Stephen Beard, Global Head of Data Centres, Knight Frank, discusses how real estate and data centres marry up, data centre growth in the UK market compared to the global picture and why Manchester should be a top priority for data centre investment.
What does your role entail and can you give an overview of what a typical day looks like for you?
I oversee a team of 14 individuals based here at Baker Street, London. We’ve got a couple of people over in Singapore in the APAC market and they’re working in conjunction with our partners in the US. So, effectively I’m responsible for running a global team on a cross-sectional basis.
From a day-to-day perspective, we’re dealing with data centre customers that are always looking for new operational, technical whitespace. We work with data centre operators that may need more land in a variety of different markets, or institutional investors looking at potentially securing equity stakes with partly stabilised or fully stabilised platforms. So a wide variety of work streams where no two days are the same, which makes for an exciting place to be.
With extensive experience in real estate, what sparked your interest in data centres?
When I finished my degree, I wanted to move to London. It seemed to make more sense to go there and I probably naively felt that becoming a West End retail agent or a central London City investment agent would be the right course of action, but it’s interesting how times change.
In a previous role, to be really candid, my main driver, which is still the case today, is making money and I had a sense that where I was at the time, I would have to wait for my director to either pass away or retire before I’d get decent exposure to big clients. At the same time, I was aware of technological advances and the growth of the digital economy and I wondered whether or not there were real estate houses that were advising within the data centre space. I was very fortunate that an opportunity presented itself at CBRE within its data centre team, so I took the opportunity and was there for three or four years. I then realised that as the industry grew, it made sense that some of the other large real estate houses could do with a data centre team. That was three years ago and it feels just like yesterday.
How do data centres and real estate marry up and how does the data centre benefit from real estate connections?
When I first started exploring the world of data centres I had no idea what a data centre was. There’s a misconception that the cloud is something which isn’t physical, that it’s just up there in the air somewhere, but data centres are physical buildings. The Internet actually has a physical home and that’s a data centre’s purpose. The whole premise here is that a data centre is just another physical building. Data centres look very similar to big industrial logistic boxes, except for the fact that instead of whatever you may have stored in an industrial building, we have lots of racks and servers and supporting infrastructure to keep the servers cool and lots of backup power like generators, UPSs and batteries, just in case the mains utility goes down. So there’s a big symbiosis between industrial and data centres. It’s much like any other asset class that location is key – the best data centres are those that are more profitable and tend to be in the better locations, i.e., those that have good access to power and fibre, so they’re no different when it comes to location. I have a saying that within data centres, location is Queen, power is King and connectivity is the Ace card.
How would you describe the data centre landscape currently, particularly in the UK, and how has it evolved over the last 12 months?
Cloud adoption combined with the huge increase in general data usage has had a massive impact on these physical pieces of real estate. When I started five years ago, the average-sized data centre in London was probably somewhere between 10 and 12 MW of IT power and for the layman, that’s probably somewhere between 100 and 120,000 square foot of gross building. Whereas today on any new site we’re selling within London, or the M25 and the surrounds, we’ve got to be able to show the prospective purchaser that the building has the ability to be five-times that size.
Another consideration is that before the cloud is adopted, data centre operators would have a multitude of different customers they’d provide a service to and what we’ve seen is mass migration to the public cloud. The likes of Microsoft, Amazon and Google now account for about 80% of total occupational take up throughout Europe. Therefore, their data centres and commitments to data centres are increasing in scale. So not only is it having an impact on being able to identify viable development – brownfield sites that can accommodate that scale – we’ve also got to find the power and that’s become a real issue, particularly in the UK.
In London currently, we’re effectively out of power. The National Grid and our infrastructure haven’t been able to keep up with the insatiable need of the end-user. I think the impact will have a positive result for those markets outside of the core London markets. We’re looking at a couple of very big sites in the south-west region – Bristol and Newport – which have the advantage of connections with the US. A lot of the data we’re transferring goes to the US via the M4 corridor out via Newport and Bristol through the transatlantic subsea cable which lands in New York. Those locations which are close to that data bullet line should in the long-term see some significant gains.
Being a Mancunian, I’ve got to promote Manchester as the second market in England. We’re seeing a lot of macro movement in terms of investment justifications in Manchester. Several major corporates have moved there such as the BBC and ITV. The fibre connectivity has improved to the extent that the latency – the speed at which you can send, process and distribute data – is just as quick as it would be from Birmingham to London. So there’s not as much of a geographical impairment from a location standpoint. We’re also starting to see the rise of Edge data centres.
With a global presence, how do Knight Frank’s regional markets differ from one another in terms of customer demand and growth and how do you meet expectations?
When I made the decision to come to Knight Frank it was on the basis that we had a very large network across Europe, Asia and the US. The first 18 months were all about educating our local teams on the ground and making sure they understood what’s required by both the data centre user and the developers. What’s happening at a macro level is that there are certain countries in emerging markets predominantly through Central and Eastern Europe from a European perspective, and South-East Asia, where they’re only just beginning their digitalisation story. The governments in each of those countries are looking at the commercial and social benefits of migrating their IT infrastructure into public cloud. So they’re tending to the likes of Microsoft, Amazon and Google and therefore, the data centre markets there are experiencing significant growth. Most of my travel tends to be in markets which are about to experience significant growth. We try and get ahead of the market – our USP is that we don’t follow the cloud, we don’t track the cloud, we get ahead of it.
Where in the UK is best for data centre investment and why?
With the majority of our population residing and working in London, it will always be the focal point. London is the second largest market in the world, second only to North Virginia, which is good news for our economy on the back of things like Brexit and COVID. This also depends on the government’s approach to supporting and aiding the digital economy and a lot of that is through protecting the growth of data centres.
In terms of investment today, if you were so lucky that you had a ready-made industrial brownfield land parcel along our golden corridor, which is predominantly Slough to Hayes and you were able to secure power by some means because it effectively is out of power, then you’re sat on a goldmine. The land price that some would be willing to pay for that type of product is astronomical.
From an investor perspective, when it comes to fixed income investments very typical to an office investment or a retail investment, the downside of data centres is the illiquidity of stock. Most data centre operators require freehold land because the cost of building out the data centre is so extortionate that they can’t be at risk of being kicked out of the building by the landlord at any point. To put it into perspective, if the average-sized data centre in the UK or London going forward is 50MW, it usually costs about £10 million per MW in terms of fitting out the facility – that’s a 500 million CapEx liability. But there are opportunities – we’ve been buying and selling a few smaller fixed income assets throughout the UK. Personally, I’d be concentrating on Manchester. If you look at every other European country, they all seem to have a very strong secondary market and I think Manchester will soon take off.
What major developments can we expect to see this year in the UK’s data centre sector?
We’re a very opaque market, which is frustrating and we at Knight Frank are trying to change that – starting with the education process. The reason why the industry is inherently opaque is because much of the stored data is sensitive. When you look at the next threat to a country, it’s more than likely going to be a digital attack rather than a physical attack. The next terrorist attack will likely be on a data centre so it’s very difficult for us to talk about major new developments, but I think there’ll be a couple of announcements about some fairly large schemes in West London and North London. There’ll be three or four announcements in Manchester too. I think the opportunity for investors is to marry their portfolio and investment considerations alongside that of this Edge rollout.
