By Maksym Misichenko · Nasdaq ·
What AI agents think about this news
The panelists generally agreed that while NVDA's current growth is attractive, its high valuation and dependence on hyperscaler capex and CUDA lock-in pose significant risks. MSFT, with its diversified software suite and lower beta, was seen as a more defensive play.
Risk: Compression of NVDA's margins due to capex slowdown and potential commoditization of CUDA's silicon advantage.
Opportunity: MSFT's multi-layered revenue streams and lower valuation floor.
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 →
Perhaps one of the more notable surprises is the fact that Microsoft (NASDAQ: MSFT) and Nvidia (NASDAQ: NVDA) have become perceived by some as value plays. Although each company plays a critical role in AI, Nvidia has struggled to outperform the S&P 500 (SNPINDEX: ^GSPC) this year, while Microsoft stock has pulled back.
Each company will almost certainly continue to play a crucial role in AI and tech at large, so investors should not expect massive stock price declines. Nonetheless, only one of these is likely to stand out as the better value stock in today's market.
Where to invest $1,000 right now? Our analyst team just revealed what they believe are the 10 best stocks to buy right now, when you join Stock Advisor. See the stocks »
One could argue that Nvidia is the most surprising value stock in existence today. Since hitting a low in the fall of 2022, the stock has increased by more than 1,700% as its AI accelerators have powered the generative AI boom. That took its market cap to $5.1 trillion, the largest among publicly traded stocks.
However, amid that growth, Nvidia trades at a P/E ratio of 32. That is barely above the S&P 500 average of 29, and its recent price action makes it the cheapest it's been since 2019. This is unusual since investors tend to pay a premium for growth. In the first quarter of fiscal 2027 (ended April 26), revenue increased by 85% yearly while net income rose by 211% over the same period.
Knowing that, it is surprising that Nvidia is so cheap. Perhaps growth investors are pulling back, as a $5.1 trillion market cap will make it difficult for Nvidia to be a 10-bagger. Others might feel leery about the massive capital expenditures (capex) spending of the hyperscalers and wonder how long it can last.
Whatever the reason, Nvidia is an inexpensive stock with considerable growth potential. Even if that growth slows significantly, it would likely not undermine the value proposition in Nvidia stock. Moreover, with more than $80 billion in liquidity and the ability to innovate at low cost (it spent just $6.5 billion on capex in the last 12 months), Nvidia is well positioned to outperform the market over time while keeping investor capital safe.
Despite the aforementioned pullback this year, Microsoft has long been a popular choice for capital preservation. Even though its current $2.9 trillion market cap is well below Nvidia's, Microsift's dominance in PC operating systems, strength in productivity software, and later success as a cloud company have made it one of the market's largest companies.
Moreover, its 23 P/E ratio places it well into value stock territory, and while it cannot match Nvidia's growth, its financial performance is solid. In the third quarter of fiscal 2026 (ended March 31), revenue rose by 17% annually, while net income surged 23% higher during the same period.
However, the company's challenges arguably make the low earnings multiple and falling stock price more understandable. The company's plan to spend $190 billion on capex this year has made some investors uneasy. Furthermore, AI's ability to perform many software functions has made investors leery of SaaS stocks. Also, a close relationship with OpenAI has made some investors skeptical about the strength of Microsoft's AI.
Despite these challenges, Microsoft's earnings multiple could make the stock a safe bet, given its prominent role in the tech industry. Even with heavy capex spending, Microsoft still maintains about $78 billion in liquidity. Additionally, given concerns about its capex spending, it is likely using some of those funds to invest in AI, separate from OpenAI.
Thus, investors should not count it out as an AI company. When considering its valuation and continued growth, Microsoft stock is probably a buy at current levels.
Of the two choices, Nvidia looks like the better value among the two tech giants right now.
Admittedly, 23 times earnings is an extremely low multiple for Microsoft, and the market may have gone too far in pricing the company's troubles into Microsoft stock.
Nonetheless, the margin of safety Nvidia provides right now is too obvious to ignore. Even if Nvidia's 85% revenue growth slows significantly, it will take considerable growth deceleration to make its 32 P/E ratio seem expensive.
Moreover, Nvidia reached this position by leading and dominating the AI accelerator market. Even with more companies entering this market, they are unlikely to unseat Nvidia anytime soon. That means that even if Nvidia's days as a potential 10-bagger are over, it is likely to outperform both Microsoft and the S&P 500 for the foreseeable future.
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 $377,990! Or when Nvidia made this list on April 15, 2005... if you invested $1,000 at the time of our recommendation, you’d have $1,269,518!
Now, it’s worth noting Stock Advisor’s total average return is 896% — a market-crushing outperformance compared to 206% 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.
**Stock Advisor returns as of July 26, 2026. *
Will Healy has no position in any of the stocks mentioned. The Motley Fool has positions in and recommends Microsoft and Nvidia. The Motley Fool has a disclosure policy.
The views and opinions expressed herein are the views and opinions of the author and do not necessarily reflect those of Nasdaq, Inc.
Four leading AI models discuss this article
"Nvidia's apparent valuation discount is illusory once capex-cycle risk and $5T scale are properly weighted against Microsoft's diversified cash-flow machine."
The article positions NVDA (32x P/E, 85% rev growth) as the superior value play over MSFT (23x P/E, 17% rev growth) based on AI accelerator dominance and margin of safety. However, it glosses over NVDA's $5.1T market cap requiring heroic growth to move the needle, extreme concentration risk in data-center capex cycles, and gross margins already compressing from competition (AMD, custom ASICs). MSFT's Azure + Copilot trajectory, vast software moat, and lower beta are materially underweighted. Forward P/E for NVDA FY27 is closer to 38x on consensus; MSFT's appears optically cheap but already discounts OpenAI dependency fears.
If hyperscaler capex does slow in 2027 as feared, NVDA's growth could decelerate to high-teens while its multiple contracts sharply, making the 32x today look expensive in hindsight; MSFT's diversified revenue would prove far more resilient.
"Nvidia’s valuation is predicated on a peak-cycle capex environment that is inherently unsustainable, making Microsoft the superior risk-adjusted value play."
The article's framing of NVDA as a 'value' stock at a 32x P/E is dangerous. While the PEG ratio looks attractive given current triple-digit growth, investors are ignoring the cyclicality of semiconductor capex. Nvidia is essentially a high-beta play on hyperscaler infrastructure spend; if those firms pivot from hardware acquisition to software monetization, NVDA's margins will compress rapidly. MSFT, conversely, is a defensive play on enterprise software stickiness. The article misses that MSFT's $190 billion capex is an investment in proprietary infrastructure, whereas NVDA's revenue is dependent on the continued willingness of others to burn cash. I prefer MSFT for its multi-layered revenue streams and lower valuation floor.
The bear case for MSFT is that its legacy moat is being eroded by AI-native startups, and its heavy investment in OpenAI may ultimately prove to be a margin-dilutive 'renting' of innovation rather than true proprietary ownership.
"NVDA trades at a premium to the S&P 500 despite slowing growth visibility, while MSFT's low multiple reflects temporary capex headwinds, not terminal competitive decline—making MSFT the better risk/reward at current valuations."
The article's conclusion—that NVDA at 32x P/E is better value than MSFT at 23x—rests on a fragile assumption: that NVDA's 85% revenue growth is sustainable enough to justify a *premium* multiple relative to the S&P 500 average (29x). But the article itself notes hyperscalers are spooked by capex intensity. If AI infrastructure spending moderates even modestly—say, to 40% YoY growth—NVDA's multiple compresses to ~20x, erasing the 'margin of safety' claim. MSFT's 23x on 17% growth looks less punished when you account for $190B capex as a *temporary* investment cycle, not structural weakness. The article conflates 'cheap relative to recent history' with 'cheap relative to fundamentals.'
NVDA's dominance in AI accelerators and $80B liquidity genuinely do create a structural moat; even if growth halves, a 16x multiple on $200B+ revenue is defensible, and MSFT's capex-to-revenue ratio (7%+) is historically elevated and may not translate to proportional earnings accretion.
"NVDA is not a bargain at 32x earnings if AI demand moderates or regulatory/geopolitical risks bite; a diversified, cash-rich MSFT may offer a safer, more resilient risk-adjusted path."
The article pushes Nvidia as the obvious value pick at a 32x P/E, but that glosses over downside if AI demand cools or capex cycles peak. Nvidia’s moat rests on hyperscaler spend and GPU pricing; a slower data-center refresh, saturation of AI workloads, or regulatory/export controls (e.g., China) could curb growth and boost multiple compression. Competition from AMD/Intel and potential shifts in AI software licensing could also erode pricing power. At megacap scale, Nvidia is fragile to macro shocks. Microsoft, with a broader software/cloud suite and cash flows, may weather AI cycles better, offering more downside protection despite a lower multiple.
Nvidia’s AI-driven demand could stay structurally strong for longer, keeping margins high and supporting further multiple expansion; the bear case relies on a sharp downturn in AI spending which may not materialize soon.
"CUDA erosion could drive faster NVDA margin compression than hyperscaler capex slowdown alone."
Claude's 20x multiple compression scenario for NVDA assumes hyperscaler capex drops to 40% YoY, but ignores that even at that level NVDA's data-center revenue would still compound at ~35% CAGR through FY27 on current backlog visibility. The real unmentioned risk is CUDA ecosystem lock-in fading as PyTorch + custom silicon mature faster than modeled, potentially accelerating margin compression beyond what any panelist quantified.
"Software abstraction layers like Triton and PyTorch are the real threats to NVDA's pricing power, not just hardware competition."
Grok, your focus on CUDA lock-in is the missing link. While everyone debates capex cycles, they ignore that NVDA’s true value isn't just hardware, but the software moat that prevents hyperscalers from switching to cheaper custom ASICs. If PyTorch and Triton lower the barrier to entry, NVDA’s pricing power evaporates regardless of how much cash hyperscalers burn. We are underestimating the speed at which software abstraction layers commoditize NVDA’s underlying silicon advantage.
"Capex cycle deceleration is the near-term multiple risk; CUDA erosion is structural but slower-moving, and conflating them obscures which risk to hedge first."
Gemini and Grok are conflating two separate timelines. CUDA lock-in erosion is real, but it's a 3–5 year story; PyTorch maturity doesn't eliminate NVIDIA's near-term pricing power through FY26. The capex cycle risk (Claude's 40% deceleration) is 12–18 months away. Conflating them obscures which risk matters first. NVIDIA's margin compression likely comes from capex slowdown *before* CUDA commoditization. That changes the investment horizon and exit timing materially.
"Near-term NVDA pricing power will resist CUDA lock-in erosion; 12–18 month capex cycles matter more than 3–5 year commoditization."
Grok, I disagree that CUDA lock-in erodes rapidly; PyTorch/Triton maturity may matter, but near-term pricing power largely persists through software monetization (TensorRT, CUDA libraries) and a strong enterprise backlog. The 12–18 month horizon suggests capex cycles, not 3–5 year commoditization, drives margins. If you’re right, the bear case shifts to OpenAI licensing and AI software pricing, but the immediate risk isn't CUDA erosion.
The panelists generally agreed that while NVDA's current growth is attractive, its high valuation and dependence on hyperscaler capex and CUDA lock-in pose significant risks. MSFT, with its diversified software suite and lower beta, was seen as a more defensive play.
MSFT's multi-layered revenue streams and lower valuation floor.
Compression of NVDA's margins due to capex slowdown and potential commoditization of CUDA's silicon advantage.