From the studio

Podcast: CIO’s Ulrike Hoffmann-Burchardi on the next market leaders | Apple | Spotify (6 mins)
Podcast: Jumpstart covering US-Iran latest, US jobs data, and debt risks (6 mins)
Video: Investors Club | LIVE from AIC HK: gold, equities, and rates (6 mins)
Video: CIO Monthly | Mark Haefele on earnings, oil, and yields (3 mins)

Thought of the day

June arrives with equity markets firmly rewarding technology leadership. The S&P 500 advanced to a fresh record high last week, the Nasdaq has gained more than 15% over the past month, and the Philadelphia Semiconductor index has surged nearly 70% since the start of April. AI-related capital spending continues to accelerate, supporting a broader set of companies across the AI supply chain, from memory and CPUs to makers of semiconductor capital equipment, as well as optical and other components.

The ongoing AI wave is not the first of its kind, nor is it the first to shake up tech market leadership. In the mainframe era, IBM dominated computing, but the shift to the PC redirected value toward Microsoft and Intel. In the internet wave, infrastructure firms helped build the network, but long-term value accrued to internet-native platforms such as Amazon and Google. In the mobile era, Apple emerged as a defining beneficiary by combining hardware, software, and a powerful app ecosystem.

The current AI transition is unfolding against a backdrop of improving monetization and persistent supply constraints. We see demand for AI tokens, the basic unit of generative AI output, continuing to outstrip supply, leaving compute in deficit even after years of heavy investment. Cloud growth for the major platforms accelerated to an estimated 40% year-over-year pace in the first quarter, up from the 34% expansion in the fourth quarter of 2025, and the aggregate backlog for the big cloud providers has risen to around USD 2tr. This combination of rising monetization, backlog growth, and constrained supply helps explain why the AI buildout still has room to run.

Without taking any single-stock views, we identify several principles from prior innovation cycles that we think can help investors navigate the current cycle:

Incumbents rarely win the next era by default. Previous market leaders rarely dominate the next era by default, in part because their focus may be on protecting existing profit pools. IBM underestimated the PC transition, just as many incumbents in past cycles were slow to adapt to new delivery models. History shows that successful pivots can happen, as with Microsoft and Meta, but they tend to be the exception rather than the rule, and often require a founder or leader willing to disrupt their own primary business.

Platforms often capture the lasting profits. Over time, durable value creation has tended to migrate from infrastructure toward the platforms and networks built on top of it. In AI, that argues for focusing not only on compute and enabling hardware, but also on companies that can turn intelligence into sticky user ecosystems, data advantages, and monetizable applications. We think the firms that reach “platform mode” fastest are likely to be best placed to capture network effects and defend margins.

When capex booms eventually fade, the pain is not equally distributed. AI capex is continuing to surge, and we now forecast that it will reach USD 820bn in 2026 and nearly USD 990bn in 2027. More than 85% of this spending is being driven by the big four technology companies. The sheer scale of this investment should support the overall ecosystem, and especially areas facing bottlenecks. However, history shows that when capacity eventually catches up, competitive hardware segments tend to feel the pressure first. Nortel ultimately went bankrupt after the internet infrastructure boom. Cisco survived, but its stock took more than 25 years to reclaim and surpass its dotcom era high. Our modeling suggests it is in the competitive hardware segments where vulnerability can be most acute when overbuild risks start to crystallize.

So, with the tech trade continuing to heat up, we think these three principles can help investors position for a potential shift in market leadership within AI. We continue to see a constructive medium-term backdrop for equities overall, but we think investors should avoid excessive concentration and remain diversified across the key AI layers and supply chain, rather than relying on a narrow set of mega-cap companies. We think owning both intelligence and applications will matter, because more users can generate more data, improve models, and reinforce the platform’s advantage over time. Alongside staying exposed to the potential beneficiaries of continued AI monetization, we also suggest investors position for a broadening equity rally across our preferred markets and sectors, while using periods of market strength to diversify portfolios and reduce single-stock concentration.