War on TV Part IV: The Biggest Pipe Will NOT Win
SEE LAST PAGE OF THIS REPORT Paul Sagawa / Artur Pylak
FOR IMPORTANT DISCLOSURES 203.901.1633 /.1634
sagawa@ / email@example.com
July 17, 2013
War on TV Part IV: The Biggest Pipe Will NOT Win
- The first 3 parts of our “War on TV” series looked at how consumer electronics innovation, improving on-line content, and ad budget shifts are catalyzing a migration of viewers and revenues from TV. As the self-reinforcing cycle of viewers to revenues to content accelerates, MSOs will be squeezed between cash strapped consumers unwilling to pay more, and networks demanding higher fees as ad sales wane. For cable operators, the presumed “ace in the hole” has been their domination of residential broadband. However, we believe plans to stymie streaming rivals via usage caps, throttling, and sharply higher broadband prices is a dangerous tack that will bring regulation and competition. Moreover, faith that the sheer capacity advantage of cable’s fiber/coax infrastructure can sustain broadband leadership is misplaced, as applications and screen resolutions requiring speeds greater than 100Mbps are unlikely to be a significant factor over the next decade – wireless residential broadband will be a very realistic alternative for most households. Rumored industry consolidation may improve bargaining power with networks, but will increase regulatory scrutiny and may hasten media’s embrace of “over the top”.
- The growing audience and revenue shift to on-line video is real. Channelized TV viewership has begun to recede with the growth of streaming, a trend that will accelerate with the spread of internet friendly consumer electronics and the improving selection of programming on-line. The growing on-line audience is drawing subscription and advertising revenue that would otherwise have been captured by TV, making the web a more attractive option for content owners and creators. The result is a self-reinforcing cycle of viewers to revenue to content to viewers that threatens the advertising sales, ARPU and subscriber rolls of traditional channelized television.
- Consumers are tapped out. After rising 6.1% per year for the past 16 years, the average monthly US video bill now exceeds $90 – more than 2% of after tax income for the typical household. Given static incomes and the pressures of rising health care (6.9% of income rising 5.1%/yr) and education ( 2.4%, rising 6.1%), consumers may no longer be willing to absorb higher cable fees. The percentage of TV owning households that do not subscribe to MSO service has risen from 14% in 2010 to 19% in 2013, with 60% of these households pointing to cost as their primary concern. The availability of viable alternatives, such as over-the-air broadcasts and on-line streaming, increases the appeal of cord-cutting or reducing services to save money. Moreover, younger consumers, highly valued by advertisers, are the most likely to shun multichannel service – including “cord-nevers” who have never subscribed – leaving MSOs more as the young cohort ages and its behaviors spread to older, more conservative consumers.
- Networks want more. With TV ad sales at risk, networks are looking to growth in fees paid by MSOs to cover the 4% annual growth of programming expenses, which make up more than 80% of total network costs. Historically, MSOs have passed the 8% annual growth in network fees on in higher prices to consumers, a tactic that would almost certainly accelerate cord-cutting if continued. The alternatives – absorbing fee increases or risking service disruptions with hard-ball negotiations – are marginally more palatable, with significant costs and risks of their own.
- Price gouging on broadband will raise regulatory scrutiny. Cable operators enjoy a monopoly for broadband service to more than 65% of US households, prompting industry executives to tout price hikes on already high margin cable modem service as their “ace in the hole”. While the cable industry has historically enjoyed regulatory carte blanche, anti-competitive behavior, such as price hikes, metered-pricing, usage tiers, or service discrimination vs. video competitors will bring new government attention. Cable and media industry lobbyists are now outgunned by their internet industry counterparts, spending big bucks to push “net neutrality” and “common carriage” regulation that would gut MSOs ability to exploit their broadband advantage. Cable operators, with their deserved reputation for poor service and high prices, are widely disliked by consumers, a further threat to the political air cover.
- Broadband speed requirements likely to plateau. Cable bulls often point to the inherent speed advantages of the industry’s fiber/coax infrastructure – CMCS demonstrated 3 Gbps download speeds at the recent Cable Show. Yet a stream of HDTV, requires just 8Mbps today and soon, 2-4Mbps with new compression technology. Even 4K TV, which faces substantial hurdles to adoption, is indistinguishable from HDTV at screen sizes below 42”, and is likely to be less than 25% of the big screen installed base a decade from now, will require just 20-30Mbps to stream. The future broadband needs for the vast majority of US households would be satisfied by speeds of less than 100Mbs.
- Wireless broadband, DSL and selective fiber builds will increase rivalry. Technology progress and high cable modem service prices will open the door to real competition in residential broadband. The LTE Advanced standard, running on cheap mini cells and ample spectrum, will allow aggressive wireless carriers to easily hit the 100Mbps mark for residential service with very competitive cost bases well before the end of the decade. An artificially high price point would bring telco-provided DSL back into play, and improve the investment payback for new build fiber networks.
- Internet players may fund broadband rivals. Would-be residential broadband competitors will also find deep-pocketed allies from the Internet industry. Google flirted with the last major spectrum auction, has begun building fiber networks in 3 cities, and has been rumored to have been in talks with Sprint regarding a role in pushing wireless broadband. Amazon, which already subsidizes consumer devices to drive consumption of its electronic media, is another obvious potential partner with strong interest in cheap, plentiful broadband. Microsoft, Apple and even, Intel, all have designs on leadership in the media world of the future and the resources to help fund the distribution needed to get there. With major spectrum auctions a year away, we expect significant developments on this front.
- MSOs will lose, on-line video, wireless operators, towers and technology will win. Cable operators are facing death by a thousand cuts – the threats of cord-cutting, declining ARPU, rising network fees, regulatory pressures and broadband competition will rise gradually, with the painful impact on earnings and cash flow still a few quarters away. Satellite TV, with their more rural subscriber bases and potential wireless play, are more insulated but not immune. On-line video aggregators, like NFLX, GOOG, AMZN, MSFT, YHOO and others, will greatly benefit. Aggressive wireless operators, like Sprint or T-Mobile, have significant opportunity in broadband, to their benefit and to the benefit of tower companies and wireless technology players, such as QCOM.
A Cable Broadband Windfall? Not So Fast!
Listen to a cable executive on the future of television. Once you get past the assertions that “cable and media are completely aligned”, that “bundling provides consumers with far more value for less money than they could ever get a la carte”, and that “customer satisfaction metrics are trending up”, you get to – “Anyway, it doesn’t matter if over-the-top video wins, since we own the broadband pipe”.
In the first three “War on TV” pieces, we described a self-reinforcing cycle – the growing on-line audience yields new revenues, which brings better programming, which draws more new viewers – that is building speed and making the slow deterioration of the traditional TV model inevitable. MSOs are pinched between networks looking to replace threatened ad sales with higher fees, and consumers with TV bills that have risen more than 6% a year for the past 16 years to reach more than 2% of the typical household’s after tax monthly budget. Video service margins, already the thinnest amongst the “triple play” services pitched by cable, are set to get even thinner.
Industry CEOs rationalize that the growing ranks of cable-cutters and cable-nevers, up to 19% of US households from just 14% in 2010, are inconsequential, since lost low-margin video revenues can be replaced by raising the price of the much higher margin broadband service. For 70% of Americans, the ONLY choice for broadband fast enough to support HDTV video is cable – if these households want to cut the cable, their MSO intends to charge them a premium for their increasing on-line video usage.
We don’t think they will get away with it. Thus far, cable MSOs have avoided onerous “net neutrality” and “common carrier” regulations, but heavy handed pricing and anti-competitive policies could bring the FCC hammer down, particularly in light of the vigorous lobbying by cash-rich Internet players and the dismal reputation of cable companies amongst consumers/constituents. Moreover, competition is coming. Household Internet speed needs are likely to plateau at less than 100Mbps, given the slow adoption of bandwidth hungry formats and the near-term availability of superior compression techniques. Wireless broadband will soon be a realistic option – an upgrade to 4G, new spectrum, cheap equipment, increased rivalry, and an influx of Internet money are all a couple of years away. High cable broadband prices would also leave a tempting price ceiling for cherry picking by new build fiber operators or even for telcos to take their DSL programs out of mothballs.
All of this is unequivocally bad for cable, but will take a couple of years to play through to cash flows. Satellite TV will have a little more time by virtue of a more rural subscriber base and a wireless broadband play. On-line video purveyors, such as NFLX, GOOG and AMZN will win, as will the newly energized 2nd tier wireless operators, i.e. Sprint, the tower companies and the vendors of wireless broadband technology, such as Qualcomm and Ericsson.
The Cycle Speeds Up
The first three installments of “The War on TV” described a self-reinforcing cycle that is slowly eroding the traditional channelized TV model. Every day, more viewers are finding and watching more video programming streamed across the Internet. This phenomenon stands to accelerate in the near future, as a new generation of consumer electronics will allow viewers to search for on-line and channelized content simultaneously, using the same tools and on the same basis, removing what has been an enormous relative advantage for the traditional TV model ( http://www.sector-sovereign.com/2013/06/june-12-2013-the-war-on-tv-the-attack-of-the-boxes/ ). The growing audience for streaming video – viewed on TVs, portable devices, and PCs – is attracting growing revenues from paid subscriptions – e.g. Netflix – and from advertising ( http://www.sector-sovereign.com/2013/06/june-19-2013-the-war-on-tv-part-ii-streaming-is-coming/ ) – e.g. Google. This growing revenue stream is, in turn, emboldening these internet video aggregators to fund new, original programming and to contract for attractive content libraries, while drawing creative talent, popular personalities and content owners to try their hands at monetizing their work on-line ( http://www.sector-sovereign.com/2013/07/july-2-2013-the-war-on-tv-part-iii-reductio-ad-absurdum/ ). The quality of this new on-line programming, like Netflix’s original series “Orange is the New Black”, popular YouTube channels like “Funny or Die” and Machinima, or exclusive re-runs of Nickelodeon children’s programming on Amazon, will then attract even more users and induce them to spend even more time watching streaming content. Viewers to revenues to programming to viewers – the pace builds with each iteration.
Exh 1: TV Households and Average Number of Televisions
At the same time, the audience for traditional television has been shrinking. The number of US households with televisions set up to received channelized programming peaked in 2010 and has turned down, despite continued formation of new households (Exhibit 1). The number of subscriptions to multichannel services declined even more sharply over the past three years, with the percentage of consumers eschewing pay TV service rising from 14% to more than 19% (Exhibit 2). The hours spent watching channelized television has remained flat since ‘10, at least according to Nielsen, but these numbers are suspect given an antiquated measurement methodology that does not distinguish between attentive and ambient viewing, and perverse incentives to participating households that may be skewing audience estimates higher than reality (Exhibit 3). The total audience for primetime shows on the big four broadcast networks are the lowest that they have been in many decades.
Exh 2: Broadcast-only US Households
Exh 3: Monthly TV Viewing Hours, All Demos 2+ 3Q2007-1Q2013
As we wrote in the previous installment of this series, television advertising revenue looks ready to turn over, with commitments for advertising in the fall season likely off $800M or more during the recent “Upfronts”. If fall ratings take a further step down, something we see as very likely given recent trends, the networks will be forced give away additional ad slots as “make goods” to their upfront advertisers, reducing inventory in what could be a painfully weak spot market. Programming costs, which make up more than 80% of the costs of a TV network, have been rising on a 6.4% CAGR (Exhibit 4). For broadcast nets, which rely on advertising for 90% of their revenues, and even for cable nets, which gain 42% of revenues from ads on average, shortfalls will be painful, particularly if they represent the beginnings of a longer term trajectory. This raises the profile of the other major category of revenues for television networks – fees from multichannel system operators (MSOs).
Exh 4: U.S. Network Programming Cost Forecast, 2009-16
Blood From a Stone
Most of the television networks carried by a typical MSO are owned by a relatively small group of media conglomerates – i.e. Disney (ABC, ESPN, Disney Channel, etc.), Fox (Fox, Fox News, FX, etc.), CBS (CBS, Showtime, The Movie Channel, etc.), Comcast (NBC, Bravo, USA, etc.), Time Warner (CNN, TBS, HBO, etc.) and Viacom (MTV, Nickelodeon, Comedy Central, etc.). Each of these companies negotiates fee agreements with each of the MSOs, receiving payments on a per subscriber basis for the networks that they provide to be broadcast on the MSOs’ systems (Exhibit 5). These agreements are negotiated on a bundled basis – an MSO must buy ALL of a media company’s networks as one. The deals have a fixed term – often as long as 10 years – and the periodic renegotiations are often contentious. If MSOs do not accede to the demands of a network owner, the channels in question may be pulled from the system, sometimes timed to block subscribers from watching popular events, such as the World Series or the Oscars.
Over the last decade, media companies have used this leverage to drive average annual fee increases of more than 8%. In turn, the MSOs have pushed these costs down onto their subscribers, raising the monthly video bill by more than 5% per year since 2000 (Exhibit 6). In the face of these ongoing price hikes, consumer demand for multichannel video service has proved inelastic, with the total number of subscribers continuing to grow every year – until 2010 (Exhibit 7). As we noted in the previous section, the percentage of US households subscribing to multichannel service dropped from 86% to 81% over the past 3 years, and number of subscribers dropped in absolute for the first time in industry history.
Exh 5: Select 2013 Basic and Basic Digital Cable Network Subscriber Fees
Exh 6: Residential Cable Subscribers versus ARPU, 2000-2012
Exh 7: US Multichannel Video Customers
This is a function of both on-line alternatives and basic economics. After rising above inflation for decades, the average monthly multichannel TV bill sits at more than $90. This is more than 2% of after tax income for the typical American family, and does not include the price paid for the broadband or telephone services added as part of the cable “triple play”. Meanwhile, median US household income, already off 8% from a relative peak in 2009, has been stagnant for two years. Out of pocket healthcare expenses, which have been rising at a 5.1% annual pace and may rise at an even faster pace for younger Americans going forward, already constitute 6.9% of the average household budget. Education spending eats up an additional 2.4% for the typical US family, and is growing at a blistering 6.0% per annum. Against this backdrop, we would expect spending on cable service to receive a lot more scrutiny during the household budgeting process (Exhibit 8).
This has been the case over the past three years. Of the 22 million US homes that have televisions but do not subscribe to a multichannel service, two-thirds are “cable nevers”, that is, they have never been subscribers in the past. As financially squeezed and digitally savvy young people form households, they are NOT signing up for cable, and may not in the future, even as their financial circumstances improve. The remaining 7.5 million households are “cable cutters”, who have had subscriptions in the past but have dropped them. Of these, more than 55% point to the cost of services as the primary reason for cancelling their subscriptions (Exhibit 9). If average cable bills continue to rise relative to household income, one can only imagine that the incidence of economic cable cutting will rise, and while many of these households might be expected to reconnect in better economic times, others may find that the alternative of over-the-air broadcasts and on-line streaming more than meets their needs (Exhibit 10-11).
Exh 8: Forecast Consumer Spending as a Percent of Median Household Income
Exh 9: Reasons US Internet users would consider cord-cutting their paid-TV subscription
Exh 10: Forecast Multichannel Penetration, 2008-2017
Exh 11: TV Distribution Sources, Number of Households
Rock, Paper, Scissors
With a suddenly price elastic subscriber base, it is no longer a no brainer for MSOs to give big annual fee increases to their media suppliers and pass them along to their subscribers. Doing so would seem ever more likely to catalyze cord cutting, a phenomenon that the industry has been able to explain away to investors as a modest and temporary reaction to tough times, but that would be perceived as a major problem if it accelerated in the face of an economic recovery. Taking on the media conglomerates to hold the line on fee increases during negotiations has its own risks. Tough negotiations could result in seeing major networks pulled from service for protracted periods, damaging the MSOs already poor reputations for customer service, reducing the value of a subscription and giving impetus for potential cord-cutters to more closely consider the alternatives.
Media companies, and their investors, are hoping that MSOs hew to a third row – give the network owners the fee increases that they want and simply eat them in the form of lower margins. While video margins are by far the thinnest of the cable “triple play” of video, internet and voice, and while all three services must support the capital and operating expense needs of the extensive fiber/coax network, operating margins on cable video are still close to 60%, so there would seem to be plenty of room to absorb further fee increases. Still, video revenues still make up 55% of sales for the average cable MSO, and 97% of sales for satellite operators, so a margin squeeze on TV would be very real, and quite possibly, unexpected by investors, if consensus forecasts are to be believed (Exhibit 10).
Exh 10: Cable MSO Sources of Revenue, 2008-12
The Ace in the Hole
This brings us back around to physics and sunk investment. The cable industry’s optical fiber and coaxial cable based infrastructure, in place and available to more than 93% of US households, is an unusually robust vehicle for carrying digital communications. Unlike the telephone twisted copper pair network architecture, which is even more ubiquitous than the cable plant, coaxial cable is shielded from interference and of significantly higher gauge. As a result, it can carry more than an order of magnitude more data over a much longer distance, giving cable modem service an enormous advantage over the telco provided DSL. All fiber networks, like Verizon’s FiOS or Google’s three city Fiber initiative, have capacity advantage over coaxial cable, but are enormously expensive to build from scratch, and as such, address a much, much smaller base of homes. As a result, for more than 60% of American households, fiber/coax based cable modem service is the ONLY available option for residential broadband fast enough to support streaming video (Exhibit 12).
Exh 12: Residential Fixed Connections by Technology
With a de facto broadband monopoly in most markets, and an increasingly cozy relationship with the telephone providers that are the only competitors in most of the others, cable executives and their supporters in the investment community, are a bit smug. If margins are squeezed in TV, or if over-the-top video inspires consumers to drop video service, cable MSOs will make it up with higher prices on broadband, which is already their highest margin service. Cable operators are already beginning to operate on this premise. Time Warner Cable and Cablevision both recently raised prices on broadband service.
Cable operators, like Comcast and Charter, have also taken to introduce usage caps to discourage their broadband customers from watching too much on-line video from competitors, often exempting their own streaming video services from counting against the cap (Exhibit 13). Other MSOs, such as Verizon and Time Warner, have been accused of throttling on-line video by limiting the size of the connection between their networks and popular streaming services like Netflix and YouTube, thus hampering the ability of subscribers to use these services during peak hours.
Exh 13: Broadband Usage Caps by Provider and Service Level
Don’t Worry About the Government
The cable industry has enjoyed benign regulation for most of its history. Originally, cable TV was granted as local franchises by municipalities looking to serve their citizens with better reception of local broadcast channels and add to the tax base. Eventually, the industry began to consolidate, and the Cable Act of 1984 allowed operators to offer unlimited options beyond the local broadcast channels. While the industry became well established as patchwork of geographic monopolies, the aggressive capital spending of the MSOs and the ongoing concern with the market power of the regional Bell operating companies put cable within the good graces of the FCC, which under pressure from Congress declined to classify cable as a communications service requiring close regulatory oversight. In contrast, telephone companies were considered “common carriers” and all pricing and service changes were subject to review and approval by the Federal regulator. Even with the introduction of telephone service and broadband, cable operators, supported by a well-funded and highly effective lobbying effort, were able to forestall movement to place them under the same scrutiny as their telco competitors (Exhibit 14). Three years ago, the FCC moved to consider reclassifying broadband as a “telecommunications service”, a designation that would have made cable subject to the tighter regulations, but the resolution was set aside for the time being without sufficient immediate support.
At the same time, Congress and the FCC also considered actions to insure that broadband providers, including cable operators would not be able to block access to 3rd party web sites for competitive reasons. This battle for “net neutrality” pit cable operators and telcos against the top Internet companies, and yielded a watered down set of rules banning the outright blocking of access to internet-based services on wired networks, with an even weaker set of restrictions for wireless operators. Feeling their oats, Comcast and Verizon have both sued the FCC over the commission’s administration of these rules.
Exh 14: 2012 Lobbying Spend by Cable and Telecom Interests
Exh 15: 2012 Top Lobbying Spenders – All Industries
Or Maybe You Should Worry About the Government
The cable industry is used to getting its own way. The National Cable and Telecommunications Association, the primary trade group for cable MSOs, has been one of the top 20 spenders on Federal lobbying in each of the past 8 years, with industry stalwart Comcast joining the list with its own spending since 2010 (Exhibit 15). NCTA and Comcast also have shown their largess in campaign contributions, accounting for about 10% of the nearly $70M contributed by the entertainment industry during the 2012 election cycle. For this money, the cable industry has the ear of legislators and air cover with FCC regulators who have eyed broadband competition with increasing suspicion.
However, the internet industry is stepping up its game. Google alone has raised its annual lobbying budget from $5M to $16M since 2010, neck and neck with NCTA amongst the top 10 spenders. The internet industry also is able to leverage the changing nature of media coverage to its advantage. The recent failure of the Motion Picture Association of America to gain Congressional support for the proposed strict anti-piracy SOPA legislation in the face of strong opposition from the internet industry and its supporters suggested that business as usual in Washington had changed, with the presence of constituents making their voices heard through social media upsetting a deal that had already been hammered out in those smoky back rooms.
Exh 16: American Customer Satisfaction Index – Subscription Television
We suspect that political support for the cable industry would disappear if the industry were to impose dramatic price hikes on broadband service or obviously stymie subscriber access to on-line video. The internet industry has deep pockets to protect its interests, and the soapbox from which to marshal public support. Furthermore, public support for cable operators is practically non-existent. Time Warner Cable, Comcast and Charter Communications all ranked in the bottom seven of 141 companies recently ranked by the American Consumer Satisfaction Index, with broadband internet service providers the single lowest rated business category (Exhibit 16). The Long Island Power Authority’s dismal post-Sandy performance was the only thing keeping perennial loser Time Warner Cable from the lowest spot.
Thus far, Republicans in Congress have blocked legislation to formalize net neutrality into law, generally spurred by a distrust of regulation in general and by the ample lobbying and campaign funds supplied by the cable industry. However, the primary argument, that regulation is unnecessary to protect the open internet, would be severely challenged by substantial broadband price hikes or network policies that hamper access to competitive video services. Given the rising lobbying presence of the Internet industry, the growing influence of social media and the public hatred of cable operators, we believe political opposition to cable regulation would disappear, giving the NCTA a nasty surprise akin to that visited on the MPAA during SOPA. Moreover, no amount of Congressional support would save cable operators from the DoJ, should aggressive anti-competitive actions yield anti-trust investigations.
Exh 17: Wireless and Wireline Advances, 2000-2015
Look Ma, No Wires!
Meanwhile, cable broadband may see competition. Today, wireless broadband is too expensive to consider as an alternative to the residential internet connection, but the same was true about cellular voice service vs. landline telephones a decade ago. The first generation of 4G LTE networks supports about 120Mbps of total user throughput in each cell, with peak speeds of more than 20Mbps available if the cell is not congested (Exhibit 17). Of course, in a busy network with many subscribers 120Mbps can be filled up quickly, making the existing service impractical for residential broadband. Moreover, the capital and operating costs of upgrading a 65,000 cell network to LTE technology make pricing mobile broadband competitively with fixed residential wired service impossibility. Finally, the US wireless industry has operated as a de facto duopoly over the past three decades, and neither Verizon or AT&T seems likely to upset the cozy détente that they have established with their cable bretheren.
But times are changing. The newest version of LTE, LTE Advanced, enables the use of wide new spectrum bands and more than doubles the spectrum efficiency of a 4G cell (Exhibit 18). With LTE Advanced and 100MHz of spectrum, a single cell could support more than 2 Gbps of aggregate capacity and offer individual peak streams of up to 100Mbps. The longer term technology roadmap for LTE goes even further, with the International Telecommunications Union setting a goal for average mobile service at 100Mbps and fixed residential wireless broadband of up to 1Gbps.
Exh 18: Wireless Data Standards Releases by Technology
At the same time, spectrum is coming available in the US. LTE Advanced will allow Sprint to use LTE in the 150MHz of spectrum that is acquiring with Clearwire, as well as the 800MHz bands that is repurposing from Nextel’s defunct network. The combination of T-Mobile and MetroPCS allows 3G service to be squeezed to a common band, opening 40MHz for LTE Advanced deployment. Verizon picked up its own new 40MHz slice in a deal with cable operators to add to the 22MHz of 700MHz band spectrum that it bought at auction in 2006. AT&T has its own 700MHz frequencies, augmented by additional blocks that it picked up from Verizon and efficiencies to be gained from the recent deal for Leap wireless. Dish Networks has 40MHz of spectrum that it hopes to combine with a controversial 40MHz block controlled by Light Squared (Exhibit 19-20). Finally, the FCC is moving toward a 2014 incentive auction of TV broadcast spectrum in the 700MHz band that could bring an additional 120MHz into use for wireless broadband. All of this will help make the speeds promised by the evolving LTE standard a reality.
Innovation in network technology will make it cheaper to deploy. Small cells, each costing less than $20K can be mounted on rooftops and light poles to project broadband capacity into congested areas. Beam forming antennae can do the same, bringing signal to areas that had been difficult to reach. Satellite TV operator Dish proposes adding receivers to its rooftop equipment to deliver wireless internet into its customers’ homes – making higher frequency networks with relatively poor indoor penetration far more viable.
Exh 19: Carrier Spectrum Holdings Summary
Exh 20: T-Mobile / MetroPCS Spectrum Holdings
Finally, the basis of competition in wireless is changing. Even five years ago, the primary measure of network quality was coverage. Verizon advertised “Can you hear me now?” while AT&T countered with red coverage maps. The ability to drive home without dropping a call, or to get signal at a vacation spot was the hallmark of good service. However, wireless voice calling is in decline as wireless data traffic is ascending on a steep, steep slope (Exhibit 21). For many users, wireless service quality is now defined by the availability of reliable, fast bandwidth for data in the places that it is needed most. It is more important to have reliable fast service at work, at home, at lunch and in the airport, than it is to have unbroken 4G LTE coverage between Miami and Portland, Maine. This plays to the advantage of secondary carriers, who, by virtue of their higher frequency spectrum holdings, cannot compete on coverage, but can project plenty of capacity into the specific areas where it is needed most. Given the aggressive plans signaled during the recent spate of wireless carrier consolidation, we believe Sprint and T-Mobile will lead a move toward cheaper wireless data that will eventually challenge for residential broadband service.
Exh 21: Average Mobile Voice Minutes of Use per User per Month
Yes, But Coaxial Cable is Still Much Faster
At the recent NCTA run Cable Show in Washington DC, Comcast CEO Brian Roberts demonstrated broadband service at a blindingly fast 3 Gbps. Never mind that the standard employed has not even been ratified by the industry association and that deployment of service with that capability is likely a decade away, the demonstration was intended to remind everyone that the fiber/coax network deployed by cable MSOs is simply faster than the twisted pair copper used by telephone companies or the LTE Advanced technology being deployed by wireless operators, and that it was capable of keeping up with the all fiber approach behind Google’s well publicized Fiber networks (Exhibit 22). Unfortunately, the future of broadband will not be a speed contest.
Exh 22: DOCSIS 3.1 Development Timeline
Of course, in recent history, it HAS been a speed contest. As recently as 2006, when YouTube was acquired by Google, video streaming was wildly frustrating experience with minutes spent staring at the spinning hour glass while the buffer filled so the playback could begin anew. While connection speeds have been increasing since the dawn of the internet, it was only some time in Q2 2011 that more than half of internet users finally had speeds in excess of the 4Mbps considered by the FCC as the minimum speed for video-rich applications. Even 4Mbps is not enough for HD video content, which streams at 5-11Mbps using the widely adopted H.264 compression standard (Exhibit 23). Given a household with multiple TVs, tablets, smartphones and PCs and 4Mbps is very far from adequate. By the end of 2012, 19.5% of US households had connections faster than 10Mbps, but even this threshold could be considered too slow for a fully connected American family (Exhibit 24).
Exh 23: Throughput Requirements by Application and Compression Technology
However, satisfaction may not be that far off. A new compression standard, H.265, that will cut the bandwidth necessary for streaming by more than half with no noticeable further loss, has been ratified and will be widely deployed within 2-3 years. With this standard, a 1080i signal, with quality equivalent to the HDTV provided over cable, could be streamed in just 2-4Mbps, with higher quality 1080p content, roughly equivalent to BluRay, taking just 4-8Mbps. In this context, even 10Mbps could be adequate, with the 100Mbps promised by wireless LTE Advanced unnecessary overkill.
Exh 24: Average Broadband Speeds and Penetration by Tier
Exh 25: HDTV Broadcast Station Analysis
Of course, the consumer electronics industry is agog over Ultra HD 4K, with roughly 16 times the pixel count of the current 1080P standard. A 4K stream, compressed using the same H.265 standard would take 20-30Mbps, still within the capabilities of that hypothetical LTE Advanced network, but pushing it if the household required more than one stream at once. However, with the first 4K sets on sale for more than $20,000, the reality is that 4K will take a long time to penetrate into more than a smattering of US homes and there is no sign that cable operators will be ready to deliver 4K signals over their networks anytime soon, particularly given that they have yet to support the 1080P standard (Exhibit 25). 4K videos are too big to be stored on BluRay disks – given the modest uptake on that technology, it seems unlikely that a more capacious disk standard will be forthcoming. Furthermore, early testers dispute that 4K picture quality is even distinguishable from HDTV on TV sets smaller than 42 inches, an important factor given that the average TV sold in the US in 2012 was less than 37 inches, and then only if the user is positioned close to the set. While SNL is forecasting 25% penetration of 4K TV sets in 2023, and IHS is predicting that 4K will comprise 50% of US TV sales by 2017, we are highly skeptical, particularly in light of the misplaced enthusiasm for 3D televisions just three years ago (Exhibits 26-29).
Our conclusion: for the better part of the next decade, 100 Mbps will be a good enough speed for broadband access. As providers tier and price services, we don’t see the average American household willing to pay a premium for gigabit internet speeds over something offered at 100Mbps. Even if cable operators were able to drive big broadband price increases without triggering regulation or anti-trust investigations, the static needs of the US consumer will leave MSOs vulnerable to competition – from wireless broadband, from fiber overbuilds and even from twisted pair DSL. Given the fixed cost nature of the network business, overall share loss would be much more painful than simple video cord cutting.
Exh 26: TV Household and HD TV Penetration Forecast, 2010-2020
Exh 27: Connected TV Penetration Forecast
Exh 28: TV Household and 3D TV Penetration Forecast
Exh 29: TV Household and 4K TV Penetration Forecast
Winners and Losers
Cable operators are facing a long slow decline that will eventually be punctuated by the death of multichannel TV as we know it. We believe that the self-reinforcing “over-the-top” cycle of audience migration, revenue growth, and improving content will drive strong growth and improving profitability for on-line video aggregators like Netflix, Amazon, and Google, pressuring media conglomerates to embrace the OTT model fully, or risk seeing their brands deteriorate as internet alternatives begin to crowd them out. This is obviously bad in the long term for cable names like Comcast, Time Warner Cable, Cablevision, Charter, and privately held Cox Communications. Industry consolidation may help leverage some bargaining power, but may be too little too late. Multichannel operators using other technologies, notably telecoms and satellite are also at risk, though the rural nature of the latter’s customer base may insulate it.
The biggest opportunities with respect to the pipe will come to new broadband challengers. We see significant opportunity for the likes of Sprint and T-Mobile to maintain their aggressive maneuvering and invest in providing fixed wireless broadband. Such services would put pressure on both cable and telecom internet providers, most notably the T/VZ duopoly. While Google’s Fiber experiment is small and under the radar, its impact on broadband prices in local markets will be telling. Some cable players have already committed to further build out faster networks in response, though we suspect customers won’t be willing to pay a premium for services beyond 100 Mbps. We also see tower companies – i.e. AMT, CCI, SBAC – as prime beneficiaries, along with wireless technology providers, such as QCOM and ERIC.
Exh 30: Winners and Losers – The Pipe