Founder's Note
Happy Labor Day. No Fed speech to parse this week, and I am keeping this one short so you can get back to the grill. Labor Day is the unofficial halfway point between summer and the year-end close, so I want to use it the way I use it in my own portfolio: as a checkpoint. Where is the market actually sitting, and what has to happen between now and December for the deals on this site to do what their pro formas say?
Where we are: jobs are fine, better than fine on Friday, with 162,000 added in August against a forecast near 55,000 and unemployment steady at 4.1%. Rates are not fine. The ten-year touched 4.81% this week and the 30-year mortgage sits at 6.71%, a 13-month high. Wages are up 3.1% while prices are up 3.4%, so the household paying rent is employed and a little squeezed. And nobody can find a framer, because construction unemployment just hit 3.1%, the lowest ever recorded.
Where we are going: four months left, and three things that matter. The August inflation print on Thursday, September 10, the Fed on September 16, and the fourth-quarter apartment deliveries that will tell us whether the supply wave really has crested. My read is that the rest of 2026 rewards owners of existing, well-leased buildings in metros where the paycheck is steady, and punishes anyone whose plan needs a rate cut or a lease-up to work. Enjoy the day off.
This Week in Markets
1. August hiring came in at triple the forecast. Wages still trail prices
Nonfarm payrolls rose 162,000 in August, well past the roughly 55,000 economists expected, and the unemployment rate stayed at 4.1% with 7.0 million unemployed. June was revised up to 31,000 and July from a loss of 23,000 to a gain of 21,000, so the three-month average now sits near 71,000 a month, up from 38,000. The gains were concentrated: food services and drinking places added 59,000, local government education 42,000, construction 22,000, manufacturing 16,000 and health care 13,000, while information shed 23,000. Participation ticked up to 61.6%. Average hourly earnings reached $37.75, up 0.3% on the month and 3.1% over the year, against a July inflation reading of 3.4%. The August inflation print arrives Thursday.
Why it matters: Your tenants are employed. They are also, on average, a little poorer in real terms than a year ago. That combination favors the unit a working household can afford over the one it aspires to, and it favors metros where the biggest employers do not run layoffs.
2. Mortgage rates hit a 13-month high and the ten-year touched 4.81%
Freddie Mac put the 30-year fixed at 6.71% for the week ending September 3, up from 6.66% and the highest in 13 months, with the 15-year at 6.04%. The ten-year Treasury climbed to 4.81% midweek, its highest since October 2023, and closed Friday at 4.78%. The two-year finished at 4.37% and the thirty-year at 5.24%. On Thursday Governor Christopher Waller said he would support holding the policy rate in September if the August inflation data show continued progress toward 2%, and would consider a hike if they come in hot. Futures had a hike priced somewhere between a coin flip and 60% by Friday's close. The decision lands September 16.
Why it matters: Nothing in Friday's jobs number makes borrowing cheaper. The long end is pricing a 4.8% risk-free rate, the mortgage market is pricing 6.7%, and both sit on the wrong side of most 2024 and 2025 underwriting. If a deal you are reviewing shows financing under 6.5%, that is the line to ask about.
3. Construction unemployment fell to 3.1%, the lowest on record
Construction added 22,000 jobs in August and the industry's unemployment rate fell to 3.1%, an all-time low, according to the Associated General Contractors' read of the government data, with member firms reporting they still cannot fill open positions. The Fed's Beige Book, published Wednesday, said the same thing in its own language: "skilled trades and technical workers were difficult to find," "significant wage increases were most often connected to demand for skilled workers in construction and manufacturing," and "residential construction declined while nonresidential construction increased on balance, with some Districts noting a high concentration of activity related to data center projects." Industry forecasts put 2026 multifamily starts near 225,000 units, the lowest since 2012 and roughly half the 2023 peak.
Why it matters: The apartment supply cliff I wrote about two weeks ago is not only a cost-of-capital story. The crews that would build the next wave are pouring slabs for data centers at wages apartments cannot match. An existing 1990s community cannot be rebuilt at 2026 labor cost, and that is most of the case for owning one.
The Take
Every pro forma has a line called rent growth, and almost none has a line called employer. That is backwards. Rent growth is the output. The input is a household that gets paid on the first and the fifteenth, by someone who is not about to stop paying them, with enough left after everything else to cover the unit. Friday's report described that household reasonably well: employed, more likely to have found work than it looked in July, and earning 3.1% more than a year ago while prices rose 3.4%.
Labor Day is a good day to say the obvious part out loud. The tenant is a worker. When I look at an apartment deal now, the first thing I want is not the rent comp, it is the employer list, and specifically how many of those employers have ever laid anyone off in a downturn. A state government, a university with 50,000 students and a regional hospital system do not run layoffs because the ten-year hit 4.8%. A metro built on those three tends to carry a lower unemployment rate than the country in good years and a much lower one in bad years, and it is the bad years that decide whether a five year hold reaches its exit.
One. The paycheck is stable, but it is not growing in real terms. 3.1% wage growth against 3.4% inflation means the household has less room, not more. Underwrite the unit they can afford at today's income, not the one the rent-growth line says they will afford in year three.
Two. The people who would build the competition are booked. A 3.1% construction unemployment rate and a Beige Book full of data center projects mean the 1990s garden community down the road is not getting a new neighbor soon. Replacement cost is a real moat right now.
Three. Ask for the employer list. If the sponsor cannot tell you the top five employers in the submarket and what share of the workforce sits in government, education and health care, they have underwritten the rent and not the renter.
None of this is an argument that jobs data replace rate data. The ten-year at 4.8% and the 30-year mortgage at 6.71% are still the numbers that set what a buyer can pay, and a hike on the 16th would push both. But rates decide the price of an asset. Employment decides whether it earns anything while you own it, and the second question is the one a five year hold actually lives with.
So the checkpoint reads like this: the paycheck is holding, the cost of money is not helping, and the builders are busy elsewhere. That is a good year-end setup for boring, well-leased, existing product and a hard one for anything that needs the market to improve. Target returns are still targets, any single deal can still go wrong, and a mostly occupied building can still lose tenants. It is just a lot harder to lose them when their employer is a state government.