July 26, 2022
Multi-residential REITs A mid-year valuation check in
Our view: In this note, we do math. We assess the reasonableness of the multi-res REIT valuation in three ways: 1) Implied cap rate spread over risk free bond, 2) Implied IRR from a private equity perspective, and 3) implied price per suite versus replacement cost. #1 suggests that multi-res REIT valuation is on the expensive side of fair while the other two suggest that they are on the cheaper side of fair. We remain constructive on the sector. We believe stability in rates and benign regulatory changes are key catalysts to a re-rating of this sub-sector.
Let’s do math
When the range of probable outcome was wide in the initial months of the pandemic, the one thing we could hang our hat on, as an analyst, was math, –i.e., constantly calculating what the market was pricing in terms of NOI decline, and then, based on evolving information, assessing, as best we could, whether the implied NOI decline looked too optimistic, too bad to be true or about right. Today, what’s uncertain is the macroeconomic environment. While the range of outcome is not as wide, there are layers of unknowable macro complexity (recession – mild/deep, inflation/stagflation, central bank policy reaction, etc...) that make for yet another challenging environment to invest in. So, let’s do math again. During the pandemic, the relevant math focused mostly on NOI, i.e., the numerator of the property value equation, simplistically defined as: Property value = NOI / Cap rate.
Today, we think getting right the math around the denominator impact will be more impactful to investor returns given the sharp reversal of multiple decades of declining rates. We believe that multi- res fundamentals are strong enough to more than withstand the macro uncertainty ahead. Indeed, in a recession, the multi-res sector should do well relative to other sectors given the needs-based nature of the asset class. Moreover, in an inflationary environment, the short duration leases allow for capturing rent inflation despite higher opex and rent regulation. So far, all point to rapidly-strengthening rent across the country.
The denominator effect
The large movement in bond yields has effectively increased cost of capital, reset return thresholds and potentially undone years of valuation increase from cap rate compression. For the multi-res sector, the market is pricing in ~30-90 bps of cap rate expansion since the beginning of 2022, such that the (simple) average implied cap rate is 4.9% for the names under our coverage universe.
To assess the reasonableness of this valuation, we suggest doing math from three perspectives:
1. Risk spread math: One way is to compare today’s implied cap rate premium over the risk-free rate with historical premiums (See Exhibit 1). Risk premiums fluctuate over time – when the market is fearless and/or expect future growth, risk premium is typically low and vice versa. Today’s risk premium of ~200 bps is low relative to the long term average of 339 bps. However, 1) the historical spread is skewed higher by mix shift in the universe; 2) the embedded mark-to-market rent or rent growth potential is likely higher today versus history, and 3) multi-res cap rates did not proportionally follow rates down post GFC (2009-2021) creating a higher than ‘normal’ spread over a relatively short time frame. From 2000-2007, spreads were tighter at 270bps when 10-year bond yield was above 4%. As such, the current risk premium looks tight but not entirely unreasonably so, in our view.
2. IRR math: Another way is to determine whether buying the REITs at the current implied cap rate results in an acceptable IRR under a private equity model. Private equity math matters because private equity players dominate the multi-res transaction market and tend to set marginal prices.
To illustrate this, let’s hypothetically assume that the five Canadian-focused multi-res REITs (Boardwalk, Canadian Apartment, Killam, InterRent, Minto) are combined into one portfolio – “APT REIT United.” Here’s what a private equity investor would be buying: A coast-to-coast Canadian portfolio at a cap rate of 4.7%, occupancy of 96%, average monthly rent of $1,325 with an estimated mark-to-market rent potential of ~14% (See Exhibit 2).
If the private equity investor has a 10-year hold period, exits the portfolio at 5.2% cap (50 bps higher than going in), leverages the portfolio at 50% LTV under a 10-year CMHC- insured debt interest rate of 4%, the investor would be able to achieve a 7%-13% levered IRR assuming NOI grows 2-5% annually.
On this basis, we believe the IRR math pencils out easily as NOI growth of 2-5% looks very achievable given the mark-to-market rent potential in the mid-teens (likely accelerating), the long term average of 3.8% while 7-13% IRR for a core or core+ portfolio is attractive, especially given that excess density/development or redevelopment upside are not factored into the math. This would imply that the public market may have priced in a bit too much cap rate increase.
Replacement cost math: Replacement cost is relevant in asset classes that are growing, and it acts somewhat as an anchor to how low values can drift. Rising construction/replacement costs bode well for existing stock because 1) rising replacement rent required on new construction should push market rents up overall and/or 2) limit new construction activity, thereby pushing occupancy/rent on existing stock.
Today, with increasing development charges by municipalities, lack of land availability, higher construction costs and shortage of labour, replacement costs have increased materially. Exhibit 3 shows the construction cost increase in 2022 vs. 2019 for select type of residential (some as high as +40%) as well as the implied price per suite of “Apt REIT united” at ~$238,000 – this is well below replacement costs which vary widely depending on geography and type of construction (a rough range is between $300,000 and $800,000). On this basis, it would be hard to argue for materially falling multi-res prices.
Bottom line: Valuation math is not an exact science. Two out of three ways of looking at multi-res REIT valuation suggest that their valuation looks to be on the cheaper side of fair while the implied cap rate spread suggests more expensive side of fair. Given this valuation range and accelerating positive fundamental outlook, we remain constructive on the sector – in other words, given where bond yield sits today, we feel that the market has priced in enough or even moderately more than enough cap rate expansion. Where the long bond yield ultimately ‘settles’ matters. In our view stability in rates (and benign regulatory changes) are key catalysts for the sector to re-rate.