Connectivity and data storage as a ‘core plus’ infrastructure asset
April 2018 | SPECIAL REPORT: INFRASTRUCTURE & PROJECT FINANCE
Financier Worldwide Magazine
April 2018 Issue
In recent years, funds investing in infrastructure have shown much greater interest in connectivity and data assets as part of the ‘core plus’ infrastructure asset class.
Drivers of infrastructure investment
There are a number of reasons for this increased interest including continued exponential growth in data use and decentralised data storage and processing, traditional telecom players deleveraging by monetising their network infrastructure, and the ever-increasing flow of money into infrastructure funds looking for new opportunities that have typical infrastructure characteristics.
Data demand is growing because methods of delivering content are becoming more online (Netflix v. DVD), consumers expect to communicate in ways that demand higher bandwidth (video v. voice, graphics and filters v. simple SMS), consumers expect to interact digitally on the move (mobile devices and networks v. PCs/TVs and fixed broadband) and machines are communicating with machines (Internet of Things, vehicle telematics and so on). As well as requiring bigger and newer networks to carry all this data, it needs to be stored somewhere. There are endless statistics and commentaries on the explosion in data consumption. One of the most often repeated statistics is that 90 percent of all data was created in the last two years. Modern networks and data storage and processing facilities require large amounts of capital to build and upgrade, and it is beyond doubt that there will be long-term demand for them.
Within the connectivity industry, this is resulting in a significant realignment of invested capital. Telcos are investing in innovation and end-customer offerings, providing a route to valuation closer to tech companies, distinct from the infrastructure-operating assets with lower but more stable valuations. This results in a shift in the investor base, reflecting the different risk appetites as assets are disintermediated.
Asset types and infrastructure characteristics
Connectivity and data storage assets include: (i) regional and specialist broadband networks; (ii) metro-area fibre networks; (iii) core telecom and internet networks; (iv) mobile or cellular network towers; and (iv) data centres and hosting providers.
The characteristics of infrastructure assets include: (i) predictable long-term demand; (ii) long-term contracted revenues; (iii) quasi-monopolistic asset presence; and (iv) limited credit risk concentration.
One of the most established infrastructure investments is cellular or mobile network tower assets, which have been packaged into sale-and-leaseback transactions by US cellular operators to such an extent that these represent the majority of tower sites in the US. Although mobile operators – being keen competitors – have traditionally shared space on their own towers on a limited basis only, the infrastructure-investor-backed tower company offers a genuinely independent service to the mobile operators. The opportunity to open up networks to other telecoms companies offers infrastructure funds access to a greater proportion of wholesale customers – moving the revenue base away from consumers and reducing exposure to changing user habits, such as switching providers. The model has also been very popular in Africa, and is beginning to become more so in Western Europe and Asia. It allows capital-constrained mobile operators to focus their investments on network and service technology, rather than on real estate, concrete and steel. These transactions allow for a steady and readily-investible revenue stream from existing networks, achieving capital expenditure (capex) and operational expenditure (opex) efficiencies from consolidating and ‘leasing-up’ sites, with the potential for growth from the roll-out of new technologies requiring denser networks. The revenue contracts (master lease agreements) are based on long initial terms (typically in excess of 10 years) for a minimum number of sites, contain indexation and are difficult for the mobile operators to relocate on expiry. There is a concentration of credit risk, with a limited number of mobile network operators in any market, but insolvencies are rare, and access to the tower sites is a must-have for the operator.
Core telecom networks
Carve-outs of data network fibre infrastructure by telecom operators pose a more difficult challenge, as they are at the core of the operations of the telecom companies, which are reluctant to cede operational control. For a carve-out transaction to be more than an alternative form of financing, the carved-out infrastructure operator needs to have majority independent ownership, and a degree of contractual independence from the anchor customer. Because of these tensions over operational control and independence, this has not yet become a popular model, though minority investments into such operations by funds, such as KKR’s investment into Telefónica’s carve-out Telxius, are a recent feature. In a carve-out, these are structured with similar long-term, minimum revenue, indexed revenues, although the concentration risk is high.
Regional and specialist broadband networks
The need to replace existing copper access networks with cable and fibre has resulted in large demands on the balance sheets of traditional telecom operators, which made their investments in the legacy networks when they had monopolies on the provision of connections to premises. This need for capex, along with market liberalisation, has provided opportunities for specialist broadband providers to raise equity and hybrid infrastructure style debt to build fibre to the premises networks in the most economically-attractive areas. Typically, these are dense urban areas, but some infrastructure investors also support the construction of new networks in the least dense areas (as a greenfield alternative to secondary transactions), where there is a realistic prospect of only one access network, with a correspondingly-high barrier to entry. There are even models where network operators offer to install high-speed fibre access networks to remote areas if a sufficient number of residents and businesses in the area will commit to taking the fibre access product, leading to demand aggregation by local communities. These networks have quasi-monopolistic features – a rural network has a natural barrier to entry and a wholesale urban access network has less incentive to duplicate – and diversified credit risk. They do not have long-term contracted revenues, regulation often mandates easy switching between providers, but long-term demand tends to be predictable – consumers and SMEs will not stop wanting high-speed data.
Metro-area and cross-continental core telecom networks saw a flood of investment in the late 1990s and early 2000s with the first internet boom, followed by many insolvencies in the early 2000s. With the second (apparently more stable) explosion of data use and decentralisation, these networks are seeing renewed interest and attracting new capital. Existing providers are being bought, sold and refinanced (the sale of Interoute in February 2018 attracted significant financial buyer interest). New models are also emerging, with the large internet content providers entering into joint ventures with infrastructure providers to build submarine and core metro networks, especially outside of the most developed markets in Western Europe and the US. These assets tend to have diversified credit risk and predictable long-term demand. Their revenue contracts may not be long-term, especially for higher-value managed capacity services where switching is more common, but the installation costs and access to wayleaves act as a partial barrier to entry.
Data centres are different from networks, and were traditionally seen as infrastructure assets, providing basic floor space with power, climate control and security. They tended to be contracted on a long-term basis, with real estate-like indexation and security of tenure. Many data centre operators have continued with this model, reflected in their investor base, but others have begun to move up the value chain from this basic infrastructure to providing data storage and computing power for hire, as part of the migration to cloud computing. With the exception of internet exchange points (where Tier 1 telecoms companies connect their networks at must-have locations), the monopolistic features of data centres are limited, although costs and risks of switching for customers are significant. Data centres generally have diversified credit risk, with multiple customers and the ability to replace a failed customer. It is generally accepted that demand for data storage and processing power will only continue to increase.
A note of caution – regulation
These infrastructure assets are in regulated businesses. Connectivity is universally regulated as an electronic communications service, with access networks being especially liable to economic regulation (non-discrimination, open-access and price control) and political intervention. Even the ‘passive’ elements of communications networks (cellular towers, underground ducts and overhead poles) can be subject to licensing and regulation, especially in emerging markets. These licences may contain change of control restrictions, of concern for the security package of lenders. It may also be difficult for an investor to assert a legal claim that it expected the regulatory environment to stay fixed over a long period of projected return.
These assets are also key dependencies for the information economy, and they are attracting political attention. National governments all around the world are increasing their focus on foreign investment and national security reviews of transactions in strategically-important connectivity assets, including data centres, access networks and core network providers. The degree of scrutiny is clearest in the US, with the Committee on Foreign Investment in the United States (CFIUS) and Team Telecom reviews, but Western European governments are beginning to exercise a comparable level of scrutiny, and less transparent foreign investment review regimes also more strongly influenced by political factors than previously.
Tom Levine is a partner at Allen & Overy LLP. He can be contacted on +44 (0)203 088 3114 or by email: firstname.lastname@example.org.
© Financier Worldwide
Allen & Overy LLP