Despite a pullback in household formation in 2022, the U.S. multifamily market remains relatively healthy on the back of booming demand for apartments in 2021. Occupancies remain healthier than pre-pandemic levels, but the market faces the largest supply wave in modern history. The U.S. Census Bureau reported 926,000 units under construction at the end of 2022, which is higher than at any point since at least 1970, and twice as large as the wave that preceded the Great Recession. However, the lengthening construction pipeline has caused construction levels and deliveries to diverge significantly, giving the market more time to absorb the newly delivered units. This supply risk is not evenly distributed – the risk of overbuilding is just as much a function of demand as it is supply.
How Many Apartments Does the U.S. Need?
In the wake of the Great Recession, housing developers pulled back significantly. Total housing construction, including single-family and multifamily units, dipped to the lowest level since 1970. At the same time, demand—particularly for multifamily—surged as Millennials entered prime renting years. The supply-demand imbalance has been exacerbated over the years, with housing shortage estimates ranging from 1-6 million units, depending on methodology. With nearly 1 million apartment units in the pipeline today, and an additional 750,000 single-family units underway, developers may exceed the lower end of the estimate in the next few years.
Constraints on Delivering Units:
Today’s construction pipeline is different from previous years, potentially smoothing out the oncoming supply wave through the next few years. As cities continue to incentivize density, the average multifamily construction project has shifted in dramatic fashion. From 2000-2015, roughly two-thirds of construction was garden-style, and one-third was mid- or high-rise. Today, just one-third of all units under construction are garden-style. Both mid- and high-rise construction projects typically take longer to complete, meaning the level of today’s construction will likely be spread out across three years compared to a much shorter delivery period before the apartment construction boom accelerated. Indeed, according to a National Association of Home Builders’ analysis of census data, the average time to complete an apartment building has increased by 5.5 months between 2013-2021.1
Construction timelines are further elongated by delays for permits, material shortages and the availability of labor. The National Multifamily Housing Council’s Construction Quarterly Survey reveals ongoing challenges across the development landscape.2 Only 13% of respondents reported facing no delays during construction in December, up from a low of 3% in the June survey. Twice as many respondents saw labor as less available in December compared to three months prior than saw labor as more available. The survey also captured that, over the same time frame, construction prices increased for all components, except lumber. These factors contribute to a longer lead time between construction starts and the date of completion.
Construction financing costs have also increased dramatically as benchmark interest rates rose 425 basis points (bps) in 2022, with BBB corporate bond rates doubling during the year and nearly 70% of senior loan officers stating they were tightening lending standards for construction and development loans.3 Rising input and financing costs have put a significant strain on development pro formas, with some uncapitalized projects unlikely to move forward. Construction starts will also likely recede, even if permits increase. In turn, the outlook will feature completions peaking this year but deliveries declining quickly in 2024 and 2025. There is some risk to this outlook as well—estimated construction delivery dates are supplied by builders who are hoping to complete projects on time, but as projects stall or remain uncapitalized, these figures could fall in the months to come.
Construction Is a Local Phenomenon:
National construction figures, while helpful for context, mask the challenges facing some markets around the nation, as construction is inherently localized at market, submarket and micromarket levels. The national market has just under 5% of its inventory under construction today, whereas some markets have much larger pipelines in the ground. Sun Belt favorites, like Nashville and Charlotte, as well as Denver will face mounting supply-side pressure on fundamentals in the coming years, whereas markets like Indianapolis, Sacramento and Cleveland have much smaller construction waves.
Though some markets have robust pipelines, not all areas within the market will face the same challenges. In Nashville, for example, much of the multifamily construction is underway downtown, where more than half of the submarket’s inventory is being built. Other submarkets across Nashville have much smaller pipelines, which will likely result in more stable performance.
Overbuilding Is a Function of Demand:
During the Great Recession, Phoenix saw one of the steepest declines in occupancy, as a result of 15,000 units delivered from 2008-2009. In hindsight, the market appeared to be overbuilding into a recession, but most projects delivered in 2008 and 2009 likely started in 2006 and 2007. From 2005-2007, Phoenix’s population grew by 3.7% on average—faster than any major metro has grown since 2010. Had population growth remained robust, apartment demand would have largely kept pace with new deliveries.
Most markets with large pipelines today are high-growth markets that have benefitted from outsized in-migration in the wake of the pandemic. These markets find themselves in a similar position as Phoenix did prior to the Great Recession; if population growth and household formation pull back, vacancies would increase significantly, and rents would fall as properties compete for tenants. With the market still experiencing ripple effects from the pandemic, the outlook for demand remains relatively uncertain. Should demand return to pre-pandemic levels, the construction risk appears more measured than metro-level construction figures would imply.
The for-sale market has long reported data on months of supply on market, which measures how many months it would take, at the current rate of sales, for the available inventory of homes to be sold. Borrowing that concept, we can look at years of supply under construction, which measures how many years would it take for the market to absorb current construction levels. Demand outlooks vary across sources, given heightened uncertainty about factors such as job growth, homeownership trends and migration patterns—but this example shows current construction relative to the average annual demand tallied between 2017-2019 (roughly 310,000 units).
With this context, some markets shift dramatically. Nashville and Austin have two of the largest supply pipelines in the nation relative to their current inventory levels, but now rank 20th and 13th respectively, using this metric. Other markets that are typically slower growth, move up the risk spectrum. Sacramento and Detroit have some of the smallest construction pipelines in the nation, but because demand has not been as robust, the magnitude of the construction pipeline is greater.
Some metros, like South Florida, saw a significant uptick in migration in the wake of the pandemic, and this baseline scenario fails to encompass that acceleration. As such, the chart below encompasses a post-pandemic migration scenario, as well as a slowing demand scenario to reflect growing recession concerns. With demand growth unlikely to match recent post-pandemic peaks, the post-pandemic series reflects a longer time frame (2017-2022—roughly 365,000 units absorbed) to normalize some of the supersized growth seen immediately following the pandemic. Should post-pandemic migration patterns hold, the superlative growth in Florida should help keep demand growth more in line with supply.
Growing recession concerns, which typically cause housing demand to pull back, are also expressed in the graph, with demand reflective of 2012-2016 levels, or about 235,000 units annually. Some markets saw stronger demand early in the 2010s, indicating that the consistent acceleration in demand leading up to the pandemic was not uniformly distributed.
The housing shortage nationally is robust, but not every market has the same degree of shortage. Overlaying the “years of construction” concept with current stabilized vacancy relative to the historical average helps illuminate the degree of shortage across markets, as well as how quickly developers are moving to meet the housing need. While Miami’s supply pipeline is robust, the market also has one of the lowest vacancies relative to its long-term historical average, indicating a lower degree of risk than the aggregate pipeline would suggest.
Not All Without Risk
The constraints on developers have shifted the construction pipeline significantly, and issues are nearly distributed across the country uniformly. With projects taking longer to complete, supply-side risk should be considered over an even longer time horizon—not all 932,000 units will be delivered in 2023, or even in 2024. However, nearly 60% of markets have more than three years of supply underway, indicating that a near-term supply-demand imbalance is likely. With a possible recession, it remains unclear what the path of demand will be amid opposing headwinds and tailwinds. But markets in the Midwest highlight tighter-than-average vacancies and limited pipelines, and a handful of markets in California and Texas appear poised to weather weaker demand if it arises.
The resiliency of the multifamily market will be tested by a burgeoning supply pipeline, but after an unprecedented run-up in rents and occupancies, the market appears set to normalize over the next few years.
1 https://eyeonhousing.org/2022/07/slightly-longer-time-to-build-apartments-in-2021/
2 https://www.nmhc.org/research-insight/nmhc-construction-survey/quarterly-survey-of-apartment-construction-development-activity-december-2022/
3 https://www.federalreserve.gov/data/sloos/sloos-202301-chart-data.htm
