By Maksym Misichenko · Yahoo Finance ·
What AI agents think about this news
The panelists generally agree that the current valuations of Oracle and Nvidia are overpriced, with potential risks including decelerating hyperscaler capex, circular financing, and debt service concerns. However, they differ on the impact of 'Sovereign AI' demand.
Risk: Decelerating hyperscaler capex and potential multiple compression
Opportunity: Potential 'Sovereign AI' demand for Nvidia and Oracle
This analysis is generated by the StockScreener pipeline — four leading LLMs (Claude, GPT, Gemini, Grok) receive identical prompts with built-in anti-hallucination guards. Read methodology →
Many artificial intelligence stocks have suffered the wrath of unhappy investors over the past several months, and Oracle (NYSE: ORCL) and Nvidia (NASDAQ: NVDA) are no exception.
Investors have become increasingly skeptical that all the spending on AI infrastructure will pay off, and that's put pressure on companies' share prices, with Oracle tumbling 28% and Nvidia sliding 9% over the past three months.
Missed Nvidia in 2009? This Rare Signal Is Flashing Again. In 2009, a "Double Down" signal flashed for a little-known chipmaker called Nvidia. For the first time in years, that same "Total Conviction" signal is flashing for a company 1/100th the size of Nvidia. Continue »
With their share prices declining and AI spending pessimism on the rise, is it time for investors to ditch Oracle and Nvidia? Not so fast.
The case for holding onto Nvidia stock
Some shareholders have grown skeptical of Nvidia's AI dealings lately. For example, Nvidia is in talks with OpenAI to provide the ChatGPT maker with $250 billion in funding to help it build data centers. Separate talks are happening about how Nvidia will supply the GPUs for those data centers.
This type of circular deal has some shareholders questioning whether Nvidia will be able to grow its sales organically or will have to rely on lending out money to other companies to build AI infrastructure that spurs sales of its GPUs.
What's more, Nvidia recently announced a $500 billion infrastructure deal with SK Group, the parent organization of the memory processor company SK Hynix. Investors pushed down Nvidia shares on that news, too, as they worried that the company is committing too much money to other tech companies.
But some of their concerns may be a bit overdone. Consider that tech companies will invest $750 billion in AI infrastructure projects this year, most of which don't involve any Nvidia investments. That spending is fueling sales of Nvidia's GPUs, which currently have 86% of the GPU data center market.
That dominance could fuel Nvidia's sales and earnings for years. Revenue rose by 85% in the first quarter of fiscal 2027 to $81.6 billion, and diluted earnings jumped 140% to $1.87 per share.
And more AI opportunities may be around the corner. Nvidia recently debuted its next-generation Vera CPU for AI, which will directly compete with Intel and Advanced Micro Devices CPUs. AI agents are increasingly relying on CPUs to power their responses, and Nvidia sees this as a potentially new growth area. Nvidia said in a press release that its Vera CPU processor offers 50% better performance for AI agents than the x86 CPU architecture used by Intel and AMD.
For current shareholders, all of the above are good reasons to hold onto Nvidia stock. The AI boom is still well underway, and Nvidia's GPU technology is a clear winner. And if it gets even a sliver of the CPU market, the company could open up an entirely new revenue stream.
Why you shouldn't cut Oracle loose just yet
Investors have punished Oracle stock lately because they're unhappy with how much the company is spending on AI data centers. And, in some regard, they're not wrong to take notice.
Oracle's capital expenditures (capex) could rise to as much as $70 billion in fiscal 2027, according to management, with $40 billion of that spending funded by debt and equity financing, $20 billion of which was previously announced and came from a share sale.
What's got investors really upset is that Oracle's already raised $43 billion in debt in fiscal 2026, and the new spending will temporarily reduce Oracle's margins, management said.
So, that's the not-so-great news. But there are some very good reasons why dumping Oracle shares right now might not be a wise move.
Most importantly, Oracle is swimming in lucrative contracts. The company had $638 billion in remaining performance obligations (RPO) -- binding contracts that represent a revenue backlog -- in fiscal 2026, a 363% increase from the previous year. About $77 billion in those RPOs will be realized as revenue in fiscal 2027, and an additional 34% will be realized over the next one to three years.
What's more, Oracle's temporarily declining margins aren't expected to be a long-term drag on the company's earnings. Management is guiding to non-GAAP earnings per share of $8.05 in fiscal 2027, representing an 18% increase over 2026.
Understandably, Oracle shareholders are seeing the spending increases, rising debt, and temporarily shrinking margins, and are panicking. But Oracle is investing in new infrastructure that will help it fulfill its sizable contracts. It's short-term pain for what likely will be long-term gain for the company and its shareholders.
Should you buy stock in Nvidia right now?
Before you buy stock in Nvidia, consider this:
The Motley Fool Stock Advisor analyst team just identified what they believe are the 10 best stocks for investors to buy now… and Nvidia wasn't one of them. The 10 stocks that made the cut could produce monster returns in the coming years.
Consider when Netflix made this list on December 17, 2004... if you invested $1,000 at the time of our recommendation, you'd have $386,727!* Or when Nvidia made this list on April 15, 2005... if you invested $1,000 at the time of our recommendation, you'd have $1,232,139!*
Now, it's worth noting Stock Advisor's total average return is 906% — a market-crushing outperformance compared to 208% for the S&P 500. Don't miss the latest top 10 list, available with Stock Advisor, and join an investing community built by individual investors for individual investors.
Chris Neiger has no position in any of the stocks mentioned. The Motley Fool has positions in and recommends Advanced Micro Devices, Intel, Nvidia, and Oracle. The Motley Fool has a disclosure policy.
Four leading AI models discuss this article
"The article understates balance-sheet strain and circular funding risks that could delay or derail the promised earnings acceleration for both names."
The article frames the 28% ORCL and 9% NVDA drops as overdone panic, citing Nvidia's 86% GPU data-center share, $81.6B Q1 FY27 revenue (+85%), $638B Oracle RPO backlog (+363%), and management's 18% EPS growth guidance. Yet it glosses over Nvidia's $750B circular funding deals (OpenAI, SK Hynix) that blur organic demand, Oracle's capex exploding to $70B with $43B already raised in debt, and temporary margin compression that could persist if utilization lags. Missing context: hyperscaler capex growth is decelerating into 2026 while GPU supply ramps; any shortfall in realized RPO or AI ROI could trigger further de-rating from current 11.6x forward P/E on NVDA and Oracle's elevated multiple.
If AI infrastructure ROI disappoints and capex growth slows faster than expected, both companies could face multi-quarter revenue misses and margin resets far worse than the article admits, sending shares another 20-30% lower before the 'long-term gain' materializes.
"The market is ignoring the risk of a significant 'air pocket' in AI infrastructure spending as hyperscalers move from initial build-out to efficiency-focused optimization."
The article conflates 'circular financing' concerns with genuine infrastructure demand. While Nvidia's 86% market share is formidable, the market is rightly fixated on the sustainability of hyperscaler capex. Oracle’s $638 billion in RPO is impressive, but it reflects a shift toward lower-margin cloud consumption rather than high-margin software licensing. I am bearish on the current valuation of both because the market is pricing in perpetual 80%+ growth, ignoring the inevitable 'air pocket' as data center build-outs hit a saturation point. Investors are underestimating the risk that these massive backlogs will be cannibalized by competitive pricing pressure from AWS and Azure as they optimize their own AI stacks.
If AI agents achieve widespread enterprise adoption, the demand for inference capacity could render current capex projections conservative, turning today's 'overspending' into tomorrow's essential utility moat.
"Both stocks are pricing in AI capex growth that must remain robust for years; any slowdown in enterprise IT spending makes Nvidia's financing deals look like demand manipulation and Oracle's debt load unsustainable."
The article conflates two very different stories. Nvidia's 86% GPU dominance and 140% EPS growth are real, but the $250B OpenAI funding talks and $500B SK Group deal are red flags disguised as bullish news—they signal Nvidia is *financing* demand rather than selling into organic demand. Oracle's $638B RPO backlog sounds impressive until you parse it: $77B realizes in FY2027, meaning 88% of that backlog is 1-3 years out. Meanwhile, Oracle is raising $40B in new debt while margins compress. The article frames this as 'short-term pain,' but if AI capex cycles slow or customers delay, Oracle's debt service becomes a real problem with no near-term revenue offset.
If AI infrastructure spending truly hits $750B industry-wide this year (as the article claims), and Nvidia captures even 20-30% of non-GPU GPU-adjacent spending via its new Vera CPU, the financing concerns evaporate—these become strategic moves, not desperation. Oracle's RPO could also prove conservative if cloud adoption accelerates faster than guided.
"The core risk to a neutral view is duration: if the AI capex boom slows or ends sooner than anticipated, Nvidia’s high valuation could compress much faster than earnings growth can offset."
Oracle and Nvidia face divergent near-term drivers. Nvidia rests on a durable data-center GPU moat and ongoing AI workloads, but the rally assumes a long-running AI capex cycle; if hyperscalers throttle spending or demand decelerates, Nvidia's growth and multiple could regress. Oracle shows strong backlog (RPO $638B) and a path to growth via AI-focused capex, yet margins are pressured in the near term by heavy investment. The article overlooks potential duration risk, regulatory/supply-chain headwinds, and the risk that AI spending peaks earlier than priced in, which could re-rate both names unfavorably.
Nvidia’s AI demand has proven sticky and could sustain earnings momentum, justifying premium multiples; the article underestimates the durability of the data-center cycle and Nvidia’s ecosystem moat. A neutral stance may miss a potential upside if AI capex persists longer than expected.
"Accelerating GPU supply against decelerating hyperscaler capex creates a utilization cliff nobody has quantified."
Claude correctly flags the back-loaded RPO and debt load, yet everyone underweights second-order supply risk: GPU supply is ramping while hyperscaler capex growth decelerates into 2026. That mismatch alone could force Nvidia utilization rates below 60% and Oracle cloud margins to compress another 400bp, amplifying any ROI disappointment into a sharper de-rating.
"Sovereign AI mandates create a geopolitical demand floor for Nvidia and Oracle that decouples from hyperscaler capex cycles."
Claude and Grok focus on debt and RPO duration, but you are all ignoring the 'Sovereign AI' wildcard. Governments are now mandating domestic compute clusters, which provides a non-hyperscaler floor for Nvidia's demand that isn't tied to commercial ROI or circular financing. This creates a geopolitical bid for hardware that persists even if Big Tech throttles capex. Oracle is the primary beneficiary here, as they are uniquely positioned to build these private, regulated clouds globally.
"Sovereign AI demand is real but too small to offset a material hyperscaler capex slowdown, and Oracle's execution risk in regulated markets is underweighted."
Gemini's 'Sovereign AI' thesis is real but overstated. Yes, governments are building domestic clusters—but these are typically lower-margin, lower-utilization workloads compared to hyperscaler ROI-driven capex. Oracle's 'unique positioning' assumes they can execute globally regulated clouds at scale; their track record on cloud infrastructure speed-to-market is mixed. This geopolitical floor exists, but it's maybe 15-20% of current hyperscaler capex run-rate, not a demand offset for a 30%+ deceleration.
"Sovereign AI demand is not a durable floor for Nvidia and won't shield the upside from hyperscaler deceleration or Oracle's debt-driven margin compression."
Gemini's 'sovereign AI' floor is an over-optimistic defense. Government-driven compute will be slower to scale, lower margin, and highly lumpy; even if it materializes, it's maybe 15-20% of run-rate, not a durable backstop. If hyperscaler capex slows or public-cloud budgets tighten, Nvidia's uplift from sovereign demand won't prevent multiple compression. Oracle could benefit some, but this doesn't fix the debt/margin headwinds.
The panelists generally agree that the current valuations of Oracle and Nvidia are overpriced, with potential risks including decelerating hyperscaler capex, circular financing, and debt service concerns. However, they differ on the impact of 'Sovereign AI' demand.
Potential 'Sovereign AI' demand for Nvidia and Oracle
Decelerating hyperscaler capex and potential multiple compression