U.S. stocks are facing the prospect of their first set of back-to-back losses since mid-May, with the S&P 500 and the tech world in retreat following a muted set of forecasts from AI chip maker Broadcom that looks to have shaken the market’s hottest trade a week ahead of its biggest IPO.

Broadcom’s forecast miss, while not massive in terms of dollars, nonetheless marks a rare retreat for a sector that has dominated markets since the end of the first quarter, driving broader indexes to record highs and boosting GDP growth estimates for the world’s biggest economy along the way.

The pullback also coincides with a market making room for SpaceX’s $1.8 trillion IPO, expected to price next week. 

It’s the largest on record, with a value three times higher than Nvidia’s, but is tied to a group of firms that have yet to turn a profit. Its sheer size and index inclusion will both seep value from other stocks in the market, while its fundamentals will make it difficult to generate post-IPO growth with a corresponding bounce in earnings. 

Morningstar says it’s “significantly overvalued.”

For markets more broadly, the picture remains just as complicated. 

Developments in the Gulf region suggest Israel has reached a cease-fire agreement with Lebanon, but a similar extension of the fragile truce between the U.S. and Iran remains elusive. 

That’s keeping oil prices higher, stoking inflation concerns, and lifting U.S. Treasury bond yields back to key support levels ahead of Friday’s crucial jobs report. 

And now the markets look untethered to the first-quarter earnings season, those developments, and big questions about the fate of the tech trade over the summer months, can no longer be brushed aside.

—Martin Baccardax

Get more of the journalism you love. Choose Barron’s as a preferred source in Google.

Broadcom beat expectations thanks to supercharged growth in its artificial intelligence chip business. CEO Hock Tan sees momentum continuing to build, with third-quarter chip revenue from AI expected to rise 200% over the year-ago period, to $16 billion. Revenue is expected to accelerate for the rest of the year.

• Broadcom reported adjusted earnings of $2.44 a share for the May quarter. Revenue rose 48%, to $22.19 billion, and revenue for the semiconductor solutions segment was $15 billion. Its second-quarter AI revenue totaled $10.8 billion, up 143% from a year ago.

• Looking ahead, Broadcom forecasts fiscal third-quarter revenue of about $29.4 billion, which would represent around 84% growth and beat expectations. It designs custom AI chip alternatives for six customers, including Alphabet and OpenAI. 

• The recent long-term agreement with Google to develop and supply multiple generations of Tensor Processing Unit chips and AI networking is a very strong agreement, Tan said. “A commitment that is very substantial in dollars.” But Google is likely to source from other companies, too, he said.

• Investors may have been disappointed when Broadcom left its 2027 outlook unchanged. It still sees guidance for AI semiconductor revenue of more than $100 billion. The shares were down 12% in after-hours trading after that.

What’s Next: Broadcom is delivering chips to OpenAI and says it’s on track for production this year. Its contract is to deploy 1.3 gigawatts of capacity in 2027, Tan told analysts. The OpenAI agreement ultimately aims for 10 gigawatts by the end of the decade.

—Adam Levine, Kit Norton, and Liz Moyer

Companies are rushing to build multi-acre battery installations that can store power for artificial-intelligence data centers and other electricity needs. The trend spans from Ford Motor, which announced a new battery storage division in May, to Fluence Energy, an upstart battery-maker collaborating with Nvidia.

• The battery industry is dominated by Chinese firms, which can make battery packs for much less money than American competitors. And the lithium-based battery product that each company puts out looks similar. These kinds of commoditized industries tend to succumb to price wars.

• A few factors have changed the calculus, causing demand for U.S. battery systems to surge and potentially keeping margins elevated. U.S. tariffs on Chinese batteries can add more than 80% to the base price. And last year’s tax bill kept domestic manufacturing subsidies in place.

• A surge in electricity demand is also boosting the battery business. Batteries have traditionally been used to store power produced by solar and wind installations. They are now put to wider uses, including as a power source on-site at AI data centers, as backup and to help stabilize power consumption.

• The rise in renewables adoption across the U.S. was already expected to cause battery installations to increase steadily, but data centers will supercharge the growth, Morgan Stanley says. Last year, 57 gigawatt-hours of storage was added to the grid. By 2030, installations could rise to 279 gigawatt-hours. 

What’s Next: Ford is spending $2 billion to build the energy storage business and expects to start making deliveries in 2027. Morgan Stanley is excited about the prospects, and predicts it can add $10 billion in value to the company once the capacity is fully built.

—Avi Salzman

The drug industry has been trying for years to develop tricks to shuttle the drugs that treat Alzheimer’s disease through the body’s protective brain-blood barrier, which blocks 99% of a dose from reaching the brain. The results of their efforts will become apparent this year.

Denali Therapeutics, which got the first U.S. approval for a shuttle-enhanced drug in March; and Roche Holding, which is in pivotal trials with its barrier-crossing Alzheimer’s antibody, are leading the pack. Success could yield annual sales above $13 billion for Roche and turn around the biotech’s fortunes.

• Roche’s anti-amyloid antibody trontinemab has a fragment that mimics the natural iron-shuttling molecule known as transferrin. Roche in September started Phase 3 trials for trontinemab in patients with early symptoms of Alzheimer’s, as well as another Phase 3 trial to treat patients before cognitive symptoms appear.

• In March, the Food and Drug Administration approved Avlayah, which taps the transferrin receptor to bring in the enzyme that is missing from the brain cells of children born with Hunter Syndrome, a rare and deadly disorder. Similar treatments for the enzyme deficiencies known as Sanfilippo, Pompe, and Gaucher diseases are in early clinical trials. 

• To get gene therapies into cells, biotech companies have used hollowed-out viruses called capsids. Both Voyager Therapeutics and Sangamo Therapeutics are developing capsids that first open blood-brain portals, then deliver genetic cargoes to brain cells. 

What’s Next: Denali recently began clinical trials of a treatment that silences the genetic instructions for the Alzheimer’s tau protein. Preclinical projects are testing vehicles for delivering amyloid antibodies, and treatments for Parkinson’s disease. The market for a successful Alzheimer’s or Parkinson’s treatment could exceed $5 billion.

—Bill Alpert and Janet H. Cho

From SpaceX to Anthropic, the pipeline of blockbuster initial public offerings is moving fast. But for quantum enthusiasts, the biggest milestone of the year has arrived: Quantinuum priced its IPO at $60 a share and it set to trade sometime today.

• The IPO was upsized to 28 million shares of its class A common stock, raising $1.68 billion. Underwriters have been granted a 30-day option to sell an additional 4.2 million shares if demand is higher than expected, the company said.

• The Honeywell-backed upstart will be listed on the Nasdaq Global Market under the ticker symbol “QNT.” The IPO caps months of uncertainty for investors looking to Quantinuum as the next big force in quantum computing.

• Formed in 2021 through a merger between Honeywell Quantum Solutions and a U.K.-based start-up, Quantinuum has built a roster of collaborators across the energy, aerospace, and finance sectors plus the backing of the U.S. government.

• As part of a $2 billion funding package, Quantinuum will receive up to $100 million from the Commerce Department to scale its trapped-ion quantum computers. In turn, the department will receive a minority equity stake in the company.

What’s Next: Quantinuum has had ample time to refine its flagship trapped-ion technology, supported by a parent with deep pockets. An investment in quantum comes with the usual disclaimers. None of the companies is profitable and all will continue to post losses for the foreseeable future.

—Mackenzie Tatananni