Lab · Review · Multifamily · 17 Nov 2026 · 6 min read

Multifamily distress by metro. In three of the twenty largest metros more than one rental home in nine stands empty, and in all three rents have stopped rising.

Loan stress on an apartment building starts with empty units and rents that stop rising. Open data does not report loan delinquency by metro, so this review reads the conditions that come before it. We set the Census Bureau’s rental vacancy rate for the first three quarters of 2026 beside the change in rent over the last twelve months and the apartment permits of 2022 to 2024, in the twenty metros with the most residents. Dallas-Fort Worth, Houston and Tampa stand apart. New York, Chicago and the California metros do not.

Cover photograph for the review
Rental vacancy, Dallas-Fort Worth, first three quarters of 2026Derived
12.7 %
the highest of twenty metros; the median of the twenty is 6.2 %, and Inland Empire is lowest at 4.4 %
Average of three quarterly readings, each with its own margin of error ·
Photo: 世品 苏 / Pexels
The argument in one paragraph

Averaged over the first three quarters of have the most rental homes standing empty of the twenty largest metros, against a . In all three the rent index moved . Add and there are five metros where vacancy is or more and rents are up or less on the year. Supply is part of the story: the ten metros that permitted the most apartments in 2022 to 2024 have a in the other ten. It is only part: .

Part 1 · Empty homes

Three metros have more than one rental home in nine standing empty, and rents there have stopped rising.

A lender sees distress late, when a payment is missed. An owner sees it early, in the units that do not fill. The Census Bureau’s housing vacancy survey reports the share of rental homes that are empty and for rent in the largest metros every quarter. One quarter in one metro is noisy, with a margin of error of one to five points, so we average the first three quarters of 2026. Dallas-Fort Worth is highest at , then Houston at and Tampa at . Atlanta is next at , and Detroit, Denver and Phoenix follow at . At the other end are Inland Empire at , San Diego at , and Los Angeles and San Francisco at . The median of the twenty is .

Then the rent. In the twelve months to September 2026 the rent index rose : the three emptiest metros are three of the six where rents moved least. Denver, at vacancy, is , and Phoenix, at , is . Those five are the metros where vacancy is .

Twenty metros: rental vacancy in 2026 and apartment permits in 2022 to 2024Sort by vacancy or by permits, then pick a metro
Phoenix
8.3 % vacancy, rent +1.0 % in the last year

Rental vacancy rate: 8.9 % in the first quarter of 2026, 8.0 % in the second, 8.1 % in the third; average 8.3 %, margin about 1.7 points. Rent index $1,726 in September 2024, $1,720 in September 2026 (−0.3 % on two years). Permits in buildings of five or more units, 2022 to 2024: 10.0 per 1,000 residents; 13,835 units in 2024, 14,800 in 2025.

Rental vacancy rate, average of the first, second and third quarters of 2026, and units permitted in buildings of five or more units in 2022 to 2024 per 1,000 residents (2024 estimate). Bars show vacancy or permits, whichever you sort by. The margin of error of each average is in the detail.

The same measure shows where little is wrong. New York, with a vacancy of and a margin of only on the average, is the most precisely measured metro of the twenty, and below the median. Chicago is at with rents , and San Francisco at with rents .

The ranking says where homes stand empty. The next question is whether new buildings explain it.

Part 2 · Supply

The ten metros that permitted the most apartments have a median vacancy of 7.4 %, against 5.6 % in the other ten.

The supply review of 14 November counted apartment permits in 2022 to 2024 as the best open read on the deliveries of 2024 to 2026. Set against vacancy, the ten metros with the most permits per residents, from Phoenix to Washington DC, have a . The other ten, from Houston to Detroit, have . Across all twenty the : more permits, more empty homes, but a weak link.

Two metros show why. Detroit permitted and has a vacancy of , with rents . Orlando permitted , the third most, and has a vacancy of . Houston permitted , eleventh of the twenty, and has the second highest vacancy. Vacancy also moves against rent: the , so the metros with the most empty homes are mostly the metros where rents have gone nowhere.

This is not a forecast. Permits in 2025 were higher than in 2024 in , the five metros above, which says the pipeline in the loose metros has not stopped. A building permitted in 2025 reaches tenants in 2027 or later.

What to take from twenty readings, and what to leave.

Part 3 · How to read it

A survey of households is not a rent roll, and a metro is not a building.

Four limits. First, the survey asks households about all rental homes, houses and condominiums as well as apartment buildings, so a metro rate is not an apartment vacancy rate. Second, every quarterly figure carries a margin of error, from to points here; averaging three quarters brings the margin on a metro to between , so a gap of a point or two between neighbours in the ranking is not a signal, and a gap of five points or more probably is. Third, the fourth quarter of 2026 is not out, and one quarter can jump: Denver reads . Fourth, vacancy is not delinquency. A building can be half empty and current on its loan, and a full building can be in default on a maturity.

For a lender, the practical question is whether the rent and occupancy in an underwriting are the metro’s or the submarket’s. For an owner, it is the gap between your own vacancy and a metro figure you cannot see inside. For a fund, it is whether a flat rent and a high vacancy in Dallas-Fort Worth, Houston or Tampa are in the assumptions for a building bought on last year’s growth.

What this page does not replace: the loan’s own payment record, a broker’s opinion of rent for the building, a licensed appraisal, a lender’s underwriting, a submarket occupancy check, or a site visit. A metro figure says nothing about one building.

Set the metro against the building in front of you.

What to do with this

If you own, lend on or buy apartments in one of the twenty: put the metro’s vacancy and rent change beside the occupancy and rent in your underwriting, and ask for the submarket, not the metro. In the five metros where vacancy is high and rents are flat, ask what the business plan assumes for rent growth in the next two years. A metro figure is a place to start, not an answer.

Sources

Found: from a dated document. Derived: calculated by us from found figures. Estimate: our judgment, with the method.

  • FoundRental vacancy rate, first three quarters of 2026, twenty metros. Dataset dis-vacancy-2026: twenty rows, one per metro. Table 4, rental vacancy rates for the 75 largest metropolitan areas, from the Housing Vacancy Survey of the federal statistical agency (Census Bureau), page revised 28 Oct 2026; the fourth quarter was not published. The metro names in the table follow the 2023 delineation. Read 17 Nov 2026.
  • FoundTypical asking rent index, September 2024, 2025 and 2026, twenty metros. Dataset dis-rent-index-2024-2026: the same twenty rows, as read for the supply review of 14 Nov 2026. A national rent index from an online listings company, metro series, single-family, condominium and multifamily rentals together, smoothed. Dollars are rounded; changes on this page are computed from the unrounded series: Dallas-Fort Worth +0.7 %, Houston +0.1 %, Tampa +0.4 %, Denver 0.0 %, Phoenix +1.0 %, Detroit +4.2 %, Chicago +4.7 %, San Francisco +11.8 % in the last twelve months.
  • FoundApartment-building permits, 2022 to 2025, twenty metros. Dataset dis-permits-2022-2025: the same twenty rows, as read for the supply review of 14 Nov 2026. Annual county files of the federal building permits survey, the column for buildings of five or more units, summed over the counties of each metro area as defined in the 2023 federal metro delineation. Permits per 1,000 residents use the 2024 population estimate.
  • DerivedThree-quarter average vacancy, its margin, and the median. Titleman calculation from card F1. The margin treats the three quarterly readings as independent, which the survey does not state; read it as a guide to size, not as a stated interval.
  • DerivedVacancy against permits and rent: medians and rank correlations. Titleman calculation from cards F1, F2 and F3, with the permits per 1,000 residents of the supply review. Medians are unweighted. Twenty observations show an association, not a cause.
  • DerivedFive metros with high vacancy and flat rents; permits in 2025. Titleman calculation from cards F1, F2 and F3. The cut-offs, 8 % and 1.5 %, are ours; a count, not a statistical test.
  • EstimateHow the average was taken and what the survey leaves out. Titleman note on scope. Ask us for the working: alexgalt@titleman.ai.

Figures are as stated in three federal and market series read on 17 November 2026 (the rental vacancy rate for the first three quarters of 2026; permit counts by county, summed to metro areas, and the 2024 population estimate; a national rent index through September 2026), and Titleman calculations from them. Medians are unweighted and cover the twenty metros only. Titleman’s opinion, not investment or legal advice.