As investors reassessed the possibility that AI-related infrastructure investments may fail to recover their full costs and that AI-driven economic growth expectations may be overstated, the Nasdaq 100 index, known as the "global tech stock bellwether," saw its total market capitalization evaporate by more than $2 trillion last month.
Amid the global tech stock rout, the latest shareholder letter from Pershing Square Inc. (PS.US), founded and personally helmed by billionaire hedge fund legend Bill Ackman, reveals that Wall Street hedge funds appear overly focused on rising AI computing infrastructure construction costs and associated AI financing costs, while significantly underestimating the long-term growth potential of large tech companies heavily investing in artificial intelligence.
Ackman's latest views are broadly consistent with those of Wall Street strategists who have shifted to a buy-the-dip stance amid the recent tech selloff. For example, Gerry Fowler, Chief European Equity Strategist and Head of Derivatives at UBS Group, said in an interview on Monday that the market is underestimating the AI super-cycle, with a potential return inflection point arriving in 2027. In his view, after the recent pullback in AI trades, valuations of hyperscale cloud service providers look attractive, because the AI super-cycle reflected in the market's latest pessimistic pricing appears shorter than it may actually last.
Deutsche Bank has shifted back to a bullish stance on US tech stocks, focusing on adjusted prices. Mike Wilson, Chief Equity Strategist at Morgan Stanley, continues to favor a rebound after risk is released, focusing on AI financing cost pressures, midterm elections, and the over-pricing and digestion path of liquidity pressures. Another Wall Street financial giant, Goldman Sachs, bases its "seven repair signals" primarily 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 stocks.
A latest equity market research report from Deutsche Bank shows that the institution suddenly turned bullish on US tech stocks, completely overturning its cautious decision three weeks ago to downgrade the entire US tech sector to "Neutral." The institution'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. Deutsche Bank strategists said tech stocks' relative performance has fallen to the bottom of their long-term range, a level that in recent years has marked attractive entry points. Deutsche Bank's proprietary statistics show that in previous instances when capital rotated back into tech stocks, the sector rose an average of about 32%, outperforming the rest of the S&P 500 constituents by about 30 percentage points. Despite robust long-term earnings growth expectations for the US tech sector, overall equity positioning remains relatively light. Deutsche Bank strategists said that if AI-related tech companies' profits remain resilient, investors still have ample room to increase exposure, and therefore Deutsche Bank strategists upgraded US tech stocks and global tech stocks from "Neutral" to "Overweight," their most optimistic long-term bullish rating.
Bill Ackman: Wall Street 'Punishes Hyperscalers' for Construction Costs, Ignores Their Growth Potential
In its shareholder letter, Bill Ackman's Pershing Square stated that Wall Street hedge funds underestimate the long-term potential of large tech companies heavily investing in artificial intelligence, focusing too much on rising infrastructure costs while ignoring future revenue growth, stronger cash flows, and investment returns. The investment firm also sees opportunities in the broader equity market. Before Tuesday, Pershing Square shares rose more than 2% in overnight trading. Pershing Square noted in its third-quarter shareholder letter that the S&P 500 rose slightly more than 2% in the third quarter, but the median constituent was down 4%. More than 60% of constituents declined during the quarter. As of September 30, the S&P 500 was up 13% year-to-date, while the average constituent gain was only 1%. Meanwhile, measured by changes in expected earnings per share for S&P 500 constituents, the median rose 4% in the quarter and nearly 10% cumulatively over the first nine months of the year. Pershing Square said this divergence has created buying opportunities: some cloud computing companies focused on AI computing resource construction and large tech giants have seen fundamentals improve while share prices remain weak.
Pershing Square, founded and helmed by Ackman, said in its shareholder letter that semiconductor and tech hardware suppliers have significantly benefited from AI-related spending. Year-to-date, the S&P 500 semiconductor index has surged 44%, while the tech hardware index has risen 36%. However, Wall Street is "punishing hyperscale data center operators" for construction costs while ignoring their strong growth potential. "AI computing-related infrastructure construction has been one of the main drivers of stock market performance this year, with companies providing the 'picks and shovels' being the biggest beneficiaries." However, large cloud computing and tech companies including Microsoft (MSFT.US), Facebook parent Meta Platforms (META.US), Amazon (AMZN.US), and Google parent Alphabet (GOOGL.US), despite earnings growth of about 27%, saw their shares rise only 8% on average, significantly underperforming the semiconductor index. "The decline in hyperscalers' price-to-earnings ratios stems mainly from investor concerns about how much return they can generate from significantly increased capital expenditures," Pershing Square wrote in its shareholder letter.
Why do hyperscale cloud providers have the potential to generate higher returns from AI investments in Ackman's view? Pershing Square believes investors are ignoring the long-term benefits of massive AI infrastructure spending. The four major tech companies' combined capital expenditures in 2026 could reach $740 billion, up from $140 billion in 2023. Wall Street analysts unanimously expect this figure to surpass $1 trillion by 2027. "We believe hyperscalers' capital expenditures are offensive, not defensive. We believe they are investing not merely to protect existing core businesses, but to accelerate growth, deploying capital into a transformative technology that directly leverages their competitive advantages." "These advantages include unmatched distribution capabilities, long-cultivated customer relationships, proprietary data and related contextual information, the ability to deliver secure and reliable services globally, and strong balance sheet strength. This financial strength enables them to invest at a scale few competitors can match."
Pershing Square said Amazon's AI investments are driving accelerated AWS growth, with the business's third-quarter revenue up 39%. AI is also improving advertising targeting precision, helping Meta accelerate revenue growth. Meta's Muse assistant surpassed 10 million downloads within just weeks, creating new opportunities for monetization through shopping, services, and enterprise tools. AI is also creating growth opportunities through Azure, Microsoft 365 Copilot, and GitHub. Some analysts expect large tech companies to continue investing heavily in AI infrastructure for years to come but do not believe these companies' revenue growth rates will significantly accelerate beyond this year and next. Pershing Square believes analysts treat these expenditures as an ongoing cost and do not fully recognize their potential to drive substantial future revenue and profit growth. "In our view, analysts have fully incorporated hyperscalers' current construction costs into their forecasts, to the point where these companies are 'punished'; but for the future revenue and earnings growth these investments are likely to generate—even exponential growth from a new round of revaluation—they give very limited long-term growth expectations."
Pershing Square said revenue growth in 2026 largely comes from AI investments made in 2024 and 2025. Therefore, the firm expects that current higher spending will drive stronger growth in the coming years, potentially significantly exceeding analysts' consensus forecasts.
Ackman Bets on Three Cloud Computing Platforms: The Market Sees This Year's Bill, But Underestimates Next Year's Business
In October, the Nasdaq 100 fell about 6.7%, and the Philadelphia Semiconductor Index dropped about 15.4%, with the Nasdaq 100's total market capitalization evaporating by more than $2 trillion last month. Additionally, last month (October), chip stocks critical to AI computing infrastructure in South Korea and Taiwan stock markets also suffered rare large-scale selloffs, especially in the South Korean market, where SK Hynix fell 20.2%—marking its third-largest monthly decline in nearly a decade—and Samsung Electronics dropped 16.2%—its fourth-largest monthly decline in nearly a decade.
Ackman's judgment at this point is that the same AI capital expenditure has already become revenue for chip and equipment suppliers at one end, but at the other end, it only becomes revenue for cloud platforms after data centers are powered up and customers begin using computing resources. If analysts immediately raise spending forecasts without correspondingly assessing how much business the new capacity can generate, they may systematically underestimate the future earnings growth trajectory of the party funding the construction.
Ackman's positive outlook and positioning on cloud computing and the tech sector come at a time when global AI computing demand continues to surge, but trillion-dollar-level AI computing resource procurement and AI infrastructure investment are generating enormous funding needs. Combined with surging AI financing costs for AI neocloud companies like CoreWeave and the two super cloud giants with long-term solid fundamentals—Google parent Alphabet and Amazon—seeing AI-related debt rise while free cash flow turns fully negative, the market is being pushed to re-examine the heavy burden of AI financing costs for AI-related tech companies and the cash recovery cycle associated with massive AI infrastructure investments.
AI computing demand is still actively expanding, but companies' ability to convert demand into deliveries, profits, and free cash flow is becoming the basis for market repricing. Google parent Alphabet's stock price plunged nearly 13% to a record-breaking level as of last Thursday's US market close, and US CPU and chip manufacturing giant Intel's stock price suffered its worst single-day drop in over forty years as of last Friday's close—both occurring at a time when AI investment cash returns are facing stricter scrutiny.
Ackman himself has expressed this positive bullish judgment through his positions. In an interview published on October 28, he said Amazon, Meta, and Microsoft together account for about one-third of his portfolio. The three purchases had different backgrounds: Amazon was bought after Trump announced "Liberation Day" tariffs; Meta and Microsoft were bought when capital expenditure commitments triggered market disappointment. This is a disclosure of existing positions and buying rationale, though there is no actual evidence that he added all three positions during the late-October decline. On November 2, Ackman and Chief Investment Officer Ryan Israel highlighted in their shareholder letter new position layouts—seven new investments made year-to-date through the third quarter, with an eighth unnamed position initiated after the quarter ended.
Nomura Securities' bullish research report on AI computing and tech themes published on November 2 effectively reveals the potential economic logic behind Ackman's trade—negative free cash flow may come from paying for construction upfront while revenue is recognized over time, and the time gap between the two cannot be directly equated with massive losses. Pershing Square cited Amazon management's explanation: data center building structures require about two years of advance investment and can be used for over 30 years after commissioning; servers and networking equipment can be procured months before commissioning, staying closer to customer demand. According to Amazon's estimates, servers pay back in slightly less than three years and have a useful life of at least five to six years. In the third quarter, AWS revenue grew 39% to $46 billion, with an operating margin of 38.7%; Google Cloud revenue grew 103% to $30.7 billion, with operating margin rising from 23.7% to 36.2%; Azure revenue grew 48%. Overall, however, these latest figures support that computing demand is being monetized, but cloud business overall margins cannot substitute for the return rate of individual new projects. The key to judging whether cloud computing giants are "wrongly sold off" in this tech stock selloff is whether, after new computing capacity is commissioned, utilization and pricing can allow projects to recoup investment and cover capital costs before equipment renewal.
Deutsche Bank Restores 'Overweight,' Morgan Stanley Waits for Buy Point: Wall Street Begins Hunting for Bottom-Fishing Opportunities
October's US stock market "invisible bear market" was a period of dramatic sector rotation. According to The Kobeissi Letter's portfolio damage gauge, AI-related US stocks fell more than 13% during the month, while the S&P constituents excluding AI actually rose about 8% and hit a new high on October 30. Deutsche Bank restored US tech stocks from "Neutral" to "Overweight" on October 30: over the past three weeks, tech and mega-cap growth stocks fell about 11%, while the rest of the S&P 500 rose nearly 6%, a performance gap of more than 16 percentage points. Deutsche Bank's basis is that tech stocks' relative performance has approached the lower bound of their long-term range while earnings are still growing. In previous rounds of capital rotation back into tech stocks, the sector rose an average of about 32%; however, this is a historical reference provided by Deutsche Bank, not this round's gain target.
Morgan Stanley's Mike Wilson favors a post-uncertainty market. Regardless of which party wins the midterm elections, he leans toward buying on dips, but maintains a path where the S&P 500 first tests 7,100 points, waits for election results, AI financing pressures, and liquidity conditions to clarify, then recovers toward 8,000 points. Goldman Sachs derivatives strategist Brian Garrett provided a trading-level check: US fundamental long-short funds' net leverage fell to 47%, in the 7th percentile of the past year; although funds net bought US stocks for the first time in four weeks, about 73% of purchases came from macro products, mainly short covering. The three institutions' positions have different emphases: Deutsche Bank has already upgraded allocations, Wilson believes further declines would improve the buy point, and Goldman Sachs sees room for retail and institutional investors to re-add positions recently, but evidence of sustained and active buying of tech stocks remains limited.
Deutsche Bank strategists see an opportunity to reallocate to tech stocks amid pessimistic market 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, having fallen to the bottom of their long-term relative performance range. For tech stocks, the most investment-significant change is the notable adjustment in the price relationship between tech stocks and other sectors: 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 over-priced by the market.
Wilson, in a major media interview on November 1 US Eastern Time, 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 driving a brief market dip; but after these pressures are concentrated and released, they may instead create opportunities for investors to re-establish equity risk exposure. He believes the S&P 500 could fall to 7,100 points, about 9% below its recent record high, then accelerate its recovery toward 8,000 points.
Another Wall Street financial giant, Goldman Sachs, reported that US fundamental long-short funds' net leverage fell to 47%, in the 7th percentile of the past year; in historical regressions, such positioning typically accompanies index drawdowns exceeding 20%, while the current S&P 500 is only about 3% below its record high. This largely means most institutional investors have already conducted considerable risk contraction in advance, and once earnings remain resilient and event shocks ease, the room to re-add allocations could become rebound momentum.
Comments