HSBC Global Investment Research maintained its ratings and price targets on big tech companies while highlighting challenges and opportunities in 2026 for the AI value chain — cloud, chips, large language models, and hardware.
The firm kept its Buy ratings on Nvidia (NVDA), Alphabet (GOOG) (GOOGL), Microsoft (MSFT), Amazon (AMZN), Meta Platforms (META), Oracle (ORCL), and Broadcom (AVGO). Apple (AAPL) has a Hold rating, while CoreWeave (CRWV) has a Reduce rating at HSBC.
Price targets: Nvidia $320, Alphabet $370, Amazon $300, Apple $250, Broadcom $535, CoreWeave $44, Meta $905, Microsoft $667, and Oracle $364.
“AI exposure lifted plenty of boats in 2025 until October when capex budgets and financing needs got ahead of first stage revenues. We retain our view that AI is showing the initial stages of a megacycle, considering the pace of AI development and the expected potential positive impact on productivity on a USD110trn+ [trillion] global GDP,” said analysts led by Nicolas Cote-Colisson.
The analysts highlighted six key points for 2026.
Strong demand means cloud capacity constraints will continue: The analysts said the three largest Western cloud players — Amazon, Microsoft, and Alphabet — all note capacity constraints, against a backdrop of strong demand and order backlogs. Given longinfrastructure project build times, the analysts do not expect a change to this narrative in 2026.
Bottlenecks: “While infrastructure is rapidly building, power and chip availabilityremain a short-term constraint to revenue upside expectations,” said Cote-Colisson and his team. The analysts added that discussions for this year should focus on power supply meeting demand. Based on Amazon’s earnings call commentary, chip availability is likely to be a long-term issue, the analysts noted.
Capital expenditure direction: The analysts said that 2025 was a year of increasing hyperscaler capex guidance, and given the capacity constraints, they expect this trend to continue. The analysts expect a total cloud capex increase of 44% in 2026.
Application-specific integrated circuits, or ASIC, chips are in vogue, but there is still room for GPUs: “Nvidia GPUs should be the preferred choice for hyperscalers. However, competition from ASICs is growing due to improving performance and the cost benefits,” said Cote-Colisson and his team. The analysts expect external chip sales to be a 2027 story. There is room for growth for both ASICs and graphics processing units, or GPUs. ASICs will increase share with cloud service providers, or CSPs, but GPU total addressable market, or TAM, is expanding through enterprise and sovereign AI, according to the analysts.
Frontier models competition: The analysts expect high sunk costs to lead to market rationalization and oligopolies with smaller specialist players. Open-weight model intelligence has been converging with state-of-the-art models, and the debate about how to charge for top-end models in 2026 looks likely, the analysts noted.
Consumer AI integration, but data center infrastructure driver for hardware: “2026 could be the year for further AI integration in smartphones and for newer form factors (i.e. smart glasses/AI devices) to potentially challenge legacy platforms,” said Cote-Colisson and his team. The analysts added that data center infrastructure is the key growth driver in the AI hardware supply chain