Nuveen tech chief Willis Tsai sees signs of investor exhaustion with tech stocks
The biggest question isn’t whether artificial intelligence will transform the world – but whether investors will stay patient long enough to find out.
Has the artificial intelligence boom created a dangerous market concentration not seen since the dotcom bubble? It’s the question keeping investors worldwide awake at night.
Willis Tsai is grappling with this daily. After two decades immersed in tech stocks – including as a semiconductors analyst at Goldman and nearly 20 years at Nuveen covering chips, communications equipment and Asian technology – he now heads up Nuveen’s global equities and tech sector coverage.
“It really is not an understatement to say there is a lot of gearing towards these mega cap companies,” Tsai says. “They are massive. They’re very profitable. They have the ability to invest.”
And the first cracks may already be showing.
Despite strong results, the market isn’t rewarding tech companies the way it did in 2023 and 2024.
Nvidia – the world’s biggest company and AI posterchild with a market cap of $US4.4 trillion ($6.65 trillion) – has gone nowhere for two weeks despite earnings and revenue guidance coming in stronger than expected.
Microsoft dipped 3 per cent on Wednesday on a report claiming the software maker cut expectations for customer spending on its Azure cloud computing division.
Microsoft said aggregate sales quotas for AI products haven’t been lowered.
“Perhaps that is a sign that there’s some investor exhaustion,” Tsai says.
The so-called Magnificent 7 tech stocks have driven 68 per cent of all earnings growth in US markets over the past decade. They now account for 35 per cent of total American corporate profits.
But unlike the dotcom era, when overvalued companies with no profits commanded stratospheric valuations, today’s tech leaders are enormously profitable cash machines.
Despite its meteoric rise, Nvidia’s shares now trade at about 25 times 12-month forward earnings – towards the lower end of its five-year range of 20-68 times.
“The earnings-based multiple doesn’t look demanding for a company like Nvidia,” Tsai says.
The worry is that those earnings forecasts assume the current frenzy of AI infrastructure spending – upwards of $US500bn for the US tech giants – continues unabated.
“The big debate is whether the consensus view of the earnings power of the company in the next 12-18 months is correct,” Tsai says. “Because right now, with the multiple where it is, there’s a portion of the market saying they are sceptical that earnings will be as durable at this level. There is concern about the pace of this spending continuing to grow as it has in recent years.”
Monetisation of AI remains largely theoretical for many. The market is taking “somewhat of a leap of faith” that meaningful revenues will materialise in the latter half of the decade, Tsai says.
But overall he’s quite constructive. He thinks the market will give AI stocks the benefit of the doubt.
“I do believe that the market has and will remain patient in terms of seeing how business models develop and seeing the absolute return from the current infrastructure investments,” Tsai says. “That is because there is a fundamental belief that as we head towards the latter half of the decade, there will be meaningful revenue generation from AI through applications that we can all see today, but also through applications yet to be unveiled.”
Nuveen, which manages $US1.4 trillion ($2.1tn) for clients globally, thinks about AI investment in three buckets: the enablers who sell the infrastructure, the builders creating AI services, and the ultimate consumers across multiple industries.
So far, market attention has fixated on the enablers. But the real opportunity lies in the companies that use AI to grow efficiency and discover new applications across sectors. “The ability to go across sector, across region really provides an interesting way to look for the best opportunities,” Tsai says.
On China he is optimistic after years of investor flight. The central government appears to be shifting away from its private-sector crackdown towards a more business-friendly stance.
Large Chinese technology companies, once targets of regulatory pressure, are attracting renewed interest thanks to their dominance and vast data sets for AI development.
“That’s drawing more interest in aspects of the Chinese market that we haven’t seen for a few years,” Tsai says.
But geopolitical uncertainty makes it tough for investors to fully embrace Chinese opportunities, even though the country still represents over 30 per cent of emerging market indices.
More broadly, he sees fascinating opportunities beyond the crowded AI trade – in Europe, Japan and pockets of emerging markets where his team are hunting for overlooked value.
“When and if there is a scenario where there’s more volatility, or a reason to think that indices won’t be driven by a small handful of companies, it actually creates a stronger case for active managers,” Tsai says.
One market dynamic particularly fascinates him. Through much of the year, highly volatile, less-profitable companies favoured by retail traders dramatically outperformed – “in a way that has not historically been the norm”. But recently that’s been reverting back towards quality stocks.
“As long-term investors, we gravitate towards quality,” he says. The shift suggests the market may be experiencing a healthy reality check after bubbly behaviour.
His top worry for 2026? The very concentration in AI infrastructure that’s driven recent gains. “There’s so much gearing to the AI infrastructure build out and positioning within a small handful of names. Results are great, but we need to be very, very keenly focused on that pocket of the market.”
Whether the AI revolution delivers on its transformative promise remains an open question. What’s becoming increasingly apparent is that separating winners from losers will prove extremely difficult – and the market’s patience may already be starting to wear thin.
Originally published as Nuveen tech chief Willis Tsai sees signs of investor exhaustion with tech stocks
Super funds hit for millions to help pay for dodgy advice compo scheme
