If you’ve been following our newsflow over the last few weeks, you will certainly have noticed a slew of high-profile developments in what, for want of a better term, could broadly be described as the digital infrastructure sector.
We are referring, of course, to Colony Northstar and Digital Bridge’s $1.4 billion first close for its digital infrastructure fund; OPTrust, GIC and Mount Elbert Capital Partners’ partnership to target North America-focused data centres; and telecoms tower operator Helios Towers Africa’s intention to list on the London and Johannesburg stock exchanges in a deal worth £2 billion ($2.8 billion; €2.2 billion).
You will also have noticed that two of the above-mentioned partnerships blur the line between infrastructure and real estate, which gives us an excuse to revisit a lingering question hanging over some of these assets: are telecoms towers, fibre-optic rollouts and data centres really an infrastructure play?
I don’t think anyone would dispute these assets are actual pieces of infrastructure, of course, but there still seem to be doubts on whether infrastructure investors should be involved in the development and operation of these assets. There shouldn’t be.
Let’s tackle data centres, since telecom towers and fibre roll-outs are, arguably, more firmly entrenched in the mainstream of infrastructure investment. It’s true that data centres have, for the past decade, been mostly a real estate play. It’s also undeniable, as Colony Northstar and Mount Elbert Capital Partners’ presence demonstrates, that real estate investors are still interested in the sector. But as those partnerships also show, they are very much not going at it alone and are increasingly teaming up with managers and investors with infrastructure expertise.
This is not surprising. Bruno Candès, a partner at Infravia Capital Partners, a pioneer in data centre investments, made a strong case in an article for us on the infrastructure characteristics and skills needed to navigate the complex and dynamic business and regulatory environments in which data centres operate. His piece is well-worth reading, if you haven’t already, but it’s also worth recapping some of his conclusions here:
“Data centres have infrastructure asset characteristics: they provide recurring cashflows on a long-term horizon due to significant switching costs; protection against inflation through the right contractual approach; and limited correlation with financial markets, given the growth of the data produced, which is independent from the economic context. Data centres may also generate yield and provide interesting diversification to a portfolio of infrastructure assets thanks to their very strong underlying growth.”
He continued: “The cloud is the emergence of a new way of delivering computing services in a most efficient way and data centres are the necessary infrastructure to implement it. This goes far beyond providing a mere ‘shell’ for cloud servers, as data centres were perceived in the past decade. For investors, this also means that their potential goes far beyond a real estate play.”
According to Cisco VNI Global IP Traffic Forecast, 2016-21, global wireless growth is set to jump from 96 exabytes per month to 278 exabytes per month in 2021; internet-capable devices are expected to grow from 15.4 billion in 2015 to 75.4 billion in 2025, data from IHS Markit shows; and the cloud-managed services market is predicted to surge from $35.5 billion in 2016 to $76.7 billion by 2021, a compound annual growth rate of 17 percent, according to financial services firm JLL.
The infrastructure to support the storage and transmission of all of that data will represent a phenomenal investment opportunity, comparable to the ongoing decarbonisation of the energy mix.
Yes, there might be work to be done to make some of the contractual structures around these assets more infrastructure-friendly; and yes, for some of these assets, having real estate nous is no bad thing (the same could be said for airports).
But if developing and operating the digital infrastructure of the future isn’t an infrastructure play, we don’t know what is.