The way tower companies have been built is with a historic view that they have to dominate that one space, according to Marc Ganzi, CEO of Digital Colony Management and president and CEO of Colony Capital, a global digital infrastructure, real estate and investment management firm with 15 digital portfolio companies, 350 employees and $47 billion worth of assets under management. However, Ganzi said that tower customers’ needs are changing, and the networks they serve are changing, too. Ganzi spoke about telecommunications towers and digital real estate at a Connected Virtual Tech Event presented by Fifth Gen Media.
Regarding towers, Ganzi said that today, site acquisition specialists do not receive from mobile network operators a search ring as they did before. “It’s not like, ‘Here’s a polygon; go work it. Go bring me back three candidates,’” he said.
Ganzi said an inquiry these days is more likely to say something like, “I have 26 polygons. My standard deviation is 500 yards. I need 14 to 18 strands of fiber. By the way, I am going to put remote access units (RAUs) on top of all this. I don’t care if it is a tower or a utility pole. Here are my coverage specifications. By the way, we are fronthauling this back. You have to find me a radio access network (RAN) hub where I am going to have 3,500 square feet. Here are my technical standards for my RAN hub. And here is another list of technical standards for my cloud partner, who, by the way, is Amazon Web Services (AWS), and you have to create an ecosystem for them, too.”
At this point, Ganzi said, if you are a tower operator, “your mind just blows, because what you have just been handed is a mesh network that is going to interface and ultimately deploy multiple-input, multiple output (MIMO) technology. The only way you can deploy MIMO is with multiple radio access point networks, where there is a small cell node, a Wi-Fi access point — whatever it is, it is all coming back to a point of aggregation. That point of aggregation is called decentralized RAN. The radio access network is now being decentralized.”
Network operators are not building $250 million Ericsson switches anymore, Ganzi said. Those were back in the days when they were building 2G and 3G networks, he explained, but that sort of construction is just not happening, anymore. “The infrastructure is lighter,” he said. “It is more nimble. The ability to design and proliferate those networks requires a totally different abacus than what we were dealing with before.”
Digital Colony is going in that direction, and Ganzi said the evidence is reflected in the work the company performs on edge computing and small cell infrastructure side, and is reflected by the amount of fiber Digital Colony bought with its acquisition of Zayo Group Holdings for $8.2 billion in partnership with EQT.
“We are prepping ourselves for a different kind of battle,” Ganzi said. “We are going to play the game differently from the way we played it before. I don’t know if our industry is adapting fast enough, which is why you are seeing more carriers self-performing than ever around their 5G architecture.”
Nevertheless, towers are still great, Ganzi said. “American Tower is not going to go out of business because they whiffed on Cloud-RAN,” he said. “Equinix is not going to go out of business because they missed on edge computing. All these are great, well-run, multibillion-dollar companies that are built for the future. However, the good growth, the high single-digit, low double-digit growth, is in a different place, now. If you want to chase that, you have to go one way. If you want 3 percent to 4 percent growth, you go another way.”
Rich Berliner of Fifth Gen Media asked Ganzi a question related to why industry analysts have more praise for Digital Colony’s fiber-optic cable assets than for those owned by tower operator Crown Castle International. Ganzi said that for Crown, the difficulty involves storytelling. He said that Crown is beginning to give a lot more disclosure about its fiber business. What many do not understand, Ganzi said, is that unlike towers, fiber has multiple different types of businesses.
“You can go invest in infrastructure, which is the part that Jay loves,” Ganzi said, referring to Jay Brown, CEO of Crown. “Infrastructure includes long-term leases, long-haul metro rings, fiber to the tower and fiber to the node. Those four verticals are performing really well for Crown. They are having a lot of success in their long-haul routes, their metro rings, and their fiber-to-the-tower business and their small cell business. Where Crown has gotten a little bit sideways is with its enterprise business.”
The enterprise fiber business is super-competitive, Ganzi said he tells everyone. He said it is hard to be successful in enterprise fiber “because there is no privacy in having that great, zoned tower where, for example, you are driving down the Garden State Parkway, and there is this one pine tree that you got zoned that no one else could get zoned. You know how that ends. It ends with four rad centers and a tower that is doing $180,000 cash flow.”
In comparison, Ganzi said, “you go run fiber down the middle of Paramus, New Jersey, and there will be four or five competitors running down that conduit. If I have a 100,000-square-foot office building, and I have a curb cut with five providers, Crown being one of them, that is just a race to the bottom. You’re pricing circuits, you’re pricing wavelengths, you’re pricing dark fiber, and it is brutally competitive.”
Although that is the part of the business that Ganzi said he did not want to say Crown underestimated or overestimated, he said that from Day One, Crown should have told the story this way: “We have this other business, and it is not like towers. Let us explain it to you. It is going to have 60 to 100 basis points of churn every month. But, here is the good news: We are growing it at 14 percent to 15 percent, so it is a nice little net 3 percent to 4 percent growth business.”
If Crown would have disclosed that, pushed it over to the left really fast, and kept its wholesale fiber business in one lane, reported small cells in the third lane, and towers in the fourth lane, Ganzi said, the company would not have had a problem. Now, he said, Crown has to unwind that story and tell the story differently. Ganzi said he understood Crown’s situation with telling it story the right way, because he observed that Zayo Group Holdings had the same problem. The reason Digital Colony was able to buy Zayo at the right price, he said, was that Zayo had a messaging problem.
“The reality is that Crown has a great management team and a great set of assets,” Ganzi said. “They are going to figure out how to tell the story. Trying to communicate with public investors what fiber is, is hard. My good friend Jennifer Fritzsche at Wells Fargo Securities used to say, ‘Fiber is just a tower turned on its side.’ It’s not that simple.”
Ganzi listed factors that differentiate the fiber business from the tower business: “Permitting. Strategic moat. Privacy of conduits. Curb cuts. Access to buildings. The Telecommunications Act of 1996 that leveled the playing field, giving equal access to all property owners. Non-discriminatory access curb cuts to the minimum/main point-of-entry (MPOE). Once that happened, building owners could not keep us out of their MPOE. An enterprise had to let everyone in to the MPOE. What an enterprise did inside its risers was up to the enterprise, but the enterprise had to provide open access for everyone. That has created kind of a free-for-all with the enterprise.”
Since Digital Colony bought Zayo Group Holdings, Ganzi said he has served on Zayo’s board of directors, and he spends three to four hours a week on matters involving Zayo. Every week, he said, he learns more about the fiber business, which he described as super-interesting. He said Digital Colony is going to do really well with Zayo, yet other companies may not do well with some of their fiber assets if they paid a multiple of 28 or 30 times for those assets, as he said he has observed some pay.
“The fiber business is hard,” Ganzi said. “You have to wake up, and you have to fight every day.”
Referring to how much of his company’s digital strategy revolves around 5G wireless communications, Ganzi spoke of the money spent on digital during 2020. “There will be about a $378 billion total available market (TAM) in digital spend,” he said, “and $211 billion of that will be fiber. Some of that has to be 5G, yes? But a lot of that just has to be bringing broadband out to homes, and a response to the COVID-19 pandemic. So, big overweight in fiber to the home, big overweight in fiber to the whatever — data center, tower, you name it — and then, fiber to the enterprise.”
In Ganzi’s view, fiber is not necessarily a 5G catalyst and, instead, fiber is the connective tissue that brings it all together. As for towers and small cells, he said they probably represent about an $80 billion TAM in 2020, in terms of what was spent, and probably another $70 billion to $80 billion in data center spend. With data center spend, Ganzi said, everything Digital Colony is doing is oriented to the cloud because most of the growth is based on cloud computing, rotation of public cloud, rotation of private cloud, enterprise workload-shifting and IoT workload-shifting.
He said Digital Colony also is doing quite a bit with artificial intelligence (AI), which is reflected in many AI workloads with Nvidia and with teaching universities such as Carnegie Mellon and Georgia Tech, where Digital Colony has data centers on their campuses. IBM defines AI workloads as applications based on machine learning and deep learning, using unstructured data and information as the fuel to drive these applications.
“That’s what we call big compute, because those types of compute loads are big power density generators.”
Under the product name of DGX Systems, Nvidia offers what it calls the world’s first portfolio of purpose-built supercomputers designed to give data scientists the most powerful tools for AI exploration that goes from the desk to the data center and to the cloud.
“You have to have a sophisticated facility to deal with those AI guys, particularly Nvidia,” Ganzi said. “Nvidia is a big customer of ours.”