Wall Street analysts are warning investors that mobile providers like AT&T, Verizon, T-Mobile and Sprint will have to spend $130-150 billion on fiber optic cables alone to make 5G wireless broadband a reality in the next 5-7 years.

A new Deloitte study found providers will have to spend a lot of money to deploy next generation wireless service across the United States, money that many may be unwilling to spend.

“5G relies heavily on fiber and will likely fall far short of its potential unless the United States significantly increases its deep fiber investments,” the study notes. “Increased speed and capacity from 5G will rely on higher radio frequencies and greater network densification (i.e., increasing the number and concentration of cell sites and access points).”

Unlike earlier cellular technology, which worked from centralized cell towers that covered several miles in all directions, 5G technology is expected to be deployed through “small cell” antennas attached to utility and light poles with coverage limited to just 300-500 feet. To reach city residents, providers will need countless thousands of new antenna installations and a massive fiber network to connect each antenna to the provider.

Telecom providers seeking financing for such networks will face the same criticism Verizon Communications took from Wall Street over the expense of its FiOS fiber-to-the-home upgrade as well as doubts about the viability of other fiber projects like Google Fiber.

Goldman Sachs told its investors back in 2012 that throwing money at Google Fiber or Verizon FiOS was not going to give them a good return on their investment. That year, Goldman was “Still Bullish on Cable, But Not Blind to the Risks.” That report, written by analyst Jason Armstrong, noted Google’s fiber upgrades would cost billions and only further dilute industry profits from increasing competition.

Goldman Sachs steered investors back to the cable industry, which gets significant praise from Wall Street for its ability to repurpose 20-year-old wired infrastructure for enhanced broadband without having to spend huge sums on a complete system rebuild.

In 2013, Alliance Bernstein continued to slam Google Fiber’s buildout as an unwise business investment:

We remain skeptical that Google will find a scalable and economically feasible model to extend its build out to a large portion of the US, as costs would be substantial, regulatory and competitive barriers material, and in the end the effort would have limited impact on the global trajectory of the business.

For example, making the far from trivial assumption that Google can identify 20 million homes in relatively contiguous areas with (on average) similar characteristics as Kansas City when it comes to the most important drivers of network deployment cost, homes per mile of plant and the mix of aerial, buried and underground infrastructure, and that Google decides to build out a fiber network to serve them over a period of five years, we estimate the [total capital expenditure] investment required to be in the order of $11 billion to pass the homes, before acquiring or connecting a single customer.

Some analysts are even questioning the relevance of 5G when providers investing in the massive fiber expansion required for 5G wireless could simply extend fiber cables directly into homes, assuring customers of more bandwidth and reliability. In many cases, fiber to the home technology is actually cheaper than 5G deployment will be.

VantagePoint released a report in February that called a lot of the excitement surrounding 5G “hype” and cautioned it will not be the ultimate broadband solution:

Undoubtedly, 5G wireless technologies will result in better broadband performance than 4G wireless technologies and will offer much promise as a mobile complement to fixed services, but they still will not be the right choice for delivering the rapidly increasing broadband demanded by thousands or millions of households and businesses across America.

Previous analysis of 4th generation (4G) wireless networks clearly demonstrated how these networks, even with generous capacity assumptions for the future, will have limited broadband capabilities, and inevitably will fail to carry the fixed broadband experience that has been and will be demanded by subscribers accustomed to their wireline counterparts. Although there is understandably much anticipation today about phenomenal possible speeds for 5G wireless networks tomorrow, they will continue to have technical shortcomings that will, like their predecessor wireless networks, render them very useful complements but poor substitutes for wireline broadband. These technical challenges include:

• Spectral limitations: 5G networks will require massive amounts of spectrum to accomplish their target speeds. At the lower frequencies traditionally used for wide area coverage, there is not enough spectrum. At the very high frequencies proposed for 5G where there may be enough spectrum, the RF signal does not propagate far enough to be practical for any wide area coverage. This is particularly important in rural areas where customer concentration is far, far less than what can be expected in densely populated urban areas where 5G may offer greater promise.
• Access Network Sharing: This is not a good solution for continuous-bit-rate traffic such as video, which will make up 82% of Internet traffic by 2020.
• Economics: When compared to a 5G network that can deliver significant bandwidth using very high, very short-haul frequencies, FTTP is often less expensive and will have lower operational costs. This is particularly true when one consider how much fiber deployment will be needed very close to each user even just to enable 5G.
• Reliability: Wireless inherently is less reliable than wireline, with significantly increased potential for impairments with the very high frequencies required by 5G.

In 2014, PricewaterhouseCoopers LLP released a report urging telecom executives to shift their thinking about telecom capital spending away from one that focuses on upgrades to deal with increasing traffic and demand and move instead to a hardline view of only spending on projects that meet Return On Investment (ROI) objectives for investors.

“The predominant task of management is to take a considered view of the future, allocate capital towards strategies that maximize value for the providers of that capital, and manage the execution of those strategies through to the delivery of returns for those investors,” wrote PricewaterhouseCoopers LLP. “For too long, telecoms have been on auto-drive for much of their capex. Departments assume if they had the money last year, they are going to get it again this year, under the premise of increasing traffic. But rarely do telecoms truly analyze that spending for its ROI or ask whether the investment should be made at all.”

In short, if a project is not certain to quickly deliver significant ROI, serious questions should be asked about whether that investment is appropriate to undertake. That reluctance is at the heart of Deloitte’s new study.

Deloitte notes if providers cannot overcome Wall Street’s reluctance to support major spending on fiber infrastructure, lack of investment will be even more costly.

It predicts falling short on fiber deployment will cause a dwindling number of broadband provider choices for consumers. Today, fewer than 33% of U.S. homes have access to fiber broadband and only 39% have the option of choosing more than one provider capable of meeting the FCC’s minimal definition of broadband – 25Mbps. As competition declines, the need to further expand is reduced while prices can freely rise.

PricewaterhouseCoopers LLP also recommends cable and phone companies partner with content providers like Netflix or Google, and let those companies take an ownership interest in return for capital investments for fiber upgrades. Those type of solutions also protect Wall Street from a feared price war should alternative providers launch in markets that are barely competitive, if at all.

Four large telecom companies won the bulk of the available licenses to operate their wireless services on the upcoming 600MHz band, once UHF TV channels occupying part of it vacate. But what exactly did AT&T, Comcast, Dish, and T-Mobile buy and where? Mosaik, a mapping firm, produced maps (courtesy Fierce Wireless) showing exactly where the four companies won 600MHz spectrum in the recent auction. The differences are striking. T-Mobile effectively won the right to launch new service almost everywhere in the country, in part because it acquired a huge number of cheap, low-demand licenses in largely rural areas.

Dish’s plans for its spectrum remain a complete mystery, while Comcast’s winning bids are entirely within areas where it provides cable service. AT&T, although already holding a large supply of low band frequencies, apparently needs more capacity in larger cities, and paid handsomely to get it.

AT&T

Most of AT&T’s winning bids cover larger cities where it already operates an extensive cellular network. Among the areas where AT&T can expand service: Philadelphia, Washington, Baltimore, St. Louis, Birmingham, Mobile, Tampa, Atlanta, Dallas, Phoenix, Las Vegas, San Francisco, Salt Lake City, Seattle, Minneapolis and Little Rock. But AT&T also grabbed licenses for rural western Massachusetts, central Ohio, and southern Michigan.

Comcast

Comcast’s winning bids consisted of 10MHz of spectrum, except in Nashville where it nabbed 20MHz. Comcast grabbed enough spectrum to cover every city in Florida except Tampa (where Charter provides cable service). The cable company focused heavily on east and west coast bids, winning spectrum across much of the Pacific Northwest, the Boston-NYC-DC corridor, and Illinois and Indiana. The only downside is that 10MHz is not a lot of spectrum to support a large wireless service, but then Comcast does not require that at this time, because it will rely primarily on a shared arrangement with Verizon Wireless to power Xfinity Mobile.

Dish Network

What Dish intends to do with its spectrum remains a complete mystery, but it grabbed a significant amount of it in New York City and its nearby suburbs, including Connecticut. It also won respectable quantities of frequencies in Alaska, California, Florida, Puerto Rico, Seattle and Portland, and several midwestern and south-central cities.

T-Mobile USA

T-Mobile published a similar map as part of its press package claiming victory in the spectrum auction. This map better highlights T-Mobile’s extensive spectrum wins in all 50 states and Puerto Rico. If T-Mobile uses it all, it will command similar coverage areas comparable to Verizon and AT&T. T-Mobile will manage this without any need to merge with anyone else, as AT&T and Sprint have historically argued in their past failed efforts to acquire T-Mobile.

T-Mobile’s 600MHz coverage map — assuming it builds out its full spectrum purchase.

One of the most consequential and visible spectrum auctions ever is over, and it will have a significant impact on broadcasters, wireless carriers, and the future competitive landscape of the wireless industry.

The world’s first “incentive auction” paid television stations to voluntarily vacate or move their assigned channels to make room for the wireless industry’s desire for more spectrum to power wireless data services. Up for bid was 70MHz of spectrum currently used by UHF television stations. A total of 50 winning wireless bidders collectively agreed to pay $19.8 billion to acquire that space. The biggest winner was T-Mobile USA, which is paying almost half the amount of total proceeds to acquire 45% of the spectrum available in the current auction. T-Mobile managed to acquire enough spectrum to cover 100% of the United States and Puerto Rico with an average of 31MHz of available spectrum nationwide, quadrupling its current inventory of important “low-band” spectrum, which is excellent for covering rural areas and inside buildings.

Consumers are likely to benefit as early as later this year when T-Mobile begins lighting up cellular service utilizing the newly available spectrum. Unfortunately, customers will have to buy new devices compatible with the new bands of frequencies.

Having the spectrum alone is not enough to beef up T-Mobile’s network. The company will have to invest in a large number of new cell sites, particularly in outlying areas, to eventually rival the coverage of AT&T and Verizon Wireless. But with an ample supply of 600MHz spectrum, T-Mobile could soon challenge AT&T and Verizon Wireless’ perceived network and coverage superiority. After this auction, AT&T continues to hold the largest portfolio of <1GHz spectrum — 70.5MHz. Verizon is second with 46.2MHz and T-Mobile has moved up in its third place position with 41.1MHz.

Although the FCC claims the current auction was among the highest grossing ever conducted by the FCC, industry observers claim companies got the new frequencies at a bargain price. A 2015 spectrum auction attracted $44.9 billion in bids, more than double the amount bid this year. The average price wireless companies paid per megahertz per person this year was just shy of 90¢, compared with $2.72 in 2015.

Where bargains are to be had, Charles Ergen and his Dish Network satellite company are sure to follow.

Few companies have as much unused wireless spectrum in their portfolio as Dish. Ergen loves to bid in auctions and has also picked up excess spectrum available on the cheap from other satellite companies that have since gone dark or bankrupt. Dish spent $6.2 billion on spectrum during the latest auction, puzzling investors who drove Dish’s share price down wondering what the company intends to do with the frequencies.

Investors were hoping Dish would eventually sell its spectrum portfolio at a profit, something that could still happen if other wireless carriers see a deal to be made. But some Wall Street analysts fear Dish might actually build a large wireless network of its own to offer wireless broadband service. Wall Street dislikes big spending projects and the competition it could bring to the marketplace, potentially driving down prices.

The other possibility is that Dish is making itself look more attractive to a possible buyer like Verizon, which could acquire the satellite company to win cheaper cable programming prices for its FiOS TV and an attractive amount of wireless spectrum for Verizon Wireless. The nation’s biggest wireless carrier notably did not participate in this spectrum auction.

Another unusual bidder was Comcast. Craig Moffett from Wall Street firm MoffettNathanson called Comcast’s $1.7 billion bid “half-hearted” and said it was unlikely to be enough spectrum for the company to begin offering its own wireless service. Comcast plans to rely on Verizon Wireless to power its wireless service, at least initially.

Comcast targeted its bids only in cities where it already provides cable service, which also nixes the theory Comcast and Charter might have been working together to form a cellular joint venture. Moffett expected Comcast would seek at least 20MHz of spectrum across most of the country. It ended up with 10MHz and only in select cities. Moffett thinks that may signal Comcast’s interest in buying an existing wireless carrier is still on the table.

With the advent of AT&T/DirecTV Now, AT&T’s new over-the-top streaming TV service launching later this year, AT&T is preparing to bury the U-verse brand.

Earlier this year, AT&T customers noticed a profound shift in the company’s marketing priorities. The phone company began steering potential customers to AT&T’s latest acquisition, satellite television provider DirecTV, instead of U-verse. There is an obvious reason for this – DirecTV has 20.45 million customers as of the second quarter of 2016 compared to 4.87 million customers for AT&T U-verse TV. Volume discounts make all the difference for pay television companies and AT&T hopes to capitalize on DirecTV’s lower programming costs.

AT&T’s buyout of DirecTV confused many Wall Street analysts, some who believe the days of satellite television are past their peak. Satellite providers lack the ability to bundle services, although some phone companies partner with the satellite company to pitch phone, broadband, and satellite TV to their customers. But consider for a moment what would happen if DirecTV introduced satellite television without the need for a satellite dish.

Phillip Dampier: The “U” in U-verse doesn’t stand for “unlimited.”

AT&T’s DirecTV Now will rely on the internet to deliver television channels instead of a satellite. AT&T is currently negotiating with most of the programmer conglomerates that own popular cable channels to allow them to be carried “over-the-top” through broadband connections. If successful, DirecTV Now could become a nationwide powerhouse alternative to traditional cable TV.

AT&T is clearly considering a potential future where DirecTV could dispense with satellites and rely on broadband instead. The company quietly began zero rating DirecTV streaming in September for AT&T Mobility customers, which means watching that programming will not count against your data plan. For current U-verse customers, broadband speeds have always been constrained by the need to reserve large amounts of bandwidth to manage television viewing. Although AT&T has been boosting speeds in selected areas, a more fundamental speed boost could be achieved if AT&T dropped U-verse television and turned the service into a simple broadband pipe that relied on DirecTV Now to manage television service for customers.

AT&T seems well on the way, adding this notice to customer bills:

“To make it simpler for our customers U-verse High Speed Internet and U-verse Voice services have new names: AT&T Internet and AT&T Phone. AT&T Internet product names will now align with our Internet speed tiers. Our voice plan names will remain the same.”

An earlier internal company memo suggested AT&T would eventually transition all of its TV products into “AT&T Entertainment” after completing a transition to its “next generation TV platform.” Increasingly, that platform seems to be an internet-powered streaming solution and not U-verse or DirecTV satellite. That transition should begin in January.

Gone by end of 2016.

It would represent a formidable change, but one that makes sense for AT&T’s investors. The transition to IP networks means providers will offer one giant broadband pipe, across which television, phone and internet access will travel. The bigger that pipe becomes, the more services customers are likely to use — and that means growing data usage. Having a lot of fiber infrastructure also lays the foundation for expansion of AT&T’s wireless network — particularly towards 5G service, which is expected to rely on small cell technology to offer faster speeds to a more localized area — fast enough to serve as a home broadband replacement. Powering that network will require plenty of fiber optics to provide backhaul access to those small cells.

Last week, AT&T announced it launched a trial 100Mbps service using point-to-point millimeter-wave spectrum to offer broadband to subscribers in multiple apartment complexes around the Minneapolis area. If the initial trial is successful, AT&T will boost speeds to include 500Mbps service to those same complexes. AT&T has chosen to provide the service outside of its usual service area — Minneapolis is served by CenturyLink. AT&T acquired a nationwide license to offer service in the 70-80GHz band back in 2009, and an AT&T spokesperson claimed the wireless signal can reach up to two miles. The company is also experimenting with new broadband over power lines technology that could offer service in rural areas.

Just like its wireless service, AT&T stands to make money not just selling access to broadband and entertainment, but also by metering customer usage to monetize all aspects of how customers communicate. Getting customers used to the idea of having their consumption measured and billed could gradually eliminate the expectation of flat rate service, at which point customers can be manipulated to spend even more to access the same services that cost providers an all-time low to deliver. Even zero rating helps drive a belief the provider is doing the customer a favor waiving data charges for certain content, delivering a value perception made possible by that provider first overcharging for data and then giving the customer “a break.”

As of mid-September, streaming media analyst Dan Rayburn noted Akamai — a major internet backbone transit provider — was selling content delivery contracts at $0.002 per gigabyte delivered, the lowest price Rayburn has ever seen. Other bids Rayburn has reviewed recently topped out at 0.5 cents per gigabyte. According to industry expert Dave Burstein, that suggests large ISPs like AT&T are paying something less than a penny per gigabyte for internet traffic.

“If you use 139GB a month, that costs your provider something like $1/month,” Burstein wrote, noting doubling backbone transit costs gives a rough estimate of the cost to the carrier, which also has to carry the bits to your local exchange. In this context, telecom services like broadband and phone service should be decreasing in cost, not increasing. But the opposite is true. Large providers with usage caps expect to be compensated many times greater than that, charging $10 for 50GB in overlimit fees while their true cost is well under 50 cents. Customers buying a cell phone are often fitted with a data plan that represents an unprecedented markup. The extent of price increases customers can expect can be previewed by looking at the cost of phone service over the last 20 years. The average, often flat rate telephone bill in 1995 was $19.98 a month. In 2014, it was $73 a month. In 2015, it was $90 a month. Those dramatically rising prices in the last few years are mostly as a result of the increased cost of data plans providers charge to clean up on customers’ growing data usage.

Both Comcast and AT&T are dedicated to a campaign of getting customers to forget about flat rate, unlimited service at a reasonable cost. Even as both companies raise usage caps, they continue to raise prices as well, even as their costs to provide the service continue to drop. Both companies hope to eventually create the kind of profitable windfall with wired services that wireless providers like AT&T and Verizon Wireless have enjoyed for years since they abandoned unlimited flat rate plans. Without significant new competition, the effective duopoly most Americans have for telecommunications services offers the opportunity to create a new, more costly (and false) paradigm for telecom services, based on three completely false claims:

• data costs are expensive,
• usage must be monetized, and
• without a bigger return on investment, investors will not finance the next generation of telecom upgrades.

But as the evidence clearly shows, profits from selling high-speed internet access are only growing, even as costs are falling. Much of the drag on profits come from increasing costs related to licensing television content. Voice over IP telephone service is almost an afterthought for most cable and phone companies, often thrown in for $10-20 a month.

AT&T’s transition puts all the attention and its quest for fatter profits on its broadband service. That’s a bad deal for AT&T customers no matter what the company calls its “next generation” network.

(Image courtesy: FCC.com)

Comcast is getting into the wireless mobile business.

Comcast CEO Brian Roberts made the surprise announcement at this morning’s Goldman Sachs Communacopia investor conference, telling attendees Comcast will offer service beginning in mid-2017.

Roberts added the service will depend heavily on Comcast’s installed base of 15 million Wi-Fi hotspots, mostly from cable modem/gateways already installed in customer homes. When away from a hotspot, Comcast’s cellular service will depend on Verizon Wireless.

The deal with Verizon Wireless was expected, because Comcast has maintained an agreement with Verizon since 2011 that allows both companies to sell each other’s services to consumers. The agreement allows Comcast to obtain service from Verizon Wireless at fixed wholesale prices.

That means Comcast can introduce its wireless service without having to build wireless infrastructure like cell towers.

“We believe there will be a big payback with reduced churn, more [customer] stickiness and better satisfaction,” Roberts said.

Comcast will continue the cable industry’s tradition of not directly competing with other cable operators and will not accept customers outside of an existing Comcast service area. Comcast will likely offer the service in a bundle with other services. This will result in a quad-play package for Comcast, bundling cable TV, internet, phone, and cellular service.

Roberts did not talk about pricing.