We had said that our previous note would be the last one of the year. But writing this from the mountains in Greece during the holidays, we felt it was worth making an exception. Not to revise a call, and certainly not to defend one, but to better understand what the data are actually telling us. Over the past days, our recession view was challenged, constructively, and that challenge is useful. Macroeconomics is not about winning a debate; it is about understanding the mechanisms at work.
The latest GDP release does not show a recession. That statement is factually correct, and analytically insufficient.
In the third quarter of 2025, real GDP increased at a 4.3 percent annualized rate, equivalent to roughly 2.3 percent year over year. At the same time, real gross domestic income increased by only about 0.6 percent year over year. This divergence between output and income is the central macro fact of the current cycle. The economy is still producing measured activity, but it is no longer generating commensurate income. Understanding where this growth comes from matters more than the headline number itself. In the Q3 GDP report, growth was driven by consumption, government spending, and net exports, while gross private investment made a broadly flat contribution. At this stage of the cycle, a flat investment contribution is not neutral — it is a warning.
Investment: the cycle is already turning
While gross private investment was flat in aggregate, non-residential investment has been contracting for three consecutive quarters. This is not noise. It is a trend.
Within non-residential investment, structures investment is now contracting sharply, after growing strongly throughout 2023 and 2024. Structures investment leads the cycle. When firms stop committing to new plants, buildings, and large projects, it signals that expected future demand is weakening.
Equipment investment does not turn independently. It follows structures with a lag. The current configuration — contracting structures, flat headline investment, and still-positive equipment — is therefore transitional, not stable. It strongly suggests that equipment investment will weaken next.
Consumption: necessity, not strength
Real personal consumption expenditures added about 2.3 percentage points to annualized GDP growth. But around 1.74 percentage points of that came from services, and more than 70 percent of services growth was concentrated in healthcare services and “other services.” This has been the case throughout the year.
This is not genuine cyclical consumption growth. It is necessity-driven spending. Healthcare and related services are largely price-inelastic and policy-driven. They do not signal household confidence or income strength; they signal cost pressures and institutional rigidities. Once this composition is taken into account, underlying discretionary consumption looks far weaker than the headline suggests.
Fiscal and trade distortions
Government consumption and investment added roughly 0.5 percentage points to GDP growth. This contribution is unlikely to be repeated. The federal deficit is already near its effective limits, and sustaining growth via fiscal impulse would require an even larger deficit.
Net exports also contributed positively, not because exports surged, but because imports fell sharply. This decline is closely linked to tariffs and trade frictions. A tariff-driven fall in imports mechanically raises GDP, even when it reflects weaker domestic demand.
One engine only: wealth and dissaving
Taken together, the GDP report describes an economy that is still growing on paper, but not growing on income, not growing on discretionary demand, and not growing on investment momentum.
The U.S. economy is therefore running on one engine only: the wealth and balance-sheet capacity of the elderly and the top 30 to 35 percent of households. The bottom half of the income distribution or the young have little or no excess savings left and are already constrained to necessities.
This is now visible directly in the saving data. Over the past year, the personal saving rate has fallen by about 1.2 percentage points, from roughly 5.4 percent to 4.2 percent. Given where savings are distributed, this decline can only be coming from the top of the income and wealth distribution.
Lower savings and higher leverage are not two different mechanisms; they are the same balance-sheet adjustment viewed from opposite sides. The negative relationship between household net worth and the saving rate reinforces the conclusion that the economy is currently being supported by wealth effects, not income fundamentals. An economy that requires a one-percentage-point-plus drop in the saving rate in a single year to sustain growth is not expanding, it is borrowing time.
Labour and production
At the same time, labour and production data already display recessionary characteristics. Hours worked are weakening, hiring momentum has slowed, job quality is deteriorating, and parts of industrial production are flat to declining. These indicators typically turn before headline GDP does.
This leads to a final clarification. A recession is not defined by two consecutive quarters of negative GDP growth. It is defined by sustained weakness in income, labour, production, and real purchasing power. On those dimensions, the economy already looks late-cycle and fragile.
GDP has not yet confirmed a recession. However, income, investment, labour, and production point in that direction.
Happy New Year.
Andre Chelhot, CFA
Editor,
The Macro Anchor