Consolidation of Fiber Networks

I’ve written a few recent blogs discussing the amount of fiber that’s going to be needed to support the 5G networks envisioned by Verizon and AT&T. This blog in particular cited a recent Deloitte study that estimates that the cost to build the fiber needed to support a ubiquitous 5G network nationwide would be $130 billion.

We know that adding fiber is now a high priority for Verizon. They announced in April a deal to buy over $1 billion of fiber from Corning over 3 years. (As an aside, all of the press releases and articles about that purchase say that amount buys 12.4 million miles of fiber per year, or 37.2 million miles of fiber. There are only a little over 4 million miles of roads in the US, so that obviously means miles of individual fiber strands. Pardon the interruption, but misleading statistics drive me up the wall.)

We can almost be certain that Verizon plans to build fiber for backhaul to cell sites. There are around 250,000 current cell towers in the country, but the deployment of small neighborhood cell sites is going to explode that number potentially by millions. Years ago both Verizon and AT&T elected to let other companies build and own cell towers, which spun off a major new industry. And in that process both companies largely agreed to lease fiber transport to reach those towers. But as the cell industry margins are tightening the companies are now looking to directly own as many of those fiber routes as possible to hold down lease expenses.

While Verizon plans to build a lot of fiber, they are also on an obvious path to buy existing fiber networks that supply transport to cell towers. Last year they purchased XO Communications and just last week announced they were buying a Chicago-area fiber network from Wide Open West.

I have seen several analysts speculate that Verizon will be considering more fiber purchases. Interestingly the analysts focus on the potential purchase of large ILECs like Consolidated or Cincinnati Bell, which both own a lot of fiber. But much of the fiber in these companies is last-mile fiber to reach customers, and it would be curious to see Verizon buy back into that business. Just last year they sold off a significant chunk of their FiOS fiber network to Frontier and it would be a major reversal of that strategy to turn around and invest this soon in last mile fiber. We’ve seen big companies pivot before, but this would be possibly the biggest such change of mind our industry would ever have seen.

I think it’s more likely that they will consider buying transport fiber networks rather than last-mile networks. The problem the company faces is that there are not that many big fiber providers left. CenturyLink recently purchased the largest such network from Level 3, which owns over 55,000 miles of fiber. The only other fiber transport networks left that own over 10,000 miles of fiber are Birch, Zayo, EarthLink, Cogent and Lightower/Fibertech. There are only another half dozen companies that own fiber transport networks of between 5,000 and 10,000 miles. I have to think that Verizon and AT&T have considered buying many of these companies over the last year or two.

There is one other set of big fiber networks that don’t get as much national attention. These are fiber transport networks built largely by consortiums of independent telephone companies. Most of these networks were constructed as a way for the telcos to gain cheap fiber transport to the world outside of their operating territories. Many of these smaller telcos were held hostage to incredibly expensive special access transport from the RBOCs which made it difficult for them to buy affordable Internet access. Since these networks were originally built a lot of them now have expanded throughout their operating regions and are now connected to cell towers, large businesses, governments, universities and other customers needing fiber transport.

Most of these ILEC-owned networks have joined together to form INDATEL. Here is a map showing the wide-spread footprint of INDATEL-member networks. Through this consortium many of these networks are now interconnected, providing a nearly nationwide fiber footprint. The various members have POPs in all of the biggest cities in their region but then also go to all of the smaller communities that have largely been ignored by most of the other fiber providers, with perhaps the exception of Level 3.

I have no idea if either Verizon or AT&T has considered buying these networks. For a company like Verizon these fiber routes would provide transport into many areas where they don’t have fiber today. The owners of these networks might want to explore the possibility of selling their networks. Now that the networks are in place the ILECs that built these networks are no longer isolated from the rest of the world. A sale would let them capitalize on their investment in fiber at a time when fiber networks have an all-time high valuation.

Of course, the downside to all of this is that if Verizon, AT&T and a few others like CenturyLink gobble up the few remaining independent fiber networks they will have a virtual monopoly on fiber transport. During the XO and Level 3 purchases there were a lot comments filed with regulators expressing concern about the negative impact on competition from fiber consolidation. I’d hate to see us go back to the bad old days where the only option for transport was a handful of the big telcos.

Too Much Fiber?

When communities consider building fiber, one of the first questions a community often asks me is how much fiber already exists in their community and how they can take advantage of it. The bad news I almost always have to give them is that their community probably contains several existing fiber networks that will be of little or no use to them. It seems there is a lot of fiber in the world that is not being put to good use.

So what do I mean by this? What I have found is that many communities have numerous existing fiber networks that have been built for one specific purpose and which can’t be used for anything else. Here are some examples:

  • K-12 Schools. School districts often own fiber networks to connect all of their schools.
  • Colleges. Colleges will often be on a different network than the other schools.
  • Traffic Lights. A number of cities now have fiber systems that feed traffic lights.
  • State Highways. They often have fiber network systems for cameras and electronic message boards.
  • Federal Highways. They build for the same reason as state highways.
  • Commercial networks. It’s more understandable why a network built by a telco, cable company, wireless company, or CLEC might not be available to a city, but most cities today contain a significant amount of fiber built by these companies.

I first ran into this issue in the late 90s when a city in Virginia asked me this question. I was helping them design a fiber network that would connect all of their government buildings. In doing so I discovered that there was already a fiber network built to traffic lights that probably already covered 80% of the network they were going to need – and they already owned it. But in looking deeper, we found that the traffic light network had been built with funds from the state highway department and that it had a prohibition in the funding language against sharing the network for other uses, including other uses by the city. That network was basically off-limits for any other use.

When you consider that building fiber can range in price from $25,000 per mile to place on poles, or $75,000 per mile to bury (in most places) or even up to $150,000 per mile in urban downtowns, it’s crazy to think that such money has been spent without considering all of the other benefits the outlay could have created.

I still see this all of the time and it is very common for a government-built fiber to be off limits to all commercial uses. But surprisingly there are often also prohibitions against other municipal uses. I can understand restrictions against commercial uses, even if I don’t like them. The fear is always there that when the government and commercial entities work together that it creates a chance for corruption. But this kind of fear should not be a reason to automatically write-off the opportunity for public-private partnerships.

I’ve always found that commercial companies are glad to share the cost of building a new fiber route. In the commercial world companies routinely share fibers and they typically create a clear division of the use of fiber pairs on a new route when multiple companies agree to share in the build costs. Governments could save a fortune if they would join into this well-established commercial practice of building fiber for more than one company.

But the restrictions of a government-owned fiber that precludes other parts of the government from using it are just wrong. When highway departments or universities or other big agencies build fiber and then don’t let other government agencies benefit from the expenditure something is very wrong and we have let bureaucracy override common sense. I often hear excuses for the practice such as the need for security, and frankly all such excuses are bosh.

I’ve told cities that there are two solutions when they run into this problem. One is to create a huge public stink so that the agency that won’t share the fibers might be shamed into doing the right thing. But the other fix is longer term, and that is to take full control of their rights-of-way. For example, one long-term fix is to require that anybody who digs a ditch in the ground must include empty conduit which will create a lot of opportunity for cheap fiber over time. But the best fix is for somebody in the city to act entrepreneurially and to get to know the fiber providers in town and develop partnerships with them. That is actually easier to do than you might think.