure at new cloud companies such as CoreWeave makes this time gap more sensitive: tightening financing conditions will first affect companies' ability to expand capital expenditure, then transmit to the industry chain through project starts, equipment deliveries, and procurement pace. Anthropic's plan to issue up to $3.5 billion of stock to Japanese investors in a US IPO reflects corporate efforts to expand funding sources; regulatory filings show the funds will be used for working capital, operating expenses, and capital expenditure. This is the portion of a US IPO offered to Japanese investors, not a separate listing in Japan. Its potential role is to use equity capital to absorb part of the long-term expansion risk, reducing pressure from relying solely on debt, while also requiring investors to accept equity dilution and endorse the strong cash return expectations from a future AI computing resource layout worth hundreds of billions of dollars.
According to the latest tech stock outlooks from three major financial giants, the most powerful bullish catalyst basis for the next repair trajectory of AI computing and semiconductor themes, and even tech stocks broadly, should be three changes occurring simultaneously: adjusted prices can compensate for financing and execution risks, customers' procurement budgets have reliable funding sources, and suppliers can convert orders into deliveries, profits, and cash. AI agent application expansion can continuously increase computing demand, but for stock market investment strategy return prospects, construction plans with financing already secured, verifiable equipment utilization, and stable payment collection are more persuasive than simply expanding capital expenditure commitments. Deutsche Bank's strategist team's latest research report provides the basis that "relative prices have already improved," Morgan Stanley senior strategist Wilson provides the path to "prepare to add exposure after risk release," and Goldman Sachs reveals the reality that "positioning has room for repair while the nature of buying is still changing."
Comments