Nvidia reassures on AI trend, but fears over tech sector overspending may not go away

Nvidia’s results on 19 November had the potential to be a make-or-break moment for global financial markets. Any disappointment could have fuelled concerns around an AI bubble poised to burst. Fortunately, Nvidia has brought the party back to life, with suggestions that everything is going fine with the business.

Demand for its products remains strong, and chief executive Jensen Huang continues to talk up AI positively.

The company has now beaten revenue estimates for 13 consecutive quarters and two in a row for earnings. Many of its products are sold out for the foreseeable future and it continues to work on new iterations of core lines and develop new ideas.

What are some potential areas of concern?

This situation is fine for now, but what happens in three months’ time when the market waits with bated breath for Nvidia’s next quarterly earnings update? Even though Nvidia’s profits and cash flow remain as healthy, there are some red flags to consider.

Inventory and receivables are growing faster than sales on a year-on-year basis, which will give bears something to chew on. This relates to revenues possibly booked by Nvidia but where payment has not yet been received.

Investors with long memories will remember how Cisco and some internet equipment giants provided so-called vendor financing in the late 1990s and how that helped to boost demand near-term but exacerbate the slowdown that followed as an investment boom turned to bust. There could be trouble ahead if customers start to slow their purchases for any unexpected reason, as unlikely as the share price seems to think that may be.

While countless companies around the world would love to deploy AI in their business, not everyone will have the capacity to build the necessary infrastructure to support it. Buried in Nvidia’s results is a section about some of the challenges it is facing which could dampen the trajectory of future earnings growth.

Even though the likes of Alphabet, Amazon, Meta and Microsoft appear to have no problem paying billions of dollars for chips, Nvidia flags that ‘less capitalised’ companies could face difficulties securing financing for large-scale infrastructure projects. Accessing adequate power to run big data centres is also a challenge for certain companies, and any setbacks could have a knock-on effect with demand for Nvidia’s chips.

Part of the reason why financial markets have been wobbly in recent weeks is down to investors paying more attention to the fundamentals of a business linked to AI, rather than focusing purely on the technology’s capabilities.

The mega cap tech firms have been able to fund AI-related spending out of free cash flow, but we’ve recently seen them start to issue corporate bonds to raise money for their AI spending spree. That’s made investors think more about whether the billions of dollars of capital expenditure among these firms is being spent wisely, and whether they might end up with more capacity than is needed.

Investors are rightfully thinking more about earnings and cash flow, which means companies large or small can no longer keep everyone happy simply by saying they’re investing in AI. Day by day, more people are asking if companies are going to make a decent financial return on the money they’re spending to support and facilitate AI. If returns disappoint in any way, investors are going to be furious.

There continues to be widespread distrust in AI’s effectiveness as a search tool, with the likes of ChatGPT and Copilot either failing to give people the information they want, or the results are not quite accurate. A human having to double check AI-generated information defeats the point of deploying technology that’s meant to increase efficiencies.

However, it would be wrong to say AI is ineffective. Nvidia gives plenty of examples where the technology has been used to help improve the processing of information and transform analytical capabilities or productivity. For example, it says Salesforce’s engineering team has seen at least 30% productivity increase after adopting Cursor’s AI coding tool.

Investors initially gave a positive reaction to Nvidia’s results, helping to push up the broader US market when Wall Street resumed trading. However, the fanfare didn’t last long, and we saw US tech stocks slide as the day progressed. The Nasdaq Composite index closed 2.2% lower on 20 November, with Nvidia down 3.2% on the day.

Once 2026 gets underway, expect investors to ask harder questions and to dive deeper into Nvidia’s figures as the number of AI sceptics grows.