By Todd Lukasik, CFA
We prefer the yield, stability, and growth prospects of the triple-net structure over RIDEA.
Until the introduction of the REIT Investment Diversification and Empowerment Act of 2007, health-care real estate investment trusts simply owned property and rented it to tenants. RIDEA, however, allows REITs to directly expose themselves to the financial performance of the underlying health-care operations, instead of just collecting rent checks. RIDEA has brought changes to health-care REITs' profitability, cash flows, and return prospects - some cosmetic and others meaningful. Although triple-net leases have dominated health-care REITs' results historically, the rapid adoption of the RIDEA structure since 2010 has resulted in portfolio exposures of as much as 34%. The new structure has been a financial boon lately, as RIDEA-structured assets have enjoyed robust growth. However, RIDEA introduces the potential for variability in health-care REITs' cash flows and makes REITs responsible for additional expenses relative to traditional triple-net assets. Also, RIDEA transactions have generally been priced more aggressively than triple-net deals, which implies that faster cash flow growth is required to make the RIDEA structure pay off. As a result, we generally favor the yield, stability, and growth characteristics of the triple-net deals.
The Basics: RIDEA vs. Triple Net
Before the introduction of RIDEA, health-care REITs stuck to a simple model: Buy property and rent it to tenants. Since 2007, however, health-care REITs have had the option to buy property and - instead of simply collecting rent from a tenant - be financially responsible for the success of health-care service operations conducted in the property.
This was a major change for health-care REITs, because - for the first time - it allowed them to be directly exposed to the underlying property operating fundamentals, including changes in occupancy and rental rates, instead of just collecting rent checks from tenants.
RIDEA allows REITs to form a taxable REIT subsidiary to oversee the management of the health-care operations within the property, which generally means hiring a third-party manager. While the third-party manager is generally motivated to run the property in alignment with the REIT's interests, it is the REIT that retains the most exposure to the financial success (or lack thereof) of the property operations. Third-party managers are generally paid a base management fee (generally 5% of property revenue) plus incentive fees (generally 1%-2% of property revenue or linked to property income growth or cash flows above a certain threshold).
While neither the triple-net nor RIDEA structure is inherently moatier than the other, we generally prefer the yield, stability, and predictable growth of the triple-net structure. Also, recent notable senior housing transactions in both the RIDEA and triple-net structures have had similar initial net operating income yields, but initial cash flow yields (returns on real estate assets) on RIDEA assets have been lower by an average of 50 basis points or so. While each RIDEA deal is unique and needs to be evaluated with regard to property location and quality, manager, and overall growth prospects, the lower initial cash flow yields of RIDEA assets require faster future cash flow growth to achieve parity with similar triple-net transactions, all else equal.
The attraction of the RIDEA structure for senior housing assets derives from health-care REITs' bullish view on the prospects for this property type, particularly in the top metropolitan areas across the United States and Canada. Strong tailwinds for senior housing are provided by the aging of the population and a recovery in home values, the stock market, and employment. The aging of the population provides incremental demand for senior housing space, while rising home and equity values provide financial resources for seniors to pay senior housing rent and fees. Higher levels of employment put adult children back in the labor force, with less free time to provide care for their senior parents.
By utilizing the RIDEA structure with senior housing properties, REITs are directly exposing themselves to the robust growth expectations associated with the category. Under the RIDEA structure, if these robust cash flow growth expectations play out, the benefit will accrue directly to the REIT as opposed to the tenant operator.
Although Ventas (NYSE:VTR) was the first health-care REIT to adopt the RIDEA structure with its 2007 acquisition of Sunrise Senior Living REIT, Health Care REIT (NYSE:HCN) has been the most active since, with RIDEA transactions involving Sunrise, Revera, Benchmark, Belmont Village, Chartwell, Senior Star, Merrill Gardens, Brookdale, and Silverado. As a result, among our coverage, Health Care REIT currently has the highest portfolio concentration allocated to RIDEA at 34% of NOI. HCP (NYSE:HCP) has not been as aggressive, but it converted a small portfolio of senior housing assets to the RIDEA structure under a new manager (Brookdale) when its original triple-net tenant (Horizon Bay) was in default on a loan. Similarly, HCP has entered agreements with Brookdale, expected to close in the third quarter, that will convert some its Brookdale and Emeritus assets from triple-net leases to RIDEA joint ventures with Brookdale.
We expect relative RIDEA exposures to remain similar, although HCP may eventually increase its exposure to 10%-15% of NOI, provided it can find the right assets with opportunistic characteristics to improve operating performance and achieve an acceptable risk-adjusted return. (We estimate HCP's RIDEA exposure will approximate 10% following the implementation of its proposed deal with Brookdale-Emeritus.)
Although Health Care REIT has called for a balance in its portfolio, we would not be surprised if it increased its RIDEA exposure beyond its current 34% of NOI, if the right deal (or deals) came along. In our opinion, Health Care REIT has been the most aggressive with the RIDEA structure to date, and we think it will remain very interested in future RIDEA opportunities.
Any future shift in RIDEA exposures may be good, bad, or indifferent for these REITs' shareholders, depending on the specifics of each transaction. Although we generally prefer the triple-net structure, the RIDEA investments of Health Care REIT and Ventas have delivered robust cash flow growth recently, and the capital these firms have invested in their RIDEA assets has been well deployed, in our opinion.
Although our take on initial return on real estate assets yields for the RIDEA deals Ventas and Health Care REIT struck from 2007 to 2011 generally fell in the range of 5.5%-6.5%, we estimate current ROREA yields are tracking in a range of 6.2%-7.9%. Thanks to continued cash flow growth expectations, we estimate five-year forward ROREA yields are likely to be 8%-9% for the majority of these deals, a range that looks attractive relative to our mid-7s estimates for these firms' weighted average cost of capital.
Financial Statement Implications of RIDEA vs. Triple Net
Most financial statement differences between RIDEA and triple-net transactions are cosmetic, in our view. Two meaningful implications of RIDEA deals, however, are increased maintenance capital expenditure requirements for the REIT and potentially increased cash flow variability.
Triple-net leases carry very high profit margins for landlords (around 100% NOI margins), because tenants are responsible for all property operating expenses, leaving minimal - if any - expense load for the landlord. Because RIDEA transactions, on the other hand, essentially make the REIT responsible for the property-level revenue and expenses of the health-care operations conducted in the property, NOI margins on these RIDEA assets are much lower (generally 30%-40%).
As Ventas (28% of NOI exposed to RIDEA) and Health Care REIT (34%) have brought RIDEA assets onto their balance sheets, for example, overall company EBITDA margin at these firms has fallen. HCP, with its minimal exposure to RIDEA (3%), has maintained a higher overall EBITDA margin.
Similarly, other health-care property types and leasing models have varying levels of NOI margin, generally 60%-75% for medical office buildings and 70%-80% for life science properties. We attribute the different levels of 2013 EBITDA margin among HCP, Health Care REIT, and Ventas mainly to their different business mixes.
While the shifting EBITDA margins due to business mix shift add some complexity to our analyses, this is largely a cosmetic issue. Deal pricing is generally based on property-level cash flow, which accounts for the lower level of profitability (and should also account for higher levels of capital expenditures) associated with RIDEA assets. Furthermore, our fair value estimates are based on company-level cash flows, which account for different levels of margins for different types of assets.
Triple-net leases require tenants to pay for property maintenance and repairs, whereas this burden falls to the REIT under the RIDEA structure. While this represents incremental cash outflow for REITs relative to triple-net leases, expected outlays for maintenance capital expenditures are generally not overly burdensome, usually falling in the range of as little as $500 per unit per year for newly constructed properties to $2,000 or so per unit per year for older, more expensively built assets.
Relative to triple-net leases, we estimate that the incremental maintenance capex for Ventas falls in the range of $45 million-$50 million (using $1,800-$2,000 per unit per year) and for Health Care REIT falls near $60 million per year (using $1,700 per unit per year).
However, maintenance capex expectations can rise over time, as properties age. For example, when Ventas acquired Sunrise Senior Living REIT in 2007, estimated capex was $750-$1,000 per unit per year. In 2010, Ventas added Atria assets to its senior housing RIDEA portfolio, with estimated capex requirements around $1,500 per unit per year. Lately, maintenance capex estimates for Ventas' portfolio of senior housing operating assets have increased to $1,800-$2,000 per unit per year. Similarly, Health Care REIT referred to maintenance capex of roughly $1,250 per unit per year on a number of its late 2010/early 2011 RIDEA deals; lately, that number has crept up to $1,700 per unit per year.
While we don't expect the heady growth rates implied by these historical expectations changes to persist, we recognize that continued high growth rates related to maintenance capex are a potential risk to the returns on the RIDEA-owned assets. And we generally expect maintenance capex requirements to increase moderately over time.
RIDEA Introduces Potential for Faster Cash Flow Growth, More Variability
Under triple-net leases, REITs generally receive 2.5%-3.5% annual rent increases, as specified by the lease agreement. With the RIDEA structure, however, REITs can benefit directly from improvements in underlying property fundamentals, such as improved occupancy and improved rental rates related to the health-care operations of the property. As a result, REITs have used the RIDEA structure to acquire assets with "upside" where they think they can capture faster rates of growth than the traditional 2.5%-3.5% escalators associated with triple-net leases. Often, RIDEA portfolios will be acquired with lower-than-normalized levels of occupancy and/or capital investment opportunities to improve occupancy and/or rates. Or - especially for portfolios with relatively high levels of initial occupancy - RIDEA assets have been concentrated in markets with favorable demographics and supply characteristics, where the owners expect better-than-average growth rates for rents over time.
To date, it appears this strategy is working, as health-care REITs' RIDEA assets have generally been delivering the highest levels of internal growth among their various property portfolio holdings.
Demographics, namely the aging of the population, support long-term demand for senior housing. We think growth at rates exceeding inflation and generally somewhat higher than other parts of health-care REITs' portfolios are likely for the next few years.
Major risks to senior housing demand, in our view, include falling home prices or a drop in the stock market, and (potentially) higher unemployment, as unemployed adult children - because they don't have to report for a job - may provide the care for their senior parents that they might otherwise get from a senior living facility. Furthermore, it is generally easier to build senior housing stock than many other types of health-care real estate, because senior housing generally lacks federal regulations and other requirements related to other types of health-care real estate. Therefore, the risk of overbuilding incremental supply is also relevant to existing senior housing landlords.
Recently, many of the health-care REITs we follow have reported same-store growth expectations for their senior housing operating (RIDEA) portfolios in the 4%-6% range, including the benefit from redevelopment efforts. Excluding the benefit of incremental capital expenditures for redevelopments, we think underlying internal growth in these portfolios is probably 3%-4% currently. While this is still respectable growth that generally exceeds our expectations for other health-care property and investment types, it is far below the high-single-digit to low-double-digit growth expectations during the initial years following the RIDEA transactions. While we think this growth trajectory is largely in line with REITs' expectations at the time of the deals, we think it shows that the easy operational gains - in terms of higher occupancy, higher rents, and related operating leverage - from these initial RIDEA transactions have largely been achieved.
While internal growth of the RIDEA portfolios of Ventas, Health Care REIT, and HCP has generally been robust since the downturn, averaging an estimated 6.5%-7.0% per year during the past three years, looking back to 2008 exposes greater variability in the cash flow streams, with a major dip in 2009. On the other hand, while triple-net portfolios have not enjoyed the same level of robust growth, the range of growth since 2008 has been much tighter, and the average triple-net growth rate of these three firms has always been positive during the past six years.
Data from the National Investment Center for the Seniors Housing & Care Industry shows the meaningful decline in occupancy senior housing facilities experienced from 2007 to 2010. Senior housing occupancy fell from roughly 91.5% in 2006 and early 2007 to roughly 87% in 2009 and 2010. In addition to falling demand due to the economic downturn, developers were completing new construction, which introduced incremental senior housing units to the market during a period of falling demand, exacerbating the occupancy declines.
Although we believe in the long-term growth drivers for senior housing units, we think the sector's RIDEA internal growth trajectory of the past few years is unlikely to continue unabated indefinitely. Although we see no immediate concerns, we think there will be years where cash flows in these portfolios decline, perhaps by a meaningful amount. Just as REITs have benefited recently from rising levels of occupancy, rental rate growth, and related operating leverage, so too can these favorable conditions reverse in a market downturn for the asset class.
If we're right, any future decline in cash flows may surprise some investors who have grown accustomed to the steadily increasing cash flows and dividends from health-care REITs over the years, which have largely been predicated (pre-2011, for the most part) on the reliability of triple-net leases and the ability of health-care REITs to accretively consolidate their industry.
RIDEA Transactions Have Pressured REITs' Returns on Real Estate Assets
With a bullish long-term view on senior housing demand, health-care REITs have bid aggressively for senior housing assets using the RIDEA structure, in an attempt to get exposure to the potential upside in the performance of the properties, in terms of higher occupancy rates and higher rental rates. In our analysis of notable transactions in the health-care REIT sector, we find that initial NOI yields have been comparable between notable RIDEA and triple-net transactions, but that initial cash flow yields, or ROREA, have been roughly 50 basis points lower for RIDEA transactions, due to the incremental capex that REITs are responsible for with RIDEA assets.
Indeed, the relatively low initial cash flow yields on RIDEA transactions depressed our estimates of ROREA for both Ventas and Health Care REIT. ROREA at both firms fell meaningfully following their initial forays into RIDEA-structured transactions.
While we expect all three health-care REITs we cover to achieve future levels of ROREA that exceed our estimates of their WACCs, our ROREA expectations are lowest for Health Care REIT and Ventas, the two with the largest RIDEA exposures. HCP, which has minimal RIDEA exposure, has had more consistent levels of ROREA over time, and we expect it to have the highest level of future ROREA across our 10-year forecast horizon. Nonetheless, we think all three enjoy sustainable competitive advantages and narrow moats derived from a combination of intangible benefits associated with owning quality properties in attractive locations and customer switching costs associated with triple-net portfolios.
With health-care REITs willing to accept lower initial cash flow yields on RIDEA deals relative to triple-net deals, they must achieve greater growth out of the RIDEA portfolios to make them pay off longer term. Using the data we compiled on recent notable RIDEA and triple-net transactions - including initial yields, expected growth rates, and initial (and future) expectations of maintenance capex - we estimate that a typical RIDEA transaction, consummated at the same initial NOI yield as a triple-net transaction, would take seven years to achieve cash flow parity.
However, we also estimate - using a 7.5% discount rate for both cash flow streams - that it would take 18 years for the cumulative present value of RIDEA cash flows to exceed that of the triple-net lease. Extending our cash flow growth assumptions over 40 years, the cumulative present value of cash flows for the RIDEA transactions would be 8% greater. However, senior housing occupancy and cash flow may decline in weak economic environments, and we think it's unlikely that the RIDEA cash flows will grow uninterrupted at a 4% rate over a 40-year period.
We Favor the Triple-Net Structure
While we think the above framework is useful for thinking about the choice between the RIDEA and triple-net structures, we recognize the limitations of this generalized analysis for any particular RIDEA or triple-net transaction. For example, this analysis does not recognize that cash flows may vary from our basic assumptions. RIDEA cash flows may grow faster, or suffer some years of slower or even negative growth, due to variability in the operating metrics of the properties and/or changes in maintenance capital expenditure requirements. Triple-net cash flows may also grow faster or slower. Many triple-net leases allow for "catch-up" rental increases after the initial lease term, which allow rents to be reset to market rates, if higher. Conversely, triple-net lease rates may need to fall at the end of the lease term, if market rates have not kept up with the rate of increase embedded in annual rent escalators. Additionally, it may be appropriate to use different discount rates for the different sets of cash flow streams.
While we have nothing against the RIDEA structure, recent triple-net transactions appear, on average, to offer a better combination of initial yield, cash flow growth, and predictability than the recent RIDEA deals, in our view. Triple-net leases provide a historically tried and tested model for generating reliably-growing cash flows from quality underwriting. RIDEA transactions promise faster growth (and have delivered to date) but also look susceptible to potential future cash flow declines in challenging economic environments and potentially higher-than-expected maintenance capex requirements. While we are not concerned about the RIDEA deals of Ventas, Health Care REIT, and HCP, we generally prefer the stability and predictability of the triple-net structure.
Health-Care REITs Are Our Most Attractive Property Sector
HCP offers investors an attractive potential total return profile in the current environment. With a 5.2% yield, an estimated 2.5%-3.0% internal growth profile, and potential additional growth through industry expansion and/or consolidation, we think HCP's total return potential looks attractive. Furthermore, its stock trades at an 18% discount to our fair value estimate. Very high customer concentration and greater exposure to government reimbursement programs represent the biggest company-specific risks, in our view.
Ventas offers investors attractive return prospects combined with the best-covered dividend and higher-than-average growth prospects. We estimate Ventas' 2014 dividend (4.4% yield) will represent 73% of our 2014 adjusted funds from operations forecast, the largest cushion of our health-care REIT coverage universe, setting the stage for potentially faster dividend growth than peers. Furthermore, we're bullish on the senior housing and medical office building property sectors, where Ventas has 71% of its portfolio.
Health Care REIT's shares are trading near our fair value estimate, so we think HCP and Ventas offer investors larger margins of safety. We have nothing against Health Care REIT, and we expect it to benefit alongside HCP and Ventas from favorable industry tailwinds, but we believe HCP and Ventas are the more attractive opportunities today.
Given health-care REIT investors' negative reaction to sharp rises in interest rates, it appears that they may be viewing investments in health-care REITs similarly to investments in long-dated bonds. On the surface, this appears to make sense, as health-care REITs' triple-net lease portfolios have very long-term weighted average lease expirations, generally 10 years or longer. Nonetheless, there are some key differences.
Despite their very long-term lease agreements (RIDEA aside), we think health-care REITs are relatively well positioned for rising rates. HCP, HealthCare REIT, and Ventas all have reasonably long-dated debt maturity schedules. Furthermore, each should benefit from both internal and external growth opportunities. In-place triple-net leases generally increase at a rate of 2.5%-3.5% per year, or the rate of increase in CPI, whichever is greater. So although the leases are locked in for long terms, there are reasonable escalators embedded in them, which - unlike a traditional bond - should increase cash flows over time. RIDEA assets may grow even faster in an inflationary environment, as the short-term contracts with individual tenants can be reset to potentially higher rental rates on a regular basis. In addition, external growth should come from both the expansion of real estate portfolios as partner operators expand their businesses and through further consolidation of what remains a very fragmented real estate sector with the vast majority of ownership in private hands.