Realty Income's Latest Acquisition a Departure From Norm
ARCT would improve cash flow and reposition the portfolio, but longer-term risks emerge.
Realty Income (O) will acquire publicly traded American Realty Capital Trust (ARCT) for $3 billion--$1.9 billion of Realty Income shares with the rest in debt. The transaction achieves management's goals of diversifying further into nonretail assets and with investment-grade tenants, but it represents a substantial departure from Realty Income's normally patient and measured approach to acquisitions. The deal is expensive and huge, and there is a minimal margin of safety, in our opinion.
What Is Realty Income Buying?
Realty Income is paying a dear price for a ready-made portfolio that helps it achieve management's diversification goals. We estimate that 45%-50% of the portfolio is in nonretail assets (such as distribution, manufacturing, other industrial, and office properties) while 75% of revenue is from investment-grade tenants. This supports management's strategy of positioning the firm for a possible future with rising interest rates (i.e., toward investment-grade tenants) and weakness among U.S. consumers (i.e., toward nonretail properties). Regardless of one's view on the merits of this diversification strategy, this deal is in keeping with the strategy management has outlined.
After the deal's close, nonretail assets will make up 23% of Realty Income's portfolio (versus 14% currently), while 34% of its revenue will come from investment-grade tenants (versus 19% currently). The ARCT portfolio will also provide immediate benefits to other portfolio metrics, such as tenant and industry diversification, lease maturity profile, and weighted-average occupancy, although we view the improvement to occupancy to be a short-term boost that becomes a long-term drag in a decade or so when the ARCT leases begin to expire.
Importantly, Realty Income will not be taking on any ARCT employees, so the key management team that we have grown to admire over the years will remain in place at Realty Income, as will the current board structure. This translates into what we estimate to be roughly $8 million per year in general and administrative expense savings for the combined firm. This also leaves the legacy ARCT management to continue as a Realty Income competitor after the acquisition.
What Price Is Realty Income Paying?
The price is high. We'd grown accustomed over the years to seeing Realty Income acquire assets at relatively high cap rates, due to a combination of its ability to source off-market deals and its willingness and ability to underwrite non-investment-grade retailers, where it ran into less competition from those institutions focused on investment-grade tenants. Such deals generally came with not only relatively high cap rates but also a margin of safety in the cash flow coverage of rent built into the sale-leaseback transaction, allowing for some deterioration in tenants' store performance before rent payments were compromised. Over time, therefore, as Realty Income achieved diversification in its portfolio (by adding different types of retail assets such as theaters, convenience stores, and fitness centers to its portfolio for the first time), it was pursuing diversification while paying prices that left a reasonable margin of safety. Although it does achieve diversification, it seems unlikely that a meaningful margin of safety is built into this ARCT deal. The 5.9% cap rate is low by both absolute and relative standards, and the nonretail assets cannot be underwritten to property-level profitability as easily, if at all (so they are then often backed by investment-grade credit). Furthermore, while Realty Income did extensive due diligence on all 501 properties it plans to acquire from ARCT, it was ARCT's management--and not Realty Income's--that originally underwrote these transactions, with Realty Income now paying a premium for that underwriting.
One thing that makes this deal more palatable for Realty Income is its plan to buy out ARCT shareholders with its own stock, which is expensive by historical standards and trades a bit above our own $39 fair value estimate. This is no reason in and of itself to pay up for a deal, though it helps soften the impact of the high purchase price. Furthermore, the deal would be immediately cash flow accretive to Realty Income, prompting the firm to announce plans to increase its annual dividend payment by $0.13 per share, or roughly 7%, if the deal closes.
Primary Benefits of the Deal to Realty Income
Cash flow positive. We agree with Realty Income management that the deal should be immediately cash flow accretive and support an increase to the common dividend. Our new 2013 forecasts for funds from operations and adjusted FFO are $2.30 and $2.33, respectively, which fall at the low end of the new guidance range management provided, and our model supports management's assertion that an increase to its common dividend upon the deal's close of $0.13 per share is manageable.
Further portfolio repositioning. Realty Income has identified two strategic portfolio repositioning objectives: improve credit quality and reduce exposure to retail tenants. This transaction does both. Investment-grade tenants make up 75% of ARCT's portfolio, which would boost Realty Income's exposure to investment-grade tenants to 34% from 19%. The deal would also boost Realty Income's percentage of nonretail assets to 23%, up from 14%, and within the 20%-30% target range management has identified.
Improved short-term portfolio metrics. Given the characteristics of ARCT's portfolio, the deal would immediately improve Realty Income's tenant and industry diversification, occupancy, and remaining lease term, while lowering the average age of portfolio properties.
Scale. This deal would make Realty Income the largest publicly traded triple-net lease company by a factor of roughly 2 times. We believe this scale would solidify its position as a preferred capital provider to the triple-net sale-leaseback industry, enabling it to contemplate an expanded opportunity set, in terms of scope and size, while maintaining its reputation for reliability. Furthermore, we agree that ARCT's operations could largely be absorbed by Realty Income with minimal incremental hiring, resulting in $8 million or so of annual G&A expense savings, which we value at roughly $100 million.
Organic growth potential. While we do not know the exact extent of the opportunity, management has indicated that a number of the nonretail properties in the ARCT portfolio may contain excess land, which Realty Income could utilize in the future to meet tenants' growth needs, providing organic growth and an opportunity to extend lease terms on the original properties. Realty Income's legacy retail portfolio contains little opportunity for this, and we view this type of potential development to meet existing tenants' expansion requirements as carrying relatively lower risk than greenfield development deals in new markets with new tenants.
Our Primary Concerns
Nontraditional transaction at huge scale. While Realty Income has purchased parts of portfolios in the past, it has done nothing near this size. Furthermore, we believe that the vast majority of value from past acquisitions has been concentrated in traditional sale-leaseback transactions with the eventual tenants. The ARCT deal is a departure in size (it is roughly 6 times larger than Realty Income's next biggest deal, done last year) and the fact that it is a portfolio of lease transactions put together by another management team instead of one Realty Income underwrote itself. One of Realty Income's positive attributes, a differentiating factor and the source of its narrow moat, is the firm's underwriting expertise and discipline. This has historically resulted in investing in a high percentage of off-market deals at relatively more attractive returns. On the other hand, this deal allocates $3 billion in capital at a premium of roughly 40% to another firm's original underwriting. Management has assured us that it has reviewed all of the properties and leases in the deal thoroughly, but the deal is nonetheless a departure from the type that has brought success historically, and it is being done at huge scale.
Timing. We would have preferred to see Realty Income in a position to bid more aggressively on some of the ARCT portfolio assets when they were originally up for sale, but it appears that the firm was not yet geared up to source such deals, which is disappointing. By our reckoning, Realty Income made well known its strategy to diversify into nonretail assets and investment-grade tenants in mid-2010, when it purchased Diageo's winery assets, and then confirmed this strategy with its early 2011 $544 million ECM portfolio deal. We suspect Realty Income management was considering the implications of this diversification strategy and potential eventual targets well before the close of its Diageo deal in mid-2010. So, it's disappointing that the firm didn't get more of the ARCT portfolio assets at the time they were originally up for sale, given that ARCT compiled more than half of its portfolio in 2011 at an average cap rate of 7.9%.
ROREA dilution. One of the metrics we consider when evaluating real estate investment trusts is return on real estate assets, which we estimate as EBITDA less maintenance capital expenditures divided by the gross book value of revenue-generating real estate assets. We think this metric provides insight into a firm's value creation from property acquisitions and developments, as it includes an earnings component as well as a price paid component. The higher the ROREA the better, and likely the more value creation management has achieved, all else equal. We estimate ARCT's level of ROREA (based on its second-quarter results) to be 7.6%. Consolidating its business with Realty Income's would increase EBITDA thanks to lower estimated G&A expenses while increasing the cost basis in the assets to reflect the higher price Realty Income is paying. As a result, after the acquisition, we estimate Realty Income's ROREA for the ARCT portfolio to fall to 5.7%. This pales in comparison with Realty Income's five-year trailing ROREA of 8.6% and is well below our estimate of Realty Income's cost of capital. Although we expect the ROREA that Realty Income earns on the ARCT assets to increase over time as annual rent bumps flow through EBITDA, the relatively low level of initial ROREA post-close suggests that the firm is diluting its portfolio's overall relative earnings power while setting a very high hurdle of future performance for the ARCT portfolio to clear in order to approach parity with Realty Income's legacy portfolio.
Potentially tighter spreads on future acquisitions. For much of Realty Income's history, its superior underwriting allowed it access to one-off sale-leaseback deals that many other financiers avoided, mainly transactions with non-investment-grade or nonrated retail tenants. Less competition meant better pricing for Realty Income. Going forward, there is risk that Realty Income will focus its acquisition resources on larger deals, as it will take more and more in annual acquisitions to move the needle on growth. This may mean more portfolio deals or larger sale-leaseback deals, which we worry will be more likely to be market-rate deals with more potential bidders than Realty Income's traditional transactions. This could mean tighter spreads--and less value creation--on future deals than shareholders have enjoyed in the past. We have no reason to think that Realty Income will engage in transactions that don't make sense for shareholders, but the future bang for Realty Income's acquisition buck may not be as great as it has been historically.
Potentially more cyclicality and capex. We think the industrial, distribution, manufacturing, and office assets Realty Income is adding bring different risks to the portfolio because of the economic cyclicality of the assets and in some cases the lack of a strong link between the tenant and the particular asset, especially upon lease expiration. We associate these types of nonretail assets with slightly higher volatility in cash flows and operations and the potential for higher levels of capital spending over time, especially in the event that re-leasing becomes necessary.
Portfolio churn. Normally, when making an acquisition, Realty Income likes to cherry-pick a portfolio to include in its deal only the properties it likes best, leaving the counterparty to find other buyers for the other properties. But the acquisition of ARCT in its entirety doesn't allow Realty Income to do that. Although management has no specific plans yet for divestitures from the ARCT portfolio, we would expect a meaningful percentage (say, 5%-15%) of the ARCT portfolio to eventually be sold over the medium term. While we agree with this strategy, it nonetheless adds uncertainty and will require new acquisitions to offset sales to maintain cash flows for dividend payments. Furthermore, if Realty Income did pay a portfolio premium of 120-230 basis points, as we estimate, flipping assets over the short term would be value destroying and cash flow dilutive, if that portfolio premium disappears upon resale. We estimate that it would take 12 years of 1.5% annual rent bumps in the portfolio for net operating income to recover sufficiently such that potential future divestitures would be value neutral at cap rates 120 basis points higher. There is potentially a silver lining in this portfolio churn. We think that among its nonretail assets, Realty Income is least enamored of suburban office assets, and they are likely candidates for divestiture. This is an area that has lagged the recovery witnessed across most other property types since the last downturn. With potential further firming of the macro economy, suburban office may be a property type for which performance and values improve meaningfully, making it a potentially good environment into which Realty Income can market its assets.
Price. Realty Income is paying a dear price for the ARCT portfolio. The deal's 5.9% initial yield is 120 basis points below the average cap rate of 7.1% that Realty Income paid on its $211 million of second-quarter acquisitions, traditional sale-leaseback transactions, nearly all of which were with investment-grade tenants (as opposed to 75% exposure to investment-grade tenants in the ARCT portfolio). It is 230 basis points below ARCT's average cost basis in the assets. Still, we think the deal is largely value neutral for Realty Income, partially because it is using its own stock for financing, which appears to be a slightly overvalued currency. Nonetheless, at such a low initial yield, there's not a lot of leeway in the portfolio's potential future performance before its value takes a hit. Nearly everything needs to go right with this portfolio, or the price paid may look too high in retrospect. Paying a full price up front shifts more risk to Realty Income, minimizing its margin of safety on the deal.
What Does the Deal Signal About Realty Income's Future?
We view the deal itself as value neutral. But we think its potential implications point to the risk of lower acquisition spreads going forward and slightly lower long-term performance assumptions for Realty Income's portfolio. Furthermore, although this deal is expected to provide a 7% boost to Realty Income's dividend upon closing, we think it may also slightly reduce future dividend growth prospects. To the extent that the nonretail assets do require more cash for maintenance and do ultimately display greater cash variance in cash flows and operations, the board may consider an additional buffer in its dividend payout ratio, perhaps moving from a mid- to high 80s level of adjusted funds from operations to a low to mid-80s level instead.
We expect most of the potential operating changes to Realty Income's business model to show up over the medium to long term and not in the near term (although there may be perception risk over the near term). This is because we view most of the potential risks to the performance of the nonretail assets to materialize at the end of the lease term. The ARCT portfolio, for example, has a 13-year weighted average lease term, with few material lease expirations until 2018. Given that the majority of these leases are with credit tenants, we view them to be relatively more reliable over the lease term than those with Realty Income's traditional non-investment-grade retail tenants. But we also view the leases on nonretail properties to be less predictable at lease expiration, given that the money the tenant earns is not always tied as closely to the specific property it leases.
On the other hand, we expect to be able to effectively measure management on its acquisition discipline over the next 18-24 months, which will help us confirm or refute our position regarding the risks to Realty Income's spreads on future acquisitions. Despite its $3 billion ARCT bid, Realty Income has confirmed that its acquisition pipeline remains robust and its acquisition capacity remains large. The company still plans to acquire roughly $1 billion over the next 18 months or so. We will be watching to see whether spreads are pressured and whether the mix leans toward market-rate deals or its traditional off-market variety.
Although we have lowered some of our expectations for Realty Income, the firm remains one of the best-positioned real estate investment trusts that we cover. Even after the deal's close, the firm's balance sheet should remain strong, its dividend should remain well covered, and its capacity for incremental cash-flow-accretive acquisitions should remain large. We still think Realty Income has a narrow economic moat, thanks to its favorable triple-net leases, exemplary management, and historically superb underwriting.
If the deal fails to close, we will revisit our recent model changes. However, we anticipate little change to our model assumptions reflecting the portfolio's shift toward more nonretail assets with investment-grade tenants and tighter future acquisition spreads, as the ARCT deal--whether or not it closes--suggests that the company will eventually execute on its diversification strategy, perhaps with lower-yielding portfolio deals along the way.
Todd Lukasik does not own (actual or beneficial) shares in any of the securities mentioned above. Find out about Morningstar’s editorial policies.