NVIDIA Reports Tonight: Morgan Stanley Says Business Remains Strong, Market Truly Nervous About Interim CEO’s Debut and Guidance Outlook

LongbridgeAII'm LongbridgeAI. Here is the key information I extracted
- Morgan Stanley expects NVIDIA's F3Q27 revenue to reach 108.1 billion dollars with an EPS of 2.47 dollars, maintaining an overweight rating and a target price of 267 dollars.
- The core market focus centers on the interim CEO's debut, future guidance, and potential changes in financing and supply chain conditions.…
Disclaimer

Zhitong Finance APP has learned that NVIDIA (NVDA.US) will release its fiscal third-quarter results for FY2027 in the morning Beijing time on Wednesday. In its earnings preview, Morgan Stanley offered a straightforward assessment: the focus of this earnings report is not on this quarter's numbers—which are already fully priced in—but on the first management appearance following Jensen Huang's stepping down, and on next-year guidance that “might not even be provided.” Morgan Stanley maintains an "Overweight" rating with a target price of $267.

This Quarter's Numbers: The Unsurprising Part

Morgan Stanley expects F3Q27 revenue of $108.1 billion, slightly below the market consensus of $108.9 billion; earnings per share (EPS) are projected at $2.47, in line with consensus. Analysts wrote in the report that channel checks indicate "business remains strong," and even though some projects in Southeast Asia have slowed, "they still cannot meet demand."

The real items to listen for on the conference call are three things from acting CEO Colette Kress: her stance on support and investment in the AI ecosystem, whether NVIDIA's innovation engine can continue, and the risk assessment regarding further tightening of export controls. Jensen Huang typically did not make opening remarks but was very active during the Q&A session—Morgan Stanley hints that this kind of "general enthusiasm about the transformative impact of AI" may marginally decrease. The report also directs readers to its November 13 special report, "NVIDIA: Which Way Forward?", arguing that the management transition itself is more worthy of elaboration than the profit or loss of a single quarter.

Divergence Beyond This Quarter: Morgan Stanley ~$70 Billion Below Consensus

Extending the timeline, the divergence becomes immediately apparent. Morgan Stanley forecasts FY2028 revenue of $623 billion and EPS of $13.37, while the market consensus stands at $694.8 billion and $15.75—a difference of approximately $100 billion, which is the result of "last week's downward revision": Morgan Stanley estimates next year's revenue growth at 52%, whereas company guidance is around 70% and customer demand increased by 100%. However, analysts believe this figure has been "fully de-risked," with surveys still pointing to sustained optimism; the real suspense lies in the fact that, with an acting CEO in place and many questions remaining, they are "uncertain whether the company will provide guidance for next year."

Focus Point 1: Demand—The Gap Between 100% and 70%

NVIDIA previously guided for FY2028 revenue growth of approximately 70%, based on customer demand growth nearing 100%. However, concerns over land, power, and server room (LPS) constraints have significantly intensified after the midterm elections: SoftBank, a partner in NVIDIA's PORTS-Pike project, has already written down $3.7 billion related to power equipment. Morgan Stanley's judgment is that the midterm elections pose a greater threat to 2028 and 2029 than to 2027—data centers under construction will not be affected by new bans, giving the ecosystem time to bypass bottlenecks; although the hyperscale cloud providers it contacted acknowledge challenges, they are "confident these challenges can be overcome."

Financing is the other side of the same coin. Jensen Huang has historically supported regional new clouds and used NVIDIA's balance sheet to help customers secure land and power resources. Whether these commitments continue after the management change will become a focal point. Morgan Stanley cites an example: three new cloud providers, CoreWeave, Nebius, and IREN, guided for combined capital expenditures of approximately $100 billion in 2027, but their combined market capitalization evaporated by $38 billion in the past month—if the statement "demand increases by 100%" partially relies on financing, then whether financing will continue to arrive, especially for long-tail small customers in the ACIE sector, needs continuous monitoring.

Morgan Stanley's bottom line is: demand from hyperscale cloud providers is sufficient to support its forecast, with shipments next year equivalent to "high teens GW," and SpaceX alone plans to deploy 10GW. The report specifically highlights a more fundamental variable: investors' willingness to invest capital in AI is a "construction trajectory risk greater than any factor NVIDIA can control"—in other words, what truly matters tonight is whether changes in the financing environment are reflected in management's wording.

Focus Point 2: Supply and Guidance—Reiteration Is Positive

Morgan Stanley does not expect changes to the full-year growth guidance, but out of caution, it has independently lowered its revenue forecast by approximately 10% due to China export control risks. On the supply side, especially memory, conditions remain "very tight," with no signs of easing—"if anything, we see tightening instead."

An intriguing detail: some investors questioned why NVIDIA has not reiterated its guidance so long after the CEO change. Morgan Stanley's interpretation is actually a positive signal—by convention, earnings reports are often accompanied by upward revisions, so pre-announcing might be "better left for the earnings day itself."

Focus Point 3: Rubin Ramp-Up—First Quarter of Mass Shipments

This is the first quarter of mass shipments for Rubin. A quarter ago, the expectation was that Rubin would account for 20% of data center revenue (approximately $20 billion). The ramp-up has not been smooth: key testing partner King Yuan Electronics has lowered its 2026 guidance for the second consecutive quarter, partly attributed to the Rubin ramp-up being slower than expected. However, Morgan Stanley does not believe this will impact NVIDIA's revenue recognition: first, shorter testing times can explain part of the gap; second, its estimate of flat quarter-over-quarter performance for the January quarter "should still be conservative," with larger volume increases for Rubin appearing in the January and April quarters.

Latest Testimony on the Demand Side

In the appendix of the report, Morgan Stanley summarized statements regarding AI capital expenditures from cloud providers' conference calls during the Q3 earnings season, which can serve as background for tonight's call: Amazon stated that "an AI accelerator is installed approximately every 14 seconds," and the payback period is shortening—its capital expenditures for 2026 are approximately $220 billion, with significantly more in 2027, and it has a "clear view of future returns"; Google stated that AI investment "brought the highest query growth rate in five years" and believes every dollar of capital expenditure "can support more revenue over time"; Oracle stated that "demand for AI is real and huge; building capacity and financing are just tasks; tasks may be difficult, but they can all be solved"; Meta raised the lower bound of its capital expenditure range from $35 billion to $37 billion; Microsoft stated that "I would be surprised if less than 13GW of computing power comes online by the end of next year"; SpaceX stated that the "probability is non-zero" for launching over 20GW of self-developed computing power next year, and plans to deploy 440,000 Vera Rubin units by the end of the second quarter to form the strongest training cluster; on the new cloud side, Nebius stated that "customers are committing to us with longer terms, higher prices, and more prepayments," while CoreWeave indicated that existing capacity "can be sold out to buyers waiting in line."

Valuation: Three Scenarios and Option-Implied Probabilities

Morgan Stanley's three scenarios: Bull case $336 (approximately 21x FY2027 EPS of $16), triggered by conditions including supply alleviation leading to data center revenue exceeding expectations, sharing agreements bringing billions in pure profit revenue, and GPU-based AI PCs opening up the consumer market; Base case $267 (approximately 20x $13.37, representing a discount to the entire semiconductor sector and a slight premium to the broader market—Morgan Stanley explains that high market share and high gross margins instead make valuation multiples "lack leverage for near-term expansion"); Bear case $130 (approximately 10x), corresponding to both major debates turning negative: supply catching up with demand faster than expected, or competitors entering the market and customers developing their own chips to divert share. Based on option-implied volatility calculations, the probabilities of breaking above the bull price, staying above the base price, and falling below the bear price within one year are approximately 0.9%, 11.2%, and 24.2%, respectively—the options market prices downside scenarios significantly heavier than upside.

Sell-side consensus distribution: 77% Overweight, 22% Neutral, 2% Underweight. In the risk-reward theme scoring provided by Morgan Stanley, the new data era, pricing power, and long-term growth are all positive.

Source: Zhitong The copyright of this article belongs to the original author/organization.

The views expressed herein are solely those of the author and do not reflect the stance of the platform. The content is intended for investment reference purposes only and shall not be considered as investment advice. Please contact us if you have any questions or suggestions regarding the content services provided by the platform.