- Over more than 20 years, some REITs have produced total returns in the 15% ballpark.
- Based on their earnings growth and dividends alone, the gains should have been much smaller.
- We consider what really happened and how to make money from REITs going forward.
REIT fans can and do boast of the excellent long-term returns REITs have produced during this century. They beat the pants off the S&P 500 during its first decade and held their own in the second even as the S&P 500 ran hard.
Emblematic of the best performance are these two REITs that can boast of a credit rating of A-. This plot is from early 1995 to the end of 2021.
Over that interval, net-lease REIT Realty Income ( O ) generated a CAGR of total return of 15.7%. Apartment REIT AvalonBay ( AVB ) generated 14.7% and is also today in a strong up cycle of earnings that could increase that. The teal line on the plot shows a CAGR of 15%.
Over 25 years ending at the same date, the REITwatch index shows all equity REITs doing 10.2%. For comparison, over the same interval the S&P 500 generated 11.0%. Of course, some individual stocks did far better than that.
By choosing an interval, one can cherry pick periods when the REIT index did better or worse than the S&P 500. But that is not really our focus here.
As that history illustrates, REITs offer some degree of decorrelation vs domestic US stocks broadly. This is a sufficient reason to hold some, but one wonders what gains to hope for going forward.
Let’s explore that.
The Problematic Details
This article germinated out of my playing around with the REITwatch tabulation of projected 2022-2023 growth of Funds From Operations, or FFO, per share. FFO is the simplest but not the best measure of REIT earnings, but is the only one for which doing really broad analysis is straightforward.
One way to plot that data is like this:
These points tend to be a little higher than is typical, as continued recovery from the pandemic is built into the estimates. There are some points off the top of the plot, mainly from hotel REITs very much still in pandemic recovery. There are a few points off to the right, too.
The median projected growth of FFO/share is 7.8%. If you add the dividend yield as of a couple weeks ago, to project total return, you get this:
The median here is 12%. It is elevated at the moment since dividend yields are up by about a quarter in the present bear market. This is consistent with a lot of my work finding total returns in the high single digits for most quality REITs in normal times.
Some REITs are above 15% on this plot, but it is mainly those still in pandemic recovery. If you look at the blue-chips, they typically are projecting about 12%.
This musing led me to wonder what the actual history of earnings growth was for REITs. It proved possible to look at that, through the lens of FFO/share as reported by TIKR back to 2005.
Now TIKR is not perfect but it usually is okay on this metric.
Here is what I found, for about a dozen REITs with long histories and blue-chip balance sheets:
The table shows values of CAGR over various time periods, all after 2004. Note that the “All Years” column includes up to 5 years of analyst estimates. Henceforth we will ignore years after 2019, from the point of view that the impacts of the pandemic are still being played out.
Ideally one might want to look at the peak-to-peak performance across cycles. But there can be idiosyncrasies that impact any specific REIT. The final column includes values adjusted by me when I think the other columns are misleading.
The clear winner here is AVB, at a robust 7.7% CAGR.
But that poses a problem. Per the Gordon growth model, add their 3.1% average dividend yield and you get an 11% total return estimate, far short of the 15% seen above.
Similarly, the expectation for most of the REITs shown, with their lower rate of growth of FFO/share, would be a total return in the high single digits.
High Single Digits Makes Sense
The basics of REIT earnings growth can be simply expressed. REITs grow earnings only in small part by rent increases. REITs grow mainly by adding to or improving their property collections.
REITs pair new equity with some fraction of debt, and so have
Return on New Equity = (cap rate/equity fraction) times (AFFO/NOI),
where the cap rate is the ratio of Net Operating Income, or NOI, to the price of the property. AFFO is FFO adjusted to take out all non-cash items and recurring capital expenses.
REITs today typically run an equity fraction of 50% to 75%. They have AFFO/NOI from just over 70% to less than 50%, depending on the magnitude of the recurring costs and other factors.
The upshot is that Return on New Equity is the cap rate, give or take about 30%. For clarity I often call this the Investment Yield.
A REIT that retains 20% of earnings and invests them at an Investment Yield of 5% to 10% grows FFO/share and AFFO/share by 1% to 2%. On top of rent increases that add perhaps 3% including a bit of debt, the total growth from these sources tends to run 3% to 5%.
Some REITs can in addition issue stock to raise capital. This adds shareholder value so long the REIT can sell AFFO (by issuing stock) for more than it can buy AFFO (from acquisition, development, etc). A lot of REITs generate an additional couple percent of growth in earnings this way but few generate as much as 5%.
What is important here is that these aspects of REIT growth make sense of the table above. Growth rates of 4% to 8%, and total returns including dividends in the high single digits, are quite natural.
This makes the long-term performance of many REITs including AVB and O quite puzzling. Let’s look more closely at those two.
Two Specific Cases
I looked at O and AVB over the period from 1996 through 2019, inclusive. After 2004, the TIKR data (along with YCharts total return) was sufficient.
For 1996 through 2005, I used data from the 10-K filings along with YCharts historical dividend yield. Overall, and not just in these two cases, REIT dividends were much higher around the turn of the century than they have been since the Great Recession.
This analysis produced the following tables, for the period of 1996 through 2019:
FFO/share (row shaded blue) for Realty Income actually decreased before the Great Recession and only exceeded its 1995 value in 2012. Perhaps this reflects the challenges faced by retail tenants during that period.
FFO/share for AvalonBay grew more rapidly in most periods but suffered more across the Great Recession. Remarkably, for both REITs the Total Return in the stock market was larger than one would expect from the Gordon Growth model.
Total Returns Beyond Growth
To my mind, the missing total return implies that the discount rate investors required to hold these companies decreased. This is one consequence of the 40-year bull market in bonds, which seems likely to be over.
The interesting disconnect is that while the Gordon Growth model predicted a larger return for AvalonBay, it was Realty Income that had the larger market returns.
The market was and remained more enthused about Realty Income than AvalonBay. Notable is that AVB fell much further in the Great Recession than O did, reflecting challenges in the housing markets across that interval.
Beyond that, in recent years O has become something of a cult stock, incessantly promoted by some authors. During 2019 in particular, O increased in price by more than many observers including me thought sensible.
My view of these two is as follows. The market overlooks the low growth produced by Realty Income and also the challenges posed by their sheer size. The market also is overly negative on the prospects for AvalonBay, fearing their presence primarily in coastal gateway cities.
Having said that, my view that AVB is today a much better investment than O will not surprise you.
Tailwinds and Headwinds
Stepping back from the specifics, it seems clear from the above data that REITs have benefited for 30 years from increases in valuation that were not fully justified by their internal economic performance. This tailwind seems to have accounted for perhaps 4% per year of appreciation, which is a lot.
Had this tailwind not existed over those decades, REITs today ought to be priced at half or even a third of their present prices. This is also, of course, true of all investments whose ultimate value is based on cash flows.
Discount rates have relentlessly plunged over that period. They are likely near their bottom.
In my view the dynamics of stock prices in the 1970s were strongly affected by increases in market discount rates. Going forward, discount rates eventually will increase, but no predictions here when that will happen.
For now, the tailwind is gone, and so investors who buy REITs at normal prices will be looking at total returns in the high single digits or a bit larger. This will meet the goals of some investors, especially including the aspect of partial decorrelation with stocks broadly.
The problem is that a period with a headwind is almost certain to follow, and to last a decade or two. Across those years, buy and hold REIT investors seem likely to me to see returns in the mid- or even low- single digits. (This problem will not be escaped by investing in most other domestic stocks either.)
This may not keep up with inflation. In my view it is likely that we will see significant inflation over the present decade. A year ago too many considered that unlikely. Today too many consider inflation certain to be really large.
It is worth noting that most REITs will benefit from inflation, through increased rents, property values, and cap rates. Because of that, inflation or other tailwinds may perhaps offset the discount-rate headwind. As a result, my view is that REITs will be a comparatively good place to invest during that next upward move in discount rates.
Note that the winds related to discount rates as such affect neither earnings nor dividend payments. Forward total returns, being based on dividends, would actually rise, but might be growing from a lower base.
How to Make Money Anyway
In a very broad investing context, it seems to me that the bond bear market, whenever it comes, will be very hard on investors. My remarks to that effect in chat rooms tend to fall on deaf ears. It is an unpleasant topic.
To my mind the point of investing in REITs, in addition to getting some level of decorrelation, it that they are an inefficient niche market. This provides more potential to find quality undervalued investments than one may see in the broader market.
The opportunity is enhanced by the obscurity of REIT financials. GAAP earnings mean little and the standardized NAREIT FFO is also often misleading.
The big opportunities with REITs, though, are provided by bear markets. Mr. Market reliably over-reacts.
In any particular case, Mr. Market suppresses the prices of some blue-chip REITs to ridiculously low levels. Buying them at those times can generate returns of 50% and more, as I discussed for the present market in two recent articles.
During 2020, many Shopping-Center REITs were opportunistic. This year, Mr. Market has been more selective. Examples at present of great opportunity include Simon Property Group ( SPG ) and Alexandria Real Estate ( ARE ).
The point in the context of the present article is this. No matter what prices may be doing as part of secular trends, bear markets will bring you opportunities for substantial gains.
I don’t favor just sitting on cash between bear markets; the numbers don’t work. My approach is to invest in a broad range of securities, so that when the bear market hits I can tap those that fall a little to fund investments in those that fall a lot. Quality REITs are often among those that fall a lot.
For further details see:
REITs: Great Gains, Or Not