The Paycheck Rose. The Economy Split.

More income reached the household. Almost none reached real consumption.

In July, personal income rose 0.4%, and disposable income after taxes rose 0.5%. Remove the month's price increase and real disposable income still rose 0.4%. The puzzle starts one step later: real consumer spending increased less than 0.1%.

That gap is the split. The household had more purchasing power and chose not to turn it into more goods and services. The saving rate edged up to 3.0%. That is not a comfortable buffer by historical standards, but it suggests some of the extra income stopped before the checkout line.

The split inside the cart matters too. Current-dollar services spending rose by $86.2 billion, while goods spending fell by $49.9 billion. Rent, health care, travel, insurance, and other services can be sticky or difficult to postpone. A household that pays more for those obligations has less room for another appliance or vehicle.

The constructive reading is restraint. Income rose, employment losses stayed low, and households rebuilt a little saving instead of chasing prices. The harder reading is that prices and borrowing costs are already forcing choices even before the labor market breaks. July alone cannot decide between them.

The business side refused to confirm a broad slowdown.

The second estimate of gross domestic product made the contrast sharper. The headline economy grew at a 1.5% annual rate in the second quarter. But real final sales to private domestic purchasers grew 4.2%.

That ungainly phrase has a simple job. It strips out inventories, trade, and government to show what U.S. households and businesses actually bought for final use. The measure says underlying private demand was moving faster than the headline economy.

Profits pointed the same way. Profits from current production rose by a $400.9 billion annual-rate amount in the second quarter, far more than in the first. In July, new orders for durable goods rose 1.1%, including a 0.4% increase excluding transportation. Companies were still placing orders for items meant to last.

Labor did not validate a recession call either. Initial unemployment claims were 206,000 in the week ending August 15, and their four-week average was 204,000. These are weekly administrative counts, not a promise about future hiring. They do show that a wave of job loss had not arrived.

Housing shows why income alone is not enough.

A home is the purchase where the overnight policy rate becomes a monthly household bill. Buyers borrow for decades, so the long rate matters. The average 30-year fixed mortgage rate was 6.65% on August 20. At that cost, a raise can disappear inside the payment before it improves the house.

New-home sales fell 10.5% in July to a 607,000 annual rate. The monthly estimate is volatile and carries a wide published margin of error, so the exact decline should not be treated as precise. The accompanying inventory evidence is harder to wave away: 488,000 new homes were available, equal to 9.6 months of supply at July's sales pace.

This is the transmission path in ordinary language. Builders and sellers set prices. Lenders attach a long interest rate. The household combines the price, rate, down payment, insurance, and tax into one monthly obligation. If the payment fails the budget, the paycheck stops before the closing table. The unsold house stays in inventory.

That pressure can remain contained. Builders can offer incentives, prices can adjust, and lower inflation can eventually pull long rates down. It becomes a broader warning if inventory keeps rising, construction slows, and labor income follows housing lower.

Investors found risk appetite, not proof of easy money.

The five largest non-stablecoin crypto assets were all higher over the seven days to the cutoff: Bitcoin, Ether, BNB, XRP, and Solana. That is useful confirmation that investors were willing to own risk. It does not identify why.

The bond market supplied a separate check. Treasury's August 25 two-year note auction sold $69 billion at a 4.204% high yield and attracted $2.60 of bids for every dollar offered. The August 26 five-year auction sold $70 billion at 4.393%, with $2.37 of bids for every dollar offered. Auction clearing means investors accepted the new debt at a price and yield; it does not mean borrowing costs became easy.

The previous Macro TLDR followed Treasury's decision to at least double the maximum size of long-end liquidity-support buybacks. That change remains real, but the larger operations do not begin until September 9. No new reserves were created by the announcement. This week's risk rally therefore cannot be used as proof that Treasury delivered quantitative easing.

Federal Reserve minutes explain the bind. Nine voters held the policy rate at 3.5% to 3.75%, while three preferred a quarter-point increase. Large firms still had accommodative financing; smaller firms and households faced tighter conditions. The official record describes the same split visible in income, profits, and housing.

Five tests can turn a July pause into a regime.

A split economy is a diagnosis with an expiration date. Business strength can pull the household side forward, or household weakness can spread back through orders, construction, and jobs.

The paycheck keeps the test simple. If it reaches the cart and the closing table again, firms had enough momentum to bridge the pause. If it keeps stopping short while profits and quarterly demand fade, July was the early warning.