Expected Returns of REITs
Lately, there has been a lot of talk about the expected future performance of real estate investment trusts (REITs) in the Wall Street Journal. Recall that REITs are companies that own and operate rental properties such as hotels, apartments, warehouses, office buildings, and shopping centers. They are required to distribute at least 90% of their profits to shareholders as dividends which are not subject to the lower “qualified dividend” rate. Even though we typically see REIT-type properties all around us, surprisingly, they make up only 3 to 4% of the total market value of US equities. Note that there is another type known as the mortgage REIT which does not concern us here.
On 2/5/2017, former Morningstar analyst John Coumarianos chimed in with “Considering a REIT? Here’s How to Evaluate Likely Returns” where he projected real (post-inflation) returns of 0 to 2%, depending on whether current “high prices” persist or revert to historical norms. Personally, I disagree with one aspect of his calculation where he deducts 2% for costs required to maintain the property at a consistent level of quality. In my opinion, the current prices and resulting dividend yields for REITs of about 4% already incorporate this this cost. Furthermore, an expected real return of 0% is a non-starter because (as of 2/21/2017) we can get a 0.4% real return with a 10-year TIPS bond at far less risk.
In Rational Expectations: Asset Allocation for Investing Adults, Dr. William Bernstein argues that because REITs can reinvest only a small portion of their earnings in their operations, their dividends grow slowly—more slowly than inflation by about 1% per year. At the time of that writing (2014), the dividend yield of REITs was about 3%, leading Bernstein to initially project a real “fundamental expected return” of 2% which he later revised down to 1%, probably due to an assumed reversion to the mean of REIT prices relative to earnings. In my opinion, however, REITs have a high amount of pricing power (ability to raise rents) to both cover future costs and compensate for inflation (see this chart for supporting evidence). Also, I do not see a need to dock future returns for a future mean reversion that may or may not happen. Thus, my forecasted long-term real return for REITs (for planning purposes) is 4% or about the same as their current dividend yield. As for the level of risk surrounding this return, I would it consider to be akin to the volatility of equities. While one may argue that this is so much lower than their historical returns, we must remember that those historical returns had a tailwind of a substantial drop in long-term interest rates. A back-of-the-envelope calculation I performed shows that over the last 24 years, annualized REIT returns were about 3 percentage points higher due to a 4 percent drop in 10-year Treasury bond yields. The real return without this drop would have been about 5%.
A subsequent letter to the editor (2/15/2017) by Brad Case of the National Association of Real Estate Investment Trusts presents a very different approach to the estimation of future REIT returns. Mr. Case dismisses Coumarianos’s analysis as “doom”, touting future real return estimates based on historical returns relationships ranging from 8.15 to 8.9%. The former derives from the current REIT stock price discount to net asset value, while the latter is based on the difference between current bond yields (both corporate and government) and REIT dividend yields. While I would love to see these 8% or higher real returns happen, it is difficult to visualize exactly how that would unfold, given our starting point of low interest rates and high valuations. To clarify, the 10-year Treasury bond yield is 2.4% while a 10-year A rated corporate bond yields 3.2% (Yahoo!Finance). Maybe I’m dense, but I just don’t see how we get 8.9% real out of those small differences. If anyone can explain it to me (as well as to Dr. Bernstein and Mr. Coumarianos), I am happy to listen.
There is no question that REITs have served investors well as part of a diversified portfolio, and at Clarity, we continue to advise our clients to have an allocation to REITs at least equal to their representation in the overall market. However, we do not expect them to perform any better than equities or bonds on a risk-adjusted basis. Anyone who expects to obtain an 8% or higher real return over the next decade (and has planned accordingly) may be in for a bitter disappointment.