Digital infrastructure roundtable: Place your bets

In a fast-paced sector, managers must take care to only invest in assets with true and long-lasting infrastructure characteristics, four infrastructure professionals tell us.

This article is sponsored by Arcus Infrastructure partners, Eurazeo, Palistar Capital and Vauban Infrastructure partners

Vast amounts of capital have been poured into the sector over the past few years, but demand for digital infrastructure still far outstrips supply.

“When it comes to towers, mobile network traffic has gone up 20 times over the past decade and the big three US carriers have spent $640 billion to meet that demand. These are eye-popping numbers,” says Joshua Oboler, investment partner at Palistar Capital. “Data centre bookings are also at record highs. AI alone will drive thousands of megawatts of absorption in the years ahead. Meanwhile, when it comes to fibre, consumer usage in the US is up between six and seven times and still only a third of the country is connected. Then there is the $65 billion of subsidies that are coming to the market to help drive connectivity. The macro is extremely healthy in this ecosystem.”

Indeed, valuations in the digital infrastructure sector have remained robust. Shirene Madani, associate investment director at Arcus Infrastructure Partners, says: “You would typically expect downward pressure on valuations in a high inflation, high interest rate environment, but the high historical book value of digital infrastructure assets, coupled with the continued trend of increased levels of data, means valuations have remained full.

Elie Nammar

“The regulatory framework in a given country can have structural implications in reducing the risk of your investment but… long-term commercial agreements with an off-taker… for example, can protect you”

Elie Nammar
Vauban Infrastructure Partners

“Infrastructure has typically been less affected by the macroeconomic environment than traditional private equity or real estate, because of the contracted inflation pass-through that many assets enjoy, but you would still expect to see highly geared assets experience a decline in valuations over the medium term.”

Elie Nammar, senior investment director and partner at Vauban Infrastructure Partners, however, is seeing opportunities to reprice. “We have been able to reprice some assets despite the macroeconomic environment where sellers have adapted their expectations to the new reality. But in other cases, new transactions are not concluding, or are sometimes not even being initiated, as sellers wait for the macroeconomic environment to evolve.”

Nammar adds that demand for data is massive and growing. We need to store, transport and compute that data and that is all very positive for digital infrastructure. But while the underlying demand is there, he also highlights a certain rationalisation in valuations, especially in the tower and fibre sectors which saw a spike in valuations post-covid. Data centres, however, continue to command top dollar in many instances given the high demand and wide asset spectrum.

Fibre fallout

The fibre industry, meanwhile, is facing an additional set of challenges. The issues of overbuild and subpar subscriber take-up have rattled some markets. “We started out investing in fibre via public initiative networks in France back in 2009,” says Nammar. “Those networks are concession-based, so we have a monopoly and there is virtually no overbuild. Take-up also benefited from a boost as covid hit and demand for work from home boomed. Other markets are more competitive, particularly those where fibre coverage has lagged, such as Germany, and are now tackled by a wide array of players.”

Laurent Chatelin

“We are not in the business of banking on what services AI will provide to customers. We are in the business of ensuring the digital infrastructure is there to enable all this to happen”

Laurent Chatelin

The regulatory environment around fibre in Europe continues to vary significantly. While France has a concession model, Spain offers subsidies that Nammar says are well structured with a clear process to tap into. “By contrast, other markets are more difficult to navigate, and subsidies take longer to secure. The regulatory framework in a given country can have structural implications in reducing the risk of your investment but it is not the only option. Long-term commercial agreements with an off-taker such as large telcos, for example, can protect you from overbuild and take-up risk as well.”

Laurent Chatelin, partner and head of infrastructure at Eurazeo, meanwhile, says the fibre space is difficult for a value-add infrastructure investor

“The European market has matured to such a degree that we would struggle to find opportunities that meet our return expectations given the amount of money that has come into the space,” he says.

“I also think that a lot of managers that are deploying capital in fibre are taking on more risk than they realise. Outside of specific situations, the market in Europe is mispriced. When you look at what is happening in Germany and the UK, it is a complete bloodbath. In Italy, you have a duopoly that is trying to become a monopoly. In Spain, most of the fibre work has been done but take-up is still challenged.”

Madani agrees. “We are cautious around investing in more crowded FTTH markets like Germany and the UK because the risk of overbuild can be high,” she says. “Instead, we look for markets with a developed open access model and, as a wholesale network operator, there needs to be a good cushion between wholesale and retail prices and a minimised threat of overbuild.

“But we are not only looking at the markets themselves. We also focus heavily on the quality of the management team, including their construction expertise, marketing capabilities and cost control experience, because cost overruns are a major theme right now. The pandemic definitely accelerated take-up of fibre but a lot of networks are still struggling. The steepness and ultimate landing point of that take-up curve forms a key part of what drives value.”

Fibre in the US has also been through a turbulent couple of years and Palistar has held back as a result. Looking forward, however, Oboler sees opportunities re-emerging. “I think the credit dynamics are a lot healthier than they were, meaning a more interesting entry point,” he explains.

“You can invest at mid to low single-digit multiples and generate double-digit, mid-teen type returns. We also think there are going to be consolidation opportunities given the number of struggling players.”

Palistar is also not only approaching the market from a buyout perspective. “There are other ways to use capital structure and structured solutions to play a constructive role amid the dislocation in the fibre market,” Oboler says. “Overall, while the fibre market in the US has seemed quite choppy for the past couple of years, the opportunity set is looking much more interesting going forward.”

Niche opportunities

Meanwhile, niche investment opportunities continue to emerge from the three core digital infrastructure verticals. “There are still pockets of attractive opportunities within fibre. Beyond fibre-to-the-home, you have, for example, fibre-to-the-enterprise or targeted parts of the fibre value chain,” says Madani.

“Investing in towers can be more difficult. We have seen some carve-out opportunities, particularly around businesses carrying more stranded asset risk, such as broadcast towers. You need to take a view on the life of those assets and capitalise on new growth opportunities utilising this infrastructure. We are also seeing a trend towards investment in ancillary infrastructure surrounding towers, such as mobile edge computing, and we may see more opportunities arise as MNOs increasingly come under pressure to fund the capex required to densify and roll out 5G networks.”

Joshua Oboler

“Investors have piled into the sector based on its growth story but that isn’t enough. You need businesses with stable P&Ls that believe in managing expenses rather than purely ploughing capex into the ground”

Joshua Oboler
Palistar Capital

Nammar adds: “New subsectors are emerging within the verticals of fibre, towers and data centres that may become core overtime – the deployment of 5G in rural areas under public initiative networks, for example. That could be considered core in the future, but it depends how the overall regulatory and commercial frameworks evolve.

“Edge computing on the data centre side could also potentially become core, depending on commercial contracts but, again, there are still a lot of ifs and maybes. Of course, it is not all that long ago that data centres and fibre were themselves not considered core. Now they absolutely are given the essential nature of those services. This is a sector that is constantly evolving.”

Oboler, meanwhile, points to interesting ancillary opportunities including spectrum, satellites, subsea long haul, wholesale contracts and power solutions at cell sites. “In addition to traditional tower assets, meanwhile, there are also opportunities to partner with railroads or cable companies, for example,” Oboler adds. “We have also spent a lot of time looking at macro wireless and believe there are lots of different ways to approach this industry, depending on your risk profile.”

Wi-Fi access is another attractive niche according to Chatelin. “Wi-Fi access is something that is being managed badly by the incumbent operators. We also see opportunities to provide indoor radio access,” he explains. “As buildings become more efficient from an energy standpoint, coverage becomes poorer. We are starting to see specialist operators emerge to deliver that service.”

In order to assess whether a new niche carries appropriate risk for the asset class, Chatelin says it is important to assess how revenues are secured and how likely the assets are to become stranded. “Before we had 4G and 5G, there was 3G, 2G and 1G. Radio access for mobile is very capital intensive but you are only ever one step away from the next generation. TETRA, for example, was the radio technology used by the emergency services but that is becoming redundant now that 5G has come along. You have to consider whether your investment is going to stand the test of time.”

The power of AI

One disruptive force that is already making itself felt in the world of digital infrastructure is artificial intelligence. The overwhelming impact that AI is having, however, is simply one of fuelling even greater demand. “We are not in the business of banking on what services AI will provide to customers,” says Chatelin. “We are in the business of ensuring the digital infrastructure is there to enable all this to happen.”

“Network architecture will evolve with the advent of AI,” adds Oboler. “Models might learn in one place and act in a more localised environment for other applications, but none of that will fundamentally change our businesses. AI is a large-scale demand driver and we are spending lots of time learning everything we can about where things are heading. There are all sorts of interesting use cases emerging, but they are for our customers, or our customers’ customers. We will provide the infrastructure that stands behind it all, starting with data centres.”

Shirene Madani

“There are pockets of digital infrastructure that seem attractive on the surface, but you need to peel right back to the fundamentals and be sure that the asset will still seem attractive when it is time to sell to a new buyer”

Shirene Madani
Arcus Infrastructure Partners

In addition to enhanced data centre capacity, AI will also fuel demand for other types of digital infrastructure, including fibre.

“With AI, data will travel back and forth between multiple data centres, rather than existing in a single compute location, thereby requiring more edge computing and also meaning backbone fibre networks will need to be capable of carrying more volume than is the case today,” says Madani.

“The additional demand that AI will put on networks is also intensifying the need for higher density processing and other ways to decrease the carbon footprint of data centres. Heritage data centres will need to be retrofitted, if possible, but a lot of this will involve greenfield development.

“High performance data centres are better placed in more economical and green locations, such as the Nordics due to the availability of green power sources and lower temperatures, while edge computing will need to take place in more urban environments, closer to the source of data consumption.”

While the splitting of data processes will help improve the industry’s green credentials, there are other steps that can be taken to make assets more environmentally friendly, including self-generation, with renewable energy sources on site and more efficient cooling.

“There is also a gap between where certain service-level agreements are today with regards to cooling requirements and the actual capacity of the equipment to withstand heat,” says Madani.

“In that regard there appears to be some low-hanging fruit when it comes to the decarbonisation of facilities. However, other levers will require a degree of compromise from all stakeholders.”

Chatelin, meanwhile, is keen to point out that the energy consumption of digital infrastructure is not as significant as it is sometimes purported to be.

“You need to put this in perspective with the wider energy market. The energy required to train ChatGPT-3, for example, is up to 1.5 gigawatt-hours, which is not huge and equivalent to the yearly consumption of about 350 EU households. And the power required is only electricity, so you can ensure a low carbon footprint by sourcing renewable energy. There are also steps you can take to promote energy efficiency. For example, any heat released into the atmosphere can be used to heat real estate instead. Most of the digital carbon footprint comes from manufacturing handsets, TVs and monitors, and only a small portion from running the network.”

Pace of change

Digital infrastructure is clearly a sector that is evolving at a remarkable pace. “In that sense, digital is a bit of an anomaly. Infrastructure typically does not change a great deal. Roads have always been roads. Power lines have always been power lines,” Chatelin says. “What is important in order to succeed in this space, therefore, is to focus on those areas that are not going to change. We will always need towers to provide radio coverage for mobile, for example. And nothing will replace fibre because nothing is faster than light, unless you change the laws of

Oboler, meanwhile, points out that this is a sector that can be very seductive. “It is vital to invest behind strong fundamentals at a company level. We saw a departure from that amidst the euphoria of the past couple of years. Investors have piled into the sector based on its growth story but that isn’t enough,” he explains.

“You need businesses with stable P&Ls that believe in managing expenses rather than purely ploughing capex into the ground. You need strong cashflow generation through the life of the asset and not just an outsized capital gain at the end. Our senior team have each been active in this space for 20 or 30 years and have seen many cycles. My advice is don’t be seduced by the macro. Focus on the micro.”

Madani adds that digital infrastructure assets also need to remain relevant over the long term. “There are pockets of digital infrastructure that seem attractive on the surface, but you need to peel right back to the fundamentals and be sure that the asset will still seem attractive when it is time to sell to a new buyer,” she says. “You have to be cognisant of the fact that there are certain dynamics that exist today, which may not exist in 10 years’ time. As a long-term investor, that must dictate the parameters of your investment.”

Nammar agrees: “This sector is evolving faster than any other area of infrastructure and so, beyond the traditional considerations for core infrastructure assets, you need to understand the technology and understand what potential replacements there could be. Make sure you are betting on what is truly infrastructure rather than simply a passing phase.”

Meet the panel

Shirene Madani
Associate investment director, Arcus Infrastructure Partners

Shirene Madani joined Arcus in 2018. She works within the investment team focusing on origination and asset management. Prior to joining the firm, she held positions in the project finance and corporate and investment banking coverage teams at Citigroup.

Laurent Chatelin
Partner and head of infrastructure, Eurazeo

Prior to joining Eurazeo, Laurent Chatelin was a partner at Marguerite, a European infrastructure manager which he joined at inception in 2010. During his tenure, he originated and executed several principal investments, primarily in the digital and energy space. He was also a principal investor in the European greenfield and brownfield infrastructure space with ABN Amro and Macquarie Capital.

Elie Nammar
Senior investment director and partner, Vauban Infrastructure Partners

Elie Nammar has been involved in various digital infrastructure deals since joining Vauban in 2021. His most recent experiences prior to joining the firm include management consulting as part of the senior leadership of the TMT practice at Oliver Wyman and as a strategist for Vodafone Group.

Joshua Oboler
Investment partner, Palistar Capital

Joshua Oboler has 16 years’ experience in finance as an equity and credit investor, and as an investment banker. He is a specialist in TMT and communications infrastructure, having previously worked in the TMT group at UBS. Oboler began his career at Rabobank International.