Home Banking The Commercial Real Estate-Small Bank Nexus

The Commercial Real Estate-Small Bank Nexus

by internationalbanker

By Philip Marey, Senior US Strategist, Rabobank





During the COVID-19 pandemic, the occurrences of remote work jumped out of sheer necessity. The technology was already available, but the pandemic accelerated its adoption, bypassing employers’ hesitations to allow people to work from home. In many cases, remote work has been successful and become a permanent feature, often in a hybrid form. For employers, it has become an employee benefit, attracting people in a tight labor market, and it saves on office-space costs. The flipside of the latter is that demand for office space has seen a structural downward shift. For example, it has been estimated (Gupta et al., 2022)1 that the underlying value of office space in New York City has permanently declined by 39 percent. This suggests that, at current prices, a bubble exists in commercial real estate. What does this mean for financial stability and the economic outlook?

Commercial real estate is heading south

If we plot the BIS’s (Bank for International Settlements’) Commercial Real Estate Price Index, it is clear that since the Great Recession of 2007-09, commercial real estate (CRE) prices have more than doubled in nominal terms (the blue line in Figure 1) but moved sideways since 2021. This suggests that prices have reached a plateau. However, in recent years, inflation has obscured CRE-price movements in real terms (the orange line), which shows a peak in 2021, but since then, a decline, almost to its level at the start of the COVID-19 pandemic. In other words, CRE prices are already failing to keep up with inflation. Is this an indication that the CRE bubble is already deflating? With nominal CRE prices remaining elevated, most of the nominal-price correction is likely still to come. If the 39-percent estimate by Gupta et al. for New York City is representative of the entire United States, we are heading for a major decline in CRE prices.

We can also plot the BIS’s index against CRE lending to show that rising prices for commercial real estate sparked a credit boom in commercial real estate (Figure 2). Given the academic literature linking financial crises to credit booms and busts (see, for example, Schularick et al., 2012)2, this should be cause for concern. Moreover, Minsky (1986)3 noted that bankers’ emphasis on collateral values and expected values of assets (instead of cashflows) is conducive to the emergence of a fragile (as opposed to a robust) financial structure.

If excess demand for office space pushed up CRE prices and that demand increased banks’ CRE lending, what would a structural downward shift in demand for office space do? If CRE prices are deflating, what does that mean for the indebted CRE sector? Will it lead to defaults? And what will that mean for the banks that issued the CRE loans? Is the CRE-bubble deflation a threat to financial stability? Also, note that due to the Federal Reserve’s (the Fed’s) steep rate-hiking cycle, some companies in the CRE sector may find it difficult to refinance their loans at substantially higher rates.

We can dig deeper by looking at the demand and supply developments in CRE lending. If we look at the Fed’s “Senior Loan Officer Opinion Survey on Bank Lending Practices” data (Figure 3), it is clear that demand for CRE loans strengthened, especially between 2012 and 2017 (lending standards loosened between 2012 and 2015). This era coincides with a strong rise in the CRE price index, which may have motivated banks to expand CRE lending. Demand for CRE loans weakened during the pandemic but bounced back as the economy reopened, then headed south again in 2022. Loan standards tightened during the pandemic and loosened when the economy rebounded, but they have tightened again since 2021. In other words, there seems to be a correlation between CRE prices and demand and supply developments in CRE lending. Currently, both are heading south if we look at CRE prices in real terms and CRE lending in terms of net demand. It seems that rising CRE prices sparked a credit boom in CRE, and now that the CRE price bubble is deflating, the CRE sector has less appetite for borrowing, and banks are tightening their lending standards.

The Fed’s incomplete stress test

CRE prices are falling in real terms, and credit for CRE loans is tightening. Does this pose a problem for the economy? Not if we believe the Fed’s June 28, 2023, press release4 that accompanied its annual bank stress test results5. The stress test looked at “a severe global recession with a 40 percent decline in commercial real estate prices [in line with the estimates by Gupta et al., 2022], a substantial increase in office vacancies, and a 38 percent decline in house prices. The unemployment rate rises by 6.4 percentage points to a peak of 10 percent, and economic output declines commensurately”. However, according to the Fed, “all 23 banks tested remained above their minimum capital requirements during the hypothetical recession”. Therefore, the central bank concluded that “large banks are well positioned to weather a severe recession and continue to lend to households and businesses even during a severe recession”. However, one line in the press release reveals the main problem with the Fed’s stress test: “The banks in this year’s test hold roughly 20 percent of the office and downtown commercial real estate loans held by banks.” So where are the remaining 80 percent held? If the stress test considers a huge decline in commercial real-estate prices, it might be relevant to know how this affects the banks that hold 80 percent of the CRE loans made by banks. Therefore, let’s take a closer look at CRE lending by large and small banks in the next section.

Bank lending: large versus small banks

We already saw in Figure 2 that increases in CRE lending accompanied rises in CRE prices until 2022. However, there is more to this story if we closely examine which banks have been lending. So far, we have looked at aggregate bank lending to the CRE sector without distinguishing between different types of banks. However, a closer look at the banking sector reveals a disturbing vulnerability that could threaten financial stability.

The Fed’s data on commercial banks distinguishes between large and small banks. Large domestically chartered commercial banks are defined as the top 25 domestically chartered commercial banks ranked by size. Small domestically chartered commercial banks are defined as all domestically chartered banks outside of the top 25. Note that according to this definition, a bank of, say, $80 billion would still be considered “small”. Figure 4 shows how CRE lending has evolved, distinguishing between large and small banks.

It turns out that CRE lending by large banks has hardly increased in the last 15 years, while at the same time, small banks’ CRE lending has more than doubled. In other words, the growth in loans to commercial real estate has come from small banks. In fact, small banks have taken over the role of the main provider of CRE loans. Therefore, the Fed’s stress test omits the most relevant section of the banking sector for commercial real estate. While large banks’ CRE lending has remained stable since 2006, CRE lending by small banks has increased rapidly. We could even call it a credit boom in CRE loans provided by small banks.

Whether the small banks’ increased share of CRE lending is a problem also depends on the relative importance of CRE loans for small banks (Figure 5). FDIC (Federal Deposit Insurance Corporation) data (Quarterly Banking Profile [QBP]) distinguishes at least three classes of asset size: more than $250 billion, $10-250 billion and $1-10 billion. The first class contains only large banks as defined by the Fed stress test; the second class is a mix of large and small banks; and the third class includes only small banks. While CRE loans were only 5.7 percent of total assets in the first quarter of 2023 for the largest banks, for the smallest banks, they were 32.9 percent! For intermediate-size banks, CRE loans comprised 18.4 percent of assets. So, not only are 80 percent of CRE bank loans made by small banks, but these loans also make up a much larger fraction of the balance sheets of small banks.

Finally, it is important to note that small banks are regional banks. In fact, the United States has such a high number of small banks because it was difficult for banks to open branches in other states for much of the country’s history. The relevant legislation has since been abolished, and the number of banks in the US has fallen, but there are still many small banks with predominantly regional clients. This means that small banks’ CRE risks are regionally concentrated. Instead of a diversified nationwide CRE-loan portfolio, a small bank tends to make loans to local borrowers. Consequently, if commercial real estate in a region turns sour, the small banks in the area will be highly exposed. Bubble or not, adverse developments in the CRE sector will hit small banks harder than large banks.

Feedback mechanisms: small banks, commercial real estate and the economy

The commercial real estate-small bank nexus brings together two vulnerable sectors that could rapidly deteriorate in a self-reinforcing loop. Problems for small banks could force them to reduce the credit supply to CRE, causing additional problems for the CRE sector, which relies heavily on small banks. The other way around, defaults in CRE will asymmetrically hurt small banks rather than large banks. Together, tighter credit from small banks and reduced activity in the CRE sector could push the economy into a mild recession. Alternatively, a mild recession caused by other factors, such as the Fed’s rate-hiking cycle, will hurt the banking sector and the CRE sector simultaneously. The self-reinforcing problems in the two sectors could further drag down the overall economy, making the initially mild recession more severe. These feedback mechanisms in the commercial real estate-small bank nexus are summarized in Figure 6.


COVID-19 appears to have had a lasting negative impact on demand for commercial real estate. Federal regulators are aware of the risks to commercial real estate, but the Fed’s stress test provides a false sense of security. The finding that large banks are equipped to absorb losses on CRE loans in case of a CRE crisis is encouraging, but small banks have provided the bulk of CRE bank loans. In fact, while CRE lending by large banks has been stable, there has been a credit boom in CRE loans provided by small banks, more than doubling the amount since 2006. Moreover, small banks are more vulnerable to the CRE sector in terms of exposure and were already hit by deposit outflows last year. The commercial real estate-small bank nexus exposes the US economy to a vulnerability that could threaten financial stability and either cause a recession or make a mild recession more severe.



1 National Bureau of Economic Research (NBER): “Work From Home and the Office Real Estate Apocalypse,” Arpit Gupta, Vrinda Mittal and Stijn Van Nieuwerburgh, September 2022, NBER Working Paper 30526.

2 American Economic Association (AEA): “Credit Booms Gone Bust: Monetary Policy, Leverage Cycles, and Financial Crises, 1870-2008,” Moritz Schularick and Alan M. Taylor, April 2017, American Economic Review,Volume 102, Number 2, Pages 1029-61.

3 Stabilizing an Unstable Economy, Hyman P. Minsky, 1986, McGraw Hill, 2008 edition, Page 261.

4 Board of Governors of the Federal Reserve System: “Federal Reserve Board releases results of annual bank stress test, which demonstrates that large banks are well positioned to weather a severe recession and continue to lend to households and businesses even during a severe recession,” June 28, 2023, Press Release.

5 Board of Governors of the Federal Reserve System: “2023 Federal Reserve Stress Test Results,” June 2023.



Philip Marey is Senior US Strategist at Rabobank in Utrecht, the Netherlands. He previously worked at Maastricht University and holds a Master of Science in Econometrics from Erasmus University Rotterdam.


Related Articles

Leave a Comment

This site is protected by reCAPTCHA and the Google Privacy Policy and Terms of Service apply.