Summary
- Multifamily REITs are rightly understood to provide excellent protection from inflation and are often priced quite high compared to REITs from other sectors, such as net lease REITs.
- They are not priced so highly at the moment, which leads some analysts to expect strong gains from here.
- From the perspective of inflation, the question is how much that protection is actually worth.
Working on a very deep look at AvalonBay ( AVB ) led me again to my conundrum about the Multifamily REITs. They are often pricey.
At the moment, the average ratio of Price to forward Funds From Operations, or P/FFO, for these REITs is 16x. For comparison, Net Lease REITs are on average at a P/FFO of 14.4x. This 11% difference is much smaller than has often been the case.
Midcycle ratios of forward Price to Funds From Operations, or P/FFO, tend to run in the low-20s over the past 15 years for Multifamily REITs. And we saw some values near 30 in 2021.
The growth rates of FFO/share over the past couple of decades for quality Multifamily REITs are near 6% overall (quite good for a REIT). Based on this, there are two ways to explain the large valuations.
One of them is to assume that those growth rates continue literally forever. Then a discount rate of 10% gets you that valuation above 25x FFO.
The other, assuming such growth for 15 to 25 years with a 3% terminal rate after that, is to drop the discount rate to 8%. So perhaps the market believes the multifamily REITs are more secure than others. But my personal minimum discount rate for any real-estate investment is 10%.
Here is similar data for three quality Net Lease REITs.
As compared with Net Lease REITs, Multifamily REITs normally cost more by about 30% in P/FFO. In terms of the more realistic P/AFFO, the difference is near 35%. (AFFO is Adjusted FFO, here adjusted mainly to take out recurring capex).
The qualitative reason for my choice to own AVB has been that multifamily REITs have the best inflation protection amongst "standard model" REITs. (Hotel REITs are different.) But is the extra cost quantitatively worth it?
We are here today to seek an answer to that question. The spirit of the inquiry is to capture the difference for typical numbers and ignore any detail that does not have a noticeable impact on the result. This is complicated enough even then.
Setting the Stage
My goal here is to model the growth of AFFO/sh for a stereotypical Multifamily REIT and a stereotypical Net Lease REIT during a sustained period of significant price inflation. I discussed long-term growth rates for a number of such REITs here .
We will take the reference growth rate for Multifamily REITs to be 6%. This is a bit below AvalonBay (6.5%) but above Camden Property ( CPT ) at 5% and Equity Residential ( EQR ) at 5.5%.
The reference growth rate for Net Lease REITs will be 4.5%. National Retail Properties ( NNN ) has done 4%. Realty Income ( O ) did 5.5%, but that was mainly by means of a lot of share issuance in a friendly stock market, which can't be counted on going forward.
The specific case modeled is 15 years of inflation at a 5% rate. This generates an increase in prices of about 2x, not atypical for past inflationary periods. We assume the inflation affects rents and asset values equally, as should be true in the long run.
An implication of this assumption is that market rents increase 5% per year throughout the 15-year period. Of course there would be local and temporal fluctuations but we assume uniform increases to capture the main trend.
We assume here that the stock markets are not friendly throughout our model period, meaning that selling stock is not an accretive option. If these markets were friendly, because stock prices are high enough, then it seems likely that both types of REIT we consider would be able to increase their earnings growth by taking advantage of it. But seeing no way to handicap this, we just ignore the possibility here.
The net impact of these assumptions for our discussion today is that the REITs of interest can grow their AFFO/sh only by reinvesting earnings or otherwise raising cash.
When REITs acquire or develop a property, the standard measure of its earnings is the stabilized cap rate. This is the ratio of stabilized Net Operating Income, or NOI, to the property price. This is quite straightforward for Net Lease REITs, as they acquire properties mainly by sale-leaseback transactions.
For Multifamily REITs, the stabilized part refers to leasing a developed property to a target occupancy or any rapid repricing of an acquired property made possible by rebranding or marking to market.
What is notable here is that price inflation as such does not alter cap rates, because NOI and Property Prices both increase in proportion. Of course, developments in the markets do lead the cap rate to change over time, but surprisingly little .
Debt renormalization is particularly important for all REITs during times of inflation. What is meant by "debt renormalization" is a change in the amount of debt to keep the ratio of debt to EBITDA constant.
The normal growth of AFFO/sh during the period before the start of higher inflation includes adjusting the debt to stay leverage neutral. It is the Incremental AFFO, beyond the normal increase, that generates the New Debt of interest here.
After establishing an initial value of the Debt/EBITDA and a reference ratio EBITDA/AFFO, we can find the New Debt as:
New Debt = (Debt/EBITDA) x (EBITDA/AFFO) x Incremental AFFO.
This New Debt generates cash that can now be invested, along with retained earnings, to generate forward growth of AFFO.
Finally, the ratio of interest expenses to NOI for new properties is the Debt Ratio for those properties times the ratio of interest rates to cap rates. In recent years, the ballpark has been that interest rates were around half of cap rates for most Net Lease REITs and somewhat larger than that for Multifamily REITs.
[Overall the (blue-chip) Multifamily REITs generate lower cap rates than the Net Lease REITs. But they also have higher credit ratings and lower interest rates. Of course, individual REITs may do better or worse.]
Impact of Inflation on Multifamily REITs
Multifamily REITs often raise renewal rates more slowly than market rates increase, but even so they mark all their rents to market over at most a few years. Here we assume that is immediate.
Such REITs generate growth in part by using retained earnings and funds from dispositions. AvalonBay, for example, uses these funds primarily to support new developments.
Quality REITs of this type have long weighted-average debt-maturities, with a typical number being 8 years. Their Debt Ratios typically run just over 1/4, so their interest costs now are in the ballpark of 1/7 of NOI.
The costs other than interest expenses run near 25% of NOI. They all respond directly to price inflation. As a result, an increase in revenues by some percentage produces an increase in AFFO by the same percentage or a bit more.
Retaining 20% of AFFO/sh and using it with a 6.5% return generates about 1.3% growth of AFFO/sh. Raising a comparable amount of funds by dispositions and working off the spread between development yields and disposition yields doubles that. It seems to me that in the long run this aspect is not clearly impacted by price inflation as such.
In round numbers, pairing just below 3% growth from development with just over 3% growth from rents has gotten these REITS to about 6% annual growth of AFFO/sh. AvalonBay has done a bit better. Camden Property ( CPT ) and Equity Residential have done a bit worse. Under inflation, though, rents will increase as a higher rate.
Here we will assume that a 5% increase in revenues due to price inflation produces a 5% increase in AFFO/sh. Taking the reference rent growth to be 3% (which could be high), this is 2% of Incremental AFFO/sh over what has been typical over the past decade. This would get us to an 8% AFFO/sh growth rate, so far.
The new investments would be made at the targeted leverage. But the 2% of incremental AFFO/sh growth represents new EBITDA, allowing the REIT to add more new debt to raise cash while sustaining leverage.
We can see here the pattern in Debt/EBITDA. It has run 4 to 5 in recent years. It seems likely to stay there to protect credit ratings.
Based on my analysis of data from REITbase one has AFFO typically 74% of EBITDA for Multifamily REITs. [It was lower for the larger collection of such REITs analyzed for this article .] The ratio EBITDA/NOI runs near 95% for these REITs.
In light of the last two paragraphs, the calculation discussed above approximates the New Debt from debt renormalization as 6 times the Incremental AFFO. This becomes cash available for investment.
For the Incremental AFFO of 2% just considered, one gets cash of 12% of AFFO. Investing this with a 6.5% return would generate a 0.8% increase in AFFO/sh.
This way of adding up the pieces gets AFFO/sh growth at 8.7% for price inflation of 5%. The real growth of 4% is comparable to the real growth of the past decade.
Impact of Inflation on Net Lease REITs
The previous material was not simple, but an evaluation for Net Lease REITs is significantly more complex. One must consider what happens to leases that are never renewed, leases that are renewed during the period of interest, and the new assets and debt added in response to the increased AFFO/sh and to debt renormalization.
Readers whose brain is near exploding (or already has exploded) might want to skip to the section "Net Lease Summary".
Existing Portfolio
We will consider a stereotypical Net Least REIT. There are individual variations.
Each such REIT holds a collection of assets. These assets are operated under very long leases. An average remaining lease duration of 15 years is common.
Ideally the weighted average maturity of the associated debt would be comparable to the lease duration. This is the case, for example, with National Retail.
It is not the case with Essential Properties Realty Trust ( EPRT ). One risk of owning EPRT is associated with what may happen when debt matures. The fact that their debt is only about 1/3 of assets does limit the risk.
The leases on the current assets include escalators, a typical value being 2% per year. That new revenue does not come with new costs of any size, and so drops straight down to AFFO/sh.
With a typical ratio of AFFO/Revenues being in the ballpark of 2/3 for Net Lease REITs, the rent escalators on retained properties produce 3% per year increase in AFFO/sh.
Here we ignore dispositions which would include properties that fail. These are small. We are seeking a model that is simple enough to be comprehensible yet complex enough to capture the main trends in reality.
If the lease term is L, and the leases are spread uniformly in time, then a fraction 1/(2L) of the leases renews each year. That fraction applies both to the reduction in assets and to the reduction in AFFO/sh.
We also make the assumption that all leases, at our starting point, produce the same AFFO per unit asset value. There will actually be variations, of course. But across hundreds or thousands of total assets, these will smooth out.
Taking the initial value of the assets to be 100 (%), the non-renewed assets in year y will be 100*(1-y/(2L), shown by the blue curve in the plot. The increase in AFFO/sh, relative to its initial value, produced by the non-renewed assets in year y will be the sum of (1.03)^y/(2 L).
Here is what we get. After 15 years, the assets that have not renewed have dropped to 50%. But thanks to the escalators, the AFFO/sh they produce has only dropped 20% relative to the initial total from all the assets.
Renewing Leases
Meanwhile the renewed leases are coming in at higher rents. We assume the rent in the renewed leases gets marked to market. It will increase 5% each year, starting from its initial value.
The rent increase on renewal also drops straight through to AFFO/sh, with no associated cost increase. From the logic above, the ratio of new to old AFFO/sh for assets with renewals in year y will be 1.5*(1.05)^y.
The AFFO/sh associated with the renewals will be the sum over years y up to Y of 100x[1.5x(1.05)^y/(2L)]. By year 15, this equals all the initial AFFO/sh, as is shown by the lower, dashed curve in the next plot. The upper curve shows the total, which climbs to 180% of the initial value.
Adding New Assets
All the above is connected with AFFO/sh from the initial group of assets. Next we need to consider new assets.
Adding assets requires finding cash to invest. By assumption, as discussed above, there is no issuing of new stock in this model.
Investing the new cash produces new AFFO/sh, in a leverage-neutral way. Finding that new AFFO/sh involves the following.
For these REITs costs other than interest expense run below but near 10% of NOI. The Debt Ratio is typically 40%. They have managed interest rates of about half (or less) their cap rates recently, so that interest expenses run about 20% of NOI.
We assume here that a side effect of inflation will be that interest rates will grow to near 2/3 of cap rates. This will push up interest expenses to about a quarter of NOI for newly acquired properties.
The AFFO produced by investing new cash is from Chris Volk's V-formula, discussed here . Based on the information above, and assuming a cap rate on new acquisitions of 7%, the yield on the new cash is 7.6%.
We will evaluate the cash available for new investments from two sources. The first is Debt Renormalization. As the AFFO/sh from the pre-existing assets grows, new debt can be issued to sustain a constant ratio of Debt to EBITDA.
Here is the historical pattern. Spirit Realty Capital ( SRC ) ran a high value for a while, as they worked their way through some problems. But they are back with the others now.
The value is 5x to 6x except in times of trouble or special circumstances. We will use 5.5x.
From the initial portfolio, the plots above show an overall growth rate of AFFO/sh of 5.3% per year. This is 2.3% above the 3% due to the escalators, discussed above.
The incremental 2.3% is available for debt renormalization at that 5.5x. This generates funds for new investments (per share) of just below 13% of the initial AFFO/sh. This in turn produces growth of about 1% of the initial AFFO/sh.
The second source of cash is retained earnings. Here we will take the fraction of the current (not the initial) AFFO to be 20%, which is a bit larger than the typical median found by REITbase. Investing this cash at a 7.6% yield generates growth of the current AFFO/sh by 1.5%.
The total growth related to new assets becomes:
AFFO/sh growth =
1% of initial AFFO/sh plus 1.5% of current AFFO/sh.
Net Lease Summary
The Net Lease REITs see four contributions to growth of AFFO/sh. 1. The original leases that do not renew grow AFFO/sh in consequence of their escalators. 2. The original leases that do renew get a one-time reset to market rents and grow with their escalators from there. 3. The retained earnings produce growth proportional to current AFFO/sh. 4. The debt renormalization produces additional growth in response to increases in AFFO/sh and EBITDA.
We can now show the Net Lease total AFFO/sh, comparing it to the contribution from original assets and to the Multifamily result:
Here you can see the multifamily AFFO/sh growing at 8.7% (green curve). The contributions to the Net Lease AFFO/sh from the original assets is shown in orange. This includes the effects of escalators and of lease renewals that mark rent to market.
The red curve shows the total AFFO/sh for the Net Lease portfolio, including also the impacts of retained earnings and debt renormalization. Overall, the Net Lease AFFO/sh produces a CAGR of 5.3%.
It is notable that the Net Lease earnings keep up with inflation. I was unsure how this would turn out.
Worth It Or Not?
We have now quantified the difference in response to inflation of Multifamily vs Net Lease REITs, at least within the confines of our model. The next question is whether this justifies the observed, midcycle difference in P/FFO that approaches 50%.
Part of this calculation is that the ratio of Funds available for Distribution and Reinvestment to FFO is different for these two sectors. In this article , I estimated the ratio as 77% for Multifamily and 91% for Net Lease. This reflects mainly the difference in recurring capex.
I included this factor in the following discounted-cash-flow model. For the case with inflation, the higher growth rate found above was applied for 15 years. Then the base growth rate was applied for another 10 years, after which the terminal growth rate was 3%.
The models that assumed no additional inflation used the base growth rate for 25 years. After that, they used a 3% terminal growth rate.
Here is what one gets. The ratio of Net Present Value to FFO is plotted against discount rate.
The sensitivity to the discount rate is remarkable, as always. That said, trends in REIT prices often seem to make sense for a 10% discount rate. Personally, I almost never invest if I see the total return as likely to be below that level.
For that discount rate of 10% or a bit more, in the high-inflation case one gets a value of the Multifamily REIT above 20x FFO. But not 25x.
For all the Net Lease growth rates (including the historical rate of 4% for National Retail), one sees FFO multiples of 15x to 18x. The Multifamily multiple for the historical growth rate also comes in a bit below 18x.
Leaving aside the exact numbers, here is what I take away from this plot. If you are certain that substantial inflation will be the story of the next decade or two, then it is worth paying more for a Multifamily REIT. How much? The plot suggests paying an FFO multiple that is about 30% larger than what you could pay for a Net Lease REIT.
But 50% or more, as was common during 2021, is definitely overdoing it. So my perspective is that quality Multifamily REITs broadly are reasonably priced for the small-inflation case and underpriced for the case of substantial and enduring inflation.
Even so, such underpricing is by modest tens of percents, not big factors. You have to decide how worthwhile that is to you. In addition, as part of your analysis, you might consider more specifically the context and details of any one of these REITs.
The above gives me what I need to go back to thinking about AvalonBay. I hope you too find it useful.
For further details see:
REITs: How Much Is Inflation Protection Worth?