CyrusOne And Data Center REITs: A Top Pick For A Turbulent 2016 Market

| About: CyrusOne (CONE)


Data center REITs have already experienced tremendous growth, and the colocation industry should continue to grow at a double-digit CAGR through 2020.

With a recent dip in benchmark bond yields and bearish expectations of FOMC rate hikes, the US macroeconomic outlook may have turned in favor of REITs.

CONE trades at an ~8% discount to my NAVPU despite rapid growth and a ~25% run-up in its stock price over the past year.

The company will continue to capitalize on the surging demand for data centers in the US and should continue to grow revenues, EBITDA and AFFO at double-digit rates.

Company Overview

CyrusOne (NASDAQ:CONE) is a real estate investment trust that owns, operates and develops enterprise and multi-tenant data centers. Based in Carrollton, Texas, the company was spun out of Cincinnati Bell (NYSE:CBB) on November 20th, 2012, when it launched its IPO on the NASDAQ. Cincinnati Bell now holds a ~9.5% interest in CONE, down from its initial majority interest of ~69% post IPO. CONE has ~921 customers concentrated in 12 markets that operate out of 31 data centers and two recovery centers. Its primary operations are located in Texas, Ohio, Phoenix, Illinois and Virginia. Recently, the company developed substantial market share in New York City with the acquisition of Cervalis. CONE's customer list boasts nine Fortune 20 members. Its revenues are very secure, with ~75% of revenue generated from Fortune 1000 companies.

Macroeconomic Outlook: Bond Yields, Oil Prices and The Strength of the US Economy

As an equity REIT, CONE is required to distribute 90% of its pre-tax earnings, making its trading value extremely sensitive to benchmark bond yield fluctuations. Changing bond yields impact CONE's cost of capital and are a comparator to its ~3.5% distribution yield. After the FOMC raised the target for the Federal Funds rate by 25 bps in December, REITs were hit hard. The US MSCI REIT Index was down ~8% after the hike, and the S&P/TSX Capped REIT index fell by a comparable level.

However, I believe that the outlook for interest rates and bond yields may have turned in favor of REITs.

The FOMC's original plan was to raise the Federal Funds rate four times this year (for an expected 100 bps worth of hikes). Given the recent turbulence in the global financial/equity markets and significant excess capacity in the global economy, the Fed is unlikely to raise rates four times this year. An 80% jump in the US policy sensitive two-year yield from October until the rate hike suggests that the market was already pricing in future rate hikes. With Fed Funds futures now pricing in little chance of a rate hike in 2016, the market has scrambled to re-price expectations of FOMC rate hikes in Treasury yields. The two-year US Treasury yield has fallen ~70% to pre-hike levels and the 10-year yield has fallen a comparable amount as investors sell off equities and pile into government bonds in a flight to safety.

If bond yields remain on their current trajectory, REITs, including CONE, should prosper due to a reduction in their cost of borrowing and the increased attractiveness of their distribution yields.

Given that ~20% of CONE's annualized rents are tied to energy companies, many of whom have weak balance sheets and severely constrained cash flows in light of low oil prices, falling/stagnant oil prices represent a moderate risk for the company. The outlook for crude prices remains volatile and gloomy. In my opinion, crude prices are likely to remain under $50/barrel over the medium term as a result of lifted Iranian sanctions, a strong US dollar, significant OPEC/shale supply and seasonal fluctuations.

Microeconomic Outlook: Data Center REITs: Favorable Demand Outlook and Inefficient Pricing

An increasing emphasis on data-driven decisions and cloud computing globally has provided data center REITs with a favorable demand backdrop over the last several years. Double-digit demand growth for data centers has been buoyed by corporate America's long-term trend towards outsourcing IT infrastructure via IaaS (infrastructure as a service), coupled with a massive increase in the amount of data available on the web.

Accordingly, a recent CNBC article highlighted the world's rapid data infrastructure growth, with CyrusOne CEO Gary Wojtaszek stating that:

"Of all the data that exists today in the world, 90% was created in the last two years."

Since the majority of CONE's customers are Fortune 1000 corporations that can afford to and are willing to pay premiums for data-related services, this environment provides the company with significant upside potential. Leases for data center services are typically 7-10 years with low churn rates and contain clauses to raise rents via escalators in areas of high utilization. Maintenance capital expenditures are surprisingly low in the sector, consuming ~3-5% of the average data center REIT's total revenue.

Data center REITs offer an attractive combination of stability and rapid growth despite minor contractions in the economy since data and IT spending will likely remain a priority in today's data driven world.

Some analysts argue that data center REITs are less efficiently priced than the broader index and traditional REITs since the sector is quite difficult to understand. REITs are complicated enough, but few people understand data analytics or the services that data center REITs offer. Valuation methods are complicated, since data center REITs lie somewhere between technology companies and REITs. With so much more to understand than the traditional real estate methodology of "location, location, location," some investors may not take the time to understand data center REITs and flock to safer, less speculative sectors. On average, research suggests that data center REITs trade at 2-4x EV/EBITDA, P/FFO and P/AFFO lower than Tower REITs despite average growth that is 300-500 bps higher. The valuation modeling contained in this article will further explore the gap between data center REIT multiples and other REIT sub-sectors.

Investment Thesis I: Rapid and Scalable Growth, Excess Capacity and Margin Expansions

Over the past two years, CONE has grown revenue, EBITDA and FFO/AFFO at double-digit levels, highlighting the company's ability to execute and grow the business profitably. Revenues for 2015 are expected to grow by ~20%, with EBITDA growth above that at ~23% and expected AFFO growth of ~39%. During the 1999-2000 tech bubble, data center supply massively exceeded demand as tech companies put the brakes on capital expenditures. As a result, real estate prices plummeted and REITs suffered. However, the nature of the industry has changed after these failures, and demand for data centers is not nearly as speculative as it once was, highlighting the shift toward stability in the industry.

CONE's growth platform is highly scalable, with enough existing building shell space to almost double its capacity and 190 acres of land under management for future development (enough to build 5 million csf worth of data centers). The highly scalable nature of CONE's platform suggests that incremental free cash flows will significantly exceed incremental costs, boosting margins as the company capitalizes on high demand for data centers over the foreseeable future. Development yields remain extremely impressive, with historical yields ranging from ~16-19%. Larger competitors, such as Digital Realty Trust (NYSE:DLR) do not have comparable growth platforms that are highly scalable.

CONE's margins also are supported by opportunities to raise rents, even as the REIT seeks to mitigate risk by lengthening its lease terms. CONE's average lease term signed over the last four quarters is 6.5 years, up significantly from its average 3.5-year lease term. Over three quarters of these leases have an escalator clause that stipulates that CONE can raise rents by ~2.6% annually. It is less common to have such escalator clauses in long-term leases. For example, Smart REIT (TSX:SRU.UN), a Canadian retail REIT of comparable scale and market capitalization, is locked into eight- to 10-year lease terms with Wal-Mart (NYSE:WMT) with no built-in escalators.

CONE also is growing its high margin colocation revenue at a rapid pace, which is discussed in Investment Thesis II.

Investment Thesis II: Favorable M&A Action and Significant Capital Returns to Shareholders

With the acquisition of Cervalis in July 2015, CyrusOne was able to diversify away from its exposure to the energy sector (since Cervalis serves mainly financial verticals in metro New York) and expand its presence in the higher margin colocation business. Annualized rents attributable to the energy sector fell from 37% in December 2012 to 20% in September 2015. Analysts also estimate that the acquisition was accretive to CONE's normalized FFO by as much as 9% per share. Further industry consolidation should prove beneficial to CONE by enabling the company to diversify its energy sector risk in light of falling oil prices and grow the business. Strong incremental cash flows from the utilization of significant amounts of existing shell space (refer to Investment Thesis I) should enable CONE to de-lever in order to make meaningful acquisitions.

Continued divestitures by Cincinnati Bell have provided CONE with a much needed improvement in its stock's liquidity. CBB's ownership has fallen continuously, from ~66% post IPO to ~9.5% at today's levels. As mentioned in Investment Thesis I and in the Investment Catalysts section, CONE's double-digit growth in revenues, AFFO and EBITDA post IPO has enabled the company to double its quarterly dividend in just two years. Comparably, AFFO payout ratios rose only ~8% from 2014-2015 and remain below the data center REIT peer median, despite anticipated hikes to CONE's quarterly distribution.

Catalysts For Future Growth

1. Further accretive M&A to diversify to higher-margin businesses or acquire strategic assets

As previously stated, the Cervalis acquisition was highly beneficial to CONE, enabling the company to expand its margins, boost normalized FFO/share by ~9% and diversify risk. Further industry consolidation/M&A at relatively low purchase multiples could enable the company to purchase strategic assets, expand its margins and further capitalize on the growing demand for data centers. Rapidly growing FCF and AFFO should allow CONE to de-lever from elevated debt levels and eventually finance acquisitions at historically low interest rates, provided that long-term bond yields remain on their current trajectory.

2. Strong earnings reports associated with price increases and expansion to higher-margin businesses

Since its IPO in 2013, CONE has beat analyst consensus earnings several times, including its December 2014 FY earnings beat by ~14%. As CONE realizes synergies and strong FFO/AFFO accretion from its Cervalis acquisition, continues to grow margins by utilizing its excess capacity and grows its colocation and interconnection businesses, the company should continue to meet/exceed earnings expectations, supporting further increases to its share price.

3. Distribution increases driven by rapid AFFO growth and FCF generation

Double-digit revenue and AFFO growth over the past two years has enabled CONE to grow its quarterly distribution from $0.16 post IPO in 2013 to $0.32 in 2015, representing a dividend growth of 100% over two years (implying a CAGR of ~41%). As CONE continues to capitalize on high demand for data center space and boosts margins and AFFO, the company should be able to continue meaningful increases to its dividend. Several research analysts have already predicted dividend growth of ~17% in 2016. Dividend increases signal management's confidence in the business and should boost the company's share price.

Investment Risks

1. Potential for oversupply of data center real estate, leading to severe pricing pressure on CONE

In the past, demand forecasting for data centers has been highly speculative. In the 1999-2000 tech bubble for example, supply rapidly exceeded demand for data centers as companies severely cut capital expenditures. While demand for data centers remains high, if a sudden demand pullback occurred or if supply significantly exceeded demand due to speculation, real estate prices could fall considerably. CONE and other data center REITs would be forced to cut rents to compete, putting considerable pressure on the company's top line, AFFO figures and payouts. Although this could negatively impact CONE's growth, the nature of data center demand forecasting has become less speculative than in the past, as previously noted.

2. Impact of below-industry-average lease lengths on CONE's churn rate

As previously noted, CONE's average lease term is ~3.5 years vs. the industry average of 7-10 years. While this likely reduces some of CONE's sensitivity to interest rates and may allow it to raise rents, it also gives customers the opportunity to leave CyrusOne and switch to a competitor, such as QTS (NYSE:QTS) or Digital Realty, driving up the company's churn rate. However, as previously mentioned, this risk is largely mitigated by CONE's one-year performance in renewing leases. Weighted average lease terms have jumped to 6.5 years while still allowing for the opportunity to raise rents annually via a built-in escalator.

3. Exposure to the energy sector

Approximately 20% of CONE's annualized rent revenue is attributable to companies in the energy sector (primarily oil and gas companies). Given the recent 70%-plus drop in oil prices over the past year and a half, most O&G companies have weak balance sheets and severely constrained cash flows. The concern is that O&G companies will cut spending on data in order to stay alive and continue production. However, it should be noted that many O&G producers have other means of protecting cash flows, such as cutting dividends. Additionally, vertically integrated producers (like Exxon Mobil (NYSE:XOM) and Suncor (NYSE:SU)) continue to generate strong cash flows from their refining operations in times of low oil prices. Many upstream producers are also locked into favorable hedges, which should provide some support for producers in the short term.

4. Sensitivity to interest rates/long-term bond yields and access to the capital markets

As previously mentioned, REITs are generally highly sensitive to fluctuations in long-term bond yields and interest rates. At a B+ credit rating, CONE is not yet rated investment grade. I performed a regression analysis in an attempt to determine the magnitude of the impact of rising US 10-year yields on CONE's share price relative to other REITs.

Relative to the US benchmark REIT index, the results of my regression analysis were indeterminate. While the CONE regression line has a steeper slope, indicating greater sensitivity to bond yields, the r-squared coefficient on the index regression was moderately higher. This indicates that a greater proportion of the variance in the MCSI index price can be explained by long-term bond yields, whereas a lower proportion of the variance in CONE's share price can be explained by long-term bond yields, all else equal.

Therefore, while CONE is certainly sensitive to long-term bond yields, we cannot determine if the stock exhibits greater sensitivity to long-term bond yields than the benchmark US REIT index.

Valuation: Net Asset Value (NAVPU)

CONE's NAVPU uses consensus 2016E NOI of $300.1 million and a target blended capitalization rate of 7.50%. My NAVPU was constructed using cap rate and NOI estimates from Jefferies and company balance sheet data as of September 30th, 2015. While some equity research analysts value CONE using a DCF, a NAV valuation seems more appropriate - since the company generates income specifically from its assets and investment properties.

Since an estimate of NAVPU can fluctuate significantly based on the target cap rate and predicted NOI, a sensitivity analysis of CONE's NAVPU is shown for a wide range of cap rates and NOI. More realistic scenarios are shaded in gray.

At my predicted target cap rate and NOI, CONE's shares trade at a ~8% discount to net asset value, despite their recent ~25% run-up in price over the past year.

Valuation: Target Price Calculation

Since the companies in CONE's comparable universe are direct competitors with similar margins, customers and corporate structures, I believe the company should trade at the median forward P/AFFO multiple of its peers. After applying the median P/AFFO multiple to consensus 2016 expected AFFO per share, I obtained a target price of ~$44/share.

After weighting the NAVPU at 40% of the target price and the relative valuation at 60%, a target price per share of ~$42/share is expected. At current prices and CONE's ~3.5% dividend yield, this implies a one-year equity upside of ~21%. It should be noted that a relative valuation was weighted slightly more heavily than NAVPU because of the significant, unwarranted discount that data center REITs trade at relative to P/AFFO levels for other American REIT sectors, despite significantly lower growth potential for these other sectors.

The sensitivity analysis for the relative valuation highlights a range of probable target share prices based on P/AFFO multiple expansions and forward AFFO/share levels. Price targets shaded in gray are more realistic estimates of CONE's potential equity upside. Note that this sensitivity analysis does not include the NAVPU component of my target price calculation.

The consensus mean price target on CONE is $41.50, with a median of $41.00 and a standard deviation of 3.59 among analysts. Of the 15 equity research analysts surveyed, the highest price target is $48 and the lowest price target is $35.50, which is roughly in line with CONE's current share price.

Valuation: Comparable Company Analysis

Based on my comparable company analysis of data center REITs and CONE relative to the aggregate American REIT universe, the company trades relatively in line with its peers on most metrics, except for P/AFFO, where it trades at a moderate discount. CyrusOne trades slightly above the median LTM EV/EBITDA peer multiple, in line with median LTM and forward P/FFO multiples and well below LTM and forward median P/AFFO multiples. However, CONE's distribution yield and forward AFFO payout ratio are slightly lower than its peers. However, CONE carries slightly more debt relative to market cap than its peers, with debt/TMC of ~45% vs. the peer median of 43%.

Personally, I believe that CyrusOne should trade at a slight premium to its peers, given its superior growth outlook and scalability. I believe that this undervaluation will soon be corrected, and that CONE's P/AFFO multiple will expand to peer median levels within a year.

Conclusion - Long CONE

The data center REIT sector remains highly lucrative, given an attractive growth and demand outlook, coupled with significant undervaluation on a P/FFO and P/AFFO basis relative to other REIT sectors. CONE is a stable, undervalued company that will continue to capitalize on surging demand from corporate America for data centers and continue to drive double-digit revenue and AFFO growth.

Disclosure: I/we have no positions in any stocks mentioned, and no plans to initiate any positions within the next 72 hours.

I wrote this article myself, and it expresses my own opinions. I am not receiving compensation for it (other than from Seeking Alpha). I have no business relationship with any company whose stock is mentioned in this article.