Good morning. Andrew here. How long can the corporate profit machine keep up? It has been in overdrive. DealBook contributor Peter Coy goes deep and has some provocative findings. We’ve also got a chart that everyone in the A.I. world is talking about. Kevin Roose, a technology columnist for The Times, walks us through its implications. And please take our quiz about Allbirds, now NewBird AI. (Was this newsletter forwarded to you? Sign up here.)
What will stop the profit juggernaut?
Times are good for big business. In the last quarter of 2025, even though economic growth and job creation slowed to a crawl, pretax corporate profits set a record, reaching their highest share of gross domestic product since record-keeping began in 1947. This year, Wall Street analysts are predicting that profits are likely to be even stronger despite dismal consumer sentiment and high energy costs. For example, Mike Wilson, the chief investment officer at Morgan Stanley, is sticking to his prediction that earnings of the S&P 500 companies will rise 17 percent in 2026. Optimism that the profit juggernaut will keep rolling drove the S&P 500 to a record this week. Will it? To figure that one out, you need to be able to explain why it’s happening in the first place, which is not easy. Experts have different ideas. However, many of them share a belief that the headroom for further profit growth is shrinking. Here are some of the leading theories for why profits have soared so high — and what could sink them. TechnologySuppliers of technology got a huge profit boost from sales of networking gear during and after the Covid-19 pandemic, and they’re getting an even bigger boost from the artificial intelligence boom. The tech sector as a whole has experienced much faster growth in stock prices and measured productivity growth than the rest of the economy, said Jim Paulsen, a veteran market strategist who writes the Paulsen Perspectives newsletter. Surprisingly, though, strong profit growth hasn’t been confined to tech. An analysis last spring by Ricardo Marto, an economist at the Federal Reserve Bank of St. Louis, found that “the recent surge in corporate profits was driven by retail and wholesale trade, construction, manufacturing and health care,” not tech. Marto said this week that the surge in profits had continued across most of these industries in the year since. One possible explanation for Marto’s finding is that tech investment is genuinely paying off outside the tech sector. If so, that would indicate that there’s room for profits to run. Artificial intelligence is beginning to boost profits by automating some jobs out of existence. And before A.I., there was the postpandemic productivity dividend. “In my previous job before Covid, I had a video monitor that was set up to do videoconferencing at my desk,” said Diane Swonk, the chief economist of the consulting firm KPMG. “I never used it once. It was kind of a big eyesore.” Since Covid, she said, “a lot of things we didn’t use suddenly got applied.” Continuous growthRecessions crush profits far more than they reduce wages, so every year without one is a year the profit share holds up. Outside of the pandemic shock in 2020, the United States has not experienced a recession since the financial crisis ended in 2009. That’s a remarkable stretch, and it could continue. The traditional preconditions for a downturn — overextended consumers, overleveraged businesses, excess staffing — are largely absent, Paulsen said. Household debt-to-income ratios have been falling since 2009, corporate balance sheets are strong, and pessimism itself has kept everyone cautious enough to avoid the kind of overreach that typically ends an expansion. “You usually don’t have a recession when everybody expects a recession, because everybody’s on their best behavior,” he said. Market powerCorporate concentration has been rising for a century, according to research from the University of Chicago Booth School of Business. Typically, fewer than 1 percent of corporations now account for more than 90 percent of corporate profits. Fewer, bigger firms tend to have more pricing power, which translates into fatter margins. Ownership matters, too. A 2025 Chicago Fed study found that a one-percentage-point increase in concentrated institutional ownership of a firm is associated with a 0.17 percent reduction in its payroll — evidence that dominant shareholders are pressing companies to keep labor costs lean. Concentration may not keep rising, though. If a Democrat wins the White House in 2028, Biden-style trustbusting could return, constraining profit growth. That’s a segue into the last theory: policy. PolicyCongress and the Federal Reserve have a lot to do with the profit surge. Over the past four decades, the federal tax rate on corporate profits has fallen to 21 percent from 46 percent. Interest rates have fallen as well. Michael Smolyansky, a Fed economist, has calculated that declining interest rates and declining tax rates “mechanically” explain more than 40 percent of real corporate profit growth from 1989 to 2019. Corporate profits have also benefited from the erosion of labor-market regulation, the decline of union membership and the broader political realignment that began with the election of President Ronald Reagan, said Michalis Nikiforos, an economist at the University of Geneva and the Levy Economics Institute of Bard College. But interest rates aren’t falling anymore, fiscal policy has become less stimulative, and the pro-business realignment under Reagan seems to be coming undone. There are lots of forces acting on profits, but one clear one is that profits this far above the historical norm tend to generate pushback. “What’s happening to workers will be a major issue in the midterms,” Swonk said. “It doesn’t matter what side of the aisle you’re on. A backlash can be quite disruptive to profits.”
Iran’s military said it had reimposed control over the Strait of Hormuz. The military said it would keep the vital waterway under its control until the U.S. ended its blockade of Iranian ports. The U.S. government wants to use Anthropic’s latest model. The company — which is locked in a legal battle with the Pentagon over the agency’s decision to label it a supply chain risk — has restricted the rollout of its Claude Mythos Preview over cybersecurity concerns. Now, the Office of Management and Budget is setting up guardrails to let some agencies start using the A.I. model “in the coming weeks,” according to a report by Bloomberg. A federal jury found that Live Nation acted as an illegal monopoly. The trial was seen as a test of antitrust enforcement under the Trump administration, which has tended to settle cases. Remedies, which could include a breakup, will be determined separately. Live Nation will almost certainly contest the verdict. Other big deals: Reed Hastings will step down from Netflix’s board. Saudi Arabia’s sovereign wealth fund is reportedly on the verge of pulling the plug on LIV Golf. And a confirmation hearing for Kevin Warsh, Trump’s nominee for Federal Reserve chair, begins on Tuesday. A.I. watchers are obsessed with this chart
From Wall Street to Silicon Valley, conversations about A.I. are increasingly dominated by one chart. Created by an obscure 30-person nonprofit called METR, it tracks the technology’s progress by measuring the length, in human hours, of a task an A.I. agent was able to complete reliably. DealBook talked with Kevin Roose, The Times’s technology columnist, about his column on the “METR time-horizon,” and why this less than straightforward approach has captured so much attention. What does this chart show, and why did METR decide this is the best way to chart A.I. progress? It’s a little complicated, but basically, they’re measuring whether A.I. systems can autonomously perform various software engineering tasks that would take expert humans anywhere from a few minutes to a few hours. They’re concerned about the risk of a so-called intelligence explosion, where A.I. systems would become capable of building better A.I. systems, sparking a cycle of self-improvement that could lead to out-of-control machine superintelligences. Measuring task length is one way to understand how close, or far away, that might be. You write that it may be only a “slight exaggeration” to say this chart is holding up the whole stock market. Really? Yes, really. The question of how fast A.I. models are progressing, and how useful they are in white-collar workplaces, is the central question behind the A.I. infrastructure build-out, and the fortunes of other software companies. Right now, there aren’t a lot of reliable, third-party evaluations of A.I. progress, and so METR’s time-horizon chart has become something of an industry standard. I have personally talked to executives who have allocated lots of capital based on their interpretations of the chart. People are using this chart to make all sorts of arguments. What do you think is safe to say it shows? I think the chart is a useful directional signal, more than a precise measurement. It shows, basically, that A.I. progress is not stalling or plateauing as some skeptics have predicted, and that it is actually speeding up along the dimensions that investors, governments and A.I. safety experts care about. I wish we had hundreds more independent studies, so we could cross-check them against each other, but for now we have the METR chart. Quiz: A.I. flyingThis question comes from a recent Times article. Click an answer to see if you’re right. (The link will be free.) The sneakers brand Allbirds announced on Wednesday that it had raised $50 million from an unnamed investor to pivot its business to A.I. Roughly how much did the price of its stock spike that day? We hope you’ve enjoyed this newsletter, which is made possible through subscriber support. Subscribe to The New York Times. Thanks for reading! We’ll see you Monday. We’d like your feedback. Please email thoughts and suggestions to dealbook@nytimes.com.
|
