Morgan Stanley Says 'Buy the Drop' on AI Stocks, Keeps $3.4 Trillion Capex Forecast

After a post-midterm selloff on data center regulation fears, Morgan Stanley argues spending is likely delayed, not lost. The winners may differ, though.

The 20-second read

1
$1.6 trillion in 2027 and $1.8 trillion in 2028. Morgan Stanley did not cut its capex forecast for the hyperscalers as a group, and expects only some data center construction to slip by a year.
2
The industry structure favors Amazon, Alphabet, Meta and Microsoft, which have the capital, the ability to secure power, and AI model efficiency.
3
The strategy is not to buy semiconductor and power stocks across the board, but to pick companies that have actually secured power and can withstand project delays.

Morgan Stanley's report, published Nov. 8, carries a blunt title: "Buy the Drop."

AI and semiconductor stocks fell sharply after the U.S. midterm elections on fears of data center regulation. Morgan Stanley judged the market reaction excessive. It sees the pullback as a chance to selectively buy high-quality AI companies.

The core evidence is one number. The bank left its combined 2027 and 2028 hyperscaler capital spending forecast unchanged at about $3.4 trillion.

Other Wall Street banks, including UBS, HSBC and Jefferies, have also said regulatory worries are overpriced. But Morgan Stanley is not telling investors to buy every AI-related stock. The report's central point is that tighter regulation favors companies that own capital, power and data centers already running.

1. The key number: $3.4 trillion has not gone away

  • 2027 expected capex: $1.6T (about $1.6 trillion)
  • 2028 expected capex: $1.8T (about $1.8 trillion)

Morgan Stanley hyperscaler capex forecast (billions of dollars)

YearCapex
2025$481B
2026$951B
2027$1,641B
2028$1,760B

Source: Morgan Stanley Research, Exhibit 1. The figures are combined company-wide capex forecasts and do not equal AI-only spending.

After the midterms, the market worried less about a Democratic win itself than about the chance that regulation could block data center construction and sharply cut AI investment.

Morgan Stanley's math differs. It estimates that about 5 GW of U.S. data center capacity scheduled for completion in 2028 could slip to 2029. That equals roughly 20% of U.S. projects and about 10% of global projects for that year.

Projects already under way for 2027 are expected to largely stay on track. Downside risk remains for 2028 spending, but a scenario in which investment plans collapse entirely is not the base case.

A canceled investment is a loss of demand. A delayed investment is a shift in the timing of supply.

Morgan Stanley argues delays could rebalance an overheated supply chain, raise rents on computing capacity already secured, and lengthen the AI capex cycle.

Delays can still hurt. Some equipment makers that counted on 2028 revenue could see weaker results. That is why growth in the AI industry should not be equated with gains in every AI stock.

2. The paradox: tougher rules favor Amazon and Google

Why does Morgan Stanley prefer the large hyperscalers?

If building new data centers gets harder, the assets of companies that already hold power, land and servers become more valuable. New entrants arrive later, and incumbents may be able to charge more for scarce computing capacity.

Big tech companies can also shift construction to other regions and raise funding on relatively favorable terms. Morgan Stanley says those strengths favor large companies such as Amazon, Alphabet and Meta.

CompanyStrength Morgan Stanley highlights
AMZN, META, GOOGLLarge capital base and ability to scale AI services
MSFTCloud and enterprise AI platforms
NVDAChips that raise AI compute per unit of power
SPCXSpaceX's separate growth story and AI infrastructure opportunities

The yardstick for AI competition is also shifting. Until now, what mattered was who secured more GPUs. Increasingly, what may matter is how much useful AI output a company can produce from the same amount of power.

Morgan Stanley noted that Meta aims to raise compute efficiency tenfold in 2027. It also cited Google's next-generation TPU as an example of efficiency gains.

Competitiveness in AI is widening from GPU holdings to productivity per unit of power and the ability to generate actual revenue.

That matters for Nvidia too. Chips that make training and inference more efficient can capture demand from companies trying to get more out of existing capacity, even if new construction is delayed.

3. Chips and power stocks are not all the same opportunity

Morgan Stanley is far more cautious on the data center supply chain.

If 2028 construction slips, demand for servers, semiconductors and power equipment could arrive later than expected. Fewer orders and rising inventory could weaken some suppliers' pricing power and margins.

By contrast, data centers that already have power connections, and companies that can deliver power quickly, may benefit.

Relatively favoredWatch with caution
EQIX: existing data center operating baseCompanies whose revenue is concentrated in new construction schedules
WULF, RIOT, HUT, APLD: relatively low Texas exposureCIFR, GLXY, MARA: risk of Texas project delays
Facilities with secured, operational powerFacilities with uncertain power connections or permits

This split is not a grade on any company's overall profitability. It is a sorting of project exposure to policy change.

The regulation-exposed regions Goldman Sachs reviewed account for about 45% of planned U.S. data center capacity, and Texas alone for about 25%. A company's data center locations and approval stages could therefore become key variables in judging its stock.

4. Morgan Stanley is not the only bull

Wall Street views compiled by Reuters on Nov. 9 are similar.

UBS said the market was too pessimistic about AI-related policy, credit risk and the possibility of spending cuts. HSBC said it is using the pullback to add to U.S. and Asian tech stocks. Jefferies also read the Nov. 5 plunge as an immediate market reaction rather than a fundamental collapse in valuations.

That does not mean Wall Street sees no problems in AI investment.

JPMorgan forecasts about $900 billion in AI-related capex for 2026. It stresses that cloud revenue to recoup the massive spending, future backlog and the ability to raise more funding all matter.

The common direction is clear. Rather than denying the AI industry's long-term growth, Wall Street is judging more strictly who can secure the capital and power that growth requires, and turn it into real profit.

Does Morgan Stanley think AI stocks have hit bottom?
  • No. Judging that the decline was overdone is different from calling the exact bottom in prices.
  • Morgan Stanley's strategy is closer to selective, staged buying.
Data centers are being delayed, so why is that positive for Nvidia?
  • Delays in installing new GPUs are a short-term burden.
  • But the need to do more computing with limited power raises the value of high-performance chips.
Is the $3.4 trillion a committed investment?
  • No. It is the sum of Morgan Stanley's corporate capex forecasts for 2027 and 2028.
  • It should be distinguished from committed AI-specific contract amounts or money already spent.
What is the biggest opposing scenario?
  • Regulatory delays drag on past 2029 and cloud revenue growth falls short of expectations.
  • In that case, capex could be cut, and earnings estimates across the data center supply chain could be revised down in a chain reaction.
What should individual investors do first?
  • Separate long-term conviction in the AI industry from the right buying price for each individual stock.
  • Distinguish companies with stable cash flow from those that depend on financing, and first check how concentrated your existing portfolio is in AI.

What to watch

Changes in Big Tech capex forecastsWhether the 2027 to 2028 forecasts actually hold is the first evidence that the investment is sustained.
Data center permits and power connectionsWatch whether approval delays in Texas and Ohio spread to other states.
Cloud revenue and AI monetizationThe key is whether service revenue and operating cash flow keep pace with the increase in capital spending.
Chip orders and marginsEven if AI demand holds, pushed-back order schedules and rising inventory would weigh on supply chain stocks.
Morgan Stanley sees regulation delaying AI data center spending rather than killing it, and recommends picking large players with power and capital.

Insight Times Editorial Desk

More by topic

Starting with this story's topics, then what readers are following now

Investor psychology, decision-making and books on investing

CapEx

This story's topic71 storiesSee all →

Money a company spends on things it will use for years — plants, equipment, bu…

Tech Selloff: Is the Bull Market Over? Wall Street's Four-Year Chart Says Not Yet

The uptrend has survived five double-digit pullbacks in four years. This time, AI companies' debt and cash flow are the new variables.

Investing

White House Swaps AI for "SI." The Real Investment Signal Is Infrastructure, Not the Name

An executive order renames artificial intelligence "Super Intelligence" and six frontier labs sign a voluntary safety accord. What investors should watch is the data center, power, nuclear and chip policy now gathered under one national-competitiveness frame.

Tech

Dark navy cover. Flow diagram: Nvidia to Nscale (credit guarantee, in talks) to Anthropic (~$45B compute lease signed in August plus a separate ~$45B, ~460MW deal in talks). Title: Why Would Nvidia Back a Customer's Loan?

Why Would Nvidia Guarantee Its Customers' Data Center Loans?

Nvidia is reportedly in talks to back financing for Nscale's West Virginia data center. Nothing is signed, and the terms are not public.

Tech

Amazon Goes After Nvidia's Strongest Moat, and It Is Software, Not GPUs

Amazon's open-source TorchNeuron aims to let existing AI code run on its Trainium chips. The contest comes down to the total cost of AI compute and the return on Amazon's $220 billion in annual capital spending.

Tech

Why the White House Swapped "AI" for "Super Intelligence"

Executive Order 14434 renames AI across the federal government without defining a new technology. The real story is about roughly $700 billion in capex, free cash flow and electricity.

Tech

AI's Next Bottleneck Is Power, Not GPUs

Jensen Huang and Elon Musk sent the same warning in Washington. The AI race is widening from building better models to supplying power and chips at industrial scale.

Tech

Memory Chip Supercycle: Not at Peak Yet, But 5 of 6 Warning Signs Are Flashing

Memory prices are still climbing, but the pace is slowing and PC and smartphone demand is starting to crack under the weight of higher chip costs. This looks less like a price peak and more like a profit plateau building toward the next supply cycle.

Investing

A building that houses servers and networking gear and supplies them with powe…

Why Nebius Is Keeping Some 2027 AI Capacity Off the Shelf

Nebius mixes long-term contracts, short-term premiums and auctions. Holding back capacity pays only if the servers stay busy, and it is not yet clear the pricing will last.

Tech

NVDA

101 storiesSee all →

Who Hits $1 Trillion Next: Reading AMD and ASML by Different Rules

One company is chasing Nvidia's GPU empire. The other sells the machine that draws the chips for every advanced chipmaker, including Nvidia. Both get floated as 2028 trillion-dollar candidates, but the numbers investors need to watch are entirely different.

Investing

Box Shows AI Its Manuals, and the AI Starts Redesigning the Workflow

Box Architect, built with NVIDIA, moves AI from reading documents to designing how work gets done. It has not yet reached beta, so the real-world payoff remains to be seen.

Tech

Nvidia and Palantir Put AI Inside the Supply Chain - The Real Shift Is That It's Learning Judgment

A single Vera Rubin rack holds 1.3 million parts. The real breakthrough in Nvidia and Palantir's new supply chain AI isn't automating that complexity, it's capturing the judgment calls that used to live only in planners' heads.

Tech

If You Only Watch Tesla, You May Be Missing Half the AI Story

Nvidia builds the AI brain. Who supplies the electricity to run it? Six power companies Tesla investors should know, and why Tesla's storage business is just one link in the chain.

Investing

Nvidia Moves From GPUs to CPUs as AI Agents Reshape a $200 Billion Chip Market

A customer's early test found Nvidia's Vera CPU handled about twice the AI agent workload of its existing chips. The bigger story is where the money in semiconductors is heading as AI shifts from answering questions to doing work.

Tech

Nvidia's $500 Billion Experiment: Wall Street Asks Who Backs the GPU, Not Just the Chip

The real issue in Nvidia's financing platform is not chip sales. It is how far Wall Street will trust GPU resale prices and rental income three to four years out.

Tech

China's AI Labs Went Open Source Because They Had to, Not Because They Wanted To

A GPU shortfall pushed Chinese AI labs toward efficiency and open weights, and that constraint is now reshaping the global competition over who controls AI's basic infrastructure.

Tech

GPU

73 storiesSee all →

A chip built to draw graphics, whose ability to run thousands of identical cal…

The Next AI Chip Race Is About How Chips Connect

Applied Materials will lift its stake in Dutch equipment maker BESI from about 9% to 20%. The bet is less about control than about a new way to link chips.

Tech

US AI Data Centers Are Headed for 90GW. Gigawatts Now Matter More Than Dollars

Goldman Sachs expects US data center capacity to grow from 64GW at the end of 2026 to 90GW at the end of 2027. The AI race is moving from buying GPUs to building power plants and transmission lines alongside them.

Tech

The Cheaper Tokens Get, the Bigger Your AI Bill Gets

As AI shifts from training to inference, the token is becoming an industrial commodity. Falling prices are not cutting demand, because agents and automation consume tokens at a scale no human chat session ever did.

Investing

Two Companies Drive 34% of S&P 500 Earnings Growth as AI Pays Sellers First

Nvidia and Micron generate a third of expected earnings growth in US large caps. Apollo says the financial effect of AI is not yet visible in the S&P 493.

Investing

AI Demand Is Proven. Now the $70 Billion Bill Comes Due

Oracle posted $19.3 billion in quarterly revenue and spent $28.5 billion on capex. The real question for the AI buildout is no longer whether demand exists, but how fast this spending turns into cash.

Tech

TSMC vs. ASML: Two AI Chip Chokepoints, but Which Is the Better Bet?

Both companies hold technical leads rivals struggle to close. A great business is not the same as a great five-year return.

Investing

Why AI Data Centers Are Starting to Make Their Own Electricity

The GPUs are secured, but the power is not. In the AI data center race, how fast a builder gets electricity is becoming a bigger variable than how cheaply it buys it.

Tech