JPMorgan Research Report Analysis: AI Enters Monetization Validation Phase, Capital Flows from Hardware to Cloud Giants
TechFlow Selected TechFlow Selected
JPMorgan Research Report Analysis: AI Enters Monetization Validation Phase, Capital Flows from Hardware to Cloud Giants
JPMorgan believes that hyperscalers with cloud platforms, data centers, and AI application ecosystems have a risk-reward ratio superior to AI hardware.
Navigating Web3 tides with focused insightsWritten by: Rita
TechFlow Guide
JPMorgan's strategy team released a report on July 20, 2026, noting significant divergence in the US stock AI sector recently, with increased volatility and fluctuating gains and losses for hardware stocks like NVIDIA and Broadcom. Microsoft, Google, Meta, and Amazon trends are stabilizing.
Market concerns whether the AI rally has peaked. JPMorgan's latest US stock strategy report gives a clear judgment: The AI super cycle is not over, but capital is changing hands. Investment logic is shifting from "competing on capital expenditure" to "competing on commercial monetization." The $870 billion annual AI capital expenditure continues to accelerate, but the market is no longer satisfied with "scale of investment" and is beginning to ask about "profit realization." JPMorgan believes that hyperscalers with cloud platforms, data centers, and AI application ecosystems have a better risk-reward ratio than AI hardware.
Capital Expenditure Still Accelerating
The core focus of this year's second quarter earnings is not profit, but AI capital expenditure.
The development of the AI industry chain highly depends on tech giants' continuous investment in data center construction. The market expects global AI-related capital expenditure to approach $870 billion by the end of 2026, a 77% year-on-year increase, with hyperscalers contributing about $750 billion.
JPMorgan internet analysts believe the market's forecast for 2027 remains conservative. Google's capital expenditure in 2027 is expected to grow by 54%, approaching $300 billion. Amazon by 42%, about $300 billion. Meta by 42%, about $200 billion.
AI infrastructure construction has not slowed down; it is still accelerating. The earnings reports of the four major hyperscalers in the coming weeks will be the core market barometer.
The Market Starts Asking: When Will It Make Money?
In the past two years, the market only focused on the scale of investment. Currently, investors are beginning to focus on the core question: When will large-scale investments yield returns?
JPMorgan believes this will be the core main line of AI investment in the coming years. Positive signals have begun to emerge. Meta is selling idle AI computing power externally, directly realizing monetization of data center resources. Anthropic's security research progress suggests that security concerns may accelerate government and enterprise workloads moving to the cloud, replacing local model deployment.
This means demand for cloud services such as Azure, Google Cloud, and AWS will continue to grow. The AI business model is being gradually validated, covering diverse paths such as model sales, computing power output, cloud services, and software licensing. JPMorgan believes that if management releases more AI commercialization signals in earnings reports, earnings forecasts and free cash flow are expected to be revised upward.
Money Is Flowing from Hardware to Cloud Giants
The biggest winners in the AI market over the past two years have been in the upstream sector. Sectors such as GPU, HBM, high-speed networking, switches, and optical modules have almost risen across the board. But JPMorgan believes sector rotation is occurring.
AI hardware belongs to typical momentum trading: positions are highly concentrated, and trading is extremely crowded. Once expectations change, volatility will be significantly amplified. The semiconductor sector recently corrected by about 15%; although crowding has eased, it is far from fully cleared. Market concerns focus on four points: sustainability of capital expenditure, capital diversion from AI company IPOs, infrastructure supply and demand dynamics, and whether return on invested capital can cover costs.
In comparison, Microsoft, Google, Meta, and Amazon have better risk-reward ratios. The aforementioned vendors control computing power, cloud platforms, and end users. Those truly realizing profits may not be the "shovel sellers," but the "gold mine operators."
Where Does the Money Come From?
Funding sources have become a new market concern: With annual investments of hundreds of billions of dollars, can tech giants' cash reserves support this?
JPMorgan specifically responded to this. The bond financing scale of the five major tech giants rose from $40-50 billion in 2022 to about $190 billion in 2026. Google completed $85 billion in equity financing this year, and Meta is also evaluating similar plans.
But JPMorgan believes this is merely a financing structure adjustment and does not represent operational deterioration. It is estimated that in 2027, tech giants' operating cash flow will still exceed $900 billion. Revenue is expected to grow at an annual average of about 17% in the coming years, with profit margins remaining high. The market is willing to provide financing to support AI construction; the core logic lies in optimism for long-term profit returns.
Regarding the inflection point of free cash flow, JPMorgan expects it to appear as early as 2027. If AI commercialization progress slows, significant improvement may need to wait until after 2028.
The Semiconductor Logic Has Not Changed
JPMorgan is not bearish on AI hardware.
Hyperscaler demand remains strong, AI lab computing power still has a gap, agent inference demand is growing rapidly, data center investment continues to increase, and new platforms such as Blackwell and Rubin further enhance demand visibility for the coming years. The focus of market discussion has shifted: previously the market focused on "the authenticity of AI demand," now shifting to "whether supply chain and power supply can match."
Besides GPUs, AI demand has spread to custom chips, HBM storage, high-speed networking, semiconductor equipment, EDA, and other fields. This year, the global semiconductor industry (excluding storage) revenue growth rate is still expected to exceed 30%. JPMorgan remains bullish on the AI industry chain, only believing that the investment rhythm has undergone rotation.
Second Quarter Earnings Remain Strong
JPMorgan is generally optimistic about the US stock second quarter earnings season. S&P 500 second quarter earnings are expected to grow 23% year-on-year, growing 19% excluding energy, with revenue growth of 12%. Almost all industry revenues have achieved growth; except for healthcare, most industry profits are moving up simultaneously.
But the growth structure is highly concentrated. Just NVIDIA and Micron contributed about 37% of S&P 500 earnings growth. Current US stock earnings growth is still mainly driven by AI.
The energy sector, driven by rising oil prices, is expected to see earnings growth of 122%, another important highlight for the season. Healthcare is the only industry expected to see declining profits, but JPMorgan remains bullish on its long-term allocation value, especially against the backdrop of AI-enabled drug development and improved medical efficiency.
TechFlow Perspective
This report juxtaposes two core variables of the AI investment chain for observation: capital expenditure slope and monetization lag. The $870 billion annual expenditure corresponds to the financing curve of the five major hyperscalers' debt issuance rising from $40 billion to $190 billion; the steepening of the curve itself is accumulating risk.
JPMorgan judges the free cash flow inflection point will appear as early as 2027, but admits a clearer recovery needs to wait until 2028 or even later. This time window hides a core bet: if monetization progress in 2027 falls short of expectations, the market's patience for a 2028 recovery may be exhausted prematurely.
The impact of changes in fair value of private investments on EPS is easily overlooked. About $4 of S&P 500 EPS in the first quarter came from non-cash revaluation, and the scale in the second quarter may further expand. A significant portion of earnings revisions is driven by accounting effects, and the actual extent of operational improvement needs prudent assessment.
Disclaimer
This article is a compilation and interpretation by TechFlow Research of a third-party brokerage research report (JPMorgan, July 20, 2026). The ratings, target prices, earnings forecasts, and related judgments cited in the text are the views of the brokerage's analysts, representing only their institution's position, not representing TechFlow Research's views, nor constituting any investment advice. The market has risks, decisions must be independent. This article should not be used as a basis for buying or selling any securities.
Join TechFlow official community to stay tuned
Telegram:https://t.me/TechFlowDaily
X (Twitter):https://x.com/TechFlowPost
X (Twitter) EN:https://x.com/BlockFlow_News
@chaoxiangooo
