Fears are rising about regional banks’ exposure to commercial real estate loans. How worried should you be?
Joe Seydl, Senior Markets Economist
Jay Serpe, Global Head of Alternative Investments, Strategy & Business Development
Ryan Asato, Manager Solutions, Alternative Credit
Renewed stress in the banking industry has a way of focusing the mind.
Since March, when the abrupt and unexpected failure of two regional U.S. banks, Silicon Valley Bank and Signature Bank, rocked U.S. markets, investors have begun voicing concerns about the balance sheet vulnerabilities of regional banks—and more specifically, their exposure to commercial real estate (CRE).
Those concerns have some merit. Compared to big banks, small banks hold 4.4 times more exposure to U.S. CRE loans than their larger peers. Within that cohort of small banks, CRE loans make up 28.7% of assets, compared with only 6.5% at big banks.1 More worrying, a significant percentage of those loans will require refinancing in the coming years, exacerbating difficulties for borrowers in a rising rate environment.
Just how worried should you be? To help cut through the noise, we delve into commercial real estate to explain ongoing industry trends and how we expect them to impact different subsectors in the quarters and months ahead. Some of those shifts, such as the widespread adoption of hybrid work patterns post COVID-19, are still creating powerful headwinds for the office sector, which is particularly—although not uniformly—vulnerable. Other trends, such as the rise of e-commerce, continue to provide tailwinds for industrial assets, including logistics properties and warehouses. Reshoring—the process of bringing production and manufacturing of domestic goods back into their country of origin—could also benefit the industrial sector.
Here’s our main takeaway: The office sector faces its own particular set of challenges. Many other sectors within CRE have stronger fundamentals. Nor do we believe that potential losses within the office sector are likely to destabilize regional banks. From both a GDP and wealth perspective, the office sector represents a small part of the overall economy.
That said, the small bank lending channel more generally does represent a macro risk, as tighter lending standards and profitability challenges in the banking sector could reduce available financing and raise the cost for small and medium sized businesses. But it’s difficult to quantify this risk precisely, and the uncertainty surrounding potential offsets is high.
Worsening conditions challenge the office sector
Make no mistake: Rising rates are piling pressure on the troubled office sector. Vacancies soared in the early stages of the pandemic and have kept rising since. Today, as shown in the chart below, the office vacancy rate of 12.5% is comparable to where it was in 2010, one year out from the global financial crisis (GFC). Office sales volume is now approaching its post-GFC lows.
Office vacancies rose when the pandemic hit, and have kept rising since
The chart describes vacancy and available rates and the gap between vacancy and available rates. The vacancy rate line started at 10.22% in March 2006 and peaked at 12.88% in September 2010 before coming down and bottomed at 9.44% in June 2019. Then it went higher again and ended the series at 12.48% in December 2022. The available rate line started at 11.6% in March 2006 and peaked at 16.07% in June 2010 before coming down to a low of 12.25% in September 2019. It ended on a higher note at 15.95% in December 2022. For the availability-vacancy rate gap, the first data point came in at 1.38% in March 2006, it climbed until it peaked at 3.23% in June 2009. Then it fluctuated near that level until the latest data point at 3.47% in December 2022.
The primary source of these challenges is the rise of remote work. Although more employees did return to offices in 2022, the outright level of remote work remains 7 times higher than it was in the pre-pandemic era. Layer on top of this the Federal Reserve’s historically rapid pace of interest rate rises over the past year, plus accelerating layoffs in professional and business services (e.g., in the technology sector) and the obsolescence of older office buildings, and it’s not hard to see pain in the office sector getting worse.
Employees are gradually returning to offices, but the level of remote work remains historically high
The chart describes the share of hours worked remotely since 2000 to 2022. The first data point came in at 1.7% in 2000 after which it trended slightly higher all the way until it reached 4.7% in 2019 before skyrocketing to the peak at 46.3% in 2020 before falling to the latest data point at 30.3% in 2022.
Investors need to remember, however, that the office sector is bifurcated. Difficulties are emerging by geography—Chicago and San Francisco are much more challenged than Miami, Raleigh and Columbus, for example—and differentially affecting property vintages. Newer office builds, particularly those built after 2010, are seeing much stronger net absorption rates than older builds.2
Inevitably, the tightening of credit availability and higher rates will spell trouble for some borrowers. Although the liquidation rate in the sector is low at present, we expect cumulative liquidations of commercial mortgage-backed securities (CMBS) for the office sector to rise to approximately 20% over the next 10 years (with cumulative losses expected to be approximately 8.5%).3
This level of distress is comparable to percentages seen in the sector in the post-GFC years (chart below), but—importantly—it will likely take many years to play out.4 In the short run, borrowers will likely take advantage of loan extension options. Looking further out, we expect to see CMBS loan maturities become more challenging by 2025–2027.5
Liquidations take time
The chart describes cumulative liquidations since 2002. Cumulative liquidations started at 0% in February 2002 then gradually climbed higher until the latest data point at 20.43% in January 2023.
The risk transmission mechanism: Rising rates can quickly erode valuations
Before putting the office sector in broader context, it’s worth reminding ourselves of the risk transmission mechanism at work in commercial real estate, because central bank policy affects even the most carefully selected properties. CRE assets are very long duration in nature, so their values are particularly sensitive to interest rates. But how, exactly, do higher interest rates—holding everything else constant—erode the value of a particular building?
To answer that question, we need to look at the impact that higher capitalization rates (cap rates) have on commercial property valuations. A cap rate measures a property’s net operating income (NOI), revenue less operating expenses, divided by its market value. A high level point-in-time metric used to value CRE is to take a property’s net operating income and divide that by its cap rate.
In the chart below, our hypothetical example shows how the value of a USD 100 million office building would decline to USD 72 million if cap rates increased from 4.5% to 6.25% (holding NOI constant). If the borrower were to look at refinancing the loan at this lower valuation, keeping the loan-to-value ratio constant, the borrower would have to contribute an additional USD 18 million in equity to get the refinancing done.
Even when net operating income remains constant, higher cap rates erode property valuations
The chart describes a cap rate valuation analysis. In the column on the left, the valuation is for a total of $100mn in which $65mn is debt and $35mn is equity. In the table below the column, the left side describes the scenario if cap rate is 4.5%. In this Original Loan situation, net operating income is $4.5mn whereas the value is $100mn. In the column on the right, the valuation is for a total of $72mn in which $46.8mn is debt, $18.2mn is rescue capital, and $7.0mn is equity. In the table below the column, the right side describes the scenario if cap rate is 6.25%. In this Pro-forma Refinance situation, net operating income is $4.5mn whereas the value is $72mn. On the top right of the chart, there is also an equation describing: Property Value = Net Operating Income (NOI)/ Cap Rate Next to the right-side column, there are two textboxes saying: 1. Additional Capital Old – New Loan ($65mn - $46.8mn) 2. New Loan Amount (65% * $72mn).
This example demonstrates just how vulnerable commercial real estate is to rising rates. If, in the months to come, the Federal Reserve (the Fed) raises rates even higher, some borrowers will likely struggle—and be unable or unwilling to provide additional capital to extend the life of their loans. Depending on the borrower and property fundamentals, however, that mismatch might not necessarily force borrowers to hand back their keys: Given current market conditions, we believe lenders are unlikely to repossess commercial properties that they would have to either manage or sell, especially at stressed valuations.
Instead of foreclosing, lenders may offer borrowers short-term forbearance or even modify their loans. After all, banks do have the flexibility to either extend a loan’s maturity, take a discounted payoff, or accept what’s known in the trade as a “deed in lieu” of foreclosure, which conveys all interest in a property to the lender to satisfy a loan that is in default.
Looking back at the early ’90s: CRE price cycles vs. GDP
Across all sectors, commercial real estate returns are currently down only about 3.5% from their peak in 2021, according to the National Council of Real Estate Investment Fiduciaries (NCREIF). The composition of CRE returns in recent quarters suggests the decline has been entirely due to higher cap rates, as opposed to weak NOI growth (chart below). To the extent that interest rates remain higher than they were pre-pandemic and economic growth continues to slow, we will likely see more price deflation in CRE going forward.
Recent CRE return patterns suggest that declines are due to higher cap rates
This chart shows the return components of the NCREIF Property Index (NPI) on a quarterly basis from Q1 2021 through Q2 2022. The returns components shown include net operating income (NOI) yield, capital expenditures (capex), NOI growth, cap rate change, and interaction. The quarterly total return is shown in a separate line overlaying the components. Total return began in Q1 2021 at 1.5% and gradually climbed until it peaked at 5.9% in Q4 2021. From there it gradually fell until it reached a low point of -3.5% in Q4 2022, the most recent data point. In terms of the components, NOI growth was a large positive contributing factor in every quarter. Cap rate change was also a large positive contributing factor through Q1 2022. Starting in Q2 2022, the return from cap rate change dropped below zero and became increasingly negative. By Q4 2022, the cap rate change alone contributed to a negative 6.14% percent in total return.
However, history suggests that CRE down cycles are less detrimental to GDP than residential real estate down cycles, given less spillover to the consumer.
The early 1990s recession offers a good example of this dynamic: Between 1989–1993, CRE prices fell by more than 20% from peak to trough, but—at the same time—the U.S. economy experienced one of the shallowest recessions in history, with real GDP falling by only 0.1% in 1991 (chart below).
In the 1990s, a large drop in CRE prices coincided with a shallow recession
U.S. real GDP vs. CRE price index (1986–1995)
The chart describes the CRE price index vs real GDP from 1986 to 1995. The real GDP started at 8231.7 in December 1986 and trended upward until the last data point at 10637.0 in December 1995. The CRE price index started at 96.5 in December 1986 and went up and peaked at 108.7 in December 1989 before bottoming at 86.7 in December 1993. Then it bounced back to 91.1 in December 1995.
Differentiating within and across CRE sectors
Within the commercial real estate complex, the office sector is unique for its high and rising vacancy rate. Occupancy in other CRE sectors, most notably industrial and retail, remains tight, with vacancy rates well below pre-pandemic levels. Construction activity in the industrial sector has been booming; the current level of construction is at a secular high relative to GDP (charts below).
Within CRE, the office sector is experiencing the highest percentage of vacancies
The chart describes commercial real estate vacancy rates from 2007 to 2022. For office, the vacancy line started at 10.04% in Q1 2007 and peaked at 12.88% in Q3 2010. Later it declined to a bottom of 9.44% in Q2 2019 before bouncing back to the height of 12.48% in Q4 2022. For retail, the vacancy line started at 5.77% in Q4 2007 and peaked at 7.23% in Q1 2010 before declining all the way until the most recent data at 4.23% in Q4 2022. For apartment, the first data point came in at 6.72% in Q1 2007. Then it stayed near the same level until it bottomed at 4.70% in Q3 2021 before recovering to 6.29% in Q4 2022. For industrial, the first data point came in at 7.98% in Q1 2007 it trended all the way downward until its most recent data at 3.95% in Q4 2022.
With the rise of e-commerce, industrial property development in the U.S. has reached a secular high
The chart describes retail & industrial CRE as a % of GDP. For the retail CRE line, the first data point came in at 0.35% in January 1993. It fluctuated and reached a peak at 0.45% in September 2007 before coming down and bottomed at 0.15% in December 2010. It stayed near the same level until the most recent datapoint at 0.17% in January 2023. For the industrial CRE line, the first data point came in at 0.45% in January 1993 after which it peaked at 0.66% in October 1996. It came down and bottomed at 0.27% in April 2004 before bouncing back to a high point at 0.55% in February 2009. Then it dropped again to 0.24% in January 2011 before climbing up to the high and recent point at 0.78% in January 2023.
Retail, on the other hand, is experiencing some headwinds, even though the vacancy rates are comparatively low. Ever since the GFC, retail construction projects have languished, in large part due to rising e-commerce activity, which creates more demand for logistics properties, such as warehouses, and less demand for brick and mortar assets.
In aggregate, retail is in a better place today than it was immediately after the GFC due to a lack of supply, but there are still challenges—especially for lower-quality assets, such as class B and C malls.6 Distress continues to be high in this subset of the CMBS market, which accounts for approximately 5%–10% of the overall market.7 Looking ahead, the resilience of the retail sector will largely depend on the confidence of the U.S. consumer over the next 12 months, despite an elevated probability of recession.
Although the apartment sector experienced tight market dynamics in 2021 and the first half of 2022, demand has now cooled: Vacancy rates are now close to where they were prior to the onset of the pandemic. This is largely a function of a slowdown in migration patterns across the U.S. Household formation has also slowed.8
We expect the apartment market to continue to loosen, especially as additional supply becomes available. Currently, more than 950,000 multifamily projects are under construction across the U.S., the highest level of activity since the 1970s (chart below). We believe demand, however, is unlikely to collapse as it has in the office sector given favorable secular dynamics, including renting trends among Millennials, who are facing the worst homebuying conditions in 40 years.9 Here again, demand is becoming increasingly bifurcated as rental markets in the South and Southwest remain strong, while coastal markets soften in the wake of recent tech industry layoffs.10
The number of apartments being built in the U.S. is higher than at any time since the 1970s.
The chart describes multifamily units under construction from January 1970 to February 2023. The first data point came in at 534 in January 1970. Soon it skyrocketed and peaked at 991 in July 1973 before dropping all the way and bottomed at 128 in August 1993. Then it climbed for some time until it reached a high point at 466 in July 2008 before a sharp drop to a low point of 166 in July 2010. Later it trended higher and ended the series at a high point at 957 in February 2023.
How would a CRE crisis impact the broader economy?
If we zoom out of CRE and contextualize the office sector in the broader U.S. economy, the backdrop becomes reassuring. From a wealth perspective, office assets represent a low share of the value of all physical capital across the country, at only 2.4% (chart below).11 Offices represent only 14% of total CRE assets in the U.S. 12 If we think about this another way, the market capitalization of one company, Apple, is—at USD 2.6 trillion—larger than the entire capitalization of the U.S. office sector, at just over USD 2 trillion.13
The private office market represents a fraction of total capital wealth in the U.S.
The chart describes the dollar value in billion of a few different things in 2018 (United States Produced Capital, Market Value of the Stock Market, Private Residential Structures, Overall CRE Value, Market Cap of Apple, Private Office Buildings). For United States Produced Capital, the value is at 86,223 USD Billion in 2018. For Market Value of the Stock Market, the value is at 35,816 USD Billion in 2018. For Private Residential Structures, the value is at 22,285 USD Billion in 2018. For Overall CRE value, the value is at 14,956 USD Billion in 2018. For Market Cap of Apple, the value is at 2,605 USD Billion in 2018. For Private Office Buildings, the value is at 2,071 USD Billion in 2018.
From a flow perspective, office construction represents an even lower share of overall U.S. GDP, at only 0.4% (chart below). Unlike single family residential investment, which soared to more than 3.3% of GDP in the mid-2000s before crashing and bottoming out post-GFC at around 0.7%, the office sector, in isolation, would likely not have anywhere near the same impact on the economy in the event that construction were to halt. It’s just not big enough.
Office construction remains a modest 0.4% of U.S. GDP
The chart describes investments in Office and Single-Family Residential as a % of GDP. The office investment line started at 0.33% in 1960 before going up to a peak at 0.88% in 1985. It then went on a downward path until the recent data point at 0.37% in 2021. The single-family residential investment came in at 2.75% in 1960 before going to a low of 1.63% in 1970. Then it went up to 3.10% in 1978. Then it declined and bottomed again at 1.24% in 1982 before going back to a peak of 3.33% in 2005. Then it troughed at 0.69% in 2011 and trended upward until the recent data point at 1.75% in 2022.
Furthermore, there are offsets to a weak office sector. In short, the weaker it is, the stronger residential and remote work related sectors will be, because the underlying source of the disruption to the office sector is technology, which is driving a shift in working patterns.14 Technological shocks usually aren’t recessionary or detrimental to aggregate demand, but they do create winners and losers.
In 2017, we made this same point with regard to the pervasive concerns at the time about brick and mortar retail assets: Throughout the 2010s, as overall GDP in the retail sector continued to rise, more of it went to e-commerce platforms while brick and mortar shrank. The upshot is that this shift represented a micro recession for brick and mortar retail, but not a macro recession for either the retail sector as a whole or the overall economy.
Small bank lending channel could potentially impact GDP
We feel confident that the office sector isn’t likely, in isolation, to be a significant source of GDP weakness. But broader and legitimate concerns about confidence in the banking system are relevant, especially questions about smaller banks and the credit they provide to the broader economy.
Inevitably, the latest spate of banking stress is a negative for economic growth because it will likely result in less bank lending. Quantifying that impact is no easy task, because the Fed will—in all likelihood—try to offset the impact of reduced lending via less restrictive monetary policy (larger, better capitalized banks may step in to fill the void as well).
From an equities and credit perspective, we are cautious on the regional banking sector. As mentioned previously, these banks have approximately 4.4 times more exposure to CRE than large banks do. In the chart below, we show CRE exposure, excluding lower risk segments (owner-occupied properties, multifamily), as a proxy for higher risk categories (offices and vintage malls). We map each bank’s exposure as a percentage of its capital. While total exposure to the weakest CRE subsectors varies by bank, those with more than 100% of their capital in these buckets are more likely to be smaller regional entities.
Although visibility is limited, we see a wide range of CRE exposures across the banking industry
This chart shows commercial real estate (CRE) as a percentage of common equity tier 1 (CET1) capital for various US banks. The banks are listed by asset size, with the largest banks to the left and the smallest to the right. The largest banks such as J.P. Morgan, Bank of America, and Citigroup have relatively low CRE as a percentage of CET1 at around 10-30%. Most of the larger banks have below 100%. Many of the smaller banks on the right side of the chart have above 100%. For example, Valley National has 396%, Synovus Financial has 250%, and Umpqua has 206%.
Market participants are trying to address a key concern: Will the potential stress in segments of the CRE market negatively impact earnings for regional banks—and create such significant drag for some that their capital could be impaired? Several factors will drive the outcome: the size of the banks’ exposure, the magnitude of their losses across vulnerable CRE sectors and the severity and speed of those losses.
Applying historical loss rates drawn from the GFC to office and retail holdings only, the impact in CRE would likely not rise to the level of a capital event for the banking industry. But we don’t yet know whether future losses will be confined to office and retail, or whether they will broaden. Our view is constructive: We see stronger fundamentals outside of office buildings and vintage malls, and we expect few losses beyond these two vulnerable CRE segments.
For a worst-case scenario, we need look no further than the Fed’s most recent industry-wide Dodd Frank Act stress test. In this exercise, the impact of a “severely adverse” market scenario would result in a 40% drop in aggregate CRE prices (i.e., a collapse in values that would not be confined to office properties and vintage malls). In this projection, the degree of capital erosion would be considerably more severe, with capital shortfalls likely becoming pervasive across the regional banking sector.
Clearly, this scenario is meant to represent an extremely bearish outcome for the U.S. economy—and does not represent our base case—but it does help answer the question “How bad could this possibly get?” Our base case assumes that aggregate CRE prices fall approximately 10%–15% in the current cycle (although for the office sector, we wouldn’t be surprised to see price declines of 30%–40% in the most stressed markets).
Investment implications
What does this mean for your portfolio? The trouble in the office sector—and its potential impact on bank lending and profitability—will contribute to an elevated probability of recession in the U.S. this year. In the portfolios we manage, we have been adding to assets that can help weather a downturn should one occur (e.g., core fixed income), while maintaining a balance across asset classes.
Within publicly traded real estate, the office sector has been one of the most popular shorts for the market since COVID-19 hit and prices already reflect material weakness. Some publicly traded REITs that focus on the industrial sector are also getting caught up in the turmoil, and we think they are approaching attractive entry points for investors who have less exposure to the space.
On the private side, the managers that we work with have limited exposure to U.S. office buildings, and opportunities may start to appear for specialists in distressed real estate in the coming quarters and years.
If you would like to discuss this topic in more detail or understand the potential opportunities for your portfolio, please reach out to your J.P. Morgan team.
1J.P. Morgan North America Securitized Products Research. Data as of March 2023.
2J.P. Morgan, Trepp. Data as of February 2023.
3J.P. Morgan Research. Data as of January 2023.
4Note: These expected liquidations/losses are for conduit CMBS exposures; Single Asset Single Borrower (SASB) exposures will likely experience lower distress given better credit quality.
5J.P. Morgan North America Securitized Products Research, “CMBS Weekly: CMBS office loans—a lay of the land.” Data as of March 2023.
6Barclays Credit Research, “Don’t Forget about the Malls.” Data as of March 2023.
7J.P. Morgan, "Assets and Liabilities of Commercial Banks in the United States - H.8". Data as of March 2023.
8CoStar, U.S. Census Bureau. Data as of March 2023.
9University of Michigan. Data as of April 2023.
10Jeffrey Langbaum, “Coastal, Sun Belt Rent Growth to Slow in 2023,” Bloomberg Finance L.P. Data as of March 2023.
11World Bank, Bureau of Economic Analysis, Bloomberg Finance L.P., Haver Analytics. Data as of 2023.
12CoStar. Data as of 2022.
13Bloomberg Finance L.P. Data as of April 2023.
14This technological shock (remote work) is also a source of the bifurcation in the office sector between newer builds and older vintages. Amid the war for labor talent, companies have tried to use newer office space in prime locations as a way to lure employees back into the office.
