The Investment Boom the Import Surge Buried

The GDP Print Said 1.5 Percent. It Was 3.9 

Vaughn Cordle CFA 

 

 

 The strongest economy in years is hiding under a headline that made it look weak. An import surge buried it. Trump’s America First policies are driving the investment boom the headline hid. The factories are hiring, and the boom is reaching the workers who need it most.

The Print Said 1.5 Percent. It Was 3.9.

The GDP print for the second quarter said 1.5 percent. The economy was exceptionally strong. Real demand grew 3.9 percent. Consumer spending rose 3.2 percent. Nonresidential investment jumped 8.4 percent, equipment 15.2 percent, intellectual property 8.8. These are blowout numbers.

The mainstream media, the AI systems, and the administration’s critics read the weak headline and blamed Trump’s war with Iran. They still do.

Roughly half the country wants Trump and the Republicans gone, starting with the midterms. The half is a coalition of tribes: government unions, Marxists, socialists, the dependent class, woke progressives rich and poor, paid and unpaid anarchists. Many are useful idiots protesting in the street. They vote against him regardless of the facts. A weak GDP number is exactly what they want to hear. So when the print said 1.5, they believed it, spread it, and blamed his war.

They never looked underneath, because the surface serves their purpose. The print is technically true. Materially misleading. The strength did not vanish. It got buried.

An import surge buried it. The capex boom built the factories. The factories are hiring the workers. The boom is reaching the bottom of the K. This is hard evidence that Trump’s America First is working.

The Equation

Start with how GDP is built. It is a sum: consumer spending, plus business investment, plus government spending, plus exports, minus imports.

GDP = C + I + G + (X − M)

Each letter is a share of GDP. C is consumption, what households spend (68%). The consumer is the engine. I is investment, what businesses spend on equipment, buildings, and software (18%). G is government spending (17%). X is exports, what we sell abroad. M is imports, what we buy from abroad. Net exports — X minus M — runs negative (−3%), because America imports more than it sells.

Imports carry a minus sign. When an American business buys a German machine, that purchase already shows up in investment. GDP counts what America produces, so the imported machine gets backed out. Subtract it, and what remains is domestic production. That is the whole trick. When imports jump in a single quarter, the subtraction grows and the number falls, even when spending and investment are rising.

The Surge

Imports did not rise a little in the second quarter. They flooded in. Two forces drove the flood, and both were temporary.

The first was tariffs. In February the Supreme Court struck down the administration’s emergency tariffs and ordered refunds. The administration bridged the gap with temporary duties, then moved to replace them with more durable ones. Every importer in the country saw the same thing: a deadline, and higher costs on the far side of it. So they pulled orders forward. Buy now, before the tariff lands. The rush compressed months of imports into a single quarter.

The second was the AI build-out. Semiconductor gear and data-center equipment poured in from overseas to feed the construction of American computing capacity. Capital goods, arriving by the shipload.

Both forces did the same thing. They pushed imports up. And imports are subtracted.

America imports more than it exports, and has for decades. Net exports usually runs near minus 3% of GDP and sits quiet from quarter to quarter. A steady drag, not a swing factor. Tariffs break the pattern. When importers frontload ahead of the duties, imports spike, the trade gap widens, and growth takes a hit. Then the reverse: the tariffs land, the buying stops, imports fall back, and the gap snaps shut. A surge, then a payback.

That is the second quarter.

 

 

Imports grew two and a half times as fast as exports. The gap widened sharply, and that widening cut 1.01 points from the reported growth rate. Without it, the 1.5 would have been higher.

Net exports contribution to real GDP growth, by quarter

 

 

The chart shows how rare a swing like this is. Most quarters, net exports barely register, the steady minus-3% drag, quiet. The big moves come with tariffs. Early 2025, importers rushed goods in ahead of the first Trump tariffs. Net exports plunged 4.68 points, then snapped back 4.83 the next quarter when the buying stopped. The surge and the payback, side by side. The second quarter of 2026 is the same pattern, smaller: a 1.01-point drag from the same frontloading, ahead of the next round of tariffs.

The raw numbers show it plainly. When imports surged 38 percent in early 2025, GDP went negative, −0.6. When imports fell back, GDP jumped to 3.8. In the first two quarters of 2026, imports rose 11.8 and 11.5 percent. GDP printed 2.1 and 1.5.

The Distortion

Strip out the trade and the inventories — the two parts that swing on timing and tell you nothing about the health of the economy — and look at what Americans and American businesses actually bought and built. The government publishes that number for exactly this reason. It is called real final sales to private domestic purchasers. It grew 3.9 percent.

The headline growth rate came in 61.5 percent below the real rate of domestic demand.

Read that precisely. Not that GDP fell 61.5 percent. GDP grew. The reported growth rate landed 61.5 percent below the rate the underlying economy was actually running. The imports did not shrink the economy. They shrank the number that measures it.

Both numbers are annualized — scaled to a full-year pace, the standard way GDP is reported. That is the only basis on which they compare.

The Confirmation

One quarter’s quirk could be noise. It is not.

The Atlanta Fed’s GDPNow model, which tracks the quarter in real time as the data arrives, puts third-quarter growth at 6.2 percent, with business investment accelerating. That number owes nothing to the second-quarter argument. It is a separate reading, taken a separate way, and it points the same direction. Frontloading borrows from the future. Imports surge in one quarter and drag GDP down. The next quarter the surge ends, imports fall back to normal, and the drag flips to a boost. The low print and the high print that follows are two halves of one event.

The economy did not weaken and then recover. It ran strong the whole time. The trade timing moved the number around. The economy did not slow. The measurement did.

The Capex Boom

The strength has a name, and the government already measures it. It sits in gross private domestic investment, the “I” in the GDP equation, and it is running at a pace with no precedent.

Nonresidential investment grew 8.4 percent last quarter, double its long-run average. Equipment grew 15.2 percent, three and a half times its norm. These are the largest gains in six quarters, and they trace to one source: the buildout of artificial intelligence.

The five largest hyperscalers plan to spend roughly $750 billion on AI infrastructure in 2026, up 77 percent from a year earlier. And that understates the total. It leaves out SpaceX, newly public, which spent $15.8 billion on AI in a single quarter, more than twenty times its AI capex a year earlier, and guides toward $45 billion this year. It leaves out the $500 billion Stargate venture. It leaves out the pure-play AI firms and the sovereign funds. Goldman Sachs projects the four largest alone will spend $5.3 trillion on capital through 2030. Jensen Huang calls it the largest private infrastructure buildout in human history. He is not exaggerating.

This is what the companies report as capex and the national accounts record as investment. The same dollars, measured twice. When SpaceX reports $15.8 billion in AI spending, when Amazon guides to $200 billion, the BEA aggregates it into the equipment and nonresidential lines that ran triple their long-run pace. The earnings reports and the GDP data are describing one phenomenon from two directions.

Two forces drive it, and both are durable.

The first is demand. The AI race rewards compute, and being short on compute is the one mistake none of these companies can afford. The buildout is committed capital: chips ordered, power contracted, data centers under construction.

The second is policy. The 2025 tax and investment law made America the cheapest place to build. Accelerated depreciation lets a company write off new equipment now instead of over years. Lower rates leave more to invest. Lighter regulation clears the path.

Demand supplies the reason. Policy supplies the location and the incentive. Together they pull the investment onshore and forward.

Committed capex is the most forecastable part of GDP, because it is contracted in advance. The policy incentives run through the end of the president’s term in 2028. Goldman’s capital window runs to 2030. Between them, the investment tailwind is in the contracts and depreciation schedules, extending across the rest of the decade.

One thing could cut it short. If the returns on AI disappoint, the whole capex stack re-rates at once, and the committed spending gets trimmed. That is the risk. But the risk runs against a wall of contracted capital and a policy regime built to keep it building.

The capex boom is the real signal. The 1.5 percent headline buried it. The 3.9 percent underneath revealed it. The 6.2 percent that followed confirmed it.

The Two Misreads

The same number, read backward, misleads two audiences at once.

The investor. A portfolio manager reads 1.5 percent, concludes the economy is cooling, and positions for it. Defensive on stocks. Betting on rate cuts. Expecting the consumer to crack. Every one of those bets is wrong if domestic demand is running near 4 percent and the next quarter is printing 6. The consensus is leaning on a misread of the tape. The edge belongs to whoever reads the composition instead of the headline.

There is a second read, just as important as the first. If the second-quarter 1.5 was distorted low by the import surge, the third-quarter 6.2 is distorted high by the reversal. The frontloading that dragged one quarter down pays the next one back. Both prints are moved by the same trade swing, in opposite directions. The consensus that read 1.5 as weakness will read 6.2 as strength, and be wrong twice for the same reason. Neither number is the economy. The economy is the 3.9 running underneath both.

The voter. The story reaching the public is simpler and more useful to one side: the economy slowed, and the war with Iran slowed it. The weak print was not war damage. It was a trade artifact, one that grew out of the administration’s tariff policies, on a timeline that had nothing to do with the war. The war did leave a mark, but not the one they point to. It pushed up oil prices, and higher energy costs fall hardest on the households already stretched thin. That is the real war effect, and it lands on the bottom of the K, not on the headline number. The critics blame the war for the wrong weakness. The number was told true and read backward, and the backward reading ran one direction, sixty days before an election.

The Two Economies

At one end, Nvidia. Years of massive gains, a valuation stretched thin on the promise of the buildout, priced for a future that has to arrive. Even after a soft 2026, NVDA and the AI names sit on accumulated wealth that has made the asset-owning half richer. This is the wealth effect in a ticker: the people who hold these shares spend, and their spending carries the consumption that drives most of GDP.

At the other end, Tractor Supply. TSCO has lost 44 percent of its value in a year, a number anyone can check by typing the ticker. Its customers are the stretched half made visible: rural and large yard households, small acreage, pets to feed, mowers and fertilizer to buy, diesel trucks to fill. When oil and inflation climbed, those customers pulled back on exactly the big-ticket purchases the company sells, and the stock absorbed the blow. TSCO is the micro face of the macro squeeze.

Two tickers, one economy, opposite ends. The market poured capital into the AI story while the companies serving stretched families got crushed. That is the fault line under the boom. It is also why energy is the risk that matters most. A sustained spike above $100 oil falls hardest on the half already stretched thin.

The Path Down

There is relief on the horizon, and it runs through the same variable that poses the risk: oil.

Before the latest round of strikes, the Energy Information Administration projected oil prices easing back toward pre-war levels. Its July 7, 2026 forecast put WTI crude at $67 a barrel by December 2026 and $57 by the end of 2027. That path still holds if the Gulf settles. Lower oil feeds directly into lower inflation, because energy runs through the price of nearly everything.

Oil and inflation, forecast to fall together

Pre-escalation forecast. As oil eases from its 2026 conflict spike, inflation falls back toward target through 2027.

 

 

And lower inflation pulls long-term interest rates down with it. That matters for the households under the most pressure. Lower rates ease the cost of credit and mortgages. Lower energy costs ease the bill at the pump and the checkout. Together they loosen the squeeze on the half of the country the boom has not reached.

So the fault line is not fixed. The same energy variable that could deepen the divide can also close it. If oil falls, inflation falls, rates fall, and the relief reaches the households that need it. The strength at the top holds. The pressure at the bottom eases. The two economies move back toward one.

The Verdict

The economy underneath the 1.5 percent print grew 3.9 percent in the second quarter and, halfway through the third, tracks toward 6.2. Put that in context. Real GDP has averaged 2.13 percent since 2007, through a financial crisis, a pandemic, and a recovery. Anything above 2 is good. The 3.9 underneath this print is exceptional, nearly double the long-run norm.

The strength has a source. Lower taxes and accelerated write-offs on new equipment lift earnings. So does the rollback of excessive regulation, whose value to business may exceed the tax cuts themselves. Together they drive corporate revenue, earnings, and stock prices. Record markets and record real estate follow, and the wealth they create flows into the biggest driver of GDP: consumption, 68 percent of the economy. The top earners, riding record asset values, are spending, and their spending is powering the strength the headline hid.

Trump’s new tariffs, under Section 301, will slow the imports that dragged the last two prints. That reverses the drag mechanically: fewer imports, a smaller subtraction, a higher number. But the mechanical effect is the small part. The deeper effect is the incentive. Tariffs give foreign and domestic producers a financial reason to build here rather than import, and the capital-spending boom already underway is the early evidence. Equipment investment ran three and a half times its long-run pace last quarter, and the ISM manufacturing index hit a four-year high. Add accelerated depreciation and lower taxes, and the reshoring is already in the data.

The critics score tariffs as a cost and stop there. They never count the investment pulled onshore, the productive base rebuilt, the growth that compounds when production comes home. MAGA is the heart of Trump’s America First policies, and the logic is plain: to make America great again, he first has to make it strong again. Trade the short-run import drag for durable domestic investment, rebuild the productive base, and the economy comes out stronger. Strength first. That is the long game.

One risk overrides the tailwinds of Trump’s policies. The strength holds unless oil breaks above $100 and stays there through year-end, driving inflation expectations higher and pushing the 10-year Treasury yield toward 5 percent. On August 3, Trump warned Iran faces “decapitation” if it refuses a deal. A strike that escalates in the Gulf is the one event that could drive oil there and hold it. That is the real threat to the economy the headline hid, not the trade artifact the critics blamed, but a sustained energy shock. Watch oil, inflation, and the 10-year Treasury yield.

The print said 1.5 percent. It was 3.9. Strong, and getting stronger. But everything from here turns on how the war with Iran ends.