NVDA Trades at a 16x Forward Multiple as Nvidia Stock Valuation Hits a Multi-Year Low
Nvidia stock valuation has compressed to levels not seen in years, even as the company’s underlying business continues to expand at a pace that most large-cap technology firms could not contemplate. NVDA has fallen 13.5% since May and now sits at the same price it traded in October 2025, trailing the S&P 500 year to date. For a stock that spent much of the past two years as the emblem of the AI trade, that is a jarring reset.
What Is Pulling Capital Away From NVDA
The immediate cause is sector rotation. Investors who accumulated large gains in NVDA and other AI infrastructure names have been redeploying into memory chips, where the supply-demand dynamic is currently far more acute. The scale of that shift is visible in the numbers: Samsung Electronics reported a Q2 2026 preliminary operating profit of 89.4 trillion South Korean won, nearly 20 times the 4.7 trillion won it earned in the same quarter a year earlier, and more than it earned across all of 2024 and 2025 combined.
That is the kind of earnings acceleration that draws capital quickly. SK Hynix, now described as the world’s most valuable memory chipmaker, saw its shares rise 5.6% on the day Samsung released those numbers. When profits in adjacent segments are growing at that rate, even a genuinely strong performer like Nvidia can look like it is standing still.
A second concern is customer competition. Amazon Web Services has been developing its own AI chips, and the relationship with Nvidia is more nuanced than simple substitution. AWS’s Trainium3-based servers each contain 144 chips and deliver more than four times the performance of the prior generation while consuming 40% less power. Its forthcoming Trainium4 will incorporate Nvidia’s NVLink Fusion interconnect technology. The relationship, then, is competitive in some respects and collaborative in others.
Nvidia has itself moved to pre-empt full disintermediation. At the COMPUTEX trade show in Taiwan, the company announced NVLink Fusion, which opens its server platform to third-party CPUs and AI chips. Partners at launch include MediaTek, Marvell, Intel, and Qualcomm. The move repositions Nvidia from a closed ecosystem to something closer to a platform, which broadens its addressable market even if it dilutes some exclusivity.
Nvidia Stock Valuation and the Forward Multiple
Against this backdrop, the valuation case is harder to dismiss than the recent price action implies. Based on consensus forecasts for FY28, which begins in February 2027, NVDA is trading at 16 times forward earnings. That is below the S&P 500’s own multiple and, by the snippet author’s assessment, the cheapest the stock has been in several years. The five-year price/earnings-to-growth ratio stands at 0.6, according to the original analysis; a reading below 1.0 is conventionally taken to indicate undervaluation relative to expected earnings growth.
The underlying business has been growing fast enough to justify that scepticism about derating. According to Nvidia’s Q4 fiscal 2025 earnings filing with the SEC, full-year fiscal 2025 revenue reached $130.5 billion, up 114% from $60.9 billion in fiscal 2024. Net income for the year was $72.9 billion, up 145% year on year. Q4 alone produced $39.3 billion in revenue, with Data Center contributing $35.6 billion of that, up 93% from the same quarter a year earlier.
The gross margin picture warrants attention. The snippet references a 75% gross margin, and that figure does reflect full-year fiscal 2025 GAAP gross margin, which was 75.0%. The Q4 figure, however, came in at 73.0%, down from 74.6% in Q3 FY25. Nvidia’s own Q1 FY26 outlook projects GAAP gross margin of 70.6%, plus or minus 50 basis points. The direction of travel is worth monitoring as memory costs and competitive pricing weigh on near-term margins.
Wall Street’s consensus price target sits at $313, representing approximately 53% upside from current levels. If NVDA were to reach that target by next summer, £5,000 invested today would become roughly £7,500. The stock may not get there on that timetable, but the directional logic is consistent with the multiple and growth profile.
Physical AI and the Adjacent Market
Beyond data centres, Nvidia has a longer-dated optionality argument. CEO Jensen Huang has estimated annual AI infrastructure spending could reach between $3 trillion and $4 trillion by 2030. Physical AI, covering autonomous vehicles and humanoid robotics, represents a distinct expansion of that opportunity. Automotive revenue for Q4 FY25 was $570 million, up 103% year on year, still a modest fraction of total revenue but growing off a low base.
Nvidia’s infrastructure credentials are already embedded at scale: more than 75% of the systems on the TOP500 list of the world’s most powerful supercomputers run on Nvidia technologies. The company is also a named technology partner for the $500 billion Stargate Project, which anchors its position in the next wave of US AI infrastructure investment.
The near-term test is whether Q1 FY26 revenue, guided at $43.0 billion, meets or beats expectations when results arrive. A miss at that level, combined with further gross margin compression, would reopen the valuation debate from the other direction.