US hyperscale cloud providers have just delivered their sixth consecutive quarter of accelerating growth, yet Intel (INTC) suffered its worst single-day plunge in half a century (-29%). Caught in the middle, hedge funds spent the entire October selling off anything with "AI" in its name.
What exactly is the market afraid of? Rich Privorotsky, head of EMEA Delta One trading at Goldman Sachs, said in his Monday morning weekly note that nobody really knows what to worry about. He wrote:
The great AI debate is becoming somewhat circular. Half the market believes we simply cannot build enough power and data centers to support the 2027 capex plans, so the growth of "picks and shovels" stocks must peak and valuation multiples will be compressed.
The other half believes we are about to massively overproduce compute, which will destroy the economics of leasing and likewise demands lower valuation multiples.
I think the bigger problem is that nobody really knows what to worry about. Strong demand is undeniable, but the return on the next trillion dollars of investment is highly uncertain.
Insufficient Compute, or Excess Compute?
Privorotsky's observation highlights the factual mutual exclusivity of the two bearish logics.
If the bottleneck is power, then compute will remain persistently scarce, leasing prices will stay firm, and the problem is simply that there is not enough electricity to connect the next GPU, flattening the growth curve of "picks and shovels" stocks. Morgan Stanley analyst Stephen Byrd estimates that the net US power shortfall will still reach 38GW by 2028.
If the industry is about to massively overproduce compute, then leasing economics will collapse, and the returns on existing chip inventories will unravel along with it. Privorotsky warned of this risk as early as June this year, when he noted that the market's core premise rests on compute scarcity.
These two scenarios cannot both be the true core constraint at the same time, yet both can serve as reasons to sell. Privorotsky wrote: "When positioning is crowded, every new worry becomes another reason to reduce risk exposure." The market does not need to pick one bearish logic; it only needs to find the most popular explanation of the day.
Demand Is Real, and So Are the Bills
Last week's earnings from hyperscale cloud providers were impressive: Azure grew 48% for the quarter, AWS grew 39%, and Google Cloud grew 103%. Privorotsky called it "extraordinary demand." Combined cloud revenue growth for Amazon, Alphabet, and Microsoft reached 54% year over year in Q3, accelerating for the sixth consecutive quarter.
But then comes the "having said that." Privorotsky acknowledged that "negative free cash flow, equity financing, and growing leverage needs have already triggered reasonable questions about return on capital."
His Goldman colleague Lee Coppersmith provided specific figures: hyperscale cloud providers' 2027 capex is expected to reach $1.2 trillion, while combined free cash flow already turned negative in Q3, burning about $31 billion in a single quarter.
The simple accounting logic: $1.2 trillion in capex combined with negative free cash flow means that the "next trillion dollars" Privorotsky worries about will mainly be financed through the bond market rather than operating cash flow. And the bond market itself is not comfortable either. The scale of AI-related off-balance-sheet debt has swelled to $4.1 trillion, and financing demand is becoming increasingly crowded.
Intel Plunges 29%
Last Friday was a typical example.
Earlier in the week, the market had still been immersed in a series of relatively positive headlines, but Intel (INTC) then delivered an "unexpectedly severe PC demand correction," sending its shares down 29% in a single day.
Privorotsky believes Intel's warning "has more to do with consumer PCs, margins, and the manufacturing business rather than AI demand itself."
But the market was clearly not that patient. Investors still traded it as yet another victim under the AI narrative.
Meanwhile, Amazon—which Privorotsky called a "crowded position"—saw its shares narrow from a 7% premarket decline to close down only 1.4%. Even though concerns about its free cash flow have not disappeared, the stock showed some resilience. The market began to price selectively. Still, Kioxia's 15% plunge on Monday served as a reminder to investors that this is not an indiscriminate broad-based rebound.
Hedge Funds Experience Largest Deleveraging Since 2015
Is it deteriorating AI fundamentals, or forced unwinding of crowded positions? Privorotsky's positioning data leans toward the latter.
Goldman-tracked systematic long/short strategies fell 2.17% in a single week, "the worst weekly performance since December 2023." Fundamental long/short net exposure only recovered to the seventh percentile in history, "mainly achieved through index short covering." Individual stocks have been net sold for seven consecutive weeks, while macro products have been net bought for two consecutive weeks. The American Association of Individual Investors (AAII) bull-bear sentiment indicator returned to negative territory, and CNN's Fear and Greed Index fell to the low 20s.
Morgan Stanley also found that hedge funds' AI-related holdings have dropped to underweight for the first time since early 2023. This follows one of the largest sustained deleveraging episodes in more than 15 years.
Privorotsky's conclusion is: "Sentiment and positioning have been reset."
But his Goldman colleague Coppersmith is more cautious. He pointed out that global risk appetite saw its largest single-month decline since November 2021 in October, but "we still have not reached the risk appetite index level that would send a contrarian buy signal." More importantly, although net exposure is already low, gross leverage remains as high as 314% (the 97th percentile over the past five years). In other words, while fund managers are reducing positions, overall leverage remains very high, and the energy for further volatility is still building within the market.
"Almost Unmanageable," but a Year-End Rally May Already Be Underway
Privorotsky had an unusually candid passage in his report:
It feels almost unmanageable. Almost every client I have spoken with is struggling. Last week felt like broad-based deleveraging... momentum factor unwind, short squeezes in the most heavily shorted stocks, leverage being pulled out of the system. Defensive stocks and high-quality assets surged, and AI valuation multiples compressed. The market is effectively pricing a sudden growth slowdown, a stall in AI capex, lower nominal growth and rates, and a premium on high-quality balance sheets.
He then pointed out that "there was nothing in last week's hyperscale cloud earnings to suggest any weakening of spending intent... if anything, the opposite." It depends on which bearish camp you belong to—this is either a reason for bulls or precisely the problem.
After getting past key events such as the midterm elections, the US Treasury's quarterly refunding announcement (QRA), and nonfarm payrolls data, Privorotsky offered his judgment:
My gut tells me we are setting up for a painful year-end rally... underpositioned, low participation, hard to chase. Friday did feel like capitulation selling, but I think we need to get past these events first before risk gets added back.
UBS also warned of year-end short squeeze risk, with short positions near record levels. Goldman derivatives strategist Brian Garrett also saw signs of "healing" in seven markets—but added that "we just need to get through 2026 first."
The conditions for a positioning squeeze do exist: net exposure at the seventh percentile in history, sentiment at rock bottom, and seasonal tailwinds approaching. But as Privorotsky himself asked: demand is real, yet the return on the next trillion dollars is uncertain. With hyperscale cloud providers' free cash flow already negative, gross leverage at the 97th percentile in history, and the 2027 capex bill about to land heavily on the bond market, any year-end rally could be a selling window.
Comments