Month: November 2021

The intersection of E5G and Digital Infrastructure

Two weeks ago, we held our first live event since March 2020: Enterprise 5G Live in Santa Clara, CA. During the two days we had:

  • More than 40 speakers and presenters in 20 workshops, bootcamps and keynotes.
  • Over 120 enterprise, systems integrators, solution vendors and distributors represented.
  • Five demo CBRS networks – we joked that our show was the only place in the country that had five CBRS networks under one roof. And as Iain can attest, coordinating the PLMNs was a chore.
  • More than 800 cookies, muffins and pieces of cake were consumed – in addition to the 390 sandwiches and salads.
  • 33 gallons of coffee were drunk, mostly by the hardy souls who traveled from Tel Aviv, Israel, to attend the show.

Some of the E5G Live’s content highlights include:

  • Talked about how their 20 kW data centers can be turned up in less than 6 hours and, along with fiber, help address the “middle mile” connectivity gap.
  • Zayo: Discussed how their fiber network links premise-based apps on edge platforms with cloud services. So doing helps create the foundation for everything from retail order fulfillment to autonomous logistics, drone landing stations and intelligent IoT sensors.
  • Athonet: Discussed how its 4G/5G core network can be deployed on premises, in the cloud or on a hybrid basis and showed how those deployments can scale across a private network installed in multiple brick-and-mortar storefronts.
  • Intel: Highlighted their work with a variety of vendors, service providers, school districts and enterprises using 4G/5G private networks to transform their operations.
  • Microsoft: Discussed their Azure-based initiatives to provide developers with the tools to transform not only the 4G/5G core, but the RAN and the applications that run on them.
  • Amazon Web Services: Talked about how their platform makes it possible to deploy a 4G/5G core network and RAN in less than 40 minutes.
  • Celona: Talked about how private networks not only need to be easy to deploy, but how they enable the uptake of new machine-to-machine communications in addition to proving essential connectivity services.

Overall, the show focused on enterprise 5G technology and use cases. One example includes Blue White Robotics, which retrofits farming equipment into autonomous farming machines. Their solution requires devices in the vehicles, compute and storage, connectivity on site and to/from the farm. Another example is Transparent Path, which offers a tracking and monitoring solution for the cold storage logistics chain. Their solution requires IoT sensors and AI, along with cellular network roaming and private networks.

At one level those use cases are all about how a private network can enable innovative enterprise solutions. But, those solutions would not function without the underlying digital infrastructure – i.e., towers and cellular networks, fiber and data center.

This time it might really be different

A few years back the economists Carmen Reinhart and Kenneth Rogoff published a wonderful economics tome called This Time is Different, about eight centuries of financial folly. Their engaging idea is that everyone always says that their financial crisis or major new trend is really different this time, but it never is. Their message is that history is forever doomed to endlessly repeat itself. These words sprung to my mind when recently writing an article for a UK financial publication on the rise – or should we say, the return of – the metaverse.

Facebook’s name change produced a collective shudder amongst investment cynics like myself who’ve also moonlighted as tech nerds. I’m old enough to remember all the failed attempts at augmented reality (AR), virtual reality (VR) and everything in between. Obviously, this time it really is different because Facebook says it is and it is pumping countless tens of billions into research to prove the point.

But the smart money knows that in many ways the meta verse is already here, it’s just that the current day experts realise that if they shout too loudly they’ll be accused of hubris. Take the endless number of virtual gaming worlds already sucking up teenagers’ attention. Or the growing B2B AR and VR implementations used by businesses built around frameworks such as Microsoft’s HoloLens. Or the fast-emerging Digital twins metaverse at the enterprise level which takes the Internet of Things (IoT) and then jams it into top gear by allowing engineers to remotely test system stresses and capabilities.

Another clue that maybe it really is different this time is that investors are taking note. Giant US investment bank Morgan Stanley has just released a spate of reports. ‘The Next Big Theme – Metaverse’ by Edward Stanley is one summing up the tone nicely – which highlight in detail why it might be different this time. That report certainly finds an echo with investment managers. I recently interviewed one of the UK’s leading technology fund managers, Ben Rogoff at Polar Capital, and he couldn’t stop him talking about the metaverse. I’ve even waxed lyrically – and cynically – on the investment case HERE.

My own central case is that it’s not really a case of this time it’s different, it’s just that existing technologies and infrastructure are finally catching up with marketeers’ exaggerated demands. Virtual worlds sounded fun initially but try getting excited about all this interaction via ADSL telephone modems. Or clunky AR glasses. Again, it sounded cool when Google brought out ‘those’ glasses but frankly the technology – and content and means of delivering the content – wasn’t quite there. In sum we’d seen the future already, but the present wasn’t a patch on the promise.

Now that promise is being delivered and the infrastructure implications are seismic, I think. But before we think through those implications let’s first get the definitional boring stuff out of the way. According to Morgan Stanley, the metaverse in a nutshell is: ‘A virtual world for immersive experiences that is persistently available and where users can explore vast numbers of experiences concurrently. Within these experiences, people across the world can meet, play, watch, trade and learn across millions of experiences.’ Or as Bart Schouw of Software AG nicely sums it up in a recent Information Age article: the metaverse mainly “refers to a convergence of physical, augmented, and virtual reality in a shared online space – or the universe of data.”

As I said there’s plenty of products already out there that are making use of these overlapping technologies. For instance, many Snowball customers today run edge-based processing with drones and then map it into 3D point clouds as they collect the drone information.

Kick the tyres and you begin to realise that in fact metaverse isn’t really a ‘product’ or ‘single technology’ as such but its own infrastructure. You can already see this new infrastructure at work via outfits such as Subspace which is helping gaming businesses reach bigger audiences by building a “parallel internet that puts high-traffic applications on their own network, ensuring the fastest and most stable path for instant experiences”. In this alternative narrative, the metaverse isn’t an application but “the next iteration of the internet that supports real-time experiences.” Or perhaps more succinctly Network as a Service model with outfits such as Subspace offering up a huge global private network for near real time applications.

OK but back in the world of hard infrastructure reality, we have to accept some real challenges, not least the obvious “interface challenges”, a wonderfully euphemistic way of saying that no one over the age of 30 is going to wear a stupid VR headset.

But there are other less obvious challenges. One school of thought reckons that for the metaverse to be really game-changing, it needs to be connected to countless tens of billions of devices via the IoT. As Bart Schouw puts it “The only way to do that will be by mass ingestion of the data coming from the Internet of Things. Only with this data will you be able to create a rich and meaningful environment. The next need after “seeing” will be “interacting,” meaning that the data not only needs to be represented in a meaningful way but also must be responsive.”

Which bring us nicely to the $64 trillion question: Latency or responsiveness needs to be absolutely minimised in order to simulate real world response times. And here we get to the infrastructure challenge. If this slightly dry subject interests you, I recommend reading a fascinating discussion between Bill Vass, VP of Engineering at Amazon Web Services, Patrick Cozzi of Cesium and Marc Petit of Epic Games. They discuss the Cloud Infrastructure for the Metaverse HERE.

To make the metaverse sing and dance, operators somehow need to find a way to overcome the big challenge: that the bridge of the network is going to cause problems. If I’m sitting at home constantly tapping away for a response and the latency starts cruising above 100 ms, then I’m off making a cup of coffee. Figuring out the solution requires a triad of solutions. Main core functions sitting in the cloud, helped along by edge services and networks all connecting to smarter devices in your pocket. As Voss says, if customers have “a really powerful thing, like an iPhone, they can run a lot there, or a car, they can run a lot there. If it’s less powerful, and they have good network, they’ll offload more and more to those tiers.”

For this lay observer, that suggests the following. The first is that to make the metaverse really work someone who is trusted to make smart handheld devices needs to run with the metaverse ball. Cue Apple. If anyone can make a device with the processing power of an old supercomputer sit on your nose, the boys and girls at Cupertino can.

BUT, and yes this is a big infrastructure BUT, that device needs to link into both services and a base layer of infrastructure that makes said Apple device work with next to no latency. And in this, I see an impending battle. We’ve already seen the likes of Google and Facebook migrate down the value chain into owning bits of digital infrastructure such as data centres and sub sea cables. But if they want to dominate the metaverse – and not leave it to Apple – they have to offer a seamless service with next to no latency into the last mile. They cannot be reliant on some backwoods telecom operator to send that final bit of data over ancient wires and cables. Or to put it bluntly I cannot see how Facebook and presumably Google can make the metaverse work unless they own the last mile into the customer, i.e the whole digital infrastructure chain.

But as they head downstream into your homes, they might run into mobile operators heading the other way, upstream. Back to Bill Vass from Amazon – “So everything you get on the cloud, you get in this (5G) hub much closer to the customer from a latency perspective. Solves the speed of light problem to a certain extent”. Already we’re seeing products emerge such as Wavelength which push the cloud to the 5G hubs, with very low latency that is ideal for metaverse applications.

Which brings me very nicely back to that Morgan Stanley report I mentioned at the beginning. If Facebook and Google start playing in the telco space, why can’t the telcos play in the cloud/content/metaverse space? This might upend the traditional model which is based around the idea of ‘Build it and they will come’.

Perhaps the better model is to build both simultaneously, i.e products and infrastructure. And telcos already know this model works. According to the Morgan Stanley Metaverse report “Vodafone saw reasonably significant volume spikes at the time of the Pokémon Go phenomenon in 2016. Vodafone noted that AR/VR is ‘sticky’ (seven out of 10 people that use Pokémon Go return to use the application the following day), while also stating that AR/VR triggers considerable other mobile usage (Facebook, YouTube, tweets, etc.) as well as fixed usage.”

So, here’s my slightly ambitious reading of one distinct possibility. Whilst most digital infrastructure investors wax lyrically about say Autonomous Cars or Smart Factories as the killer apps for 5G, what about thinking of AR/VR as “the medium-term killer 5G app” (Source: Vodafone CTO, Technology Day, Q32021)? In other words, the launch of stand-alone 5G networks in 2023 could drive a wave of AR/VR innovation and the mobile telcos could be in the lead pushing new products and services as part of an expanded metaverse. If you already provide infrastructure via a network, why not take Facebook at their word and play them at their own game? This way the telcos can square the circle of mobile volumes continuing to rise by ca +30-40% YoY whilst being constantly undercut by per unit deflation (price per Gb). For the telcos, the metaverse could be the solution to their revenue growth rate challenge. The challenge then becomes how to move faster than Facebook or, should we say, Meta!

ColoHouse Expands Footprint With Acquisition of Quonix

Source: BusinessWire

MIAMI–(BUSINESS WIRE)–ColoHouse, a full suite provider of colocation, cloud and managed services, announces its acquisition of Quonix, including Data102, and Turnkey Internet data centers, further expanding the company’s market footprint and services. This acquisition expands the ColoHouse portfolio by three locations and over 300 colocation customers in Albany, NY, Philadelphia, PA, and Colorado Springs, CO. In addition, Turnkey Internet, a leading provider of bare metal and services, will add another 1,500 customers to the ColoHouse base, fortifying the company’s bare metal capabilities.

“2021 continues to be the year of formulating the perfect group of exemplary companies and teams to create a best-in-class colocation, cloud, and services company,” said Paul Bint, CEO of ColoHouse. “The Quonix Group acquisition includes ColoHouse’s first foray into building ownership with the Turnkey Internet facility in Albany, NY. This is an exciting milestone for us as a growing team. In the last year, ColoHouse will have grown its data center footprint by 600%in the United States, supercharging our service offerings. The leadership teams are trailblazing our organization into an enterprise brand, using our multiple locations, talent, and service portfolios to be best in class.”

“ColoHouse’s acquisition of the Quonix family gives our customers the resources, structure, and geographic footprint necessary to meet their growing technology needs with one provider,” said ColoHouse’s new SVP, Managed Services, Randal Kohutek. “ColoHouse is leading the way by offering one-stop colocation, cloud, and now bare metal services, across numerous locations. The vision for ColoHouse is extraordinary, and I am looking forward to our customers, both new and existing, growing with us. It’s an exciting time.”

“With the addition of the Quonix Group, ColoHouse will be significantly expanding its bare metal offering. ColoHouse is no stranger to the bare metal game,” said Turnkey Internet’s Adam Wills. Since 2018, ColoHouse has offered bare metal and dedicated servers as part of their portfolio. “Turnkey Internet, as one of the leaders in the bare metal space, will be able to provide the technical expertise, personnel, and strategic planning to scale into our other US locations including Miami, New York, Dallas, and Phoenix alongside our entry into the European market. Our existing customers will continue to work with the same people they always have. The Turnkey Internet team is looking forward to providing nothing short of amazing solutions across a diverse geographic network and data center provider.”

Most recently, ColoHouse was funded by Valterra Partners to create a dominant data center and cloud solutions company that is ready to address the complex needs of their customers. ColoHouse and its investors will continue to grow the business through strategic acquisitions to create a single provider option for its client base.

About ColoHouse

ColoHouse is a worldwide retail colocation, cloud, and managed services provider with 26 locations in 21 cities in North America, Europe, and Asia. Our full suite of colocation, cloud and managed services gives our customers the flexibility to customize their IT infrastructure needs to meet their business objectives. We focus on delivering quality infrastructure, services, and support, giving our customers the ability to allocate more resources toward their core business. For more information, please visit

Digital Infrastructure has more to gain from the Infrastructure Bill

Everyone interested in digital infrastructure knows that when Congress finally passed the Infrastructure Investment and Jobs Act last Friday, it included $65 billion to improve the country’s broadband infrastructure and increase Americans’ access to high-speed Internet. It might not be as clear that many of the other line items in the bill will also provide opportunity for significant digital infrastructure investment. This article breaks apart the spending in the bill as we understand it at present and highlights the opportunities for DI investment. Obviously, having the potential to get the funding and actually receiving the investment are two very different things!

The bill has set aside $110 billion for roads, bridges and other major transportation programs. While the roads and bridges are being constructed, fiber and conduit could be laid, environmental sensors could be installed in the bridges, radios and small cells could be deployed, and edge data centers could be built to process the increased volume of data. In short, this could be the basis of smart road and smart city infrastructure. The cost of deploying this as the bridge/road is being updated would be far lower than installing after the fact.

Public transit will be modernized with $39 billion. Part of that sum will be used to repair more than 24,000 buses and 5,000 rail cars, which could obviously be upgraded with Wi-Fi and cellular broadband. Thousands of miles of train tracks will also be modernized, which means more fiber, the possible installation of small cells along the tracks to provide coverage for neighboring communities, more sensors, and edge data centers.

$66 billion is earmarked for high-speed rail, improvements to safety, Amtrak grants, and the modernization of the route between Washington, D.C. and Boston. It is not a stretch to assume that improved Wi-Fi and cellular broadband will be part of the train improvements, while fiber and sensors could be installed with the new tracks, along with small cells and edge compute infrastructure. The exact opportunity here will depend on the degree of improvements to the physical track infrastructure compared to improvements in operations, staffing etc.

The nation’s electricity grid will be upgraded with $65 billion and will include thousands of miles of new transmission lines and funds for smart-grid technology. In addition, all of the nation’s lead pipes will be replaced and service lines will be upgraded to provide clean drinking water at a cost of $55 billion. Both the electric grid and the water infrastructure could be retrofitted with sensors, remote switching and AI, which would work in conjunction to provide power optimization/water efficiency, usage behavior analysis, fault diagnosis/leak detection, water quality analysis and preemptive maintenance. Again, the size of the opportunity depends on the amount of new electricity/water infrastructure, but it would appear this could be a sizeable opportunity for digital infrastructure.

The bill also includes $7.5 billion for a network of electric-vehicle chargers along highway corridors to support electric cars and buses. Each charging station will need to be connected by fiber and will need power and broadband, including a small cell to provide a good signal for the passengers patiently waiting for their car to charge, as well as to monitor and control the station. Edge compute could be used to provide the necessary processing for billing and other applications, including entertainment. In essence, an EV charging station looks a lot like a small cell installation but without the pole! In addition to funds for the charging stations, $5 billion is slated for zero-emission buses (including thousands of electric school buses) and $2.5 billion for ferries, which will both more than likely be upgraded with Wi-Fi and cellular broadband.

More than $25 billion is included in the infrastructure bill to modernize America’s airports, which have already implemented in-building wireless (IBW) systems and the applications they support. With the funds, the airports will have more – more 5G and CBRS to support improved customer mobile experiences, airline private networks, smart applications for the automation of food and beverage, smart parking, enhanced security, etc. AI and edge compute, powered by data from sensors and cameras, will also be part of the airports, which are basically just small smart cities. Assume that the $17 billion for ports, which are also mini smart cities, will include similar infrastructure and applications.

$11 billion has been set aside for road safety, including transportation safety programs to help states and local governments reduce crashes and fatalities in their communities, especially among cyclists and pedestrians. In other words, $11 billion for V2X infrastructure and applications. The “everything” in V2X includes the cellular network, traffic signal equipment, vehicles and pedestrians with their mobile devices. V2X and 5G can provide automatic braking, accident avoidance, road condition monitoring and warning, intersection collision avoidance, approaching emergency vehicle warning and pedestrian awareness, which are the intended goals of the bill. Note that according to iGR’s analysis, $11 billion is not sufficient to build the necessary V2X infrastructure across the U.S., but it could be a ‘down payment’ on significant infrastructure to kick start the industry.

In summary, the Infrastructure Investment and Jobs Act includes many opportunities for digital infrastructure investment, in addition to the $65 billion slotted to improve broadband infrastructure. The funding included in this bill will have a positive impact on an array of digital infrastructure providers, from fiber installers to small cell vendors. Digital infrastructure will be used to improve all of America’s infrastructure.

Should We Party Like It Is 1999?

Boom!  It is finally done.  On Friday (11/5) night, Congress passed the long- awaited Infrastructure Investment and Jobs Act.  This bill allocates $43 billion for broadband infrastructure, and an incremental $23 billion for “projects and funding bearing upon digital inclusion.” $65 billion earmarked for broadband in total.   No surprise, this represents the largest single federal investment EVER in broadband.  
So the question becomes – now what?  A lot it seems.   One of the common threads in our write up has been that the pivot to fiber is happening at tremendous pace.  Giants such as AT&T have more than woken up to the fiber theme (note: the word ‘fiber’ was mentioned 37 times in its Q3’21 earnings call).  The smaller players have showed their fiber enthusiasm as well (note: Shentel’s recent move to ditch FWA in favor of a more ambitious fiber build).  
While some of these are recent pivots, the fact is the industry have been moving down this fiber deep broadband path for the past several years.  According to USTelecom’s data, US Broadband Provider Cap-x for the last three years has been in the range of ~ $80B EACH year….even in the face of a pandemic.  By layering on the significant funding coming from this bill, what will the broadband world look like years from now with the pedal-to-the-metal at breakneck speed? 
A company CEO recently said to me when observing all this fiber push – “I am saying this because you may be one of the few that remember….we are entering 1999 again.”  It was a curious comment because not only do I remember 1999 but I (vividly) remember the telecom balloon going ‘pop’ two years after.  And it was….ummm…not fun. 
For those who don’t remember…the catalyst for the telecom bubble burst was the discovery of Worldcom’s fraud but there really was much more carnage looming about.  During that time it seemed like anyone with a pretty PowerPoint and snazzy marketing campaign could raise capital to build broadband.  Companies such as Covad, PSINet and Rhythms NetConnections were Wall Street darlings in the late 1990s (I remember being an associate analyst following them!).  At the time, these players were perceived to be viable alternatives to the Baby Bells (there were still about six of them at the time) for services like high-speed Internet access.  But all these players (as well as a host of others) ended up filing for bankruptcy protection in 2001.   
While the CLEC problem was they had to lease copper lines leased to them by larger carriers, the bigger picture issue was it was somewhat of a cowboy ‘build it and it will come’ mentality.  There are many facts which suggest that this time is much different.  Data from AT&T’s Q3’21 trending schedule shows that the only broadband adds that are in the black are the fiber ones (Non-Fiber and DSL continue to bleed red).    AT&T is not alone in this regard.    The demand for faster broadband speeds is very much there.
So can we just calmly say “That was then, and this is now?”  The answer: maybe.  
I fully acknowledge the idea of a broadband stimulus is nothing but good.  Who does not want to close the digital and homework divide?  We all know how heartbreaking it was seeing the pictures of kids sitting outside fast food restaurants at the height of COVID to grab on to their WiFi connection.  But the question to watch is how this money gets allocated and does it get into the right hands?
In my view, the key horses to watch will be the ones doing ‘think outside the box’ type moves.  Watch the likes of Conexon and other models that partner with the Co-ops (both electric and telecom) to become a more meaningful player in the market.   Another important trend to watch will be open networks.  The Europeans have embraced this model – yet, over on this side of the pond, they have been welcomed with much less fanfare.  But with billions of dollars raining down from the broadband gods….one has to wonder if these type of models finally find their sea legs.    If the capital is allocated in a way which will drive effective broadband builds in both underserved rural and urban (note: I live 14 miles north of Chicago and basically have the choice of one broadband provider!)  then a 2001-like event won’t happen.
However, if the ‘horses’ I describe above or ones like it get trampled or don’t get a chance to gallop out of the gates and the capital gets allocated to players who have the ‘broadband white board’ and not much else….then the ghosts of 2001 could come back to roost. 
So Washington can have a chest thumping moment for now – but how execution of the plan goes from here will be critical to say the least.

The importance of narratives

An investment columnist’s job always involves the juggling of narratives. The economist Robert Shiller, rightly in my view, thinks that investors are far less focussed on hard numbers than they say they are and much more impressed by financial narratives. Take for instance a boring measure such as price to earnings ratios based on forward earnings ratio. In the good/bad old days, if that number was much above 20, investors would frown and mutter ‘expensive’ under their breath. But once the narrative of technological disruption becomes attached to the PE ratio and the investors can combine the narrative of tech change and corporate adaptation (currently called digitisation), then suddenly 20 looks laughably cheap.

And so, narratives are important. Which brings me to this column. I was toying with the idea of investigating the wave of M&A activity within the digital infrastructure sector and the wall of private equity money heading into the sector. But another narrative screams out at me: market mayhem or, less elegantly, the possibility of an imminent equities sell off.

As I write this, the US markets in particular seem to be shaking off the growing number of alarm bells. As a literate audience you’ll be able to tick off the obvious ones such as (in no order of importance): China; inflation; Covid not quite going away; faltering recovery as stimulus dies down. I could also mention valuations, slowing earnings growth projections for mid-2020 and a few other bond market flashing sirens, but for me the most obvious factors are the least discussed, notably faltering momentum and declining global liquidity flows.

For this observer, the logic of market gains over the last decade isn’t terrifically complicated. Central banks have flushed money into the financial system, bought out bond holders and encouraged investors to take the view that there is no alternative except to invest in equities. This liquidity driven push into equities takes on a form of momentum over time, pushing yet more money towards sectors, sometimes unconnected to fundamentals.

Quite a few hedge fund-oriented research outfits keep a close eye on these flows and many are now saying that their liquidity measures are flashing red. For instance, one London based outfit called Cross Border Capital only this week observed that “both US private sector liquidity and US cross-border capital inflows have already turned lower from exceptionally high levels, whereas US Fed liquidity remains unusually elevated”. If the US Federal Reserve was to start to aggressively taper, they think we’re guaranteed a big stock market correction.

It’s also the case that as liquidity tightens, equity market momentum falters. In that scenario the marginal buyer of equities – the ones who felt they had no choice other than to invest in equities – head for the hills, sell equities en-masse and start to re-appraise their bond portfolios. We are now at one of those possible turning points.

Most measures of relative strength within US equities are beginning to point downwards. That said, one shouldn’t get too carried away with crystal ball gazing. Market volatility levels are still low – the classic measure, the VIX, is languishing below 15, significantly below the 25+ point mid-September, though it’s also worth noting that that level is still quite elevated compared to FX volatility (CVIX Index) and rates volatility (MOVE Index). As for earnings, Corporate USA Inc still looks in good shape as we head into another earnings season. According to BlackRock “companies representing more than half of the S&P 500 Index market value have reported third-quarter earnings. Over 80% of them have beaten expectations on profit and more than three quarters have exceeded revenue estimates”.

Obviously, we need to take all this debate around signals with a large household cup of salt but when you combine it with the macro and geopolitical warning signs I’ve already listed, then maybe it’s time to sit up and take notice. Which prompts my next question – what happens to the publicly listed digital infrastructure space if we do head into a downturn?

As luck would have it, I’ve been crunching numbers across a number of sectors in recent weeks looking at what happened in past stock market sell offs. The big table below contains the fruits of those efforts.

I’ve looked at six S&P 500 sell offs over the last 15 years or so since QE properly started. Declines in the US benchmark index have varied between just under 10% through to 48% in 2008/9. I’ve also pulled out the biggest stock market listed digital infrastructure businesses – which happen to be the leading holdings within the Digital Infrastructure Investor Index (DIGITAL) – and analysed how they performed during those periods of maximum distress.

I’ve also included peer group data for general tech stocks – I’ve used Microsoft as a surrogate – and traditional infrastructure funds, in this case two long established UK listed investment funds, HICL and INPP which have both been around on the London market for two decades. These two funds invest in ‘classic’ infrastructure such as public private partnerships and they are the very definition of what it means to be an alternative fund.

First the good news. By and large the leading digital infrastructure businesses – mostly US listed of course – have fallen by less than the market index (the S&P 500). Unfortunately, that relative out- performance isn’t actually that impressive. Take the February/March 2020 sell off – the S&P 500 fell 33% but my average for 11 large cap digital infra businesses declined 23%. By comparison Microsoft was down 23% while even my two ‘classic’ infrastructure names were down around 20%. Or take the 2008/2009 sell off. The S&P 500 was down 48% but digital infra stocks were down 44%.

Now the bad news. I’ve also included measures of share price volatility which clearly show that the daily dispersion of returns is much, much greater for my digital infrastructure businesses. It’s also on average in excess of Microsoft and the ‘classic’ infrastructure plays.

Stepping back from the wall of numbers I’d make the following observation. Many big institutional investors choose to invest in the digital infrastructure for two reasons. The first, much voiced by the sector, is a thematic one based around growth. I would argue that an equally important reason is that investors like the bond-like dependability of the sector: steady cashflows unaffected by the business cycle should provide some diversification benefits in the event of a sell off. In effect this pigeonholes digital infra stocks as a form of low volatility, quality stock. But the numbers above don’t really support this notion of digital infra stocks as being even remotely alternative. In reality, digital infrastructure stocks display very few of the qualities I’d expect from a genuine alternative – in fact they sell off nearly as aggressively as growth stocks in a high volatility regime. If the current market moves from its current levels of market volatility to a proper sell off – a big if I grant you – then my worry is that digital infrastructure stocks will provide no safe haven.