Commercial real estate (CRE) is real estate held to generate income or used as an input into production by firms. It is notably different from other asset classes of a similar magnitude in that CRE is traded in private, illiquid markets. CRE is a hugely important asset class that has received less attention from the academic literature than asset classes that rival CRE in terms of sheer value. Yet pension funds, life insurance companies, sovereign wealth funds and other institutional investors seek the diversification benefits provided by CRE’s unusually steady income flow. The paper, “Commercial Real Estate as an Asset Class,” by Andra Ghent of UNC Kenan-Flagler Business School, Walter Torous of the MIT Center for Real Estate and Rossen Valkanov of UCSD’s Rady School of Management provides a much-needed overview of the CRE literature thus far, focusing on its attributes as an asset class.
In terms of share magnitude, the estimated value of the stock of CRE has varied over time from 1.2 times GDP in 1980 to 0.9 times GDP in 2016, which translates into more than $12 trillion of outstanding commercial real estate today. For comparison, the only other major asset classes larger than CRE in 2016 were residential real estate and common stock. The value of the stock of CRE is larger than both the stock of U.S. treasuries and U.S. corporate bonds.
Commercial real estate is held by both non-financial corporations and financial institutions such as real estate investment trusts (REITs). As of 2016, non-financial corporations own an estimated 64% of the total stock of commercial real estate, primarily as an input into production. The remainder is held by corporate financial firms (including REITs) at 7%, private partnerships at 7%, sole proprietorships at 5% and nonprofit institutions at 13%.
The trend over time has been for more businesses to rent their space rather than own real estate. For example, in 1960, non-financial corporate businesses held an average of 45% (40% for non-corporate businesses) of their balance sheets as real estate. Today that number is down to 30% (25% for non-corporate).
While commercial real estate held by non-financial corporations is indubitably important, it makes up less than 3% of all total commercial real estate transactions from 2001-2015.1 Thus, the rest of this discussion will focus on commercial real estate held by firms to lease.
Commercial real estate can be disaggregated into privately traded CRE and publicly traded CRE held by REITs. The former has limited data available to the public, while there is ample data available for REITs as they are required by law to report their holdings.
For privately held CRE, annual returns are estimated to be in the ballpark of 9-12%, while returns for publicly held CRE are between 10-12%. The estimated standard deviation of annual returns for privately held CRE is between 4.2-5.2%, and for publicly held CRE between 16.5-17.4%. Returns to CRE are negatively skewed, suggesting that there is significant downside risk from periods during which prices collapse.
Perhaps the most attractive quality of CRE as an asset class is that returns from rental income are remarkably stable at 7% annually for both privately and publicly held CRE, with income returns also positively skewed over the sample. On the other side of the coin, this means that it is fluctuations in the price of CRE that make it a risky investment. Average returns from price appreciation alone are between 2-4.5% for privately held CRE and 2.9-5% for publicly held CRE. This evidence suggests CRE is a much more attractive asset to investors with a long-term investment horizon, such as pension funds.
Lastly, returns for privately held CRE are persistent over time, meaning that today’s return is a good predictor of tomorrow’s returns. On the other hand, publicly held CRE returns today do not predict tomorrow’s returns. This is likely due to an inefficiency in the market for privately held CRE not present in the market for publicly held CRE. For example, privately held CRE is a far less liquid asset than publicly held CRE, which becomes liquid through REITs.
The paper classifies properties into core and non-core property types following earlier research. Core includes apartment buildings, freestanding retail, industrial, office, regional malls and shopping centers. Non-core includes medical buildings, lodging resorts, manufactured homes and self-storage.
Core properties are generally considered to be less risky than non-core properties, but the data does not support that perception. The average annual return for core properties is between 10-15%, with a standard deviation between 18-29%. Meanwhile, average annual returns for non-core properties range from 9.4-16.4% with standard deviations between 18-30%.
Overall, non-core properties are less sensitive to cyclical variations (such as expansions and recessions) than core properties. Of the core properties, industrial and office tend to be more cyclical than apartments and retail. Of non-core properties, lodgings are highly cyclical, suggesting the lodgings are a luxury good that get hit particularly hard during recessions.
Note that due to the lack of data on privately held CRE, it is too difficult to say much about property-level returns. However, current research suggests that transaction risk – the risk associated with the costly nature of executing a CRE transaction – is one of the largest sources of risk associated with CRE2.
Unlike residential real estate, mortgages make up only a small share of financing associated with CRE transactions. For example, commercial mortgages have never amounted to more than 25% of the value of the total stock of CRE. Corporations are more likely to borrow money with unsecured debt as opposed to using a mortgage. Research has found that non-financial corporations only constitute one-third of all commercial real estate mortgages outstanding, suggesting that these corporations use other unsecured debt instruments to raise funds.
For the remaining non-corporate borrowers of commercial mortgages, the literature has found that U.S. REITs, on average, operate under a leverage ratio between 35-50%, while the average leverage ratio for mortgages associated with privately held CRE is around 30%.
The sources of CRE debt also differ from that of residential mortgages. From 2005-2012, only 50% of commercial mortgages came from traditional banks, with the other 50% made up of commercial mortgage-backed securities (CMBS), life insurers and non-bank lenders who specialize in CRE loans.
CRE loans are also far less likely to be securitized than residential mortgages, with CMBS constituting only, at peak, 30% of all CRE mortgages. One of the major reasons is that securitized loans typically place strict conditions on prepayment, which can be risky from the perspective of the developer. Yet those who seek to borrow from a securitized lender can borrow at a higher leverage ratio.
Lastly, CRE loans tend to have shorter maturities of 7-10 years, as opposed to 15-30 years for a residential loan. CRE loans are also not fully amortizing, meaning missed payments do not necessarily increase the amount borrowed.
This research paper is the first to summarize the literature on commercial real estate from the perspective of an asset class. The authors point out that CRE is a hugely important asset class valued at around $12 trillion. They document ownership patterns over time, showing that fewer corporations today decide to own the real estate within which they conduct their business. The authors show CRE returns exhibit a surprisingly stable income component, and highly volatile price appreciate component. Lastly, they document how mortgage securitization has affected the financing of commercial real estate differently thanit has for residential real estate.
1 Ghent, A. C. (2019): “What’s Wrong with Pittsburgh? Delegated Investors and Liquidity Concentration,” Tech. rep., University of North Carolina, Chapel Hill.
2 Sagi, J. (2019): “Asset-Level Risk and Return in Real Estate Investments”, Tech rep., UNC Chapel Hill.