Google CEO Sundar Pichai gestures during a meeting with French President Emmanuel Macron on the sidelines of the AI Impact Summit in New Delhi on February 19, 2026.
Ludovic Marin | Afp | Getty Images
Wall Street analysts estimate that total AI capital spending in 2027 could now rise to over $1 trillion after hyperscalers revealed even bigger spending plans in tech earnings on Wednesday.
Both Evercore and Bank of America have capital spending of over $1 trillion after 2027 earnings releases, with estimates for 2026 rising to $800 billion to $900 billion.
“Cap-Ex continues to rise as demand exceeds supply and prices rise,” analysts at Jefferies said in a note to investors Thursday.
This year’s spending forecasts were positive across the board, particularly at Google’s parent company alphabet increased 4% to $185 billion, Amazon increased by 1% to $200 billion, Meta increased 8% to $135 billion, and Microsoft According to a Bank of America balance sheet, the value rose a whopping 24% to $190 billion.
Tech CEOs exude confidence in their investments in artificial intelligence as evidence of monetization such as: B. rising cloud revenues, but the huge expenses still raise skepticism among investors.
Amazon CEO Andy Jassy said the company is “confident in the long-term capital investments we are making” and forecasts a $200 billion expansion for the year.
Alphabet’s cloud revenue rose 63% year-over-year in the first quarter, driving its stock up about 10%. Chief Financial Officer Anat Ashkenazi said on Wednesday that investment plans would be increased to meet “strong demand.”
Investors are hunting for returns
The overall cost of AI expansion is causing excitement, but analysts say they are seeing a shift from investment to revenue as valuations and market caps rise.
“Cap-ex continues to rise, but [return on investment] “The ROI is reflected in an order backlog of approximately $2 trillion and accelerated cloud growth,” analysts at Jefferies said. “Margin levers remain intact for the hyperscalers despite AI investments, highlighting the structural structure.” [operating expense, or] Opex discipline.
Confidence in monetization is particularly high at Alphabet, as backlog growth underpins computing inventories and expansions.
The “backlog supports [the] “Cap-Ex supercycle,” Brian Pitz wrote for BMO Capital Markets on Thursday. “Google’s backlog nearly doubled quarter-over-quarter, up 400% year over year to $462 billion.”
“The majority of the backlog is for core Google Cloud Platform contracts… and Google expects to recognize just over 50% of this as revenue over the next 24 months,” he added.
While Google’s cloud revenues impressed analysts, Meta’s expansion plans worried investors who wanted to see a higher return on the investments made. The shares recently fell by around 8%.
“Meta will likely remain in the penalty box until a clear investment ROI emerges,” analysts at Jefferies wrote in a note Thursday.
The company spent $72 billion on capital expenditures in 2025 and expects that number to double to $125 billion to $145 billion in 2026. That’s up from $115 billion to $135 billion previously.
“We are increasing our infrastructure investment forecast for this year,” Meta CEO Mark Zuckerberg said on Wednesday. “Most of this is due to higher component costs, particularly memory prices. But every indication we see in our own work and across the industry gives us confidence in this investment.”
Meta’s free cash flow has declined, falling from $26 billion in the same period last year to just $1.2 billion in the first quarter.
Analysts at Bank of America expect revenue and free cash flow across the sector to improve in 2026, helping to support spending.
Who benefits?
The continued investment growth is good news for chipmakers and equipment providers. They are customers of the hyperscalers, and analysts took note of that on Thursday. CPU maker’s first quarter earnings Intel were particularly strong because AI expansion requires more than just graphics processors or GPUs.
There is a “strong and growing demand for various custom-made products”. [application-specific integrated circuits] programs (TPU, Trainium, Maia and MTIA),” said analysts at Evercore, who touted “an acute focus on agent AI as a key use case that we believe will lead to a CPU renaissance in the next few years.”
RBC Capital Markets maintains its positive ratings Nvidia, Micron technology, Marvell, Astera Labs, Arm stocksAnd Lattice semiconductors.
“Strong investment trends should also bode well for industry-rated AVGO, AMD, SNDK and INTC,” RBC said. “AI demand is driving double-digit growth in the wafer fab.”
Choose CNBC as your preferred source on Google and never miss a moment from the most trusted name in business news.