AI infrastructure suppliers shed $4.2 trillion in market value | Magica
Data center stocks shed $4.2 trillion despite strong supplier orders
Editorial Team
••📖6 min read
Manufacturers, utilities and miners linked to data center construction have shed $4.2 trillion in market value since June. Strong supplier orders contrast with financing concerns, project delays and political resistance.
Data center-linked manufacturers, utilities and miners have lost $4.2 trillion in market value since June.
The retreat reverses almost half their preceding year's gains; it does not measure canceled construction spending.
Strong reported sales and backlogs leave the durability of future orders as the central test.
Global manufacturing, utility and mining companies tied to data center construction have shed $4.2 trillion in market value since June, according to an analysis of the sell-off. Investors are retreating from the physical suppliers of the AI expansion even as those businesses report substantial demand.
The reversal challenges the investment case for buying the generators, electrical equipment, cooling systems and fiber optics behind AI computing. These companies offer exposure to infrastructure spending, but their shares remain vulnerable to doubts about customers' financing and their ability to secure permission to build.
Magica chart: the loss since June erased almost half the preceding year’s gains, according to Financial Times analysis. Bars show magnitudes of market-value changes, not canceled construction spending. Source: Financial Times.
Almost half the gains, not half the companies' value
The $4.2 trillion decline follows a $9.4 trillion increase in the same companies' market value over the previous 12 months. The loss therefore represents roughly 45% of those earlier gains, calculated by dividing $4.2 trillion by $9.4 trillion. It is neither a 45% fall in the group's total market capitalization nor a tally of money withdrawn from construction projects.
The analysis found that companies more frequently associated with data center construction in media coverage suffered the biggest declines. That measures a relationship between market performance and association with the boom; it does not establish how much of each company's revenue comes from AI infrastructure. The published account does not provide a full constituent list or enough methodological detail to independently reproduce the aggregate.
Caterpillar, founded as Caterpillar Tractor Co. in 1925, makes construction and mining equipment as well as engines and industrial gas turbines, as described in its annual filing. Its role as a major gas-generator supplier to data centers gave investors another route into the AI build-out.
RIKEN has stopped Fugaku’s research jobs after an intrusion into its operational network. Up to 4,597 users’ account information may be affected, and intruders tried acquired credentials on other systems; research-result leakage remains unconfirmed.
Editorial Team
The company lost $214 billion in market capitalization, or 44%, from its June 30 share-price peak, according to FactSet figures cited in the account of the analysis. Nuclear power developer Oklo fell 49% despite signing a contract with technology company Meta in January. Heating, ventilation and air-conditioning supplier Comfort Systems and fiber-optic manufacturer Corning each lost around half their value over the period described in the report.
The weakness reaches beyond individual stocks. Between June 30 and November 13, the S&P 500 industrial sector fell 9%, compared with an 11% decline for technology. Energy rose 22% over that interval. The distinction matters: losses among energy businesses associated with data centers do not imply that the broader energy sector fell.
The customers' risks reach the suppliers
Concerns center on escalating debt-financed spending by hyperscalers—the large technology companies operating vast computing infrastructure—and growing US opposition to data center construction. Both pressures feature in the account of the reversal.
For suppliers, the link is practical. A customer's financing difficulties or a construction restriction can affect whether equipment is needed and when it can be delivered. A large market for computing does not by itself settle the timing or location of the buildings needed to serve it.
There is also evidence of projects falling behind schedule. Dan Thompson, an analyst with 451 Research at S&P Global Energy, estimated that roughly two-thirds of the projects S&P Global was tracking faced delays, ranging from construction problems to community opposition, prospective regulation and moratoriums. The group was tracking 157 projects that had been paused, severely delayed or abandoned, according to a report on construction pressures citing his remarks in an earlier account.
That report also cites Sightline Climate analysts' estimate that 30–50% of US AI data centers planned for 2026 would be delayed or canceled. These estimates concern particular project populations; they do not establish cancellation rates across all global data center investment or how much of suppliers' current backlogs is exposed.
Deutsche Bank senior US economist Brett Ryan argued in the original analysis that institutional buying had helped produce overly enthusiastic valuations. He also said equity markets were catching up with concerns already evident in credit markets about risky construction financing. That is an interpretation of the repricing, rather than evidence that all the projects behind it have become uneconomic.
Magica chart of company-reported figures: Caterpillar’s total backlog, Siemens’ data center electrical-equipment backlog in 2026, and Vertiv’s third-quarter sales. These are distinct measures, not equivalent order totals. Source: Seeking Alpha.
Orders remain stronger than share prices
The strongest challenge to a straightforward bust narrative is the reported operating performance. US data center construction spending reached $96 billion in the 12 months through September, according to government figures cited in the original analysis. That figure excludes chips and IT equipment and cannot be compared directly with a global stock-market loss.
Company figures relayed in that analysis show that Caterpillar reported a record $78 billion backlog and said power-generation sales rose 75% year on year, driven by data center demand. The backlog is the company's reported total; it should not be read as $78 billion of data center orders.
Electrical equipment supplier Siemens reported that its data center electrical-equipment backlog rose 60% in 2026 to €7 billion. Eaton, another electrical equipment supplier, reported orders doubling year on year. The account does not isolate a data center-only order total for Eaton. These measures describe different businesses and order categories, rather than a single comparable growth rate across the industry.
Power and cooling supplier Vertiv illustrates how sharply the market and operations can diverge. Its third-quarter sales rose 61% to $4.31 billion and orders more than doubled, yet shares fell 23% in the week after the US midterm elections, according to the report on infrastructure suppliers.
Dell'Oro Group analyst Alex Cordovil argued that record backlogs supported revenue growth through 2027, with electrification and supply-chain restructuring also supporting industrial firms. But the original report carried a counterargument from venture capitalist Paul Kedrosky: customers placing orders with several suppliers to reserve scarce capacity could exaggerate underlying demand. His warning identifies possible duplication; it does not quantify inflated orders. Strong order books offer visibility, not certainty that every reservation will become revenue.
What must hold for the backlogs to pay off
The next test is whether suppliers convert orders into deliveries while customers navigate financing, construction delays and political restrictions. Restrictions on new starts need not have the same effect as interruptions to work already under way. Richard S Warzala, chief executive of industrial-controls maker Allient, argued on a November 5 earnings call, as relayed in the original report, that installations already started still needed to be completed. His view comes from a supplier with a stake in continued construction.
The geographic effect also remains unresolved. Citi researchers expected tighter regulation to increase volatility but considered widespread construction restrictions unlikely, according to the account of their assessment. Ryan argued that restrictions could move projects to more accommodating jurisdictions rather than eliminate demand.
Evidence on cancellations, duplicate orders and actual deliveries will help establish how dependable the backlogs are. Financing terms and enacted construction rules will help determine which later projects can proceed, where they can be built and when suppliers can recognize the resulting sales.
DeepSeek users report daily caps, shorter inputs and restrictions on resuming inactive conversations. Reported financing and a planned data center do not establish when more capacity will reach the free chat service, whose quota rules remain unclear.
Google found an attacker framework configured to compare software patches, develop exploits and execute them without human intervention. The finding sharpens the pressure on defenders to patch quickly, but does not establish that entire attacks now run autonomously.
Corma’s $125 million Series A brings total funding to $185 million as the defensive cybersecurity AI lab confronts a computing bottleneck. Its specialized-model strategy faces existing security agents and an unresolved test of performance and deployment costs.
Ask Maps is adding shareable travel schedules, saved-list availability searches and U.S. appointment finding. The update extends an existing AI planning service rather than introducing trip planning to Maps.
SemiAnalysis has reportedly cut its 2030 global AI accelerator spending forecast by about 40% to $923 billion, with Nvidia facing the largest revisions. Lower price assumptions and financing constraints help explain the reduction, while customer deals show both the strength and financial cost of competing chip platforms.
Federal Reserve Governor Michael S. Barr says banks report AI agents attempting account actions without verifiable permission. Autonomous deposit transfers could accelerate future bank runs, but their scale and speed remain unmeasured in his speech.
Amazon-owned Zoox is opening its Las Vegas robotaxis to Uber riders while keeping its own app. The November launch follows an earlier summer target; a joint MGM Resorts agreement addresses resort pickups, while fleet size and financial terms remain undisclosed.
Germany’s reported military AI shortlist excludes Palantir, while France and the Netherlands pursue European alternatives. The challenge is to secure national control over data and infrastructure while delivering usable systems that work with allies already adopting the US supplier.
China aims to make school AI education broadly available by 2030. National guidance already calls for pupil assessment, but an investigation finds common measures of AI literacy still lag an uneven rollout spanning courses, tutoring and student monitoring.
Drive Capital led Zus Health's $33 million Series B, extending a shared-record strategy backed by earlier financing and EHR partnerships. The company is expanding AI workflows and pursuing a more direct role in national data exchange, while disclosed activity figures leave customer returns unmeasured.
Atomarine’s $19 million seed round and Mocean Energy’s wave-powered computing proposal target land, water and electricity constraints. Earlier underwater projects show technical promise, but do not establish the costs or performance of these new floating designs.
Tencent’s new Hunyuan model charges 2 yuan per million output tokens and targets everyday office work. Its claimed 95% task-cost saving against Hy4 comes from an internal benchmark, with a trade-off on long coding tasks.
The Indian chip startup is funding programmable image sensors and targeting production in 2027. Its promised system savings remain unproven, while Sony's earlier designs challenge broad claims of novelty.
Baidu’s AI cloud infrastructure revenue nearly doubled in the third quarter, cushioning a 22% advertising decline. But AI applications were flat, app users fell, and the return to reported profit obscured lower adjusted earnings and a wider annual free-cash-flow deficit.
The Trump administration reportedly plans to designate more than two dozen AI companies to help win overseas business. Existing export rules support integrated technology packages, but guarantee neither financing nor licenses and leave room to support competing suppliers.
Singular leads the Danish accounting software company's seed round as it builds a London commercial hub and targets Benelux. Its pitch rests on automating preparation while accountants retain sign-off, with claimed efficiency gains still awaiting independent verification.
Alibaba says Qwen4-Max reduces average task costs by 62% on its workplace benchmark. The hosted launch builds on its predecessor’s agent capabilities as Alibaba expands a growing cloud business and bears the expense of AI infrastructure.
The FBI warns that ARTEX can accelerate intrusions, with some reaching privileged access in under three hours. Microsoft also documents unsuccessful mass targeting, while compromised backup-service credentials opened a separate route into customer environments.
Jensen Huang has reportedly discussed returning to Nvidia three days after resigning. Supporters want reinstatement during an investigation, while directors have installed interim leadership and begun a permanent CEO search.