Global manufacturing, utility and mining companies have shed $4.2 trillion in value since June as the AI “picks and shovels” trade swings into reverse, the Financial Times reported on 15 November. An FT analysis found that the companies more frequently linked by media coverage to data centre construction were hit hardest, and that the losses wiped out almost half of the $9.4 trillion the same companies gained in the previous 12 months.
Our own check uses public prices. An equal-weighted basket of the nine US-listed companies labelled in the FT’s chart fell 39.4% from its 22 June peak to the 13 November close, by our arithmetic on Nasdaq official closing prices. That gives back about 82% of the basket’s 2026 gain; the SPY S&P 500 ETF is up 10.6% this year on the same closes.
Business Pill · Backlog Explained
An evergreen Business Engineer explainer on backlog, revenue that is signed but not yet earned, the measure at the centre of the debate the FT reports. It teaches the general idea only and does not discuss these companies.
The key insight: As we read it, the FT’s numbers and ours point the same way from two directions: a $4.2 trillion loss across a broad global set since June, and a 39.4% fall in a nine-stock US basket from its 22 June peak, by our arithmetic, while the companies’ reported backlogs, per the FT, kept growing.
What the FT Reported
The FT cites Caterpillar, “a leading supplier of gas generators to data centres”: its market capitalisation fell by $214 billion, or 44%, since its shares hit an all-time high on 30 June, according to FactSet data cited by the paper.
It says nuclear power company Oklo plunged 49% despite signing a contract with Meta in January, and that Comfort Systems, an HVAC maker, and Corning, a fibre optics maker, both lost around half of their value over the same period. Industrials, it adds, have been the second-worst-performing sector in the S&P 500 since the summer, after technology.
The paper ties the sell-off to investors questioning whether debt-financed capital spending at hyperscalers and political opposition to data centres in the US could lead to a bust. The $4.2 trillion loss is 44.7% of the $9.4 trillion gain, by our arithmetic, which matches the FT’s “almost half”.

Our Basket, by Our Arithmetic
The FT did not list every company in its analysis, so we built our own basket from the names labelled in its scatter chart that trade in the US: Caterpillar, Vertiv, GE Vernova, Eaton, Comfort Systems USA, Corning, Oklo, Constellation Energy and Bloom Energy. Mitsubishi Heavy Industries, Siemens and ABB, also labelled, are not US-listed and are left out. The selection is ours, not the FT’s.
- The basket. Each stock re-based to 100 at its 31 December 2025 close, then averaged with equal weights. The basket peaked at 191.9 on 22 June and closed at 116.3 on 13 November, a fall of 39.4%.
- Caterpillar checks out. Its 30 June close of $1,064.90 was its 2026 high; the 13 November close of $600.20 is 43.6% lower, in line with the FT’s 44%.
- The deepest falls. From each stock’s own 2026 high close: Oklo down 77.3% (high on 9 January), Corning down 55.4%, Vertiv down 46.3% and Comfort Systems down 46.2%. Seven of the nine are more than 30% below their 2026 high.
- The exception. Constellation Energy is 12.8% above its 30 June close; its 2026 high close came on 2 January. The other eight are below their 30 June close, and the SPY ETF is up 1.0% over the same span.
The Case on Each Side, in the FT’s Reporting
Paul Kedrosky, a venture capitalist and fellow at the MIT Institute, said investors had been wrong to start treating old-school industrial firms like semiconductor stocks. He told the FT: “It had become endless in discussions with hedge funds and family offices, this idea that everything was now a growth stock — which is ludicrous, of course.”
Brett Ryan, senior US economist at Deutsche Bank, said the stock market was “finally benchmarking to what credit was already telling you” about the growth in risky financing for data centre construction.
The other side rests on orders. Alex Cordovil, an analyst at Dell’Oro Group, said “The build-out is not decelerating as of yet”. Per the FT, Caterpillar reported a record $78 billion backlog and 75% year-on-year growth in power generation sales. Siemens’ data centre backlog grew 60% in 2026 to €7 billion, and Eaton reported orders had doubled year-on-year, a pattern in line with the power-equipment orders we tracked in October.
US spending on data centre construction, excluding IT equipment and chips, hit $96 billion in the 12 months to September, the latest government data showed, per the paper.
Kedrosky cautioned that backlogs and orders are trailing indicators, with people “double and triple ordering to protect their position in the queue given perceived scarcity”. Constellation Energy’s Joseph Dominguez told investors the post-election move had been an “overreaction”, and Citi researchers said widespread restrictions on construction remained unlikely, the FT reported.
Two of the names have their own recent threads in our coverage: Corning’s analyst target versus its June peak and Oklo’s fuel deal tied to Meta.
The Structural Read
As we read the FT’s reporting, the dispute is about what the order books mean: the bulls quoted point to record backlogs and orders, while Kedrosky calls them trailing indicators inflated by double and triple ordering. Our basket shows the share prices have moved toward the second reading since June; it does not show which reading the order books will bear out.
Brett Ryan, senior US economist at Deutsche Bank, quoted by the Financial Times, 15 November 2026
“At the end of the day, if there is demand for these things, they will get built somewhere”
Three Implications
FOR SUPPLIERS Per the FT, Caterpillar reported a record $78 billion backlog and Eaton said orders doubled year-on-year, even as our basket fell 39.4% from its June peak.
FOR FINANCING Deutsche Bank’s Brett Ryan told the FT the stock market was catching up with what credit markets had been signalling about risky data centre financing.
FOR SITING The FT says the impact of the moratoriums backed in this month’s midterms is still unclear; Citi researchers said widespread construction restrictions remained unlikely.
The Business Engineer Lens
This story maps onto the Business Engineer framework The AI Capex Map & The State of AI Hyperscalers.
The framework argues: “This is not a continuation of the cloud capex cycle. This is a different cycle, with different physics.”
As we read it, the FT’s $4.2 trillion figure is the market repricing the suppliers of that cycle, the generator, cooling, power and fibre makers, rather than the hyperscalers that fund it. The framework predates this sell-off and does not discuss it.

What Is Not Established
The $4.2 trillion and $9.4 trillion figures are the FT’s own analysis of global manufacturing, utility and mining companies; the article we read does not list every company or set out how media coverage was scored. Our nine-stock basket is a different, much smaller set, chosen by us and weighted equally rather than by market value, so its 39.4% fall is not comparable to the FT’s dollar loss.
Our figures run to the 13 November close and can differ from the FT’s for the same company: Oklo is 54.3% below its 30 June close on our data, against the FT’s 49%; the article does not give the end date of its measurement. The impact of the data centre moratoriums backed by US voters in this month’s midterms is still unclear, the FT says.
The Bottom Line
The Financial Times reported that global manufacturing, utility and mining companies have shed $4.2 trillion since June, almost half of the $9.4 trillion they gained in the previous 12 months, with data-centre-linked names hit hardest. On Nasdaq closing prices, our equal-weighted basket of nine US-listed names from the FT’s chart fell 39.4% from its 22 June peak to 13 November, giving back about 82% of its 2026 gain, by our arithmetic.
95,000+ executives read Business Engineer for the AI strategy frameworks cited by ChatGPT, Claude, and Perplexity.
A note on sourcing. We read the Financial Times article by Clara Murray in full; every figure and quote about the $4.2 trillion loss, the companies it names and the people it quotes comes from it. The basket, its index, each stock’s change from its 2026 high and from 30 June, and the 44.7% and 82% ratios are our arithmetic on Nasdaq official closing prices to 13 November 2026. The choice of the nine companies is ours.
Sources: Financial Times, 15 Nov 2026 · Nasdaq historical closing prices (CAT, VRT, GEV, ETN, FIX, GLW, OKLO, CEG, BE, SPY)









