top of page
Writer's pictureArjun Infrastructure

Digital infrastructure: The art of not over paying

Data centres are commanding eyewatering prices, while some fibre assets are crumbling under overbuild pressures. Still, there are ways to play both sectors – and win, seven industry insiders tell Amy Carroll and Anne-Louise Stranne Petersen


An overwhelming 97 percent of respondents to a recent CBRE survey said they plan to up their exposure to data centres this year. Appetite for the sector is clearly sky high. So too are prices.


“We are seeing a huge amount of demand for data centres driven by ever - growing data consumption and AI,” says Elie Nammar, senior investment director and partner at Vauban Infrastructure Partners.


“Finding assets that fit your risk return criteria is challenging, however, given the variety of investment opportunities that exist, which can attract investors from real estate, private equity and infrastructure. We’ll always look for brownfield assets that allow for long-term, steady cashflows with a solid capability to grow in terms of land and energy availability. We also try and avoid the large processes which tend to push prices up, favouring bilateral discussions with industrial partners.”



“The best protection against obsolescence is to have management teams capable of anticipating it. That ensures you… can keep up to date with what is going on in each industry that you are investing in”


STÉPHANE CALAS

Cube Infrastructure Managers



Staying away from highly intermediated situations is certainly one way to mitigate the extremes of this red-hot market. Stéphane Calas, partner at Cube Infrastructure Managers, advocates becoming partners with sellers and founders, so that they are not incentivised to maximise value at the point of entry.


“If you can convince sellers to create additional value alongside you so that you can both benefit at the point of exit, that is ideal,” he explains. “It’s not always possible, but if you invest below the radar of the largest funds, you can find opportunities to co-invest with founders.”


Enrico Del Prete, fund partner at I Squared Capital, focuses on acquiring smaller, 2MW to 10MW sites. “Oftentimes, we are redeveloping old facilities, and expanding the historical customer base,” Del Prete says. “It is a more complex area, both commercially and operationally, but we have built an entrepreneurial team that has the necessary industrial expertise and capabilities. That team has now deployed this strategy across 35 sites in Europe.”


Another option, of course, is to pursue a greenfield approach. “That is the easiest way to avoid overpaying,” says Qilan Tang, managing director, portfolio oversight at CIM Group. “Regardless of whether you are executing a greenfield or brownfield strategy, having a clear understanding of maintenance capex, including retrofitting and replacement costs, is critical, which means that some kind of development mindset is needed. The bottom line is that data centre investment is typically a growth play. This isn’t a static industry, so you need those development skills.”




Whether you are executing a greenfield or brownfield strategy, having a clear understanding of maintenance capex, including retrofitting and replacement costs, is critical”


QILAN TANG

CIM Group




Far from being static, the data centre world is being turbocharged by the rapid escalation of artificial intelligence. AI is of course driving demand for data centre capacity. But it is also raising important questions about the risk profile of some counterparties. Matteo Colombo, head of digital infrastructure at Legal & General is unperturbed, though.


“AI is not a customer. It’s a technology, which all customers, including financial institutions like ourselves, are starting to use. The largest users of AI are the hyperscale tech players, which are clearly investment grade. And when you look at a lot of the newer specialist AI businesses, most have hyperscale clients sitting behind them.”


Colombo does concede, however, that lenders tend to have strict criteria when it comes to knowing your customer base. “We have some businesses where we’ve got some interesting contracts that lenders won’t lend against so naturally the cost of capital is higher.”




Valuing digital assets


But, value in the world of digital infrastructure can sometimes seem like a nebulous concept. As a relatively nascent sector, historical transaction data, in particular, is often in short supply.


Joshua Oboler, investment partner at Palistar Capital, refutes this argument. “A lot of these assets have actually existed for a very long time,” he says. “AT&T was founded in 1885. There are decades of history around frameworks for valuation. Take the fibre space, for example. The valuation of fibre assets has always been a function of growth, churn, capital intensity and cost of capital. That’s an equation that spits out a defendable fundamental number, and we’ve been doing that for 30 or 40 years. Some of the dynamics may be new, but there is a good amount of precedent. The frameworks are well established.”


Rohini Pahwa, managing director at Arjun Infrastructure Partners, adds that historical multiples can be misleading. “This is especially true for a heterogeneous sector such as this. In any case, as a firm, we tend not to rely solely on historical multiples as a guide for future valuations in any sector.


“For instance, with a data centre asset, we will look at the quality of the counterparty, length of contracts, size of the asset, stage of development and whether it is a platform or single site play. Locations are also super critical, as is power supply, particularly in Europe. Regulation and permitting plays into valuation as well. Every asset is different and needs to be analysed independently, rather than relying on what has come before.”


“Multiples are used as a sense check, but they are not the basis of how we value an asset,” agrees Calas. “You have to look at the specific characteristics of the asset you are valuing and the framework for doing that is not all that different to other sectors. What can be more delicate is the parameters you use, given the level of growth that the data centre industry, for example, is experiencing right now. That can be challenging but is a core part of our team’s competence.”


Del Prete adds that infrastructure investors are not used to the fact that digital assets have long payback periods but a mostly uncontracted or shortterm contracted profile. “We are used to long payback periods, but those tend to be backed by contracts that match that duration,” he says. “That’s the biggest valuation challenge that we see. Competitive dynamics also tend to be more intense in digital infrastructure. That doesn’t make these assets less suited to our mandate, it is just something that needs to be factored into valuations.”





“We [don’t] rely solely on historical multiples as a guide for future valuations… Every asset is different and needs to be analysed independently, rather than relying on what has come before”


ROHINI PAHWA

Arjun Infrastructure Partners




Colombo, meanwhile, adds that capital intensity also needs to be incorporated into valuations. “This is something that is sometimes overlooked,” he says. “You need to think about replacement cost and depreciation. Is it EBIT or EBITDA that you should be valuing? And, of course, not all digital subsectors have the same depreciation. There are subtleties that need to be taken into account.”




Data centres’ sustainability challenge


The sustainability credentials of digital infrastructure and data centres, in particular, are coming under increasing scrutiny


Enrico Del Prete of I Squared Capital believes that infrastructure managers are ideally placed to carve out a greener future for the sector. “As infrastructure investors, collectively, we are uniquely qualified to develop sustainable solutions that don’t get in the way of progress. We have decades of experience dealing with resource management and engaging with local communities.”


As well as electricity consumption, water consumption has become a hot topic. “We see water as the biggest bottleneck, particularly in the Western US, where we are focused,” says Qilan Tang of CIM. “Waterless cooling has become critical in these drought regions. With waterless cooling, our facilities require only 12,000 gallons of water in a single loop as opposed to 200 million gallons of water annually for a facility of the same size. We have seen permitting situations where big tech tenants have tried to take the land but have been blocked because of water usage. It is an important differentiator.”


As well as innovations such as waterless cooling, geographical destinations have a role to play in improving data centres’ sustainability profile. “We have invested in Iceland, where data centres benefit from cheap renewable energy and cool weather with limited fluctuation,” says Vauban’s Elie Nammar. “You might miss out on very limited opportunities where latency matters, but there are plenty of clients for whom this location is great.”


In addition to mitigating negative externalities, there are also opportunities to promote positive impact. “We own a data centre in the Western part of Sweden that is injecting heat into the city’s district heating network,” says Stéphane Calas of Cube Infrastructure Managers.


Legal & General, meanwhile, has combined the resources and expertise of its infrastructure and real estate divisions, siting data centres next door to residential developments. “Despite misperceptions, data centres aren’t noisy. They can also give you heat, so they are good neighbours,” says the firm’s Matteo Colombo. “And I, too, would agree that as a group, infrastructure managers are well placed to address this difficult issue. It is a collaborative effort with industry groups helping to provide solutions. Our customers are also very proactive in this area. Sustainability is top of their agenda. Together we will solve these challenges.”




Fibre fall-out


One sector that has attracted a degree of buyer’s remorse in some quarters, in the wake of high prices paid, is fibre. Pahwa, however, believes there are still attractive opportunities if you target the right markets.


“Different countries have different models,” she says. “We have two fibre platforms, one in France and one in Spain, both of which are established markets with strong fundamentals. In Spain, there is a proven wholesale market, which helps take care of the demand side. In France, we have invested in Altitude Infra, which operates under a concession and so benefits from strong monopolistic features.


“The UK and Germany are a strong contrast, with no real wholesale market. We are also seeing high rollout costs in Germany and a land grab in the UK.” Portugal, meanwhile, is another fibre favourite. “Portugal is a perfect example of a market where you can operate in a de facto monopoly situation in a number of rural areas and where you can have all the ISPs on your network, including the incumbent,” says Calas. “In fact, it’s even better than the concession model available in France, because it is an endless concession. You don’t have to return the network to any local or national authority at the end of the day. You own it.”





“[Capital intensity] is sometimes overlooked. You need to think about replacement cost and depreciation. Is it EBIT or EBITDA that you should be valuing?”


MATTEO COLOMBO

Legal & General




“While the fibre market is more mature than data centres, it is still possible to find opportunities to invest in fibre with good visibility and stability in future cashflows,” adds Nammar. “Typically, we’ve seen different ways of achieving such stability, be it through a concession, high barriers to entry, subsidies or through long-term contracts with off-takers.”


Oboler, meanwhile, points to opportunities to provide more structured solutions. “We like partnering with strategics, providing interesting, structured solutions including wholesale agreements. We are at a point in the cycle where that type of capital can be very supportive of business plans.”


The general perception, of course, is that fibre itself has no real competitors. Del Prete, though, believes that the disruption risk that satellite technology could present is not being factored in adequately.


Pahwa disagrees: “The threat posed by satellite technology was something we looked at very carefully when we made our first fibre investment. But when we looked at the impact of cost on the demand side, as well as satellite reach, we came to the conclusion that it is not as compelling as the unit economics of fibre.” “Satellite has a role to play in the ecosystem. It is ideal for far-flung places that are not economic for terrestrial networks,” says Oboler. “But at a minimum of $120 a month, it is very expensive for what is provided in terms of capacity, speed and performance. I see it as a complement but not a threat.”




“We are used to long payback periods, but those tend to be backed by contracts that match that duration. That’s the biggest valuation challenge that we see”


ENRICO DEL PRETE

I Squared Capital







Stranded assets


Being alert to potential disruptors such as satellites is, of course, paramount if digital infrastructure investors are to avoid the spectre of stranded assets.


When it comes to data centres, at least, Tang says the key to avoiding obsolescence is to be forward thinking and flexible about design. “Today, everyone is talking about AI. A few years ago, it was all about transition to the cloud. Tomorrow it might be quantum computing. There is always going to be something that could result in a stranded asset.


“To help avoid that, think carefully about design. For example, if you are doing build-to-suit for a certain hyperscaler, consider the likelihood of other hyperscalers leasing the space based on a competitor’s design if the tenant terminates.


“Three things are certain in a technological society. Death, taxes and data centres. As long as there is more data, there will be more data centres. This is probably the beginning of a long-term, decades-long rush to control the new technological landscape. Building data centres with the most adaptability is the best path, rather than singular-focused data centres. Meanwhile, core markets will always be more valuable and desirable. They are harder to develop in, with greater scarcity, rules and regulations, but quality and design still matter. Rework is expensive. Downtime is expensive. Reputation is priceless.”


For Pahwa, meanwhile, the best way to avoid being stranded is to have a disciplined approach to investing, partnering with experienced operators while backing seasoned management teams. “That is one ‘soft’ aspect that sometimes doesn’t get the attention it deserves. We’re not investing in technology. We’re investing in infrastructure that enables technology. As that technology advances, the infrastructure will need to be upgraded. Working with the best management teams means you will always be on the front foot when that happens.”


Calas agrees. “The best protection against obsolescence is to have management teams capable of anticipating it,” he says, adding that being focused on a limited number of sectors within digital infrastructure is also important. “That ensures you yourself can keep up to date with what is going on in each industry that you are investing in.”


“Towers have virtually no technology risk inherent to them. In fact, towercos will benefit from mobile network technology evolution such as 6G and higher frequency spectrum usage, where more towers will be needed to achieve the right coverage and meet the capacity requirements,” says Nammar.




“As mid-market players, we are conscious of not getting too big. That helps maximise exit options in terms of strategic, financial and potentially public markets buyers”


JOSHUA OBOLER

Palistar Capital




“With fibre, we invest in the passive infrastructure with a wholesale monetisation strategy, where the client brings his own active equipment. It is very much future-proofed as there is no medium that allows for similar speed and latency, and the upgrade happens at the active equipment level.


“Data centres have a higher level of technology risk that needs to be monitored, as we seek to optimise usage per square metre and energy consumption, for example,” he adds. “That is where having sufficient modularity in the building’s set-up, good management and a good understanding of technology play a key role to allow for future-proofing.”


Oboler, meanwhile, says that in addition to obsolescence risk, it is important to be mindful of exit risk. “As mid-market players, we are conscious of not getting too big. That helps maximise exit options in terms of strategic, financial and potentially public markets buyers.


“In a world where people are putting huge quantums of capital to work, there are businesses in this sector that may be performing very nicely but that will struggle to find a new home. The ability to return dollars to investors is vital in today’s market, but I question the monetisation path for some of these very large assets.”




“While the fibre market is more mature than data centres, it is still possible to find opportunities to invest in fibre with good visibility and stability in future cashflows”


ELIE NAMMAR

Vauban Infrastructure Partners








Extract from Infrastructure Investor November 2024 issue magazine

コメント


コメント機能がオフになっています。
bottom of page