The S&P 500 index just posted its fattest profit margins in a generation, and the market is paying up for them. The index's blended net margin reached 15.7% in the second quarter of 2026, the highest that FactSet has recorded since 2009. Post-tax corporate profits now run about 12.4% of U.S. gross domestic product, well above the long-run average of 7.3%. Investors are underwriting these levels as the new normal.

Some companies deserve the praise and attention that investors are giving them. The question is which ones, because growing margins alone don't tell us much.

When margins expand, we should ask what else moved. Did the business grow? In other words, did the firm create more value? Or did the firm's value remain flat, and its conversion of payroll into profit generate the extra margin? The first is growth. The second is redivision. The first is durable since the gain is manufactured, not transferred. The second is fragile, for two reasons every macro investor should recognize.

Margin growth that relies on cutting jobs invites mean reversion. A margin built on a falling labor share rests on a claim that someone else will eventually contest, through wage pressure, turnover, or the simple exhaustion of how far head count can be cut. And it invites policy intervention, because rising profit shares sitting on top of falling labor shares is the cleanest political target around. Everyone can sense the economic and financial vulnerabilities it creates.

Balanced growth can't move the profit share of gross domestic product. When output and pay rise together, profit share sits still. A profit share running at about 12.4% of GDP, five points above its post--World War II average, isn't what an economy full of sustainable stories of scale looks like. It is what redivision looks like in aggregate. The task for an investor is to find the firms that are the exception.

Watch a company's profit margin -- its profit as a percentage of revenue -- against its own history. And watch its labor share -- its compensation as a percentage of the value the firm creates -- against its own past peak. Together, these numbers reveal what margin numbers alone can't: where the gain came from.

A margin that rose while the labor share held wasn't carved out of pay. A margin that rose while the labor share fell is the redivision case, where the extra profit and labor's loss are the same dollars. You don't need to know which technology the firm used or how it was deployed. Two lines, both reported, separate the durable margin from the borrowed one.

This matters more now than ever because artificial intelligence is about to run the experiment across the whole market at once. It's already starting: Layoff notices in finance, consulting, and media now routinely cite AI. In its latest public filing, Oracle disclosed that it cut about 21,000 jobs, roughly 13% of its workforce, over the past year. It named AI deployment as a cause.

When a company like Oracle swaps a worker for AI tokens and books the difference, its income statement improves and the labor-share reading falls in the same quarter. The margin looks better.

Index concentration intensifies this effect. The top 10 companies in the S&P 500 are generating about 34% of the index's profits, roughly double their share in the mid-1990s. If you strip out Alphabet, whose second-quarter 2026 net income was swollen by about $77 billion of unrealized gains on its stakes in SpaceX and Anthropic, the index margin falls from 15.7% to 14.4%.

A market this top-heavy is one where a handful of labor-share readings carry the whole index's durability, and where the gap between real operating leverage and a one-time markdown of the wage bill decides whether today's multiple survives.

Eventually, policy will press on this. The profit-versus-labor gap is quickly becoming the most legible grievance in politics. Just last week, Rep. Greg Casar (D., Texas) introduced a bill taxing large AI companies on the tokens they sell, with the revenue funding public jobs, and Sen. Ron Wyden (D., Ore.) proposed new taxes on data centers and supports for workers. Sen. Elizabeth Warren (D., Mass.) has floated a similar bill. More will follow.

A portfolio positioned as if record margins are permanent and untaxed ignores both mean reversion and legislative risk. Both have the same underlying reason: The margin was borrowed from labor, and the market and the state tend to collect on that eventually.

That isn't to say one should avoid high-margin companies altogether. The lesson is to think more critically about eye-popping margins. Run the two-number screen across a portfolio, and the holdings split cleanly into firms compounding output with their labor share intact -- with margins as real as the price implies -- and firms whose recent gains are a wage transfer wearing an operating-leverage costume.

In a market paying peak multiples for peak margins, knowing which is which is an edge. Wages in the U.S. ran near $13.2 trillion at an annual rate in early 2026. That is the base the extracted margins came out of, and where they can return.

Guest commentaries like this one are written by authors outside the Barron's newsroom. They reflect the perspective and opinions of the authors. Submit feedback and commentary pitches to ideas@barrons.com .

Amir Goren is an independent researcher and investor. He holds a doctorate in economics from the University of California, Irvine, and is the author of The Labor-Share Screen. He holds no position in Oracle or Alphabet.