Should you include Reits in Your Portfolio? And if so, which ones?

Many investors have real-estate investment trusts in their portfolios because they provide exposure to the real-estate market without having to buy property directly. But which of the many types of REITs on the market have performed best over the long run, and do they provide the housing-like returns investors seek?

To study this issue, my research assistants (Alice Nguyen and Diego Alcantara Leon) and I pulled all REIT mutual funds and exchange-traded funds that are publicly listed on U.S. markets. We then separated our REITs into five broad categories—commercial; infrastructure; residential; data-center-focused; and general, where all types of REITs are packaged in a single product. We then collected the average return, including dividends, for all REITs over five years and 10 years and measured their average volatility, or the standard deviation of returns, over the past five years.

The first interesting finding is that over a five-year horizon, general REITs and data-center REITs have had the greatest average returns, yielding 15.18% and 15.57% a year, respectively. As for risk over this period, the general REIT category averaged a relatively high volatility of 21.52% and the data-center REIT category averaged a volatility of 21.20%. To put this in perspective, the S&P 500 delivered a 15.89% return a year over the same period, with lower risk (18.63% volatility).

On the other end of performance, we found that commercial REITs fared the worst over the past five years, with an annualized return of 7.83% a year and the highest risk as well (25.62% volatility).

When we look at the 10-year horizon for volatility and returns we see similar rankings of all REITs as when we looked at the five-year horizon. Again, general REITs and data-center REITs fared the best with annualized returns of 9.80% and 9.50%, respectively. And commercial REITs did the worst over the 10-year horizon with an annualized return of 7.43% and with the highest risk out of the full group (25.62% volatility).

Next we explored how well REITs track housing prices. This is especially important for renters who want to gain access to housing-like returns without having to buy a house.

The bottom line: The REITs have low correlation with a common housing-price index, which suggests that investors aren’t going to get housing-like returns by buying a REIT. On average, REITs are close to uncorrelated with actual housing-price movements (correlation of just 0.10). While on the other end of the spectrum, REITs are highly correlated with equities (a correlation of 0.89, where 1 is the highest correlation you can have).

It should also be noted that correlation with the housing market might not be the only thing that a REIT investor is looking for when investing. REITs can pay out significant income to their owners in the form of dividends, which serve as a benefit to those that are in a low tax bracket or those who own the REIT in a tax-sheltered account.

All in all, commercial REITs have fared the worst since the pandemic as well as over a 10-year horizon and data-center REITs have fared the best. And if you are a renter who wants something that tracks housing prices, REITs operate much more like stocks than a tracker of housing prices and hence aren’t the best way to get a piece of the housing market.

Derek Horstmeyer is a professor of finance at Costello College of Business, George Mason University, in Fairfax, Va. He can be reached at reports@wsj.com.