It’s a clash of the titans as Nvidia and Federal Reserve Chairman Kevin Warsh vie for the stock market’s attention this week.

The market needs them both to deliver if it’s to avoid back-to-back weekly losses. 

The artificial-intelligence trade—and chip stocks in particular—have faltered recently and need a favor from the world’s most valuable company. The PHLX Semiconductor Index, or SOX, is down 18% since the start of July.

The mood isn’t great to start the week—and Asian trading is partly to blame. Alibaba’s $10 billion share sale to fund its AI spending spooked the market, while Samsung stock tumbled 9% after its buyback plan disappointed investors.

Nerves are setting in, ramping up the pressure on chip maker Nvidia. The tech trade needs its biggest name to provide a steady hand on the tiller when it posts earnings after the bell Wednesday. Any cracks in its figures will spike fears about AI spending and a possible bubble.

The broader stock market needs the same reassurance from the Fed’s Warsh. Since he took over in May, the central bank’s leader has retreated from giving forward guidance—in contrast to predecessor Jerome Powell—but the market really needs to hear how he plans to tackle inflation at the annual Jackson Hole central bank symposium. If core PCE, the Fed’s closely watched inflation metric, comes in hot Wednesday, it will turn up the heat on Warsh.

Meantime, bond market fears are still lurking. Rising yields hit stocks last week prompting intervention from the Treasury, which only paused the bond selloff.

Nvidia earnings and Jackson Hole butt heads annually but are rarely this intertwined. Surging bond yields are the tech trade’s nemesis, particularly with companies borrowing so much to fund the explosive AI spending boom.  

It’s not so much Nvidia vs. Jackson Hole this year. The stock market needs a boost from both—or else.

—Callum Keown

Barron’s Live: Join Barron’s Editor in Chief Ben Levisohn and Investor Circle Newsletter Editor Josh Schafer today at noon when they speak with Michael Cuggino, president and portfolio manager of Permanent Portfolio Family of Funds, about the outlook for financial markets, industry sectors, and individual stocks. Sign up here.

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

Billions of dollars are pouring into the AI buildout, racing ahead of Wall Street forecasts and leaving analysts playing catch-up. A Barron’s analysis of more than a decade of analyst forecasts shows capital expenditure estimates for hyperscalers including Alphabet, Amazon, Microsoft, Meta, and Oracle are being raised at historic rates.

• Those miscalculations can cost investors. Last month, Alphabet stock fell 7% after it announced plans to spend as much as $205 billion this year, up from $190 billion. Other hyperscalers fell, too, as investors grappled with how much Big Tech will have to spend to keep pace in the AI race.

• Hyperscalers collectively spent $330 billion in this year’s half, and analysts estimate an additional $469 billion of capital expenditures in the second half. Analysts expect even more spending next year—more than $1 trillion of capex across major data center operators. Their initial fiscal 2027 forecasts have increased by an average of 183% across the five companies, compared with an average revision of just 33% over the past decade.

• Even with major revisions to their initial estimates, experts say analysts are still way off. I/O Fund’s Beth Kindig calls this year’s initial estimates “laughably low,” and next year’s projections “still too low.” Goldman Sachs strategists agree.

• That spending has more than a few investors nervous because they haven’t seen any return on those billions. But economist Jessica Wachter thinks the concern is overblown. “Of course, it’s not without risk, but…it’s more like the baseline level of risk that we all accept when we invest in the stock market.”

What’s Next: Companies that supply the hyperscalers stand to benefit from their massive infrastructure buildout. Evercore ISI analysts identified five—CoreWeave, Western Digital, Seagate Technology, Arista Networks, and Vertiv Holdings—with “high exposure” to AI capex. Their average return this year is 100%, compared with the hyperscalers’ -3.6%. And investors should expect that dynamic to keep going.

—Elijah Nicholson-Messmer

The race to build advanced AI tools has spawned a new market for real-world data. Google’s $10 million outlay for data from the defunct Spirit Airlines is just the latest example. And it has taken on more urgency now that scraping the internet has become more legally fraught.

• Google won access to Spirit’s software code, internal messages, and financial and operational data in a bankruptcy auction this month. It hopes to use the data to improve its products and AI models, a spokesperson said. The industry’s demand for data is linked to how AI large-language models are created and refined.

• AI systems ingest huge volumes of human information and use it to mimic human processes, such as writing reports, interpreting and responding to messages, and recommending things. More data gives the models more information to study. In response, public companies are weighing whether to license their data.

• The five largest cloud providers are spending a combined $800 billion in 2026 on data centers, and AI labs are paying hundreds of billions of dollars to use them. Millions of dollars—with an “M”—mean very little when creating new AI tools, but that’s big money for data sellers.

• In January, Cloudflare acquired Human Native, a marketplace connecting content creators with AI developers. Two months later, News Corp, the owner of Barron’s publisher Dow Jones, struck a licensing deal with Meta Platforms, where the publisher sold access to its content for $50 million a year.

What’s Next: Reddit already has agreements with OpenAI and Google to license conversations from its forums, bringing in around $60 million a year from each lab. The social-media platform is now in negotiations with Google, and some investors and analysts expect a much larger deal.

—Nate Wolf

The Trump administration may be considering a proposal to cut capital-gains taxes by indexing gains to inflation. Experts say the change would overwhelmingly benefit wealthy Americans. The concept is that only returns above inflation count as profits, meaning lower tax bills for individuals who report gains.

• National Economic Council Director Kevin Hassett recently told Fox News that President Donald Trump is looking for new policy ideas that could give voters more incentives to support Republicans in November. Fox host Larry Kudlow said he recently discussed capital gains indexing with Trump.

• Proponents argue that the policy, which has been floated by Republicans as far back as President George H.W. Bush, would benefit middle class taxpayers. But the math suggests that the vast majority of the benefits would be enjoyed by the wealthiest Americans.

• About two-thirds of all capital gains reported in 2023 came from households with at least $1 million in income, according to IRS data. Households earning $75,000 to $100,000—the typical American income range—accounted for just over 1% of total capital gains. That gap has major implications.

• Last year, Republican Senators Ted Cruz and Thom Tillis introduced legislation that would index capital gains to inflation on assets held for more than three years. The bill was referred to the Committee on Finance, but has yet to receive a vote.

What’s Next: Sen. Tillis says the “bill brings much-needed fairness to the tax system and helps families keep more of what they earn.” But critics say the policy would provide little benefit to most Americans.

—Elijah Nicholson-Messmer

China has spent five years trying to clean up the wreckage of its property bust, triggered by the giant developer Evergrande. And it’s not done yet, despite helping build about 7.5 presold homes last year. That big number is still only 15% to 40% of the country’s total stock of presold but unfinishing housing, according to the World Bank.

• Chinese developers traditionally sold apartments before construction was complete, using buyers’ money to help finance building. Evergrande’s 2021 default shattered confidence in that bargain.

• Beijing has thrown considerable money at the problem—more than 7 trillion yuan, or about $1.04 trillion, in loans toward viable projects under construction. Yet, property investment broadly is shrinking. Residential construction starts dropped 24.6%, completions fell 25.5%, and home sales by floor area declined 12.7%, according to the government.

• That is the backdrop to last week’s sentencing of Hui Ka Yan, founder of Evergrande. He received life in prison for an assortment of crimes, including fraudulent securities issuance. Another court accepted a bankruptcy-liquidation petition against Hengda Real Estate, Evergrande’s main mainland property unit.