Nvidia’s earnings growth challenges AI bubble comparisons

Key Points

1. Nvidia trades at about 17 times forward earnings while profits are expected to rise approximately 70% next year.
2. DBS investment chief Hou Wey Fook contrasts that valuation with Cisco’s roughly 100-times multiple before the dot-com crash.
3. AI infrastructure spending now reaches chips, cloud services, data centers, power generation, networking and cooling.

The latest

Nvidia’s current earnings outlook makes the chipmaker difficult to classify as a traditional speculative bubble, DBS Group Chief Investment Officer Hou Wey Fook said. The company, a leading supplier of graphics processing units for training and running advanced AI models, sits at the center of a global infrastructure buildout. Hou said semiconductor and AI investments continue to benefit from significant structural tailwinds, while arguing that portfolios should not be concentrated entirely in technology stocks.

Details

• Earnings support: The market is increasingly valuing Nvidia against revenue and profit already generated by AI infrastructure, rather than solely on expectations of future profitability. Rapid earnings growth can support a high share multiple; equally, even a lower multiple can become expensive if earnings forecasts deteriorate.
• Critical supplier: Nvidia’s GPUs provide much of the computing power needed to train and operate large AI models. That role has made the company a critical supplier to technology giants and cloud providers investing billions in new capacity, transforming its financial profile and placing its earnings at the center of the market debate.
• Dot-com benchmark: Hou cited Cisco Systems, which traded at roughly 100 times earnings before the dot-com crash, far above Nvidia’s current forward multiple. Many late-1990s technology companies had limited or no profits, unlike leading AI businesses now producing substantial revenue and earnings.
• Capital-spending reach: Hyperscalers and technology groups are directing billions of dollars into data centers, advanced semiconductors, networking equipment, electricity generation and cooling systems. The resulting investment cycle links chip demand with cloud capacity, energy supply and physical infrastructure across a widening economic ecosystem.
• Demand-side risks: The bullish case depends on customers continuing to order computing infrastructure and on AI demand translating into durable profits. Slower capital spending, reduced hardware orders or technological advances that weaken demand for current-generation chips could pressure earnings expectations and alter the valuation argument.
• Portfolio implications: Hou favors a barbell strategy combining growth-oriented technology shares with investment-grade fixed income, preserving exposure to AI’s upside while seeking to reduce overall volatility. He identifies hedge funds and gold as possible diversifiers in the portfolio’s middle, addressing risks created by capital concentration in a relatively small group of technology companies.

Between the lines

The central economic question is whether productivity gains from AI will justify the vast expenditure on computing capacity. Nvidia’s results strengthen the case today, but the investment cycle ultimately rests on sustained demand converting infrastructure spending into earnings.

What’s next

Investors will track Nvidia’s projected approximately 70% earnings growth next year alongside hyperscaler capital spending, customer orders for computing infrastructure and demand for current-generation GPUs. Those indicators will determine whether revenue growth continues to support the sector’s valuation and planned data-center investment.