Morgan Stanley raises tech giants capex forecast to $805 billion for 2026

Shoppers of data centres and investors alike are circling the new capex story: Morgan Stanley has lifted its combined hyperscaler capital-spending forecast to $805bn for 2026, a signal that big tech is doubling down on AI infrastructure , and that shift could matter for Microsoft, Meta and the wider chip and cloud supply chain.

Essential Takeaways

• Bigger number: Morgan Stanley now projects about $805bn in capex from major hyperscalers in 2026, signalling a sharp uptick in infrastructure spending.
• AI-first spending: The rise is tied to heavier investment in AI compute, data centres and related hardware, which smells like longer workloads and denser racks.
• Winners in focus: Markets see this as supportive of Microsoft and Meta’s growth trajectories, though odds differ , Meta’s price targets appear especially buoyant.
• Policy and supply-chain angle: US industrial policy, including the CHIPS Act, is a backdrop , expect onshoring and semiconductor security to shape where cash flows.
• Watch the signals: Quarterly capex announcements, earnings talk and geopolitical moves on US-China tech will be the next obvious catalysts.

Why $805bn is a louder buy signal than it first looks

The headline figure is arresting because it nearly doubles recent year-on-year levels and concentrates capital where it’s most visible: AI compute and data-centre expansion. That creates a distinct sensory image , thicker rows of servers, louder cooling systems and shipping manifests full of GPUs and custom silicon. According to Morgan Stanley commentary, the revision reflects companies preparing for sustained AI workloads and the need to be self-sufficient on compute. For investors and customers, that suggests capacity growth that could underwrite product rollouts and lower latency for AI services.

How this ties into tech-policy and the chip story

Spending on compute doesn’t happen in a vacuum. The CHIPS and Science Act and other US policy moves are nudging semiconductor investment onto domestic soil, which in turn makes hyperscalers more willing to lock in partnerships and long-term orders. That means capex isn’t just racks and power; it’s also deals with fabs and power projects to feed denser hardware. So if you’re tracking where the money lands, follow the fab announcements and energy investments as closely as the server orders.

Who benefits , why markets are cheering Microsoft and Meta differently

Markets are pricing nuanced outcomes: Microsoft is seen as a principal beneficiary of broader AI infrastructure, but current prediction markets put its chance of being the biggest company by December 2026 at a modest probability. Meta, by contrast, has very aggressive market pricing around a specific price target, reflecting investor conviction in its AI investments. The difference boils down to product mix, monetisation pathways and investor storytelling , Microsoft’s cloud and enterprise exposure versus Meta’s ad-revenue-plus-AI play. For portfolio-minded readers, that means distinguishing durable infrastructure value from headline-grabbing growth narratives.

Practical tips for investors and customers watching capex rollouts

If you want to read the tea leaves, start with company-level capex disclosures and the pace of server and GPU orders reported in earnings calls. Watch for concrete signs , land purchases for data centres, power deals, or long-term chip supply agreements , rather than vague promises. For customers, larger capex can mean better service reliability and geographic expansion; for investors, it’s about timing: capex often precedes revenue gains, so patience matters. And if you care about risk, don’t ignore geopolitical exposures and the supply chain pinch points that could delay deployments.

What could derail the momentum , and what to watch next

The obvious risks are policy shocks, a cooling macroeconomy that tightens corporate budgets, or prolonged US-China frictions that complicate supply chains. Analysts will also be scanning energy markets, since denser compute needs more power and cooling, and that’s where Morgan Stanley’s energy outlook pieces come into play. Keep an eye on quarterly revisions to capex plans and any comments at tech conferences where CEOs and CFOs detail hardware priorities; these are the moments plans become commitments.

It’s a big, loud bet on AI infrastructure , and one small shift in policy or supply can change the tune.

Source Reference Map

Story idea inspired by: ^[1]

Sources by paragraph:

Noah Fact Check Pro

The draft above was created using the information available at the time the story first
emerged. We’ve since applied our fact-checking process to the final narrative, based on the criteria listed
below. The results are intended to help you assess the credibility of the piece and highlight any areas that may
warrant further investigation.

Freshness check

Score:
7

Notes:
The article was published on May 3, 2026, and references a recent upward revision by Morgan Stanley of hyperscaler capital expenditure forecasts to $805 billion for 2026. This information aligns with recent reports from reputable sources such as Tom’s Hardware and FXStreet, indicating that the narrative is current and not recycled. However, the article’s reliance on a single source, BitcoinEthereumNews, which aggregates content from other outlets, raises concerns about originality and source independence. Additionally, the article includes a ‘Source Reference Map’ with links to other articles, suggesting that the content may be derivative. The earliest known publication date of the substantially similar content is May 1, 2026, from Tom’s Hardware. Given these factors, the freshness score is reduced to 7.

Quotes check

Score:
5

Notes:
The article does not provide direct quotes from Morgan Stanley or other primary sources. Instead, it paraphrases information from secondary sources, which diminishes the ability to verify the accuracy of the statements. The lack of direct quotes or citations from original sources raises concerns about the reliability and verifiability of the information presented. Therefore, the quotes score is 5.

Source reliability

Score:
4

Notes:
The primary source of the article is BitcoinEthereumNews, which aggregates content from other outlets. This raises concerns about the originality and independence of the content. The article also references other sources, including Tom’s Hardware and FXStreet, but does not provide direct links or citations, making it difficult to assess the reliability of the information. Given these factors, the source reliability score is 4.

Plausibility check

Score:
6

Notes:
The article discusses Morgan Stanley’s upward revision of hyperscaler capital expenditure forecasts to $805 billion for 2026, a figure that aligns with recent reports from reputable sources such as Tom’s Hardware and FXStreet. The discussion of increased investment in AI infrastructure and its potential impact on companies like Microsoft and Meta is plausible and consistent with industry trends. However, the lack of direct quotes or citations from original sources raises concerns about the accuracy and verifiability of the information presented. Therefore, the plausibility score is 6.

Overall assessment

Verdict (FAIL, OPEN, PASS): FAIL

Confidence (LOW, MEDIUM, HIGH): MEDIUM

Summary:
The article presents information about Morgan Stanley’s upward revision of hyperscaler capital expenditure forecasts to $805 billion for 2026, aligning with recent reports from reputable sources. However, the reliance on aggregated content from secondary sources, the lack of direct quotes or citations from original sources, and the absence of original reporting or analysis raise significant concerns about the originality, source independence, and verifiability of the information presented. Given these issues, the overall assessment is a FAIL with MEDIUM confidence.