Rajesh Sennik, Senior Managing Director at FTI Consulting, talks to Kezia Joseph, Editor, Tech Finance, about the increased investment into Edge computing, how M&A activity is looking in the sector and what this new technology means for datacentre and cloud providers.
Sennik will be continuing the conversation at TMT Finance World 2019, Nov, 27, where he will be chairing the Edge Computing, Cloud and Datacentres breakout session.
TMT Finance: How can datacentre operators and cloud services providers benefit from Edge computing and in what way can Edge computing be best deployed and used efficiently?
Sennik: We are at a unique time in history with phenomenal demand. Just to give you some stats, if you look at workloads on the consumer side, such as search and social, and on the enterprise side such as ERP and analytics, we are seeing 30% CAGR increase in storage and 20% CAGR increase in compute requirements for the next few years. This is based on the research we have done.
That demand for storage and compute translates into datacentre and colocation requirements and the rising wave is also having an impact on the Edge. For example, previously low utilized city centre-based, older datacentres, we are finding now have a new lease of life being recategorized as Edge data centres.
I would characterize the current situation as what I would call Edge 2.0 whilst Edge 1.0 was driven mainly by CDN demand (which largely under delivered for various reasons).
Edge 3.0 will potentially drive additional demand and we are now seeing a lot of interest in this through interesting use cases such as IoT, distributed computing, digital industry, connected/autonomous vehicles and gaming.
In terms of how best to deploy Edge. It is consistent with the fundamentals of data centres i.e. location, location, location; carrier neutral facilities in city centres with low power usage effectiveness. But I believe players also need to think more about the service layer and whether to extend to solution design and architect types of services.
TMT: How important are partnerships when it comes to deploying Edge computing and what opportunities are there for partnerships between datacentre operators and cloud providers? Are we already seeing this in the market, and can you give some examples?
Sennik: There are two basic types of partnership, the supply side and the revenue side. On the supply side, clearly that's important to help with identifying locations, the construction and kitting out data centres with HVAC etc. These relationships are important since they help deliver speed to market which can be a differentiating factor.
Partnerships on the revenue side focuses more on cloud providers and this, if you mean Hyperscalers, is critical. As we all know, Hyperscalers require rapid access to scalable co-location. With these partnerships, it is critical to show how you fit into their technical architecture.
Amazon Web Services for example have regions and within regions they have availability zones and within each availability zone are specific data centre requirements. These are optimized to, for example, mirror data between datacentres. This technical design leads to strict requirements around for example latency. Being able to deliver those as part of a data centre solution is critical.
Interestingly, what we are now seeing, is the AWS deployment of CloudFront. This is basically a content delivery network for local caching of content driving Edge co-location demand.
TMT: What do you think future partnerships in this space will look like? i.e. any new types of companies moving into this space
Sennik: If we view it through the lens of market structure and partnerships then I think it will be both the same and different. What I mean by that is clearly you will have established players like Equinix, which is a phenomenal business, very innovative and a compelling business model. They have had what, 50 straight quarters of growth? So, they will be a key portion of that market structure. Combined with that, you'll see infrastructure investors increasingly present. The recently announced NLDC deal is a good example of that. At the same time, we’re seeing a gradual retrenchment from telcos. Telecom Italia, for example, is divesting datacentres and there is a strong rationale for that.
I believe those dynamics will continue to play out. In parallel, given the fragmentation we have, I believe we will see more roll-ups and companies looking to build pan-geographic platforms where they are combining datacentres. When you combine datacentres across geographies and apply an orchestration layer across those, you give customers the ability to optimize their workloads and achieve a lower cost overall.
TMT: Tell me about the M&A activity in the cloud infrastructure space? What do you think this market will look like in the next few years?
Sennik: In one word, hot. FTI have worked on several deals this year in Europe alone and we don't see any sign of M&A activity slowing. Just look at current processes that are underway, then you've got the publicly announced pipeline of deals, such as the Telecom Italia divestment.
However, “there are a lot of deals but there a few good deals” is what I often hear from my clients. I think you will increasingly see innovations in the business model with for example players who specialise more on specific customer segments or eco-systems.
TMT: What are some of the key geographies seeing M&A activities in the Edge computing sector?
Sennik: We’re seeing sustained activity in what I refer to as the 3 pillars of data centre activity. That’s the Golden Triangle, of New York, Virginia, and San Francisco; the European FLAP markets and the APAC corridor of Singapore, Hong Kong and Tokyo. They will continue to dwarf other cities.
FTI have just completed a study of the global datacentre market and what's quite interesting in the future is that APAC will be the single largest geography for data centres, bigger than the US and that's driven by the sheer volume of eyeballs and amount of investment going into digital infrastructure in that region. For example, outside China, look at all the activity in Mumbai.
Closer to home in Europe, we are also starting to see a lot of interest in what I'd call the three M's - Milan, Madrid, Munich and similar midsize but significant cities.
TMT: How have multiples in the sector changed and what you think is driving this?
Sennik: EBITDA multiples tend to be around the 12 times mark for a carveout or more complex situation and 16 to 20 times for a pure play.
Having said that, I think the more interesting question is 12 times what? We’ve certainly seen a lot of businesses where you see these projections and the EBITDA but then you probe where it's coming from and the off-take/demand is not locked in. So, it’s interesting to look at how you value that and how aggressive you want to be. We do see some investors sitting on the side in those situations and waiting more patiently.
What is driving the multiples? I previously mentioned some of the secular market drivers, including 30% CAGR growth in workloads and 40% CAGR growth in IP traffic.
Coupled with that data centres have strong infrastructure characteristics such as long term contracts, strong revenue visibility, high margins, the potential to ratchet pricing and the opportunity to deploy capital. That appeals to infrastructure investors. Then there are existing participants who are constantly looking to expand their business.
Plus it's a simple business model if you get it right, it's all about fill rate, pricing and construction costs. That all translates into competition for assets and the current multiples.
