Twenty-seven minutes after this email lands, the Bureau of Economic Analysis is going to publish three things simultaneously: the second estimate of second-quarter GDP, the first read on corporate profits for that quarter, and the July personal income and outlays report with the PCE price index attached.
That's an unusual amount of information in one drop. It's also the release where the difference between reading a headline and reading a report costs you the most, because the headline for the second quarter is genuinely misleading, and it's misleading in a direction that would make you defensive at exactly the wrong time.
The advance estimate had the economy growing at a 1.5 percent annualized rate in the second quarter, down from 2.1 percent in the first, and below the roughly 2.1 percent forecasters expected. Every outlet ran some version of "growth slows." Fair enough as far as it goes.
Now look one line down. Real final sales to private domestic purchasers, which is consumer spending plus private fixed investment, grew at 3.9 percent. In the first quarter that same measure grew 1.7 percent. So the part of the economy that's actually made of American households buying things and American businesses investing in things more than doubled its growth rate in the exact quarter that headline GDP decelerated.
Both numbers come from the same report. Neither is wrong. They just measure different things, and only one of them is telling you about the underlying economy.
Here's the mechanical reason for the gap, and it's worth understanding properly because this same distortion will show up again.
Gross domestic product is designed to count domestic production. But the data the government collects is mostly spending data, and Americans spend money on foreign-made goods. So the accounting handles it by adding up all the spending and then subtracting imports at the end, because those imports were already counted in the consumption and investment lines and don't represent domestic output.
The consequence is that a quarter where businesses import heavily looks weak in headline GDP even when domestic demand is roaring. That's precisely what happened. Net trade subtracted a full percentage point from second-quarter growth, driven by strong imports as companies bought technology equipment and continued rewiring supply chains around a tariff landscape that now looks permanent rather than temporary. Inventories took another 0.7 points off. Real final sales, which strips inventories out, grew 2.2 percent.
Then there's the government line, which produced the single best example of why you should read footnotes. A large share of the reported decline in federal spending came from sales of oil out of the Strategic Petroleum Reserve. When the government sells oil, the accounting shows up as negative federal expenditure, but the oil gets bought by somebody who shows up as a positive somewhere else. The BEA states plainly in the release that this has no direct net effect on GDP. It's a bookkeeping entry, not a spending cut. Anyone who read that line as evidence of fiscal austerity read it wrong.
Strip all of it out and the picture is a private economy running considerably hotter than the headline suggests. Real consumer spending grew 3.2 percent. Durable goods outlays surged 6.9 percent, led by autos and furniture. Nondurables grew 4.4 percent. Services grew 2.2 percent, with recreation and food service leading. Business investment held up on equipment and intellectual property, tied heavily to artificial intelligence buildout.
Which brings up the honest caveat. A 6.9 percent quarter in durable goods is not a sustainable run rate. Durables are lumpy, they front-run price increases, and a tariff environment gives households a rational reason to buy the car and the couch now rather than later. That kind of pull-forward borrows from the second half. So the correct read isn't "the economy is stronger than you think, everything's fine." It's "the economy is stronger than the headline, and part of that strength was purchased from the future."
Now the inflation side, which is where Wednesday gets more interesting than usual.
The advance report showed the PCE price index rising 5.1 percent at an annual rate in the second quarter, up from 4.6 percent in the first. Core PCE, which excludes food and energy, rose 3.4 percent, down from 4.4 percent. The broader price index for gross domestic purchases jumped to 5.7 percent from 3.6 percent.
Headline accelerating while core decelerates tells you the pressure is concentrated in food and energy rather than in the sticky services categories the Fed cares most about. That's a meaningfully better composition than the reverse. It's also a distinction most people miss, because they see a five-handle on inflation and stop reading.
Here's the piece that trips up almost everyone, and it's worth ten seconds of your attention. The 5.1 percent figure is a quarterly rate annualized. It answers the question "if the last three months repeated for a full year, what would inflation be." The year-over-year figures, which is what you actually experience, are different numbers: June PCE ran 3.7 percent headline and 3.3 percent core against the prior year. Quarterly annualized rates are noisier by construction, and using them interchangeably with year-over-year rates is how people end up convinced inflation has doubled when it hasn't.
The July monthly data arriving Wednesday will update that year-over-year series, and core PCE has been grinding in a narrow band around 3.3 percent for months. Sticky, above target, not accelerating.
But the number I'd actually stare at on Wednesday isn't the price index. It's the personal saving rate.
In June it was 2.7 percent. That is very low by historical standards. It means American households are funding that impressive 3.2 percent consumption growth out of an extremely thin buffer, and a thin buffer is what turns an ordinary slowdown into a sharp one. Strong spending with a healthy saving rate is durable. Strong spending with a 2.7 percent saving rate is a high-wire act.
You can pull the saving rate series yourself from the Bureau of Economic Analysis in about thirty seconds, and I'd argue it's a more honest read on household health than any confidence survey.
The third item Wednesday, and the one that only appears in the second estimate, is corporate profits. The advance report doesn't include them. This is the first official look at what companies actually earned in the second quarter, on a national accounts basis rather than a reported-earnings basis.
That distinction matters. Reported earnings from public companies are shaped by share buybacks, accounting choices, and one-time items. The BEA's profits-from-current-production measure strips out inventory valuation and capital consumption distortions and gives you a cleaner read on whether the corporate sector is actually generating cash.
Profits are the input to almost everything you care about. Companies hire out of profits. They invest out of profits. Equity returns over any horizon longer than sentiment are ultimately a claim on profits. When profit growth stalls, hiring freezes tend to follow within two to three quarters, and capital expenditure plans get quietly trimmed before anyone announces anything. If you want an early read on whether the labor market is going to tighten against you, corporate profits are a better place to look than the unemployment rate, which tells you about a decision that already happened.
The last thing to understand about Wednesday is why a second estimate exists at all. The advance number is published roughly four weeks after the quarter ends, which means the BEA is working with incomplete source data and has to fill gaps with assumptions. The second estimate incorporates the Quarterly Services Survey, more complete international trade figures, and updated inventory data. Revisions are routine and occasionally large. There have been quarters in the last two years where a component moved by a full percentage point between estimates.
The practical implication is straightforward: the advance estimate is a draft, and treating a draft as a decision input is a mistake people make constantly. If a 1.5 percent print made you nervous four weeks ago, understand that the number you were reacting to was provisional and the underlying demand figure was 3.9 percent the whole time.
So what does this actually change for you?
Start tracking real final sales to private domestic purchasers instead of headline GDP. It's published in the same release, it's one line down, and it answers the question you're actually asking, which is whether American households and American businesses are still spending. Headline GDP answers a national accounting question that gets distorted by trade flows, inventory timing, and oil reserve transactions. If you only have room for one number, that's the one.
Use the saving rate as your fragility gauge, and then apply it to yourself rather than to the market. A household saving rate near 2.7 percent nationally doesn't tell you what to do with your equity allocation. It tells you what happens to consumer-facing businesses if income growth stumbles, and it tells you that your own buffer is your own responsibility because the aggregate one is thin. If you can't answer how many months of expenses you're currently carrying, that's a more urgent gap than anything in Wednesday's report.
Watch corporate profits as a leading indicator for your income, not for your portfolio. This is the reframe I'd push hardest. Most people read profit data as a stock market input, which it partly is, but the more actionable use is career and cash flow. Profit compression is the earliest reliable warning that hiring is about to slow in your industry. That's a signal to extend your cash runway and to have the conversation about compensation while the numbers still support it, not after.
And build the mechanics so that none of this requires you to act on release day. The reason people trade the headline is that their systems make trading easy and waiting hard. Reverse that. Automatic contributions on a fixed schedule and a target allocation that rebalances on rules rather than on impulse means a 1.5 percent GDP print does exactly nothing to your portfolio, which is the correct response. Setting that up once at a platform like M1 Finance takes an afternoon and removes the decision permanently.
Being able to see your own numbers is the other half. Aggregate data is only useful in contrast to your own position, and most people can't produce that position quickly. A free account at Empower connects your accounts and gives you net worth, allocation, and monthly cash flow in one view, which is what turns "GDP slowed" from an anxiety into an arithmetic problem you can dismiss in a minute.
For the release calendar itself, I run mine on autopilot. The BEA, BLS, and Conference Board all publish their schedules in advance, and pulling them into a single feed with alerts means I read the reports that matter on my schedule rather than discovering them through a headline. I use Make for that, and the whole workflow took under an hour to build.
Wednesday morning, someone will write that the economy grew more slowly than expected while inflation ran above target. That sentence will be accurate and will describe an economy that doesn't exist. The one that does exist had private domestic demand accelerating, core inflation cooling, households spending out of a dangerously thin cushion, and a corporate profit figure that nobody had seen yet.
Read the second line, not the first.
I put together a one-page decoder for the GDP release: which line items get distorted and why, how to convert quarterly annualized rates into something comparable to your lived experience, which components historically get revised the most between the advance and second estimates, and the four numbers worth pulling out of every quarterly report in under two minutes.
Reply with the word REVISION and I'll send it to you.
If this saved you from misreading a headline, pass it along. Three referrals unlocks the Money Systems Lab playbook library. Ten unlocks lifetime premium access.
Taylor Voss
Money Systems Lab