Corporate Profit Margins Are Near a Record While Consumer Sentiment Is at a Record Low

By Yuriy Matso · The Trading Tools · August 26, 2026 · 9 min read

Research note: BEA nonfinancial corporate profits and value added, University of Michigan sentiment, BLS productivity and labor costs, and Robert Shiller’s monthly S&P prices. Computed August 26, 2026. How we checked it · How we use AI

Q2 2026 margin

15.2%

Historical record

15.5% · Q2 2021

Profit growth, quarter over quarter

+9.0%

Value-added growth, quarter over quarter

+2.8%

Q2 sentiment average

54.4

Next margin release

Sep 30, 2026

Why are corporate profit margins so high?

Start with the measure. The numerator is after-tax profit for nonfinancial corporations, adjusted for inventory values and capital consumption. The denominator is the same sector’s current-dollar gross value added. A 15.2% reading means roughly fifteen cents of adjusted after-tax profit remained from each dollar of value produced.

The latest jump came from the numerator. Profits rose 9.0% in Q2 while value added rose 2.8%. Over the year, the split was 17.4% against 8.3%. That lifted the margin by 0.9 percentage points in one quarter, from 14.4% to 15.2%.

Labor provides one useful clue. Productivity was up 2.2% from a year earlier and unit labor costs were up 1.4%. Firms were producing more per hour without a matching increase in labor cost per unit. That is consistent with margin support. These series cannot tell me how much of the jump came from pricing, taxes, interest expense or industry mix, so I would not assign percentages to those stories.

Households are describing a different economy

The University of Michigan series is a mood measure. Its twelve-month average smooths the noisy monthly survey and stood at 53.6 at publication, the lowest continuous monthly reading back to 1978. The raw survey has also been plumbing its postwar lows.

I covered the hard household data in June. Spending was still rising despite the miserable survey. The corporate data adds a second contrast. Businesses were retaining an unusually large share of value added while households described an unusually bad economy.

The gap is the original finding

Dollars and survey points cannot be compared directly. Their positions in history can. For each complete quarter, I place the corporate margin and the quarter’s average sentiment reading within their own records. Subtracting the sentiment rank from the margin rank gives one common scale.

The spread reached 98 points in Q2 2026, the widest of 190 quarters since 1979. All 10 of the ten widest readings occurred in 2022 or later. That makes the divergence a multi-year condition. It also leaves us with one regime and no honest way to say what the combined signal predicts.

This measures profitability against perception. It does not show who received the economy’s income. Corporate margins are size-weighted aggregates; sentiment gives each surveyed household one response. A distribution claim would require compensation share, household income and ownership data. I am keeping that claim outside the evidence on this page.

What did very low sentiment mean for stocks?

The margin-sentiment gap has no independent historical sample. The older question we can grade is narrower: what happened after sentiment first entered the bottom tenth of its record?

The short horizon offers little comfort. The four episodes split evenly between gains and losses after three months, and their median return was +1.7% against a +3.1% all-month baseline. Twelve months changed the picture. All 4 episodes were positive, with a median +25.0% return against +12.3% normally.

Each episode begins with the first bottom-decile reading. Because the final survey arrives near month-end, the entry price is the following month’s average. Returns use Shiller’s monthly S&P composite price series and exclude dividends. The all-month medians use the same release-aware timing.

Four positive twelve-month outcomes sound stronger than they are. A random month in this record was positive after a year 79% of the time. If four observations were independent, four gains would occur about 38% of the time. My read is modest: deep pessimism did not identify the exact bottom, but it aged better over a year than over a quarter.

What would change my mind

• • The Q2 margin jump disappears. If the September 30 BEA release revises or replaces the latest ratio below 14.4%, profits are no longer outrunning value added in the way this article describes. Watch Corporate Profit Margin.
• • Labor pressure returns. If unit labor-cost growth rises above 3% while productivity growth falls below 1%, the contained-cost clue turns into a margin warning. The next BLS revision is September 3. Watch Unit Labor Costs and Labor Productivity.
• • The gap leaves its exceptional range. A spread below 79 points would take it out of the widest 5% of quarters. Either lower margins or better sentiment can get it there. Watch the two live pages above.
• • The next sentiment episode loses money over a year. That would break the current four-for-four record from first observable entry and make the twelve-month pattern less interesting.

How we checked it

The corporate margin divides adjusted after-tax profit by current-dollar gross value added for the same nonfinancial corporate sector. Keeping both sides in current dollars prevents inflation from lifting the ratio by itself. For the comparison chart, each quarter uses the average of its three monthly twelve-month sentiment readings. A percentile says where a value ranks in its own history; subtracting the two ranks puts otherwise incompatible series on one scale.

The stock study takes the first month sentiment entered the bottom 10% of its full record. Another episode begins only after twelve months without a qualifying reading. Returns start with the month after the survey release, avoiding a price from before readers knew the result, and every horizon had to be complete by August 26, 2026. The Shiller series is a price index, so dividends are absent. Four episodes leave plenty of uncertainty.