While CIO sees further room for the current tech rally to run, they think investors should review and optimize their exposure to tech to protect against potential declines and benefit from opportunities beyond big tech. (UBS)

The index has gained in 16 of the last 18 weeks for the first time since 1971. Since a low point in October, the S&P 500 is now up 24.8%, with leading technology companies in the vanguard. The FANG+, a gauge of the 10 most traded US technology firms, is up 42.7% over the same period.

With generative AI looking set to be the growth theme of the decade and US large-cap tech companies leading the revolution, we continue to believe that US tech stocks should make up a substantial portion of investors’ equity allocations. This means building up a strategic US large-cap holding for those underinvested.

But investors with excessive exposure should consider diversification, in our view. Other emerging trends in the industry also speak to broadening one’s tech exposure to capture the next growth opportunities. We think Asia is a compelling destination for tech diversification, and see further potential in AI customs chips and foundries.

The AI custom chip market is set to thrive. AI custom chips mostly refer to the accelerator chips designed by big tech and other startups as a complement to the expensive graphic processing units (GPUs) manufactured by companies like Nvidia and AMD. Through the application specific integrated circuit (ASIC) routes, these chips can be cost-effective for the leading platform companies in special cases like inference and training smaller large language models (LLMs) even if their performance lags that of AI GPUs. We expect custom chips to account for 8–10% of AI computing over the next five years, from 2–3% today, as leading platform companies try to diversify their AI computing resources. This would benefit foundries with additional AI chip opportunities, semiconductor equipment companies, and other memory companies globally.

AI edge computing should emerge as one of the fastest-growing segments, benefiting Asia supply chain names. While we believe the majority of near-term investments in generative AI computing will focus on data centers, investors should not overlook AI edge computing, in which AI processes are conducted by a local device like a cell phone instead of by a distant server. This technology can provide faster and more personalized generative AI services that are less resource intensive, and are therefore easier to integrate inside end-devices like smartphones and PCs to perform tasks like basic image generation and translation services. In the medium to longer term, other consumer and industrial devices, like household appliances and autos, could also integrate AI edge chips to take advantage of the proliferation of generative AI applications. Overall, the substantial demand we see in this area suggests revenues could grow from USD 4bn this year to USD 18bn in 2027, with the potential to reach USD 30bn in 5–10 years. In our view, the AI edge computing trend is a significant tailwind for Asia’s supply chain names.

Valuations are lower in Asia with strong earnings growth. While the forward price-to-earnings ratio of global tech and AI-related stocks has fallen from a peak of around 34 times in April 2022, to around 27 times today, it is still over 20% higher than its 10-year historical average. We think such valuations are justified given rapidly growing AI demand, but those for Asian AI-related stocks are cheaper, at 18–20 times earnings. In addition to a compelling value proposition, we also see robust growth ahead for select companies in the region. In fact, both sales and earnings growth are estimated to come in well above those for the Magnificent 7 for the next two years.

So, while we see further room for the current tech rally to run, we think investors should review and optimize their exposure to technology to protect against potential declines and benefit from opportunities beyond big tech.

Main contributors – Solita Marcelli, Mark Haefele, Sundeep Gantori, Daisy Tseng, Vincent Heaney

Read the original report : Diversified tech exposure offers benefits, 4 March 2024.