To The Moon

Nvidia’s stock has risen around 14% this year, only slightly ahead of the S&P 500. For a company at the centre of the AI boom, that is a surprisingly modest performance.

But the investment debate is changing. Instead of chasing every company linked to AI spending, Wall Street is becoming more selective, paying closer attention to debt, valuations and whether today’s massive infrastructure investments can generate sustainable returns.

That shift raises an interesting question: Could Nvidia become a relatively defensive choice for investors who still want exposure to AI?

1. Nvidia sells the picks and shovels of the AI boom

One argument in Nvidia’s favour is that it does not need to predict which AI company will ultimately win.

Whether cloud providers compete to build better models or AI developers race to attract customers, many still need Nvidia’s chips to train and run their systems.

Its dominant market position, strong demand visibility and product roadmap give it advantages that many smaller AI companies lack.

The transition from Blackwell to the upcoming Rubin platform could provide another growth driver. If customers continue investing heavily in computing infrastructure, Nvidia could benefit from successive generations of increasingly powerful chips.

There is also a valuation argument. The article puts Nvidia at approximately 15 times estimated earnings over the next 12 months, making it one of the cheaper stocks in the Philadelphia Semiconductor Index by that measure.

For a company expected to deliver substantial growth, that valuation could attract investors looking for AI exposure without paying the highest multiples in the sector.

2. Nvidia’s biggest customers could also become its biggest weakness

The risk is that not all AI spending is equally sustainable.

Neocloud companies such as CoreWeave borrow heavily to purchase Nvidia’s GPUs and rent computing capacity to customers. Their business models depend on continued demand, strong utilisation and the ability to service their debt.

If AI demand grows more slowly than expected, these companies could face pressure from both sides: lower revenue growth and substantial financing obligations.

That could eventually affect Nvidia itself.

If customers cut back on orders, Nvidia could experience slower sales growth. If it offers more attractive financing arrangements to support customers, it could also face greater financial exposure and pressure on margins.

In other words, Nvidia’s customers are a source of growth, but their financial health is becoming an increasingly important part of the investment case.

3. The infrastructure boom has physical limits

AI expansion requires more than chips. It also needs electricity, data centres, cooling systems and the capacity to connect everything to the grid.

If power shortages and labour constraints slow construction, planned computing capacity may not translate into actual demand for Nvidia’s products as quickly as expected.

There is also a longer-term concern: what happens when the industry eventually builds enough computing capacity?

Nvidia’s growth expectations depend partly on continued investment by cloud providers and AI companies. If that spending eventually slows, the market may reassess the earnings growth currently built into its share price.

These are not necessarily immediate problems, but they matter when investors are valuing a business on years of expected expansion.

4. Is Nvidia really a defensive stock?

Calling Nvidia defensive requires some qualification.

Compared with smaller AI companies that depend on external financing or a single product, Nvidia has significant competitive advantages and a more established business.

However, it remains a semiconductor stock with exposure to capital spending cycles, competition, customer concentration and changing market expectations.

It is not a substitute for Treasury bonds or traditional defensive investments.

Its potential defensive appeal is relative: investors may prefer owning the company supplying essential AI infrastructure rather than taking larger risks on individual AI applications or heavily indebted infrastructure operators.

My takeaway

Nvidia may offer a more established way to participate in AI growth, but that does not make it immune to the risks building across the sector.

The key distinction is between AI demand continuing to grow and AI infrastructure spending remaining profitable. The first may be true even if some companies struggle to earn adequate returns on the billions they invest.

For Nvidia, the next stage of growth will depend not only on producing better chips, but also on whether its customers can turn their enormous investments into sustainable businesses.

For discussion purposes only.