From the studio

Video: The AI Show | Opportunites amid software disruption and margin squeeze (4 mins)
Video: Investors Club | Staying positioned in US financials, US, and China tech (9 mins)
Video: Portfolio positioning amid US-Iran talks (5 mins)

Thought of the day

Artificial intelligence headlines this week have offered fresh evidence that investor interest and customer demand remains strong. Amazon said it would invest up to USD 25bn more in Anthropic, while the AI lab in return committed to spend more than USD 100bn over the next decade on Amazon’s cloud technologies. Reuters reported that a separate Bezos-backed AI lab is close to raising USD 10bn at a USD 38bn valuation. The Financial Times reported that Recursive Superintelligence, a four-month-old frontier AI lab, has raised at least USD 500mn at a USD 4bn valuation. That follows a first quarter in which AI-related venture activity helped lift global start-up investment to a record USD 300bn, according to Crunchbase data.

The hardware side of the ecosystem continues to demonstrate strength. The Philadelphia semiconductor index has rallied for 15 consecutive sessions, advancing more than 30% during this period. South Korea’s Kospi again hit new record highs on Wednesday, encouraged both by positive developments in the Middle East conflict and bullish exports data for the first 20 days of April, which showed semiconductor exports surged more than 180% year over year. At the same time, some large technology companies that entered 2026 under more pressure on AI have begun to respond. Apple has elevated its hardware chief to CEO to replace Tim Cook, cementing its product-centric approach to adapting AI, while Adobe launched a new enterprise AI suite focused on digital marketing and customer engagement to fight back against encroaching AI rivals.

For investors, we think the next phase of the AI trade will increasingly reward execution and returns, not just exposure to the theme:

Supply-demand imbalance still supports elevated hardware pricing. North Asia’s strong export data, ongoing custom-chip initiatives, and continued reports of tight leading-edge capacity all point to a supply chain that remains under pressure from sustained AI demand. We remain constructive on memory, advanced semiconductors, and enabling infrastructure, supported by visibility through 2026 and into 2027. That said, we would not assume leadership stays static. Some parts of the hardware chain have already re-rated sharply, and rotation within the theme is likely as investors weigh capacity, pricing, and execution.

Adaptation will separate legacy beneficiaries and laggards. For software and consumer hardware companies, the challenge is shifting from simply having an AI story to proving that it can support products, workflows, and returns. We have argued that widespread disruption in software is more likely a long-tail scenario than an immediate one, especially for enterprise-facing and mission-critical providers with sticky customer relationships. Very rapid revenue growth for leading AI coding tools has also helped ease some investor concerns around elevated data center spending. We do anticipate the gap between resilient and vulnerable business models is likely to widen. That means investors should focus on data advantages, switching costs, and monetization pathways rather than on headlines alone.

Private capital still sees a long runway for AI. The latest venture and strategic funding announcements reinforce our view that AI remains a multi-year investment cycle, not a short-lived burst of enthusiasm. They also suggest private capital is still willing to back both foundational infrastructure and newer research-led entrants at scale. However, our private markets analysis also highlights several risks, including concentration in popular AI themes. We suggest a selective, diversified approach to private exposure rather than broad optimism.

Taken together, this week’s developments reinforce that AI is still broadening across funding, hardware, and enterprise adoption, even as the market becomes more selective about where returns will accrue. As earnings season approaches, we think investors should watch three things: the strength of the core businesses supporting AI investment, any increase in capex plans, and managements' outlook on returns on capital. Our preference remains for a diversified approach across the AI value chain, consistent with our AI TRIO positioning, balancing strong long-term fundamentals against geopolitical risks, rising competition, and free cash flow pressures. More broadly, we retain an Attractive view on US equities overall, with an S&P 500 target of 7,500 for December, and continue to favor the consumer discretionary, financials, health care, industrials, and utilities sectors. We continue to see a role for selective private equity exposure linked to infrastructure, resilient enterprise software, and broader AI-enabled value creation.