Deutsche Bank: Why the AI boom Risks a US$800bn Shortfall
The AI boom cannot continue without exponential increases in technology spending, Deutsche Bank warns – as new research reveals an US$800bn gap between projected AI revenues and the computing power needed to sustain growth.
The bank’s analysis shows that AI capital expenditure has reached such massive levels that it is single-handedly preventing the US economy from entering recession.
Without this tech-driven investment surge, America would already be in economic contraction, the bank’s researchers conclude.
George Saravelos, Head of FX Research at Deutsche Bank tells clients that “AI machines – in quite a literal sense – appear to be saving the US economy right now.”
He says that “in the absence of tech-related spending, the US would be close to, or in, recession this year.”
The maths problem
This assessment comes alongside Bain & Company’s projection that the AI sector faces a mathematics problem.
The consulting firm estimates that meeting anticipated AI demand by 2030 will require US$2tn in annual revenue to fund the necessary computing infrastructure.
Even accounting for AI-driven cost savings across industries, the world remains US$800bn short of this target.
The warning follows Nvidia’s record US$100bn investment in OpenAI, highlighting the enormous sums flowing into AI infrastructure.
Nvidia now supplies the graphics processing units (GPUs) that power everything from ChatGPT to autonomous vehicles.
The impact of the “magnificent seven” driving market gains
This concentration of AI investment has created significant market distortions.
Deutsche Bank’s Head of Macro and Thematic Research, Jim Reid, notes that “the S&P 500 is now up 13.81% so far this year, whereas the equal-weighted version is only up 7.65%.”
The difference reflects how heavily the index relies on just seven technology giants: Apple, Microsoft, Alphabet, Amazon, Meta, Tesla and Nvidia.
Jim explains that “it’s been the Magnificent 7 driving the gains,” while the remaining 493 companies in the index have seen far more modest returns.
This divergence has created concerns about market stability and whether current valuations can be sustained.
Apollo Management’s Partner and Chief Economist, Torsten Sløk, reinforces this view, noting that “the upward consensus revision to 2026 earnings for the S&P 500 since Liberation Day comes entirely from the Magnificent 7.”
Meanwhile, he observes, “earnings expectations for the S&P 493 have remained suppressed and are not moving higher.”
Goldman Sachs offers a more optimistic perspective, with analysts led by Economist Manuel Abecasis projecting significant long-term benefits.
“We expect productivity gains from AI to boost GDP significantly, by about 0.4% through the next few years and 1.5% cumulatively as adoption rises over the long run,” he says.
The Goldman team argues that “once it is widely adopted, AI is likely to allow workers and firms to produce more output for a given set of inputs, which will raise total factor productivity growth.”
The key area of current AI growth
However, Deutsche Bank’s analysis suggests the present economic impact comes not from AI applications themselves but from the massive infrastructure projects required to support them.
Companies are pouring billions into data centres, specialised computing hardware and the power systems needed to run AI operations.
George points out that “growth is not coming from AI itself but from building the factories to generate AI capacity.”
This distinction matters because infrastructure spending cannot increase indefinitely at current rates.
Goldman Sachs estimates that AI capital expenditure reached US$368bn through August alone, with hyperscalers like Amazon Web Services (AWS), Microsoft Azure and Google Cloud leading the investment surge.
The sustainability question hinges on whether technology companies can maintain these spending levels.
George warns that “in order for the tech cycle to continue contributing to GDP growth, capital investment needs to remain parabolic.
“This is highly unlikely,” he says.
He suggests that Nvidia’s role has become central to US economic performance.
George notes that “it may not be an exaggeration to write that Nvidia – the key supplier of capital goods for the AI investment cycle – is currently carrying the weight of US economic growth.”
Meanwhile, Bain & Company’s report acknowledges potential cost savings but remains pessimistic about the revenue gap.
“Two trillion dollars in annual revenue is what’s needed to fund computing power needed to meet anticipated AI demand by 2030,” the report says.
“However, even with AI-related savings, the world is still US$800 bn short to keep pace with demand.”
