Interview with Oliver Bradley, Global Head of Digital Infrastructure Investments at Macquarie Capital

Oliver Bradley: Looking at 2021 to now, the differences can be categorised as (1) macro influences that impacted everyone, but particularly digital infrastructure, and then (2) changes that were more specific to the digital infrastructure market.

The obvious macro changes in the last three years relate to interest rates; at the end of-2020 and into 2021 we were in a zero interest-rate environment. That had a dramatic impact on the types of investors looking at the space. If you think about a digital infrastructure network, it is characterised by having a lot of CapEx and not particularly high earnings for the first few years, but then cash flows later.

When interest rates are more or less zero, discount rates are low and people are attracted to leveraged equity returns of 7-8%. Looking at some of the deals that were completed in Spain, base rates were negative 0.5% for a time and all-in borrowing costs were around 3%. Taking into account equity return expectations, there was probably an all-in project cost of 6-7%. If one is discounting future cash flows at those levels then it doesn’t matter as much that cash flows take longer to come through in projections.  

Now that interest rates are a lot higher, close to 5% in the UK and 2.5-3 % in Europe, your all-in cost of borrowing has probably gone from being 3% to being 6-7%. Your leveraged equity cost has gone from being 7-8% to 12-13%. And the total cost of capital for the project has gone from 6% to 10%, for example. That's quite a dramatic difference and when you compare the returns of some of these projects with the risk-free rate, they don't stack up so well. That generally affects all infrastructure investments, but there are a couple of reasons why digital infrastructure investments have been particularly affected.  

First, digital infrastructure is particularly sensitive to interest rate fluctuations because you see several years of heavily negative cash flow before positive returns are achieved. You're delaying the cash flows out further than you might do for other infrastructure assets which exasperates the cost of capital issue.

The second challenge is that when funds invest, they’re not just discounting future cash flows, they’re having to think about what those future cash flows are. I think the assumptions that people made during COVID were rosier than how things have turned out. For example, during COVID, when everyone was at home, you saw the share prices of Ocado, Peloton and Zoom skyrocket, but they haven't lived up to the hype since. During that time, in the case of digital infrastructure, investors made overly optimistic assumptions about how quickly fibre would be adopted and how easy it is to build a fibre network. Building fibre networks to budget and at scale is actually harder than people realise, and there are some operational challenges that were maybe underestimated.  

So, again speaking very generally, discount rates and the cost of capital have gone up a lot and these businesses also have less attractive profiles in the current market. This is combined with the fact that business cases were probably too aggressive a few years ago and people have realised that development is a bit harder and adoption is taking longer. Overall, this has had a negative effect on sentiment. A lot of businesses are doing pretty well, actually, but the few businesses that have done badly have really shaken the market. This has been reflected in both debt and equity and it's very hard to get people to take a fair look at digital infrastructure. People want to see much more proof of scale and of earnings because people have lower risk tolerance these days. There’s an element of overcorrecting from a few years ago and that's contributed to a fairly challenging market with regards to raising capital or M&A for both good businesses and for bad businesses, because ultimately there is a sentiment problem.

Oliver Bradley: I think a lot of differences stem from how markets are set up and how companies are used to operating within them. I would argue that it is important to consider the ‘culture of wholesaling’ in a particular market. You've got some markets where there's no wholesaling and companies just build their own networks as there is no culture of wholesaling. You see that a bit less in Europe, but you see it quite a lot in America. ARPU and capex are both quite high so the operators want to keep their network for themselves and don't want to let their competitors onto a network. They've never done it and there’s not that culture of wholesaling in place. 

Then you've got markets, more commonly in Europe, where you've had an incumbent which has led the way in wholesaling. In this case, the incumbent has built a network and people wholesale off that network. The incumbent may well have started life as a nationally owned monopoly. While it's a big wholesale market, there's still not a big culture of wholesaling because everyone wholesales off one player. The UK is the most obvious example of this, where BT has dominated the network side of things and also has a retail division. The other big ISPs, competitors of BT at a retail level, wholesale off the same network despite it being owned by their competitor. They are not building networks themselves to wholesale off each other. You still have a culture of wholesaling, but it's a culture of wholesaling from one party, not from each other.

Markets like Spain are, in my view, one of the most constructive markets for wholesaling in Europe as there's a really good culture of wholesaling from each other. There is still a dominant incumbent who led the way in wholesaling but, crucially, other big players have built networks and have wholesaled to others. So it's not just everyone wholesaling from Telefonica, everyone is wholesaling from each other. This created a good opportunity for us to create an independent wholesale platform, Onivia, which now has approximately 50 ISPs who use it, including all the large ISPs, because they understand how to do it and see the benefits of an independent wholesale network.  It might sound like a trite observation, but I think it's quite important. When you talk to Tier 1, Tier 2 and Tier 3 ISPs, they know how to wholesale and they know what should be in a contract - they’re used to it. We're speaking the same language when we talk about how to do it, and they understand the benefits of a shared network. It’s enabled us to grow Onivia to be the third national network in Spain and the largest independent wholesale network in Europe. And that means there is a really dynamic competitive environment in Spain, which has driven fibre adoption and helped contribute to Spain being one of the first countries in Europe where all old copper exchanges have been decommissioned.

In the UK, it’s taken a lot longer for large operators to get comfortable wholesaling from networks other than BT Openreach, certainly at scale. Although it is starting to happen now. 

When you're looking at a market, you do need to take account of the familiarity and the culture of wholesaling in place. You can't just assume that network sharing will work because it rationally makes sense. You have to look at some of those softer barriers: Are people used to doing it? Have they got their heads around it? In the US for example, you’ve seen some of the big national players begin to look at wholesale adjacent arrangement for the first time, but often with quite a lot of conditionality about the extent to which they will be truly open access networks.  It’s a journey!

Oliver Bradley: My expectation is that investor sentiment will shake itself out and fix itself in the next few years. The characteristics that attracted infrastructure investors to the sector in the first place still very much remain, and digital infrastructure is a core infrastructure sector. I think you'll see consolidation, you'll see fewer businesses, you'll see businesses coming together achieving operational synergies. But the business cases for fibre still very much makes sense.

Markets like Germany, in particular, have struggled because it's taken a lot longer to get people to sign up to networks and so investors have found they have to spend a lot more on OpEx whilst they wait to fill their networks. This was compounded by having some of the highest capex costs in Europe. That is, in my view, a timing issue and it will fix itself.

People will move to fibre. The economics of fibre do make sense. It’s very low cost to run, it's future-proof. What we all need to do is think about how we can get adoption to be a bit higher more quickly. But once you start to get speedier take-up, and you start to have businesses that are actually making money, then I think people will appreciate that fibre networks do indeed make very good infrastructure investments. They represent a core infrastructure sector as much as any other utility.

There does need to be a bit of shaking out and a bit of settling. You have a bit too much uncertainty about competition and the speed of take-up, and some uncertainty about long-term markets, but those are lesser risks. It's going to take a little bit of time and for some it may be a painful process, but for others it may be ok. It's hard to predict who the trouble will apply to. Looking forward a few years, I think you'll see some consolidation, fewer networks and more network sharing. These developments will in turn enable a greater capacity for innovation amongst service providers on the network. They will be able to provide better services, better prices and better customer service across the same network.

Oliver Bradley: You have seen people trying sharing in different countries in different ways, with different levels of success. For example, in the UK and other markets, there've been trials around using sewers to lay fibre cables. I’m not an expert on cable laying, but I believe one of the reasons it hasn't worked so well in the UK is because of where the entry points are for sewers, making it less practical.

I've seen this being rolled out more effectively in other countries such as Austria and Germany. Some people are partnering with District heating suppliers and those ducts are being used to lay fibre with some success in terms of bringing down overall costs. 
There's no technical reason why you can't lay fibre next to high-voltage electricity cables and provide the sort of infrastructure improvements that we need for EV charging for example, in a coordinated way. If you dig up a road to lay fibre cables, you should be putting in cables that will help to upgrade our electricity. That requires a lot of coordination, but there's absolutely no reason why a high-voltage electricity cable can't run alongside fibre. Indeed, in Ireland and in Israel, and certain other markets, we've seen experimentation with laying fibre along overhead power lines. This is technically quite complicated and it’s more of a specialist job dealing with electricity power cables rather than dealing with ducts, however anywhere there’s already a duct or pole, there’s opportunity for fibre. Fibre is inert and it’s pretty thin so you can run it through almost any duct, and if you can avoid digging up a road, which is often quite costly, this is a better approach. 

Oliver Bradley: It’s a really interesting time and we are generally thinking about how to move beyond fibre. I think the good news from a European perspective is that in the next few years most people will have access to fibre. It’s been a bit messy, but I think it's been pretty successful, looking at the amount of private capital and the public support for fibre. The good news is that most people in Europe over the course of this decade will probably have access to high-speed internet through fibre.

Then we have to think about what is next in terms of the digital infrastructure ecosystem and ask questions about the impact of AI on data centres, how the networks will adjust to mobility, Internet of Things and Smart Cities for example. What role will Satellites play. These are the things that we are now thinking about at Macquarie Capital.