Why Digital Infrastructure is Real Estate 2.0

A lot of readers here at Dgtl Infra have been asking us…

• Why do you call Digital Infrastructure, “Real Estate 2.0”?
• Why is Digital Infrastructure considered Real Estate?
• What makes towers, data centers, fiber, small cells, and distributed antenna systems like Real Estate assets?

We discuss precisely why Digital Infrastructure is “Real Estate 2.0” and how it will re-shape the future of real estate. Specifically, we analyze the characteristics that make Digital Infrastructure real estate-like, why Digital Infrastructure is becoming more relevant than traditional real estate, and the key demand drivers in-place for Digital Infrastructure’s growth.

Real Estate vs. Digital Infrastructure

Real Estate asset classes are really struggling to survive in today’s world. Traditional real estate’s, growth is questionable, even with interest rates at all-time lows and public equity markets near all-time highs. However, every category of real estate is incurring massive amounts of capital expenditures to remain relevant. Therefore, the notion of physical, financial, and functional obsolescence of traditional real estate exists.

In retail, Amazon and Shopify are questioning whether brick-and-mortar retail should exist anymore. In office, WeWork and Spaces are providing high-density office space on short-term leases, pressuring the duration on traditional office leases. Additionally, the work-from-home environment, is raising the question as to whether people need to even return to offices again, or at least whether the same amount of people do. Finally, in hospitality, Airbnb and Vrbo are adding supply to the market. At the same time, Expedia is offering transparent pricing, further pressuring traditional hotels.

Overall, these themes have caused year-over-year declines in the revenue of traditional real estate verticals. Tenants are putting pressure on their landlords, whether it be shopping malls or office buildings, to lower rents. For hotels, occupancy and thus revenues have declined dramatically.

However, long-term contractual revenues from creditworthy counterparties are attainable, and much more secure in the “digital” frontier than the “legacy” real estate space. Digital infrastructure’s characteristics include stronger:

• Lease Tenure: long-term rental contracts
• Covenant Strength: high-concentration exposure to investment-grade counterparty risk
• Churn Rate: low churn
• Organic Cash Flow: cash flows that are growing with customers that are growing – which is the most important factor

Organic revenue growth occurring in the main verticals of digital infrastructure is much stronger. Specifically, towers have grown at 5% year-over-year, data centers have grown at 10% year-over-year and fiber has grown at 5% year-over-year.

Digital Infrastructure – Real Estate Characteristics

Each of the Digital Infrastructure sectors of towers, data centers, fiber, small cells, and distributed antenna systems, originated within the telecom services sector.

However, in many ways these sectors have evolved to be more comparable to the real estate sector, with specific examples including:

Characteristics of Digital Infrastructure and Real Estate

Pricing of Assets

To determine the correct selling price for assets, investors price both real estate and digital infrastructure, based off of stabilized yield trends. Additionally, in digital infrastructure, pricing for assets increases, in-line with inflation, which is akin to other real estate and infrastructure sectors. This is particularly true because the lease contracts (e.g., towers) have embedded rental escalators built-in, ranging from 1% to 3%.

Contribution Margin

The tower business model demonstrates significant operating leverage as tenancy increases. Specifically, the incremental return on the second, third and fourth tenant added to the infrastructure, ranges from 80% to 98% pure profit. Operating leverage is very much a characteristic of real estate as well.

Lease Duration and Quality of Covenant

Long-term leases of 10- to 15-years or more, are very common on tower, small cell, and dark fiber assets. There is also increasing recognition that hyperscale customers are investment grade counterparties and sign long-term leases. Typically, hyperscale lease contracts have a term ranging from 10 to 15 years.

Land and Building Ownership

Digital infrastructure owners and real estate owners typically own the land and building that houses the tenant (e.g., data center). Expenses associated with digital infrastructure and real estate properties are also similar. Examples include ground rent, monitoring, insurance, real estate & municipal taxes, utilities, and site maintenance.

Further, tenant-related capital expenditures are quite similar. For example, data centers have much longer sales cycles and need to be leased before installing equipment for tenants. This is similar to the concept of needing a pre-lease on an office building before fitting it out for a tenant.

Digital Infrastructure Companies are Real Estate Investment Trusts (REITs)

8 of the 15 public companies described as Digital Infrastructure are real estate investment trusts (REITs). Therefore, these companies have met the legal guidelines established by the IRS, and include:

• Towers: American Tower, Crown Castle, and SBA Communications
• Data Centers: Equinix, Digital Realty, CyrusOne, CoreSite, and QTS Realty Trust

In addition, several tower, data center, and fiber companies are now a part of various real estate indexes. Vanguard, for example, added towers to its key REIT ETF in 2018 and aligned it with the Global Industry Classification Standard (GICS) methodology for classifying global market sectors used by S&P and MSCI.

Overall, the reduction in returns on new assets being generated, and thus higher pricing for existing digital infrastructure assets reflects investors increasing recognition of the sector as infrastructure and real estate-like. Whereas previously, it was being considered more of a telecom and technology sector. This positive investor perception reflects the maturation and validation of the sector as a lower risk asset class.

Key Demand Drivers In-Place for Digital Infrastructure

Demand drivers including i) mobile video usage, ii) cloud computing, iii) a handset replacement cycle, iv) the Internet of Things (IoT), and v) artificial intelligence, are driving more of a need for Digital Infrastructure and in turn adoption of mega trends such as 5G technology. As more digital infrastructure is needed, there will be less demand for traditional real estate.

Mobile Video Usage

Video will account for ~70% of mobile network traffic in 2022, up from only 4% in 2015. Video’s rise as a percentage of network traffic is being driven by the continued proliferation and use of video streaming apps, also known as over-the-top (OTT) media services. Examples of OTT services include Netflix, Hulu, Prime Video, Disney+, Peacock and major TV stations and sports leagues placing their content online via apps. Additional drivers of mobile video usage are the proliferation of videoconferencing, telecommuting, fitness online, telemedicine and distance learning.

Beyond the content and forms of video use, quality is also driving increased mobile video consumption. Higher quality video streaming, such as 4K ultra-high-definition video uses 2.5x more bandwidth than 1080p full high-definition. This means that more data needs to travel over the network as users upgrade their viewing experience.

Additionally, the increasing proliferation of uploaded live video to platforms such as Facebook Live, Instagram Live and YouTube Live, is growing rapidly. Live video streams are being uploaded by celebrities and social media influencers who are streaming, on a live basis, to 100s of thousands or millions of their followers simultaneously.

Based on all of the above trends, mobile video data traffic globally is projected to grow 9x from 6.8 exabytes per month in 2017 to 60.9 exabytes per month by 2022. This represents a compound annual growth rate (CAGR) of 55%.

Cloud Computing

Global enterprise IT spending over the last ten years (2009 to 2019) shows that annual spending on cloud infrastructure services has gone from virtually zero to over $100bn. For, Q2 2020 spend on cloud infrastructure services was ~$30bn for the quarter, an increase of $7.5bn from the second quarter of 2019. This signifies that the annualized run-rate for cloud infrastructure services is currently $120bn. Key cloud infrastructure companies include:

• Amazon Web Services (AWS) which has a 33% market share, equating to a $40bn run rate
• Microsoft Azure which has an 18% market share, equating to a $22bn run rate
• Google Cloud which has a 9% market share, equating to an $11bn run rate

An example of how this relates to Digital Infrastructure is that Hyperscale companies (e.g., Amazon Web Services) will need more data center infrastructure. Infrastructure is needed in order to continue to fuel the rapid increase in revenues from their cloud infrastructure services.

Overall, cloud traffic globally is expected to grow 3.3x from 6.0 zettabytes per month in 2016 to 19.5 zettabytes per month by 2022. This represents a compound annual growth rate (CAGR) of 27%.

Handset Replacement Cycle

Two key trends are pushing the handset replacement cycle. First, old 4G phones create an appetite for an upgrade cycle. Second. the proliferation of unlimited data plans, will push more users to 5G-based phone plans. 5G will accelerate smartphone replacement cycles, particularly with consumers that have been pushing out purchases in anticipation of new 5G smartphones. Ultimately, this will result in a multi-year tailwind for the handset industry.

The average age of the smartphone installed base has increased by >3 months from 16.2 months to 19.5 months in the last three years. Over the same period, the implied age at replacement of devices (when consumers are asked how frequently they plan to replace their devices) has increased by 2.5 months from 26 to 28.5 months. 5G smartphones will help to unwind these trends in favor of a more rapid handset replacement cycle.

Internet of Things (IoT)

Internet of things (IoT) is a network of physical objects (or things) containing embedded technology to communicate and interact with their internal states or the external environment. “Connections” in a 5G world are no longer just about phones and tablets, but will largely involve low-cost, low-battery power, and low-complexity sensors & devices to support a massive Internet of Things environment. Examples of Internet of Things applications include smart cities, agriculture, asset monitoring and fleet management.

The Internet of Things has two sub-groups. Firstly, personal devices such as wearables and smartphones. Secondly, devices with industrial use applications, known as the Industrial Internet of Things (IIoT), which is based on sensors and actuators to collect digital data.

1. Personal device adoption is accelerating as consumers use Internet of Things products in their home. Examples include: smart speakers (Google Home, Amazon’s Echo and EchoDot, Apple’s HomePod), security cameras, smart plugs, smart lighting, smart air conditioners and smart appliances
2. Industrial Internet of Things (IIoT) technology will heavily rely on 5G given latency improvements which enable Industry 4.0 and robotics in manufacturing

Additionally, a paradigm shift has occurred allowing previously un-addressable applications to become both serviceable and profitable with the Internet of Things. Specifically, the Internet of Things is growing as a result of low-battery power devices that can operate in the field for more than 10 years.

At the end of 2018, there were 9.1 billion total connections in the Internet of Things. By 2025, total connections grow to 25.2 billion worldwide, a compound annual growth rate (CAGR) of 14%.

Artificial Intelligence

Information-intensive industries will see the most productivity from artificial intelligence improvements initially. Nevertheless, artificial intelligence will eventually apply to almost every sector. It is worth noting that “artificial Intelligence” incorporates the fields of machine learning and deep learning.

Artificial intelligence is complementary to humans, and it works best in augmenting human tasks. In turn, this automation will drive a step function increase in productivity for many industries. Specifically, artificial intelligence will automate mundane tasks, with robots in manufacturing and self-driving cars – considered the most promising use cases. Additionally, artificial intelligence has uses for legal, financial, medical diagnosis and tax preparation services. In these applications it will allow professionals (e.g., lawyers, doctors and accountants) to focus on more value-added tasks.

Examples of Consumer-Facing Artificial Intelligence Use Cases include:

• Amazon, Netflix and Spotify recommendation engines
• Smart assistants (e.g., Alexa, Siri and Google)
• Transcription (e.g., Google), which recognizes 120 languages and automatically identifies spoken language
• Smart phones have voice recognition, predictive texting and directions built-in
• Mapping applications such as Google Maps and Waze
• Google Search and Mail

Artificial intelligence is projected to add $13 trillion to the global economy by 2030. This is equivalent to 25% of GDP growth per year, which is mostly derived from these massive productivity improvements.