Ever since Tesla rolled out its autopilot capabilities three years ago, the prospect of self-driving cars has captured the public’s imagination. The gating item that most prominently stands in the way, however, has little to do with the technology. A more significant barrier is the communications infrastructure, which in its current state is insufficient to support the data needs required for mainstream adoption.
When investors consider the infrastructure opportunity, the country’s crumbling roads and outdated electrical grids command the most attention. Communications infrastructure, however, is emerging as a priority. The ongoing transition to 5G, for instance, will one day enable a range of new technologies. Outside of consumer applications, the administration’s move to block the Broadcom/Qualcomm merger underscores a view that leadership in this area is indeed an issue of national security.
Investors have taken notice. Three years ago, more than $5 billion flowed into private-market communications investments. The capital invested in 2018 is on pace to roughly double that. Success for investors will rest on their ability to identify geographies with the most pronounced expansion needs and excess capacity, while recognizing the latent growth potential of the customer base to manage risk as valuations increase and ensure long-term and stable cash flows over time.
According to the Cisco Visual Networking Index, global Internet Protocol (“IP”) traffic is expected to increase by 24% annually through 2021, while traffic of mobile data is expected to increase 46%. According to current projections, the Asia Pacific region and North America will account for more than 75% of the total IP demand, globally.
Three years ago, more than $5 billion flowed into private-market communications investments. The capital invested in 2018 is on pace to roughly double that.
This creates an opportunity across the three primary subsectors: wireless infrastructure, wired fiber networks, and data centers. Each segment already has acute needs for further capital investments. But as IP traffic continues to grow, advancing capabilities – from connected devices to self-driving cars – will create an enormous need for additional infrastructure yet to be built.
An institutional investor’s allocation to communications infrastructure can be diversified by location, subsector, and vehicle type. Within private capital allocations, communications exposure can be obtained through open- or closed-end vehicles. Sector specific managers have also emerged that are offering closed-end funds to access communications infrastructure through a focused portfolio. While several investment options exist, it helps to understand both the drivers and risks inherent to the various categories.
Macro cell towers, small cells and spectrum comprise the wireless infrastructure segment. Macro cell towers are the large-scale vertical structures typically found in rural areas and along highways. Wireless carriers originally owned and operated their own cell towers. As demand and costs grew, independent tower companies absorbed more and more marketshare and, today, will lease tower space to as many as five carriers at a time. Lease contracts typically run five to 10 years and provide a dependable revenue stream.
Small cells, in contrast, are low-powered antennas typically installed in densely populated, high-usage areas that require improved signal and speed. Small-cell equipment is often situated on large office towers or utility poles.
To put the market opportunity in perspective, with a 4G network, five small cells can produce the same output as one macro cell tower; with 5G, it will require 25 small cells to equal one macro cell. Moreover, current estimates predict that each self-driving car will generate the same volumes of data as approximately 3,000 people based on current usage patterns. This helps explain how 5G deployment could see the 50,000 outdoor small cells currently in operation grow four-fold over the next five years.
Spectrum, meanwhile, refers to the radio frequencies in the electromagnetic field within which wireless-communications signals travel. Spectrum is a finite resource as the same frequency cannot be exploited by different users in the same geographic area. The two ways that carriers can acquire spectrum is through government auctions, overseen by the FCC, or secondary-market sales in which acquired licenses are sold or leased to carriers.
Wired Fiber Networks
Fiber refers to the bundle of glass threads that connect the entire communications infrastructure. Fiber is the primary mode of transport between the internet, wireless towers, and data centres across the world, extending across continents through metro networks, subsea cables, wirelines, and long-distance cables. The prevailing strategy of operators in this segment is to continuously build new networks or expand the coverage of existing networks to reach economies of scale.
Fiber can be monetised through sales of either “dark fiber” or “lit fiber.” The sale of dark fiber generally involves larger customers who make significant up-front commitments and then control the data transported over the networks with their own optical gear. Lit fiber, alternatively, is generally sold as a service, allowing smaller customers to transfer data across the shared network for a recurring fee.
Current estimates predict that each self-driving car will generate the same volumes of data as approximately 3,000 people based on current usage patterns.
Finally, data centres are the specialised buildings that house mission-critical IT infrastructure. Investments in data centre infrastructure can target either hyper-scale data centre access or enterprise data centres.
The former, typically the lowest-cost option for large customers, refers to facilities in which a large tenant will lease entire buildings to run back-office, cloud and content infrastructure, as well as data analytics and web-hosting services. Renewal rates for leases that can range from five to 20 years are exceptionally high due to the prohibitive switching costs. This generally lends to low-but-steady cash yields for investors.
Enterprise data centres, meanwhile, cater to smaller businesses that may require less space but more service. The smaller businesses utilising enterprise data centres typically pay a premium rate compared to hyper-scale customers, but enterprise data centres generally manage the servers on behalf of their customers.
Staying Attuned to the Risks
As data demands grow and IP data usage escalates, each of these segments should benefit from an enduring tailwind. That’s not to say that these investments don’t carry certain risks.
Carrier consolidation, for instance, would impact the underlying economics of small cells and macro cell towers, which could also be affected by the expansion of small-cell networks in certain geographies. Spectrum, similarly, would be affected by end-market consolidation and typically carries additional risk around entry valuations given the complexity of FCC auctions.
Fiber, meanwhile, has to contend with supply-side threats and the potential for overbuilt conditions, particularly in densely populated areas. Permitting delays can represent another challenge.
And within the data centre segment, hyper-scale operators are susceptible to customer-concentration risk, which can be offset with the longer lease contracts and price escalators. Enterprise data centres, on the other hand, face competitive threats, as Amazon, Google, Microsoft, and now IBM have all made inroads in the segment.
But the opportunity for investors – particularly ahead of 5G deployment and data intensive innovations – increasingly outweighs the risks. Like any private asset category, however, access and experience will be key to both locate and fund compelling assets and, ultimately, back management teams with the requisite skillset to capture the industry’s growth while staying attuned to the risks.
Adam Toczylowski and Lisa Bacon are vice president and senior vice president, respectively, at Meketa Investment Group.