Category Archives: The Industry

Is There a Spectrum Shortage?

I’ve regularly seen editorials and blogs that claim that the U.S. is far behind in providing spectrum to wireless carriers. The purpose of these articles seems to be to lobby the FCC to make more cellular spectrum available, particularly mid-range spectrum. That’s spectrum between 1 GHz and 6 GHz, which is the sweet space for 5G cellular coverage. Spectrum in this range is perfect for mobile cell coverage – the spectrum penetrates foliage and other impediments and also carriers for a decent distance of a few miles from cell towers.

Part of the reason for the recent spate of articles is that the FCC has lost the ability to hold spectrum auctions. In March 2023, Congress declined to renew the agencies ability to hold auctions, which has been the common way to decide how to allocate spectrum – to those who are willing to pay the most to use it.

A common claim made is that the U.S. is far behind China, which is the bogey-man for everything wireless. I recall an op-ed in the Wall Street Journal that claimed that China has 245 5G base stations per 100,000 people, Europe has 103, and the U.S. has only 51. While this may be true, this is more of a case of the willingness of U.S. cell carriers to invest in new cell towers and sites. In U.S. cities, cellphone download speeds are regularly reported to be over 100 Mbps download, and cell carriers here don’t see any reason why they need to beef up speeds more than that.

The main difference between the cellular networks in U.S. compared to the rest of the world is that a lot more people here rely on landline broadband and WiFi for their broadband needs versus using the cellular networks. In pars of the world where wireless carries a much larger share o broadband, it’s far more important to have robust wireless networks.

I’ve seen a lot of wireless experts that claim that the big U.S. foray into FWA wireless is due to carriers finding a way to monetize the excess spectrum that cell carriers already own. It’s an interesting choice of how to use spectrum, because from a monetary perspective, cell carriers are selling a gigabit of FWA broadband at a tiny fraction of what they charge for a gigabit of cellular broadband. It doesn’t make sense that carriers would pursue FWA so aggressively if they were feeling spectrum constraints.

Another reason why U.S. cellular carriers have excess spectrum capacity is that the many projected uses for 5G that were touted a decade ago never materialized. They had originally predicted that a lot of the devices in our homes and outdoor sensors would all be subscribed to cellular – but it never happened. In the home, everything moved to free WiFi, which became overabundant when ISPs increased home broadband speeds, and the FCC released a gigantic pile of new WiFi spectrum in the 6 GHz band. Connecting outdoor sensors to cellular never took off because it required deploying batteries that have to be replaced.

This is not to say that we won’t ever have a shortage of spectrum. The use of wireless broadband is growing at a faster pace than home broadband. While it’s still only a tiny fraction of all broadband usage, the inexorable impact of annual growth will mean that new spectrum will be needed.

One of the biggest controversies with mid-range spectrum is that much of it is use by various parts of the military and government. The process of relocating government users to different spectrum bands is costly and takes many years. Government users often make a convincing case why they should be allowed to continue to use the spectrum they have.

If there is any place where there is a spectrum shortage, it’s in rural America. This is not due to a lack of available spectrum – but of spectrum that they owners are willing to share. Carriers generally buy spectrum to satisfy their needs in urban areas. Bands of spectrum that might be full and business in a city likely sit fallow and unused in rural areas where there is no demand. There are huge amounts if mid-range spectrum that are not in use in most rural markets.

Cellular carriers could share spectrum. We see that happening with CBRS where spectrum holders in rural areas are required to allow others to use the spectrum when it’s not in use. But spectrum owners won’t do this voluntarily because it adds to their costs. The FCC has to act to require spectrum sharing on any given band of spectrum. In FCC filings, the cellular carriers argue strongly against requiring more sharing – and they would like to curtail what is already allowed.

Is BEAD a 10-Year Program?

Politico ran an article earlier this month by John Hendel that noted the slow pace of the BEAD grant program. It’s a newsworthy topic because during the current election cycle I’ve heard Democrats mention fixing rural broadband as one of their accomplishments.

There is one paragraph in that article that instantly caught my eye. The government is executing BEAD “on the 10-year timeline Congress intended,” a spokesperson for the National Telecommunications and Information Administration, the Commerce Department agency tasked with overseeing the program, told POLITICO.

This is the first time I’ve heard that BEAD was intended to be a 10-year plan. I can recall numerous occasions during the BEAD process when the public was told that the goal was to solve rural broadband. I sat through numerous public meetings where State Broadband offices met with the public and the message was always that rural broadband is a huge problem and the goal was to start fixing it as quickly as is practical.

There were a lot of other big infrastructure programs funded by the IIJA that are moving faster than BEAD. Consider the Bridge Investment Program (BIP). This is a $40 billion program for bridge replacement, rehabilitation, preservation, and protection. The U.S. Department of Transportation started to accept BIP grant applications in December 2023. That’s roughly a year faster than BEAD. There are many other examples of spending that have already been made from the same infrastructure legislation.

I have a hard time believing that it’s faster and easier create the engineering studies needed to build or fix a bridge than it is to design a broadband network. Any local government that applied for a BIP bridge grant will already have done the detailed engineering analysis needed to quantify the amount of grant needed.

Unfortunately, BEAD is almost turning into a 10-year program. States will be making BEAD awards starting sometime in 2025. A State can’t make a grant award until they find an ISP to serve every unserved and underserved location and also gotten NTIA approval for the full pile of awards. I don’t expect more than perhaps a few token customers to be connected to a BEAD-constructed network in 2025. ISPs will have up to four years after grant contracts have been signed. This means that if everything works as planned, some BEAD projects won’t be completed until sometime in 2029 – the eighth year after the BEAD program was created by Congress.

There would have been a huge outcry inside the industry and from the rural politicians if NTIA had formally announced years ago that BEAD was intended to be a 10-year grant program. There have been countless opportunities for NTIA to have made this claim, and it is only now arising in a quote given to a reporter from a political site. We can’t know who provided the quote to Politco, but if this is an official line, then NTIA is throwing up a smoke screen to cover for the slowness of the program. It’s disappointing to know that a rural household without good broadband will see a student who was in the sixth grade when BEAD was announced graduate from high school without seeing any improvement from the BEAD program.

How Does U.S. Broadband Stack Up?

FarrPoint, an international digital consultant from Edinburgh, Scotland, recently released its International Digital Connectivity Readiness Index that rates the G7 nations on a variety of factors that it believes defines a counties success in implementing digital technologies. FarrPoint says it consults with 4G, 5G, NTN, digital connectivity, enterprise networking, and technology-enabled care. The G7 nations include Canada, France, Germany, Italy, Japan, the UK, and the United States.

The analysis ranked each country on a scale of zero to one hundred on twelve different factors that contribute to overall digital readiness. The report ranks the U.S. as tied with Germany for second in overall readiness with an overall score of 83 – behind number one UK, with a rating on 84. Italy was seventh with a rating of 74.

Looking at individual rating is interesting and verifies some of the issues that are widely discussed about the state of U.S. broadband.

  • Japan got a rating of 100 for Gigabit Connectivity. France was close behind with a rating of 97. The U.S. was last in this category with a rating of 48, with the second lowest rating for Germany with a 62.
  • In the category of Decent Fixed Connectivity, France was last with a rating of 82, while all of the other countries had ratings between 91 and 94.
  • When looking at 4G Connectivity, four countries – Canada, France, Germany, and the U.S. got a rating of 100. The lowest rating was Italy at 88. I’m curious about the 100 rating for the U.S. since the FCC just announced a $9 billion 5G Fund to bring cellular coverage to rural areas that have none.
  • The U.S. was number one for 5G Connectivity with a rating of 82. Germany and Japan were close behind the U.S., with the other countries lower, and the lowest being Italy with a rating of 57.
  • The report combined the above four categories to create a rating for Digital Connectivity Infrastructure. The leaders were Japan (94) and Canada (93). The U.S. was next to last with a rating of 87.
  • In assessing Online Households, the leaders were Canada and the UK with a rating of 89. The U.S. was fifth with a rating of 84. Interestingly, Japan was sixth with a rating of 79 even though the company had a 100 rating for gigabit availability.
  • In the category of Digital Skills, the leaders were Germany and the UK at 78, with the U.S. third at 76.
  • Another category is Online Security. No country was rated as great, with ratings ranging from 67 for France to 77 for Canada. The U.S. was rated at 71.
  • There was also a close knit grouping of ratings for Online Wellbeing. Canada led this category with a rating of 76 The U.S. and U.K were second with a rating of 72.
  • An interesting category was Accessible Public Services. The U.S. led this category with a rating of 93, followed closely by Japan and the UK.
  • In the category which is the most glaring weakness for the U.S, we ranked low for Affordability, with rating of 56. The lowest rating was Japan at 37, and the highest ratings for affordability were in Germany (80) and the UK (79). The U.S. dropped from a rating of 60 in 2023 – indicting that broadband is becoming less affordable – perhaps due to the end of the ACP program.
  • There was a category for having a thriving Digital economy, and the U.S. and UK led this category with a rating of 92. Lowest was Italy (66) and Canada (69).
  • The final category was Innovation. The U.S. led this category with a rating of 91. There was a big gap to the lowest countries of Italy (52) and Canada (53).

Most ratings were similar to FarrPoint’s 2023 ratings. The biggest one-year changes were UK’s 5G Connectivity (58 to 68), France’s 4G Connectivity (90 to 100), France’s Accessibility of Public Services (78 to 87), and the U.S. Gigabit Connectivity (41 to 48).

Has Your ISP Been Hacked?

As if we didn’t have a long enough list of problems to worry about, Lumen researchers at its Black Lotus Labs recently released a blog that said that it knows of three U.S. ISPs and one in India was hacked this summer. Lumen said the hackers took advantage of flaws in software provided by Versa Networks being used to manage wide-area networks.

The hacks were described as zero-day hacks, which describes a software or hardware vulnerability that is unknown to the vendor and for which no patch or other fix is immediately available – a vendor has zero days to prepare a patch when a vulnerability is exploited. While developers have a  goal of delivering products with no vulnerability, virtually all software and hardware contain bugs of some sort.

Lumen said it is moderately confident that the attacks were originated by a group known as Volt Typhoon, a Chinese state-sponsored set of hackers. This is the group that U.S. intelligence officials said had been trying to penetrate and hack American ports. Lumen describes the attacks as highly significant.

CISA (the Cybersecurity & Infrastructure Security Agency) leads the effort to protect the country against malicious hackers. The agency’s mission is to “lead the national effort to understand, manage, and reduce risk to our cyber and physical infrastructure.”

Unfortunately, the U.S. is in the bullseye for ransomware attacks. The following chart, which comes from NTT Data shows the industries that are the targets of hacking in 2024. Manufacturing has overtaken the technology sector as the most attacked, while the financial sector has climbed to third.

Hackers constantly change tactics to stay ahead of the effort to block them. In 2024 there has been a decline in banking trojans that has been offset by a rise in info-stealers and penetration testing tools.

The big concern in the security industry is that hackers are starting to deploy AI to aid in hacks. AI is particularly useful in finding vulnerable code that is buried in lengthy software programs. AI is likely going to be used as a tool to develop a custom approach to hacked any given entity.

I know it sounds like old and trite advice, but ISPs need to adopt robust cybersecurity systems to try to keep most hackers out of your network and software.

How Big is the Broadband Industry

I constantly see articles that make claims about the percentage of homes that have broadband, cable TV, or telephone service. I remember an FCC report a few years ago that claimed that 88% of homes at the time had home broadband. Any time I see a statement like that, I ask the question – how many total homes are in the U.S. – a number that is needed to calculate a penetration rate. There doesn’t seem to be any consensus on that question.

Let me provide some examples.

  • Cartesian recently released a report in conjunction with the Fiber Broadband association that said as of June 2023 there were 54.1 million fiber passings, which the report says is a 45% national market penetration. That math suggests there are 120 million total U.S. possible passings.
  • RVA released another report in conjunction with the FBA that said as of September 2023 there were 69 million fiber passings, which equates to 128 million total U.S. possible passings.
  • Leichtman Research Group has for years published a list of the claimed broadband customers of the largest ISPs. As of the end of end of 2023 the biggest ISPs collectively claimed 114.7 million broadband customers. Leichtman said these ISPs represent 95% of the broadband market, which implies that there are more than 120 million broadband customers at the end of 2023. That implies a much higher number of total possible passings.
  • The FCC broadband map counts BSLs (broadband serviceable locations), which are places where a customer can buy broadband. The FCC fabric says there were 114.4 million BSLs at the end of 2023. It’s hard to know what to do with this number.

Over the years, I’ve tried to use the U.S. Census to figure this out. According to the U.S. Census Bureau, the total potential U.S. broadband market at the end of 2023 was something like the following:

Residential living units           145.8 million

Business establishments         13.6 million

Total                                         169.4 million

The business count comes from subtracting 19 million home-based businesses from 32.6 million total businesses in the Census.

Consider some other statistics that probably have some impact on figuring out the denominator of potential broadband customers:

  • CoStar, which tracks the housing industry, says there were 18.4 million apartment units at the end of 2023.
  • A recent survey from GOBankingRates said that 40% of families have a second or vacation home. Counting vacation homes is a controversial issue with the FCC maps. Vacation homes can very from multi-million dollar homes at the beach to a hunting cabin with no electricity.
  • A recent study by Lending Tree showed there are 5.5 million vacant homes just in the largest 179 metropolitan areas that are used as short-term rentals and vacation homes.

I’ve tackled this exercise periodically and have never gotten close to having a satisfactory answer. I think many of the folks who cite penetration rates are off base – as you can see from this blog, that’s not surprising. Who knows what ISPs are counting when they claim fiber passings? I have an idea that a lot of ISPs are claiming the same passings, particular for multi-dwelling units. Are folks counting businesses properly? Are they accounting for abandoned and empty homes? What’s the right way to count second homes?

The bottom line for me is to assume any statistic that claims a national penetration rate for a  broadband statistic is not accurate – possibly badly so. This doesn’t imply that the folks making a claim are being deceptive – I think it just means that there is no consensus for the denominator used to calculate a national penetration rate. There are folks in academia and government who might be able to shed light on this – but I would expect that different experts will come up with different methods of counting.

CNET Survey of Broadband Rates

Joey Supan wrote an article for CNET with the headline of Internet in the US Costs $63 a Month. The research for the article is based on the advertised prices charged in different parts of the country by 27 of the largest ISPs. The headline is for the median price found in that research, meaning half of the prices were higher and half lower. The $63 price assumed that customers would take advantage of any auto-pay discount.

The article went on to add that the ISPs add another average $15 to rent the equipment needed for broadband, meaning the average total monthly cost of broadband in his analysis is $78. The report also assumed people pay the advertised marketing rates but that many ISPs jack up advertised rates in the second and third years. The medium price before equipment charged increased to $76 in the second year and $80 in the third year. With equipment fees added, the second and third year median rates increase to $91 and $95.

That is higher than what I’ve been seeing from surveys we’ve been conducting over the last few years in rural counties across the country. Our surveys have shown average broadband bills (not median prices) range from $75 to $80. The differences in average prices is county-specific based on the ISPs operating locally.

I think the difference between my surveys and the results of the CNET article is that a lot of folks are willing to accept slower speeds, and inferior broadband performance to save money. The CNET price assumes that people buy the best broadband available to them.

This article also shows what most other analysis shows – people in rural areas pay more for broadband than people in towns and cities. In many counties Starlink has done well, but they are more expensive than other broadband. Subscriptions to high-orbit satellites or cellular hotspots are also incredibly expensive due to overage charges due to low data caps – something this study couldn’t quantify from a rate sheet.

To offset what CNET reports and we’ve found, prices are currently dropping for lucky rural customers within range of the FWA broadband from Verizon or T-Mobile – but the coverage areas of this new technology is generally not large in the rural counties I’ve examined in detail.

The study only found a few examples of broadband junk fees in broadband prices. This included a $12.97 network access and maintenance fee from Astound, a $12.95 tech assurance fee from Metronet, a $3.99 Internet cost recovery fee from CenturyLink, and $2.97 broadband cost recovery fee from Consolidated Communications. We also see that many ISPs also hide or disguise their equipment fees when they advertise low rates.

The profile of available prices will change in rural areas when new broadband networks are finally built from the many grant programs. That’s not always going to mean lower prices, but it should mean higher-quality broadband. I expect customers paying a lot for satellite broadband or traditional cellular hotspots today will flock to a new grant-funded network. I expect the advent of grant solutions will allow telcos to finally shut down DSL networks and force households to some other technology. In urban areas, competition from fiber and FWA is pushing down prices in neighborhoods that are lucky enough to get a technology alternative.

Consumers have been yelling for broadband competition for many years, and we’re finally seeing it. Unfortunately, it’s happening neighborhood by neighborhood and not industry-wide, but it’s a start. I expect that if CNET repeats this survey in a year they will find even lower rates for advertised broadband.

Telecom Strikes

As I write this, there has been a strike by 17,000 union member of the Communications Workers of America (CWA) against AT&T that started on August 16. Like most strikes that last for more than a week, there is some rhetoric flying from both parties accusing the other side of negotiating in bad faith.

Telcom strikes have been part of the history of the industry. One of the first big telecom strikes was in 1947 with the union at that time being the National Federation of Telephone Workers. AT&T was determined at the time to bust the union, and instead of negotiating nationally, reached agreements with regional subsets of workers, which quickly broke the union.

The CWA was born out of the aftermath of the 1947 strike. CWA workers went on strike over the years with regional strikes against carriers like Southern Bell and General Telephone Company. CWA had a nationwide strike against the Ma Bell AT&T in 1968, 1971, and 1983. After divestiture there were strikes against AT&T in 1986 and  Verizon in 2000 and 2011.

The nationwide strikes have generally been resolved fairly quickly. The 1968 strike lasted eighteen days, the 1971 strike for seven days, and the 1983 strike for twenty-two days. Other strikes lasted a lot longer, such as a strike against General Telephone in Indiana, Ohio, and Kentucky that lasted for five months. As I am writing this, the current strike against AT&T has already lasted for twenty-two days.

I decided to write about this because I was part of the 1983 strike against AT&T. I was in management at Southwestern Bell and was informed at midnight on a Saturday night to report immediately to be an operator.

Being on the management side of a strike makes you quickly appreciate what the union employees do for a living. A handful of other management people and I walked into an operator center in St. Louis that probably had sixty operator positions. The boards were flashing with waiting calls and we got a crash hands-on course in being an operator from an operator center manager. It was largely a disaster for the first day or two since the quick training could not teach us all of the nuances of being an operator and operating a switchboard.

I ran into a situation that really shook me. I got a call from a woman in a phone booth whose husband was having a heart attack and needed an ambulance. Unfortunately, that phone booth was located at an intersection where St. Louis and two suburbs all came together. I had no idea that each of these towns had a different 911 system, and I guessed wrong and had to try all three towns to find a dispatcher who would send an ambulance. That was ten minutes of panic on my part – and I never found out what happened to the heart attack victim. A few years after the strike, the region started working on a unified 911 system that eliminated this kind of jurisdictional nightmare. The union operators would have known this kind of nuanced information, but I had no clue.

A lot of my colleagues were assigned to fix outdoor problems. The 1983 strike started on August 7, and in St. Louis, that means thunderstorm season. I recall storms hitting almost immediately after the strike started and then regularly through the strike. My friends working outdoors said they fell behind immediately and the backlog of open issues increased every day. They got better at making repairs, but they never came close to catching up. This strike happened long before broadband, and my experience is that broadband customers today are far more demanding than voice customers were in 1983. I can’t imagine what would have happened if there had been a major hail storm or something that knocked out a lot of the area all at once.

I grew up in a union household, and I saw first-hand how being in a strong union benefited workers and their families with good wages and a solid retirement plan. Getting a firsthand look at the day-to-day work of running a giant telco during a strike gave me an increased appreciation of the great work the union folks at the company did every day.

Verizon to Buy Frontier

There are not many telecom headlines that give me an instant smile, but the announcement of Verizon’s offer to buy Frontier Communications is one that did. What amused me was that Frontier grew to its current size by buying cast-off copper networks from Verizon.

Frontier got its start as Citizens Communications but adopted the Frontier name in 2008. Frontier paid a lot for Verizon properties. In July 2010, Frontier paid $8.6 billion for Verizon properties in 14 states that included Fort Wayne, Indiana and West Virginia. In April 2016, Frontier paid $10.5 billion for Verizon properties in Florida, Texas, and California. The company made other acquisitions over the years from GTE, Commonwealth Telephone Company, and AT&T properties in Connecticut.

The current sale is estimated to be valued at around $20 billion. That’s a little less than what Frontier paid over the years for copper networks. However, Verizon is buying Frontier for its fiber portfolio. Frontier had 6.5 million fiber passings at the end of 2023 and has announced plans to add 3.5 million more passings by 2026.

This sale raises some interesting questions. Verizon is clearly interested in having a bigger fiber portfolio. I have to wonder if Verizon will be in favor of Frontier’s remaining fiber expansion plans since much of it is in rural and less densely populated portions of the Frontier portfolio.

Frontier also seems to have a fairly aggressive plan to pursue BEAD grants, particularly in rural areas of its existing footprint. It’s going to be interesting to see if Verizon influences those plans which would come to fruition before the close of a sale.

I’ve always been fascinated seeing ISPs merge and deal with different prices. Prices usually stay the same for a while after a merger, but eventually are brought into synch. Following are the list fiber prices for the two companies. This could mean future catch-up rate increases for Frontier customers.

There must still be many employees at Frontier that came from one of the Verizon acquisitions. It’s likely going to feel really odd for them to end up back where they started.

I’ve heard a lot of folks speculating that the next few years might be full of fiber acquisitions and mergers. T-Mobile recently announced the purchase of Metronet, which followed the purchase of Lumos.

This acquisition brings Verizon close to AT&T’s and would give the company 14.5 million broadband customers, counting at the end of the second quarter of 2024. That’s 7.7 million Verizon broadband customers, 3.8 million Verizon FWA wireless customers, and another 3 million customers from Frontier. The three entities are growing quickly and added 442,000 broadband customers in the second quarter of this year.

Frontier has come a long way since the bankruptcy in filed in 2020. The company emerged from bankruptcy in 2021 with a clean balance sheet, cash to grow, and a new management team.

Broadband Usage 2Q 2024

OpenVault recently published its Broadband Insights Report for the end of the second quarter of 2024. As usual, OpenVault is documenting the continued growth in broadband usage by U.S. households.

I think one of the most useful statistics from OpenVault is the average monthly broadband usage per customers in gigabytes. Below is the trend in average monthly U.S. download and upload volumes since the second quarter of 2020. These are the average amount of data used combined for all residential and small business customers. In the four years, average monthly download usage has increased 52% while average upload usage has increased by 80%.

The average U.S. broadband customer used 46.3 more downloaded gigabytes and 5.4 uploaded gigabits per month than a year earlier. This means continued pressure on broadband networks because if we assume roughly 120 million broadband subscribers, this equates to over 6.2 billion more gigabytes of data used each month than a year earlier.

The growth in what OpenVault calls power users is even more dramatic than overall bandwidth growth. Below are the percentage of broadband customers who use more than 1 terabyte of data per month and those using more than 2 terabytes. OpenVault predicts that more than 20% of all customers will use more than a terabyte per month by the end of this year. These numbers show the potential harm created when ISPs place data caps on monthly usage.

OpenVault always includes other interesting statistics in its quarterly reports:

  • The percentage of subscribers purchasing speeds lower than 400 Mbps is now down to 41% of all customers, a 33% decrease over the last year.
  • Median household broadband usage was 382.6 gigabytes – half of homes use more broadband than the median, and half use less.
  • The report includes a section that shows there is no longer a strong correlation between faster speeds and higher data consumption as was seen in past years. OpenVault looked at 15 different ISPs in detail and found that the average monthly downstream consumption was similar for ISPs with average speeds between 100 Mbps and 700 Mbps.

FWA Wins 2Q 2024

Below is a table showing the broadband additions for many of the largest ISPs for both the first and second quarters of 2024. This list is not complete because privately held companies like Cox,  Mediacom, Windstream, and Breezeline don’t announce total broadband customers. However, the carriers on this list represent a huge percentage of the broadband industry, so seeing their performance gives us a peek at the trends in the industry. This group of companies picked up 70,000 more new customers in the second quarter than in the first.

It’s clear that cable companies are starting to lose an increasing number of customers. The losses in the table don’t yet reflect much, if any losses from the end of ACP that died in the middle of May. Most of the companies on this list announced a soft landing for ACP customers for at least a few months, and we probably won’t see the impact of ACP until fall.

To the dismay of other ISPs, the three FWA cellular customers did better in the second quarter than in the first, and collectively picked up 933,000 net customers in the quarter, including 139,000 from AT&T, the newest entrant into the market.

Telcos collectively showed a small gain for the quarter, although all of the telcos reported significant fiber sales. It’s going to obviously take a while to shake free of DSL losses on copper. Lumen continues to underperform the rest of the telcos, which all show modest gains.

Not included on this list is the fiber networks being purchased by T-Mobile. The company recently announced it’s buying into Lumos and Metronet, and those sales will take a while to close, But those acquisitions give T-Mobile a sizable, and quickly growing fiber business.