Rents have stopped sliding. Apartment List's National Rent Report for September 2026, published this week, puts the national median rent at $1,388, down 0.1 percent on the month and down 0.4 percent from a year ago. That annual figure bottomed at minus 1.6 percent in April, so the trend line has been grinding back toward zero for five months. Apartment List also has the national multifamily vacancy rate at 7 percent in September, down from its peak earlier this year, and units leasing in an average of 34 days.

Meanwhile the for-sale side moved the other direction. The National Association of Realtors reported August 2026 existing-home sales of 3.98 million with 1.62 million homes on the market — 4.9 months of supply, which NAR called the highest in more than ten years. Two exits, two directions. If you are buying this month, that split should decide which exit you underwrite as primary and which one you treat as the fallback.

What the rent numbers actually say

They say stabilization, not recovery. A median rent that is still negative year over year is not rent growth. It is the end of the bleeding. Vacancy at 7 percent remains high against long-run norms, and 34 days on market is slower than is typical for early autumn. What changed is direction and the supply picture behind it: fewer new units are delivering into the market than in 2024 and 2025, which is what pulled vacancy off its high.

Single-family tells a similar story with better numbers. Cotality's Single-Family Rent Index had detached and attached rents up about 1.5 percent year over year in June 2026, with the Midwest leading — Chicago at 5.0 percent, Detroit 3.4 percent, Philadelphia 3.2 percent. The split by price tier matters more for most investors: high-end rents were up 2.4 percent while low-end rents rose 0.4 percent. If your rental strategy is workforce housing in a Sun Belt metro, the national average is not your market. Pull comparable leases signed in the last 60 days and use those.

Collections are improving, slowly

Chandan Economics' Independent Landlord Rental Performance Report for September 2026 put on-time rent payments at independently operated properties at 83.2 percent, a 91 basis point improvement from a year earlier and the strongest year-over-year gain in more than three years. The forecast full-payment rate, which accounts for payments that arrive late, was 96.2 percent for September, up from 95.6 percent in August.

Read both numbers together. Roughly one in six tenants at small properties is paying late, but most of them eventually pay. That is a timing problem, not a loss problem, and timing problems are what kill leveraged deals. If you are underwriting a two to four unit with a monthly debt service payment, the relevant assumption is not 5 percent vacancy and 100 percent collection. It is that some months arrive short and get made whole later. Reserves cover the gap. Optimism does not.

The resale exit got harder to price

NAR's 4.9 months of supply in August 2026 means buyers can negotiate and wait. The new-construction side is worse. The Census Bureau and HUD reported new-home sales at a seasonally adjusted annual rate of 684,000 in August 2026, with 483,000 homes for sale — 8.5 months of supply. The average new-home sale price was $478,700, down 8.8 percent from a year earlier, and Census flagged that price drop as the one statistically significant change in the report.

Builders discounting by that much are competing directly with your renovated flip in the same price band, and they can buy down a rate in a way you cannot. If your ARV comps lean on sales from the first half of the year, they are stale.

What this changes in the underwrite

• Price the rental fallback before you buy, not after the listing goes stale. With rents flat to slightly negative nationally and vacancy still at 7 percent, a hold is survivable but not free. Run the rent number at today's signed leases, not at a forecast.
• Discount forecast rent growth heavily. Zillow's September 2026 forecast raised its fourth-quarter multifamily rent growth call to 2.1 percent year over year, after starting the year expecting 0.3 percent. That is a large revision in nine months, in the friendly direction. It could revise back. A deal that only works on year-two rent bumps is not a deal.
• Budget more vacancy days than last year. Thirty-four days to lease is a national average that includes tight Midwest markets. Two months of carry on a turn is a defensible assumption in an oversupplied metro.
• Extend the sale timeline in the carry model. At 4.9 months of supply, the difference between a 60-day and a 120-day marketing period is real money on a short-term loan. Model the longer one and see if the deal still clears.
• Stress the debt service coverage at flat rent. Rent growth near one percent a year does not rescue a property that only covers its payment on paper at origination. If coverage is thin at today's rent, it is thin, and the rent line will not bail it out in twelve months.

The practical read

The rental side is the more predictable of the two exits this month. That is a low bar. It is predictable because rents are roughly flat and vacancy is improving slowly, not because cash flow got better. The resale side carries the bigger price risk right now, with a decade-high supply of existing homes and builders cutting new-home prices by high single digits.

The deals that work in that environment are bought at a basis that survives both. If a property only pencils on a fast retail sale at a spring comp, it is a bet on the softer of the two markets. If it also pencils as a rental at a September lease rate with a late-paying tenant and two months of vacancy, it is a deal.

Figures come from Apartment List's National Rent Report for September 2026 published this week, Chandan Economics' Independent Landlord Rental Performance Report for September 2026, Cotality's Single-Family Rent Index for June 2026, and August 2026 releases from the National Association of Realtors and the Census Bureau.