HSBC reiterated a generally bullish stance on US equities in a note issued on Friday, maintaining a mild overweight position supported by resilient economic growth, broadening earnings across sectors, and continued leadership in artificial intelligence — a theme that has been a primary driver of US equity returns over the past year.

The bank addressed the recent weakness in semiconductor and memory stocks, characterising the sharp pullback as rotation rather than a genuine retreat from the AI trade. According to HSBC analysts, the move reflects investors reassessing whether profit growth can keep pace with elevated expectations, rather than a decision to exit the sector altogether.

The scale of the chip-stock decline has been notable. Semiconductor and memory names including Samsung, SK Hynix, Intel, and Micron have each fallen by roughly a third over the past month. Meanwhile, software stocks — an area HSBC describes as sometimes unloved — have come back into favour. The Morningstar Global Software-App Index has rebounded approximately 16% from its June low, and companies such as Salesforce, Workday, and ServiceNow have seen their share prices rise sharply over the past week. That rebound reinforces the bank's view that capital is rotating within the broader technology trade rather than leaving it, which could limit further downside in semiconductor names once portfolio reshuffling runs its course. The rotation also highlights a structural difference in AI monetisation: hardware revenues are tied to cyclical capital expenditure, while software firms can embed AI features into recurring subscription models, offering more predictable revenue visibility.

HSBC said it continues to view the structural AI theme as intact, maintaining positions across the AI ecosystem in Asia, spanning power, infrastructure, and industrial automation. The bank noted that competition among AI models is intensifying and that pricing pressure is building as the era of subsidised AI access fades, pushing providers toward monetisation approaches such as Model-as-a-Service — a transition that will shape which companies succeed in converting AI development costs into durable revenue streams.

In China specifically, HSBC observed renewed investor preference shifting toward biotech, internet platforms, hyperscalers, and electric vehicle makers.

Underpinning the bank's broader conviction is its outlook for AI capital expenditure. HSBC projects AI capex rising from below $400 billion in 2025 to more than $1 trillion by 2028, a trajectory it expects to generate meaningful new revenue opportunities for businesses across the AI supply chain.

HSBC also pointed to Asia's growing role in the global data centre build-out, forecasting that the region's data centre capacity will more than double by 2030 to account for approximately 40% of global capacity. That expansion should support demand across the full data centre supply chain, including chips, semiconductor equipment, cooling systems, servers, commodities, on-site power generation, and energy storage — a thesis the bank has framed as its High Conviction call on Asia's data centre boom.

Taken together, HSBC's continued overweight on US equities, paired with selective conviction in Asian AI infrastructure, suggests the bank views the current volatility as a repositioning phase within a still-intact growth narrative rather than a turning point.