The U.S. Economy’s 1.5% Growth Problem Is Bigger Than the Headline
Second-quarter growth slowed while core inflation stayed at 3.3%. The overlooked issue is that AI investment is propping up demand—but imported hardware is masking its impact in GDP.
The number that matters is not 1.5%—it is the combination
The U.S. economy grew at a 1.5% annualized rate in the second quarter of 2026, down from 2.1% in the first quarter and below the pace economists had expected. On its own, that is not a recession signal. It is a slowdown.
But the market-relevant fact is that the slowdown arrived without the clean inflation relief policymakers need. June’s core PCE price index—the Federal Reserve’s preferred underlying inflation gauge—was up 3.3% from a year earlier. Headline PCE inflation was 3.7%.
That is the uncomfortable mix now defining the macro outlook: slower real growth, still-elevated inflation, and an economy increasingly dependent on a narrow capital-spending boom centered on AI infrastructure.
I would not call this stagflation. That label is too blunt and too dramatic for an economy still expanding and a consumer still spending. But I would call it a constraint economy: one in which growth is too soft to celebrate, inflation is too high to declare victory, and interest-rate relief remains difficult to justify.
For markets, this is a much harder backdrop than the old soft-landing script. Investors have spent much of the past year looking for a combination of durable growth, cooling prices and eventual monetary easing. The latest data says the first condition is weakening while the second remains incomplete.
GDP slowed, but the consumer did not disappear
The first instinct after a 1.5% GDP print is to assume households are pulling back. The underlying data does not support that simple conclusion.
Personal consumption expenditures rose 0.3% in June in nominal terms, while real consumer spending increased 0.4% for the month. Disposable personal income rose 0.2%, and the personal saving rate fell to 2.7% from 3.0% in May.
That tells a more complicated story. Consumers are still spending, but the cushion is thinner. A lower savings rate can support demand in the near term; it cannot be treated as a permanent growth engine. If income growth stays modest and households continue financing consumption by saving less, the consumer will become more vulnerable to any fresh rise in energy, food, borrowing or housing costs.
The second-quarter GDP number therefore should not be read as a collapse in demand. It is better understood as evidence that the economy is losing momentum even before the consumer has clearly broken.
That distinction matters for operators. If you run a consumer-facing business, the data does not argue for panic cuts. It argues for tighter segmentation. Higher-income households, asset owners and customers tied to strong labor markets may continue spending. More rate-sensitive and budget-constrained customers are likely to become increasingly selective.
The broad consumer is holding up. The marginal consumer is becoming more expensive to serve and harder to retain.
Inflation improved month to month—but the Federal Reserve cannot ignore the annual trend
There was genuinely constructive news in June’s inflation report. The overall PCE price index fell 0.1% from May, while core PCE rose only 0.1% for the month. That is a far better near-term rhythm than the year-over-year figures alone suggest.
Still, the annual measures remain uncomfortably high. Headline PCE inflation at 3.7% and core PCE at 3.3% are well above the Federal Reserve’s 2% objective. The central bank can take comfort from a softer monthly reading, but it cannot confidently declare that price stability has been restored.
That creates a policy asymmetry. A weak GDP headline raises the case for easier policy. Persistent core inflation raises the cost of easing too early. When those forces collide, central bankers usually choose patience—particularly when the labor market and consumer spending have not clearly deteriorated.
This is why the market should resist treating slower growth as automatically bullish. In the old playbook, weaker activity meant lower yields, faster rate-cut expectations and higher valuations for long-duration equities. In this playbook, weaker growth can also mean lower earnings expectations while inflation prevents the policy response investors want.
The result is not necessarily a broad market selloff. It is more likely to be greater dispersion: companies with pricing power, reliable cash flow and low refinancing needs will be valued differently from companies that depend on cheap capital, aggressive multiple expansion or a rapid return to lower rates.
AI is supporting the economy—and distorting the way we read it
The overlooked angle in this report is not simply that AI investment is large. Everyone knows that. It is that AI’s role in the economy is increasingly difficult to see clearly in the headline GDP number.
Federal Reserve economists estimated that AI-related investment—including software, data centers, power infrastructure and computing equipment—added about 0.73 percentage point to first-quarter GDP growth. Yet the same investment surge is heavily dependent on imported chips, servers, networking equipment and related hardware. Those imports subtract from GDP under the national-accounts framework.
Axios highlighted the scale of that effect: net imports of computers, peripherals and parts subtracted 0.45 percentage point from first-quarter growth. AI-related products accounted for 23% of all U.S. imports in 2025, up from 15% in 2023, according to research cited by the publication.
This is not a statistical glitch. It is how GDP is designed. GDP measures domestic production, not simply domestic demand or corporate investment. If an American company builds a data center using imported servers, the construction, installation, software work and domestic services count positively. The imported hardware offsets part of the investment surge in the trade calculation.
That means a lower GDP number can understate the intensity of domestic capital spending. It can also overstate the degree to which the U.S. is converting its AI buildout into domestic production.
This is the central macro tension: AI is giving the economy a real investment impulse, but much of the physical equipment is sourced abroad. America gets the deployment, some of the high-value software and services, and much of the financial upside. Other countries capture a substantial share of the manufacturing activity embedded in the buildout.
For investors, that distinction matters. A GDP slowdown does not necessarily mean the AI capex cycle is fading. It may mean the cycle is being reflected more in import growth and corporate orders than in a clean, headline GDP acceleration.
The contrarian view: slower GDP may be less alarming than the composition of growth
My contrarian take is that the 1.5% figure is not the most worrying part of this release.
A moderate growth rate can be sustainable if it comes with broad productivity gains, healthy household balance sheets and lower inflation. The bigger concern is concentration. The economy’s most visible investment engine is a capital-intensive AI buildout led by a relatively small group of hyperscalers, chip firms, data-center developers, utilities and equipment suppliers.
That is powerful, but it is not the same as broad-based expansion.
If consumer demand weakens, housing remains constrained by high financing costs, and smaller businesses continue to face expensive credit, then a handful of giant balance sheets can keep investment data respectable while the rest of the economy feels much softer. That is a classic recipe for a widening gap between market performance and lived economic conditions.
There is another risk: the AI investment cycle is unusually import-heavy and unusually energy-intensive. The former complicates GDP measurement. The latter creates exposure to electricity costs, grid bottlenecks and energy-price shocks. A surge in AI spending can boost growth and corporate revenue while also adding pressure to the real-world inputs that shape inflation.
That is why the market should stop asking only whether AI is productive. The more immediate question is whether the investment boom can broaden into domestic supply, software productivity and new business formation before the cost pressures become binding.
What this means for you
For investors, this is not a moment to make an all-or-nothing call on either recession or reflation. It is a moment to prioritize balance-sheet strength, pricing power and valuation discipline. A slower economy with sticky inflation is hostile to businesses that need lower rates to justify their multiples or refinance their debt.
For operators, plan around rates staying restrictive longer than your optimistic case assumes. Audit customer sensitivity, borrowing costs, supplier exposure and energy inputs. The businesses that win in this environment will not be those waiting for macro relief; they will be those redesigning operations for a world in which capital remains expensive and demand is uneven.
For policymakers, the message is equally direct. The country cannot rely indefinitely on importing the physical foundation of its most important investment boom. If AI is to become a durable national growth story rather than a narrow market story, the United States needs more domestic capacity in power equipment, grid infrastructure, advanced manufacturing and the industrial supply chain around data centers.
The second-quarter data does not say the expansion is over. It says the margin for error is shrinking. Growth is slowing, inflation is still above target, and the most important investment boom in the economy is harder to read through conventional statistics than many investors realize. That combination deserves more attention than a single GDP headline ever can.