From the studio

Podcast: Signal over Noise | The open model AI ripple, on Apple and Spotify (6 mins)
Video: Top of Mind in APAC | The current bull market and risks to watch (5 mins)

Thought of the day

Microsoft’s shares rose nearly 9% in after-hours trading on Wednesday as its cloud unit grew at the fastest pace in four years, and the company projected further acceleration for the current quarter. Meta, meanwhile, fell 7.5% after it reported a 91% plunge in second-quarter free cash flow to the lowest level since late 2022.

Without taking any single-company view, the divergence in share price performance underscores investors’ continued focus on monetization of AI investments. Alphabet’s shares fell last week after the company reported its first-ever quarter of negative free cash flow while raising its estimate for 2026 capital spending.

Ahead of Amazon’s earnings due later today, we discuss our takeaways from recent quarterly results across the major US hyperscalers:

Near-term AI capex remains strong, and monetization is becoming more visible. Following Alphabet’s higher capex guidance, Meta raised the lower end of its spending outlook, projecting 2026 spending of between USD 130bn and USD 145bn. Microsoft said its investment expectations for this year remain unchanged, although an accounting tweak means some of its capital expenditures will shift to operating costs. With both Alphabet and Microsoft signaling further capex increases for 2027, we expect near-term AI spending overall to remain strong. Separately, Microsoft indicated further acceleration in cloud revenue growth for the current quarter, while Alphabet reported an increase in advanced cloud orders that have yet to be recorded as revenue. This adds to encouraging monetization trends, in our view.

But pressure on free cash flow could constrain AI spending beyond 2027. While capex is likely to rise further next year, and monetization is picking up, the latest results showed that hyperscalers’ heavy spending is increasingly weighing on their cash flows. We estimate that these hyperscalers’ operating cash flows will be overtaken by their cash capex requirements in the current quarter, and this means that the risk of AI spending in 2028 coming in below the 2027 level has risen. A potential capex slowdown could lead to weaker prices for semiconductors, weighing on parts of the sector where strong pricing power has driven most of the recent rally.

The AI investment opportunity has become more differentiated. Given the backdrop of near-term strength and risks on the horizon, simply owning companies linked to AI infrastructure spending is unlikely to benefit from the next phase of AI growth. We believe investors should be more selective in their exposure to semiconductors, as the industry is composed of diverse segments each with its own cycle and dynamics. We also believe the broader semiconductor and hardware complex now offers a more balanced risk-reward profile following the strong rally in the second quarter of this year. In our view, investors should retain exposure across the entire AI value chain, and we recommend a barbell positioning that complements holdings in semiconductors with more defensive areas of tech. We favor semi-cap equipment, foundries, and compute names within semis, and see opportunities in smartphone makers, payment networks, data center REITs, and select consumer electronics.

So, we remain constructive on the AI growth story, but believe investors should manage concentration risk by broadening their exposure to defensive tech stocks. We also continue to recommend a broadly diversified equity allocation that spans sectors and geographies.