Why Uncertainty Won't Slow AI Hardware Investment

Our Head of U.S. IT Hardware Erik Woodring gives his key takeaways from Morgan Stanley’s Technology, Media and Telecom (TMT) conference, including why there appears to be a long runway ahead for AI infrastructure spending, despite macro uncertainty. 

----- Transcript -----

Welcome to Thoughts on the Market. I’m Erik Woodring, Morgan Stanley’s Head of U.S. IT Hardware Research. 

Here are some reflections I recorded last week at Morgan Stanley’s Technology, Media, and Telecom Conference in San Francisco. 

It’s Monday, March 10th at 9am in New York. 

This was another year of record attendance at our TMT Conference. And what is clear from speaking to investors is that the demand for new, under-discovered or under-appreciated ideas is higher than ever. In a stock-pickers’ market – like the one we have now – investors are really digging into themes and single name ideas. 

Big picture – uncertainty was a key theme this week. Whether it’s tariffs and the changing geopolitical landscape, market volatility, or government spending, the level of relative uncertainty is elevated. That said, we are not hearing about a material change in demand for PCs, smartphones, and other technology hardware. 

On the enterprise side of my coverage, we are emerging from one of the most prolonged downcycles in the last 10-plus years, and what we heard from several enterprise hardware vendors and others is an expectation that most enterprise hardware markets – PCs , Servers, and Storage – return to growth this year given pent up refresh demand. This, despite the challenges of navigating the tariff situation, which is resulting in most companies raising prices to mitigate higher input costs. 

On the consumer side of the world, the demand environment for more discretionary products like speakers, cameras, PCs and other endpoint devices looks a bit more challenged. The recent downtick in consumer sentiment is contributing to this environment given the close correlation between sentiment and discretionary spending on consumer technology goods. 

Against this backdrop, the most dynamic topic of the conference remains GenerativeAI. What I’ve been hearing is a confidence that new GenAI solutions can increasingly meet the needs of market participants. They also continue to evolve rapidly and build momentum towards successful GenAI monetization. 

To this point, underlying infrastructure spending—on servers, storage and other data center componentry – to enable these emerging AI solutions remains robust. To put some numbers behind this, the 10 largest cloud customers are spending upwards of [$]350 billion this year in capex, which is up over 30 percent year-over-year. Keep in mind that this is coming off the strongest year of growth on record in 2024. Early indications for 2026 CapEx spending still point to growth, albeit a deceleration from 2025. 

And what’s even more compelling is that it’s still early days. My fireside chats this week highlighted that AI infrastructure spending from their largest and most sophisticated customers is only in the second inning, while AI investments from enterprises, down to small and mid-sized businesses, is only in the first inning, or maybe even earlier. So there appears to be a long runway ahead for AI infrastructure spending, despite the volatility we have seen in AI infrastructure stocks, which we see as an opportunity for investors. 

I’d just highlight that amidst the elevated market uncertainty, there is a prioritization on cost efficiencies and adopting GenAI to drive these efficiencies. Company executives from some of the major players this week all discussed near-term cost efficiency initiatives, and we expect these efforts to both help protect the bottom line and drive productivity growth amidst a quickly changing market backdrop. 

Thanks for listening. If you enjoy the show, please leave us a review wherever you listen and share Thoughts on the Market with a friend or colleague today.