BlackRock Strategist Favors AI 'Scarcity' Stocks, Shorter-Term Bonds, and Infrastructure — Barrons.com

By Reshma Kapadia

As investors debate whether artificial intelligence is in a bubble or has the potential to substantially push economic growth higher, Wei Li, global chief investment strategist for BlackRock Investment Institute sees some of the best opportunities in companies providing inputs that are scarce but critical for AI — a reason BlackRock still favors U.S. stocks, infrastructure plays, and shorter-term bonds as interest rates settle higher.

On Tuesday, the Institute released its midyear outlook, in which it highlighted a difficult investing backdrop for asset allocators amid a series of big unanswered macroeconomic questions — such as, Will the cost of AI be deflationary or inflationary?, and, Can it push economic growth beyond its long-run trend of 2%?

Whereas investors in the past were able to build a base case, Li says the answers to today's questions result in wildly different scenarios that are equally likely. That makes it hard to find an anchor for portfolios.

BlackRock highlighted three themes against this backdrop: AI scarcity, which results in a portfolio overweight in U.S. equities with a focus on companies geared to supply-chain bottlenecks, or inputs in short supply, such as in power, batteries, and industrial manufacturing; durable income with a focus on short- and medium- term bonds, whose income is supported by clear cash flows as higher rates take hold; and a more nuanced approach to diversification as AI concentration is embedded in an array of asset classes.

While there are growing concerns about how much spending is needed for this transformation, Li tells Barron's she still sees AI in the early stages of a super cycle. Incremental margins across the AI value chain — power, data centers, software, and hyperscalers — are still higher than operating margins, suggesting that companies are still generating positive returns on their spending, she adds.

Li says it is still too early to pick AI winners and losers, a reason the team is focusing on bottlenecks. While the closure of the Strait of Hormuz highlighted the importance of these chokepoints, Li notes that they extend beyond just physical to financial given the financing needs for the AI buildout, energy transition, and supply chains.

The U.S. has an advantage on multiple fronts, with leading chips, state-of-the-art AI models, and energy independence, but also with the depth of its capital markets needed to finance the AI buildout. The U.S. stock and bond markets make up 90% of global equity and bond markets, with Japan and China next in line at about 10%, Li said in an interview.

"Having deep capital markets is a significant advantage — and a chokepoint. Europe has significant savings pool but its ability to mobilize it has been less efficient," Li says. "It's difficult to pick winners because it's still early but when you think about the advantages the U.S. has — physical in terms of chips, leading models and the depth of the capital markets and energy independence — together with the innovative spirit and willingness to take risk, chances are that many of the winners in broader AI adoption will be in the U.S. market."

China also has an edge on some of these bottlenecks, including AI batteries, industrial manufacturing, robotics, and rare earths, which it used as leverage in U.S.-China tensions last year. But Li favors a more selective approach versus a broad overweight to China, citing flatter return on equities and intense price competition, which can mean market-share gains don't translate to profitability.

Though BlackRock strategists see a trickier second half for stocks and bonds broadly as inflationary pressures in the U.S., Japan, and Europe feed central-bank hawkishness and potential interest-rate hikes, Li doesn't see higher rates derailing the opportunity for industrial and other companies geared toward addressing some of the bottlenecks and capital spending for AI, the energy transition, or supply-chain rewiring. If growth can offset the drag from the higher discount rate, Li says the asset class can do well. So far, earnings growth from the U.S. market broadly, and from industrial companies in particular, is outrunning rates, she says.

One note of caution for investors: Diversification will become harder as AI concentration spreads across asset classes and as some of the attractive opportunities, like infrastructure, don't fit neatly into one asset class.

While some are describing small-cap and value stocks as "anti-AI" trades, Li says half of the gains for the global small-cap index this year and three-quarters for the global value index have come from memory stocks even though the index weighting to AI-advantaged stocks is in the single digits.

For those looking for investments not tied to AI, Li says European financials are attractively priced, and her team is finding opportunity in healthcare in the U.S. "There's an element of AI transforming it but conviction is mostly on valuation, and we are starting to see earnings upgrades but performance has lagged behind," Li adds.

Another potential ballast for portfolios: Infrastructure, which Li says sits at the crossroads of several of its major themes beyond AI, like an aging population, geopolitical fragmentation, and the energy transition. Li thinks it could be closer to 20% of investors' portfolio allocations if they look to get that exposure through various asset classes — via bonds and equities, but also private equity and investing directly in infrastructure.

Write to Reshma Kapadia at reshma.kapadia@barrons.com

This content was created by Barron's, which is operated by Dow Jones & Co. Barron's is published independently from Dow Jones Newswires and The Wall Street Journal.