cloud convergence - zacks investment research

EQUITY RESEARCH

INDUSTRY UPDATE

Oppenheimer & Co Inc. 85 Broad Street, New York, NY 10004 Tel: 800-221-5588 Fax: 212-667-8229

Timothy Horan, [email protected]

Jonathan [email protected]

Oppenheimer & Co. Inc. does and seeks to do business with companies coveredin its research reports. As a result, investors should be aware that the firm mayhave a conflict of interest that could affect the objectivity of this report. Investorsshould consider this report as only a single factor in making their investmentdecision. See "Important Disclosures and Certifications" section at the end ofthis report for important disclosures, including potential conflicts of interest. See"Price Target Calculation" and "Key Risks to Price Target" sections at the end ofthis report, where applicable.

March 17, 2015

COMMUNICATION AND CLOUD

Cloud ConvergenceAnd the Googlization of NetworksSUMMARYWe believe that Internet companies are looking to more vertically integratecommunications and cloud infrastructure and over time offer a fully integratedsuite of applications to control the customer relationship. This is occurring bothin the consumer and enterprise markets. We see Google as wanting to dominateevery aspect of consumers' daily life. Currently the carriers control the customerrelationships and critical last-mile infrastructure. Carriers will need to invest in newvirtualized and a dozen other technologies to retain this leadership. Luckily theseinvestments should leverage existing infrastructure and dramatically lower costs froman automated network and end-user experience. Regardless, companies with uniqueinfrastructure—spectrum, towers, fiber, carrier-neutral datacenters, and managedservice—are well positioned.

KEY POINTS

We believe Google is looking to create a fully integrated, low-cost wireless/cloud service. Google announced two weeks ago that it is entering the wirelessmarket as an MVNO after expanding its fiber rollout. Initially with tablets, but wesee the smartphone as the primary integrator of this converged world, with theability to enable true over-the-top communications and aggregation of the Internet ofThings. This will place pressure on the networks that were not designed to handleupstream traffic, and massive video over-the-top, which we now see as startingto cannibalize paid TV. The cloud is the nexus for this real-time communicationsand computing, something that has never occurred before. There will clearly bestandalone applications like Uber and Netflix, but the ability to integrate theseapplications into an intelligent bundle will be a huge competitive advantage.Internet companies are driving the transformation of communications/cloud ina more vertical way as the suppliers have been lagging, which is usual for anew industry. Google and others are building their own networks (leased fiber/builtdatacenters) with their own virtualized optronic and switching equipment. The largestInternet companies with the most scale have been doing this for some time withinand between datacenters.The incumbent telcos' core competency is control of critical last-mileinfrastructure, but they are using partially outdated technology. We believe to retaintheir dominance they need to deploy the same virtualized open-source equipment,and become more horizontally structured. Positively, we see the telcos' infrastructureas highly undervalued, but this needs to be managed under a better capital structure.We see TMUS, EQIX, CCOI and AKAM as the main beneficiaries of this networktransformation within our coverage universe. We believe that mobility is theprimary driver of this convergence; TMUS has considerable M&A and can leverage itsspectrum, and AKAM/EQIX can efficiently deliver traffic. Last, this is consistent withOpco analyst Jason Helfstein's Perform rating on GOOG, based on a pattern ofcontinued infrastructure investments that will likely weigh on margins.

2

Disruptive Technologies Driving the

Googlization of Networks/Services

We believe that networks are now converging with cloud computing and there will be a

battle to control the customer. Currently this relationship is controlled by the carriers with

the combination of network and billing, but this is under threat. We believe that Internet

companies are looking to converge their cloud applications with networks to dominate

every aspect of a consumer’s daily life—social, entertainment, information, commerce,

etc. with a fully integrated platform. They are seeking to use automated customer care

and integrated real-time cloud data to create a unique, new Uber-like cloud service.

Internet companies (Google, Apple, Facebook, Amazon, NFLX) have virtualized

computing to enable the cloud and want to do the same with network technology. The

second half of this report goes through much of the expected changes in network

technology, essentially virtualization (SDN/NFV) enabling on-demand network and

customer functionality, but also fiber closer to customers with small cell sites. The first

half of the report looks at the new services and expected response from the incumbents.

Positively, the incumbent telcos can use these same technologies to greatly improve

network quality/capacity and automate provisioning and care on a much lower cost basis.

Ultimately, everyone’s goal is to create an Uber-like on-demand service for cloud and

communications, and the network to do so will need to emulate GOOG’s (highlighted in

the Appendix section of this report). Regardless of the outcome, companies in our

coverage universe with unique infrastructure should continue to benefit from strong

volume growth and pricing power. These include AMT, EQIX, CCOI, and AKAM, among

others.

Essentially we see Google looking to create a high-quality, low-cost converged

wireless/cloud service that will lock in and expand its customer relationships. Google

announced its intent to enter the wireless market as a service provider two weeks ago in

Barcelona, although this is still in very early stages. We see the smartphone as the

primary integrator of this new converged world, with applications global and primarily

mobile and truly over-the-top communications and video. As the smartphone becomes

an aggregating device of upstream traffic, it also will be the driver of the Internet of things,

which will increase demand for a more intelligent, predictive converged service for

consumers.

The cloud is the nexus for this real-time communications and computing, something that

has never occurred before. There will clearly be standalone applications like Uber and

Netflix, but the ability to integrate these applications into an intelligent bundle will be a

huge competitive advantage. There are a dozen or more technology/services companies

looking to do the same thing. The holy grail of cloud computing will be the ability to be the

dominant operating system that every application and piece of hardware is standardized

around, but the winner of this operating system, if there is just one, is not yet clear.


3

Exhibit 1: The Battle for the Consumer

Software(GOOG, MSFT)

Consumer

Hardware(AAPL, Samsung)

Social Media(FB, TWTR)

Commerce(AMZN)

On-Demand Apps(Uber, NFLX)

Wireless(T, VZ, TMUS, S)

Cable(CMCSA, CVC)

Source: Oppenheimer

There have been many examples of Internet companies attacking the consumer from

different angles, attempting to bundle services, and drive consumer-facing applications.

AMZN has released hardware (phone, tablet) to drive its e-commerce platform and on-

demand apps (customer receives one year free of AMZN prime and unlimited cloud

storage for photos with a purchase of the AMZN Fire phone). FB has attempted to release

a phone on HTC hardware that drives its social media platform, and its division Whatsapp

is set to launch an over-the-top voice service. MSFT has introduced hardware (tablets,

Lumia phones with the NOK acquisition) to drive its operating system and attract

application developers, as well as drive its own applications and cloud storage. APPL has

the hardware, software, and is moving toward on-demand apps with its iTunes/TV

offering. However, the most evident of companies attempting to bundle the consumer

experience is GOOG. GOOG’s open operating system (software) approach allowed it to

be placed in hardware devices (driving smartphone pricing down, i.e., Moto for $180

subsidized), alleviating the need for it to build its own. Having the hardware relationships

along with the software, it has created a platform to drive its search engine, on-demand

applications (GOOG Now/Android TV in set-top boxes in France with Iliad), social media

(GOOG +), and is now looking to completely drive its applications through a

wireless/wireline integrated offering.

The Internet companies are moving much faster than the service providers and are

looking to acquire basic fiber/datacenters and possibly even spectrum to be made

compatible with their own equipment. The largest Internet companies with the most scale

have been doing this for computing, datacenters and transport within and between

datacenters using virtualized technologies and are clearly looking to do this on a fully

integrated basis through to the customer connection.

The incumbent telcos’ primary core competency is that they do control the critical last-mile

infrastructure (and customer billing), but they are using outdated technology in some

instances. To retain their dominance in infrastructure, we believe they need to focus on

being the low-cost, high-quality network providers through a more horizontal structure.

We believe that the Internet companies will drive them to this new structure and

technologies faster than most observers believe. Positively, we see the telcos’

infrastructure (specifically their spectrum) as highly undervalued as currently operated and

capitalized, but this still needs to be managed.


4

Internet Companies Driving Convergence

Anyone who has gone from using a traditional taxi or car service for transportation to

Uber, or from linear viewing of television to Netflix, understands the transition that we will

see throughout the economy. This will be a highly organic process, but the broad trends

are in line with our long-held horizontal segmentation thesis. We think we are entering the

sweet spot of this industrywide restructuring, and we expect to see more change in the

communications/cloud industries in the next five years than we have seen in the last 15.

As usual, this is being driven by technology; right now wireless broadband is seeing an

explosion in productivity, but the same is happening in wireline optronics and switching.

We believe that new communications companies (Internet companies) are using the

power of the cloud and wireless broadband to drive their ecosystem of applications to

deliver a superior on-demand end-user experience to take full control of the customer

relationship. The next stage in the evolution will be seamless integration of these assets

bundled with some services at a low price. For example, GOOG’s applications and

products, from YouTube to Google Docs, Google Games and Television, are further

driven to the end user through its (open) Android wireless operating system, whether on

smartphones, Chromebooks, or set-top boxes (i.e., with Iliad in France). Having taken

control of the customer experience through applications, what better way to further

penetration of applications and brand equity than taking direct control of the network by

operating as an MVNO, where the consumer interacts solely with GOOG for

access/application needs?

We think this is a threat to the wireless carriers longer term, and believe other Internet

companies (AMZN, AAPL, FB) with consumer-facing applications will eventually head

down the same path, leveraging their superior capabilities in compute, storage and

networking to drive customer engagement and lock in. This competitive threat, in our

opinion, will force the incumbent carriers to upgrade their networks to offer the most

capacity and best coverage either to compete on the consumer side or act as wholesale

connectivity providers in the long term. Given the threat to their business models, we do

not believe that the wireless carriers would be willing to wholesale their networks in a way

where Google could use software-defined networks to switch between different networks.

We discuss the battle for the consumer in more detail below.

Disruptive Technologies

Ultimately, technology is driving this process, with competition and regulators having

impacts on the trajectory at the margins. We see the combination of very high-speed

wireless and wireline broadband enabling true cannibalization of legacy voice, data, and

video applications for the first time. The communication networks have historically been

built using purpose-manufactured hardware with integrated software. This is set to move

to standardized white boxes with the software layer separate from the hardware. This is

an extremely difficult transition for incumbents, which is why it is being driven by the

Internet companies, which have lower quality/regulatory requirements.

We are in the third major wave of new access technologies, dial-up, wireline broadband

and currently wireless broadband. Each new wave of access technologies has created a

few new dominant content companies and has been more positive for certain segments of

the communications food chain than others. Wireline broadband speeds have doubled

about every three years or so in a fairly consistent manner over the last 20 years.

Similarly, wireless broadband speeds have also been doubling every three years, but


5

more recently in a major step-function fashion, with speeds increasing about 10x in the

last two years. This productivity growth is consistent with Moore’s Law, and we believe

this is set to continue with major network improvements driving cloud and communications

convergence:

Major Improvements/Developments Will Drive Communications Convergence

1. Cloud—Shared virtualized computing

2. Internet of Things

3. Integrated Circuits of Optical Components by Infinera (optical multiplexing) OFDM

4. 100 Gig wavelengths to metro upgrades with standardized boxes

5. LTE DAS (small cells are much more efficient than expected), using

unlicensed spectrum with advanced carrier aggregation

6. Shutdown of 2G and 3G spectrum to migrate to LTE (only 30% of spectrum has been built out for LTE)

7. Broadcast spectrum auctions (game changer for Sprint/TMUS or Dish)

8. Voice over LTE enabling lower cost handsets and higher quality

9. Remove cloud-based LTE radio access devices (RADs) 10. OTT Unified Communications and integration of wireless and wireline 11. DOCSIS 3.1

12. VDSL using vectoring

13. Video OTT and multicast video on wireless networks

14. Virtualization of networking

15. Open compute of network hardware

With continuous improvements on the wireless side, we believe wireless broadband,

leveraging the capabilities of the cloud, is driving new applications and content focused on

the real-time flow of information among people and devices. The access to real-time data

flowing in both directions is an economy-wide game changer, allowing for the Internet of

Things and affecting every vertical of the economy from health care to financial services.

Exhibit 2: Three Major Waves of Access Technologies

Access Technology Dial-up Wireline Broadband Wireless Broadband

Network Providers: Incumbent Telcos Telcos and Cable companies Telcos, Cable, and possibly

Internet Cos

Dominate Internet Companies: AOL, YHOO GOOG, FB AAPL, GOOG, NLFX, Uber

New Applications:Second Lines, E-mail, Web

Browsing, Instant Messaging Social Media, Search, Streaming On-demand apps, OTT Video

Disrupted: Traditional mail services Dial up internet service providersTraditional Linear Content

Providers

Source: Oppenheimer

The legacy telecom operators benefited from the addition of second phone lines, and the

Internet was born with new companies such as AOL and Yahoo benefiting. When wireline

broadband arrived, new Internet companies such as GOOG and FB were able to leverage

faster speeds and more ubiquitous connectivity. In the current wave, wireless broadband,

we see the on-demand, over-the-top applications as well as the basic infrastructure


6

providers benefiting the most, such as AAPL, AMZN (AWS for infrastructure), Uber, etc.

For the wireless carriers there is strong demand for their services with limited supply.

Pricing has been aggressive by Sprint and TMUS in the last year, but we do not believe

that they can reduce prices further.

These new applications are stressing the legacy wireline networks, and they need to be

able to deliver on-demand broadband virtually anywhere in a similar way that AWS has

delivered on-demand computing and Netflix on-demand video. Given recent advances in

network access technology over the past couple of years (wireless broadband/LTE, dual

SSID wireless routers/WiFi hotspots, and SDN/NFV), we believe the communications

industry is poised for a tectonic shift in terms of both broadband deployment and video

consumption.

In this report we go over these overarching impacts and also discuss how the incumbent

broadband providers (AT&T, Verizon, Comcast, etc.) can partially offset rising competition

by leveraging wireless technology and software-defined networking (SDN). It also touches

upon the viability of OTT video and how the rising costs of linear TV (outpacing inflation by

~4x) are creating pent-up demand for lower cost Internet video. The success of OTT video

is more of a wild card given that the content providers hold all of the cards and have

incentives to keep the status quo (specifically $3B in annual re-trans fee; expected to grow

to $6B by 2018). That being said, if the FCC forces the content providers to sell

programming to the OTT video providers without discrimination, it could open the

floodgates for the entire OTT video industry. We discuss this possibility further below.

Incumbent Response

Over the next year we will be examining the reaction of each incumbent and each

segment of the market in more detail. There are almost too many variables to give an

accurate prediction to this organic process, but here are a few thoughts:

The structure of the wireless market could largely be determined by outsiders—Dish and

Google and possibly cable. Cable companies will need some mobile capabilities, and

they will need to offset declines in video revenue with business and broadband, which will

be a challenge. We have written about Dish’s options with its spectrum, and where this

ends up will largely dictate future industry profitability. However, Google’s entry into

wireless is equally important. Google, as discussed, is creating essentially a smartphone

than can elect to roam on the best wireless network (in this case, Sprint, TMUS, or WiFi),

and could be highly deflationary to the wireless industry.

The response of T and VZ to this trend will be interesting. T is trying to create a

converged product on its own with the digital home/car/video over-the-top, but as

mentioned, we believe an Internet company with a global reach will be able to implement

this better. At some point the incumbent telcos will come to this realization, and we

believe they will look to break themselves up into several companies; as we have written

previously, their spectrum alone could represent the value of the entire company. In the

meantime, we look for them to aggressively deploy new virtualized technologies to enable

more seamless consumer interactions and dramatically lower operating/capex costs.

Ultimately, we see these advances having major impacts to the industry, including:

1. The demand for truly converged/integrated services, not just applications


7

2. The fight between seven sub-segments of the market to create this converged

service

3. The encroachment of cable/Internet companies into the wireless space

4. Some consumers replacing wireline broadband with wireless broadband

5. The proliferation of over-the-top services/the Internet of Things with

cannibalization of the paid video market

6. Transformation of legacy networks using more fiber and LTE combined with

SDN/NFV to better leverage compute and storage capabilities, effectively driving

convergence between cloud and the network

7. Continued telco spin-out of assets but also further consolidation. The most

important consolidation for the industry is Sprint/TMUS, but we expect more

datacenter and fiber consolidation also.

Wild Cards for the Industry

1. Does Dish build a wholesale wireless network, or does it lease spectrum to the

telcos, cable or Internet companies to build?

2. Are Sprint and TMUS willing to separately wholesale their networks to GOOG?

3. If GOOG enters the wireless space as an MVNO, do other Internet companies

need to follow suit?

4. Does Google or Apple look to use smart SIMs to have instant roaming between

different cellular networks and WiFi?

5. Can the telcos update their networks to enable the “Uberization” of the customer

experience before Internet companies take control of the network to drive a

superior end-user experience?

6. Does AWS get spun out and/or acquire AKAM as cloud computing begins to

mature and content delivery, acceleration and security become even more

crucial?

These trends will have an outsized impact on the P&L statements of the communications

industry, with scale being the dominant factor. On the revenue side, we expect to see

pricing pressure in voice and video applications driven primarily by over-the-top providers

(RingCentral in business voice, Netflix, etc.), helped by the FCC’s net neutrality order,

which will in turn drive growth from innovation/new applications. On the cost side, we see

SDN, virtualization, cheaper optronics and LTE technologies (which are seeing massive

productivity improvements) driving lower procurement and operating costs.

We expect to see increasing competition in video similar to what we’ve seen with

voice over the past ten years. Initially, we saw the wireless cellphone gradually

substitute and then altogether replace home phone lines as we moved from regional 1 and

2G networks, to nationwide coverage on 3G and LTE networks. More recently, we have

seen the impact of cloud computing on the voice industry, where Skype, RingCentral, and

other VoIP providers have entered the market with lower price points, but like Uber,

dramatically improved automated/integrated services and a simplified business

proposition. These business models were able to rely heavily on cloud computing and

improved last-mile wired and wireless broadband, enabling them to provide services

running over the public Internet, instead of having to spend heavily on upfront CAPX


8

related to on-premise PBX or spending on a home phone line for consumers. In this

regard, legacy voice revenue of the carriers has continuously been declining in the 7-9%

range in the last decade.

The virtualization of applications and networks and seamless integration with cloud

computing needs to happen, which will be the most dramatic change since the Internet

started to become adopted by consumers 20 years ago. Positively, we estimate that the

industry could save 20% or more off its CAPX spending and 40% off its OPEX if it

automates network and customer care. The question is whether the incumbents can do

this themselves or will start up Internet companies do it.

At this point it is too early to say whether the carriers can offset cost savings from network

productivity enough to outweigh the effects of OTT competition, but this will largely be

determined by the number of players in the industry. However, the last-mile assets are

unique infrastructure, and if the telcos/cable companies remain price disciplined will see

very strong volume growth from this convergence and good returns on invested capital for

scarce infrastructure. The highest returns for infrastructure are generally recognized

through scale from horizontal structure in a more consolidated environment. We believe

players at the wholesale layer are in a very strong position (EQIX, AMT, AKAM), but in the

enterprise and consumer layer, the industry is more fragmented, with a battle for the end

user being waged. The wireline carriers are likely in a strong position in the enterprise

market, but we believe they need to figure out a way to integrate networks with cloud more

efficiently.

Exhibit 3: Horizontal Segmentation

Wholesale(I.E. AMT, CCOI, EQIX,

Dish?)

Enterprise(I.E. T, VZ, CRM, MSFT,

RAX)

Consumer(I.E. GOOG, AAPL, FB,

AMZN)

Source: Oppenheimer

Wireless Technology Driving Technological

Convergence

LTE is driving massive improvements in mobile broadband speeds, coverage and

capacity, and this will be improved with small cells (in many forms) and unlicensed

spectrum. Consumers will get used to and demand mobility for all their applications. As

we’re beginning to see across the industry, Comcast and Cablevision are gradually taking

steps to enter the wireless industry, as is GOOG, while the wireless carriers are


9

increasingly looking to LTE-U and small cells to densify their networks and increase

spectrum capacity.

We’ve seen these developments accelerate over the past several months with Comcast

turning in-home routers into public WiFi hotspots, Cablevision introducing its WiFi-only

service Freewheel, and GOOG expected to enter the market as an MVNO. On the

wireless side we’ve seen AT&T and VZ make a strategic shift from copper-based DSL

toward fiber and fixed wireless LTE. Broadly speaking, we see the convergence of

communications technologies into a single last-mile access industry that is primarily

horizontally segmented, but with the network providers looking to source OTT revenue

from the content providers. We believe this trend will lead to additional consolidation as

broadband access becomes the truly meaningful differentiator, and greater scale becomes

needed to compete.

Exhibit 4: NC Computing Drives Convergence Among Devices, Media,

Communications and Computing

Source: Oppenheimer & Co.

We believe that this will be the next major driver of consolidation in the US. We expect to

see three to four players ultimately competing on a national level with a quadruaple-play

service for both the enterprise and consumer markets. As a result, we would expect to

see competition between the cable companies, satellite and telcos heating up, helped by

technological innovation and convergence.

The technologies driving this competition include:

1. LTE Advanced with up to 50 Mbps speeds

2. WiFi hotspots leverage the consumers’ broadband connection

3. Cloud, leveraging compute/storage capabilities inside virtualized networks

4. More fiber pushed direct to consumer premises in 50% of the most dense homes

and the rest split between LTE and DSL/Vectoring

Overall, we believe that given the increased importance of mobility, the cable industry will

enter the wireless space more aggressively, and new entrants in the enterprise market will

emerge due to improvements in their own technology, the cost of upgrading their plant as

well as distprutive technologies. The above trends will be positive for content

providers, and we believe companies with unique infrastructure or applications are

well positioned. In infrastructure─spectrum, towers, fiber, datacenters and

managed services─have strong competitive advantages and barriers to entry due to

huge economies of scale. This infrastructure is currently dominated by VZ/T, AMT,

CCOI, EQIX, and AKAM, respectively. The telcos have spun out towers and

outsourced some fiber, datacenters and CDNs, but this is a trend that needs to

accelerate to lower costs.


10

The Battle for the Consumer

Exhibit 5: Current communications ecosystem dominated by ISPs

AT&TVerizon CenturylinkComcast

Consumer Consumer Consumer

Enterprise

Telcos/MSOs leverage last-mile broadband connection and spectrum to maintain control over the end users.

Web applications & Content (Google, Facebook, Amazon)

Web Companies , telcos and enterprises utilize proprietary datacenters, wholesale datacenters and carrier-neutral facilities to store and process data.

Public Cloud (AMZN, GOOG)

Carrier Neutral Facilities (EQIX, INXN)

Towers (AMT, CCI)

CDNs (AKAM)

Fiber (LVLT, CCOI)

Enterprise

Wholesale Datacenters(DLR, DFT)

Towers & CDNs offer uniquely positioned infrastructure to facilitate deployment of communications services.

Enterprise

CLECs, Long-haul Operators offer low-cost data transit services and compete against ILECs to provision voice and broadband for enterprise customers.ILECs in SMB &

Enterprise

Enterprise

Internet companies , have control of the customer on the application side, but not on the network access side. Business model is advertising based.

Source: Oppenheimer

The current communications ecosystem is dominated by the last-mile wireless and

wireline broadband providers, with the cable companies and VZ having the best

infrastructure. This enables them to control the consumer relationship. Google and Apple

and others are challenging this by focusing on the entire consumer experience. Initially

this was driven by separate and distinct apps which are now converging. For example,

GOOG Now provides users updates on weather when they wake up, suggestions for

groceries they may need to re-order (and provides a link to do so at the click of a button,

or press of a screen), the traffic conditions and the time it will take to travel to work, etc.

GOOG is essentially a full personal assistant, providing real-time, instantaneous

information and services at the click of a button (not to mention social networking with

GOOG+).

Through Android, GOOG’s open operating system, it is essentially driving its search and

consumer-facing apps (YouTube, Docs, Games, etc.). Other Internet companies are

attempting to do the same thing. AMZN released its Fire phone to drive use of its e-

commerce platform as well as its instant video service (the phone came with a one-year

free subscription to Prime and free unlimited cloud storage for photos). FB in conjunction

with HTC attempted to release a phone that utilized GOOG’s Android OS, but drove FB

applications. To state the obvious, APPL has been very successful in driving its full

ecosystem of hardware and content. The list goes on, with the success of some initiatives

questionable. These companies are aiming to become the communications companies of

the future, with several initiatives and acquisitions in recent years highlighting their efforts

including:

MSFT buying Skype


11

FB buying WhatsApp

GOOG Talk/Hangouts

GOOG buying YouTube

GOOG becoming an MVNO/laying fiber

GOOG deployed in Iliad’s set-top boxes

AMZN introducing Instant Video

AMZN introducing Fire TV

AAPL TV

APPL buying Beats Audio/Electronics

Given the compute, storage and networking capabilities Internet companies have built out,

they are eager to deliver an over-the-top, automated, on-demand end-user experience,

but traditional networks are not built to handle the growth in bandwidth demand and

volatility of usage and latency requirements to do so. We believe that in response, these

companies will look to further their direct control of the consumer experience by offering

not only the applications, but the network. The most visible of these plans is from GOOG,

which is building out fiber on the wireline side to entice carriers to upgrade their networks,

but is also planning on becoming an MVNO, to take control of the wireless network and

the consumer relationship.

GOOG is looking to become a wireless provider in a unique way, with an MVNO

agreement with S and TMUS for cellular service migrating between both networks for

better coverage and capacity. This can be accomplished with software-defined sim cards,

making the GOOG service better than S or TMUS can provide on a standalone basis. We

also believe it will look to partner with a cable company to offer WiFi offloading, leveraging

hotspots so that it has the ability and technology to offer seamless transition of

connectivity to whatever network is the strongest and most appriopriate for the application

being run (whether it’s over-the-top voice, video or data). The cable companies may also

look to build out more LTE hotspots using consumers' broadband as backhaul to increase

their coverage. We believe this is a potential threat over the long term to wireless carriers,

so we would be surprised if they enabled this in a very large way, but this said, Sprint is

fairly desperate. Also, markets outside the US have much larger wholsale components

and are oftentimes more fragmented, so it may be easier for GOOG to deploy its “smart”

wireless MVNO outside the US initially. If GOOG can integrate cheap wireless serivce

with its over-the-top applications, its offering could be extremely attractive. Like most

disruptive technologies, it would likely start on one device and penetrate the rest of the

market over time.

We think if GOOG partners with both S and TMUS, this will lead to the commoditization of

wireless carriers, where the only way they can compete is on network quailty and

coverage. To combat this, we think S and TMUS should form a partnership before

entering into an agreement with GOOG, to drive pricing power and maintain control and

leverage over their last-mile advantage.


12

Exhibit 6: Possible Communication Ecosystem Transition

Consumer ConsumerConsumer

Enterprise

Telcos/MSOs compelled to sell last-mile access to the internet companies, who in turn sell a complete cloud/communications offering direct to the consumer.

Web applications & Content (Google, Facebook, Amazon)

Internet Companies , leverage proprietary datacenters and 3rd party facilities to seamless store, process, and provision, web applications.

Carrier Neutral Facilities (EQIX, INXN)

Towers (AMT, CCI)

CDNs (AKAM)

Fiber (LVLT, CCOI)

Enterprise

Wholesale Datacenters (DLR, DFT)

Towers & CDNs offer uniquely positioned infrastructure to facilitate deployment of communications services.

Enterprise

CLECs, Long-haul Operators offer low-cost data transit services and compete against ILECs to provision voice and broadband for enterprise customers.ILECs in SMB &

Enterprise

Enterprise

Internet companies have control of the customer on the application side, and now purchase access capability to reach the customer directly on the communications side.

Last Mile Access (T, VZ, CMCSA)

Public Cloud (AMZN, GOOG)

Google Facebook Amazon

Source: Oppenheimer

Shifting Competitive Landscape

The communications competitive landscape is undergoing a dramatic shift driven by

technological advancements and the demand for high-speed/reliable connectivity. Trends

in end-user behavior including moving from fixed connections to mobile and from owning

computing resources to sharing those resources are shaping the landscape. Those trends

are driving network changes and the necessity for incumbent carriers to upgrade

infrastructure and utilize new technology to meet demands.

Exhibit 7: Shifting Trends in End-User Behavior and Networks

Aspect From To

Connectivity: Fixed Mobile

Provisioning: Call Centers Click and Install

Computing: Owned Sharing Resources

Aspect From To

Wireline: Copper Fiber

Voice: TDM IP

Wireless: 3G LTE

Architecture Focus: Hardware SoftwareSource: Oppenheimer

End User Behavior

Networks


13

On the wireless side LTE is a truly revolutionary product, enabling fast-speed mobile

broadband, and we believe, over time, it will lead to some cannibalization of wireline

broadband. It is a largely open standards-based technology that has been wildly

successful for the telcos (something that is needed in SDN). To solidify their positions, the

incumbent wireless carriers are increasingly looking to small cells and LTE-U to increase

speeds/capacity. By using QCOM chips on small cell base stations, incumbents can

greatly increase wireless speeds and reduce congestion on networks. By doing so,

wireless broadband could progress to the point where it becomes a suitable replacement

for wireline broadband. To this point, data capacity constraints have limited the

effectiveness of fixed wireless LTE as a suitable replacement for wireline access, but we

believe that as wireless networks continue to densify, this will all change.

Exhibit 8: Internet Access Price Points for Standalone Broadband

Offering Speeds Monthly Cost Start-up cost Data Cap

Verizon Fixed Wireless LTE 5-12Mbps $60 $0 10GB



Centurylink DSL (Phoenix) 7 Mbps $49 $160 250GB

Centurylink DSL (Phoenix) 12Mbps $54 $160 250GB



Windstream DSL (Lincoln) 6Mbps $57 $60 Unlimited

Windstream DSL (Lincoln) 24Mbps $67 $60 Unlimited

Comcast Coaxial Cable (Brooklyn) 3Mbps $50 $40 Unlimited




Source: Verizon, Centurylink, Windstream, Comcast

Note: Start-up cost assumes purchase of modem plus installation

Note: Start-up cost waived with 2 year contract for fixed wireless LTE

Note: Monthly cost ignores promotional pricing

Note: Activation fees not included

Note: Comcast Windstream pricing includes a $10/month and $7/month modem rental fee respectively

While cable has dominated last-mile high-speed broadband in the last ten years, the cable

companies are not sitting idle; they are looking to enter the wireless business leveraging

WiFi hotspots. While we believe WiFi-only services’ impact, such as CVC’s Freewheel, is

currently minimal, given no backup cellular service without a cellular network, it highlights

the threat dense hotspot networks could be, especially if cable companies turn home

routers into dual-access points for personal/public use.

On the wireline side, if the incumbent telcos want to compete, they must realize that

there is enormous demand for fiber connections in a distributed environment. However, it

is difficult to deploy infrastructure for their own usage, when it is much better to share the

infrastructure. This is one of the primary reasons the tower and datacenter companies are

in existence. It is also one of the reasons CCOI and Level 3 are doing so well, but they

are also benefiting from increased enterprise outsourcing of datacenters/computing.

Regardless, at the last-mile level we do believe that this is more of a duopoly environment

and that the incumbent wireline carriers should aggressively deploy fiber. It has proven

time and again that there is relatively strong demand for high-quality communications

infrastructure. Previously, upgrading wireline infrastructure may not have been cost

effective, but now the technology is cheap enough to do.

The incumbent telcos in particular have historically underappreciated the competitive

advantage of having premium speeds. In effect, they have been slow to upgrade their

plant and to share infrastructure. Speeds have been slower than those of cable

companies, and more investment is needed to close the gap. Several major forces will


14

(should) drive the incumbent telcos to upgrade their plant more aggressively, mostly due

to strong demand, lower costs, improved capabilities and most importantly competitive

necessity:

Demand remains very strong. We estimate that global IP traffic will continue to

grow at a 30% CAGR through 2018.

Cost to provision fiber closer to the home has plummeted. We believe that

fiber can be strung on aerial lines for $20,000 per mile. With the average suburb

in the US at more than 20 households per mile, the home passed cost is only

$1,000, and the connection cost another $1,000; at 50% penetration the total

cost is $3,000 per home. The $2,000 in total costs compares to Comcast’s

current firm value of $5,000 per home passed. We believe that even with 50%

penetration of homes passed, the cost per home is still in the $3,000 area.

Double-play bundled services can support $1,500 per year in revenue and

$1,000 in EBITDA, for a payback period in the three-year range. The added

benefit is that homes and businesses not connected directly with fiber should

also have much shorter loop lengths with the ability to upgrade to 100 Mbps

easily. At this point labor costs remain the most expensive aspect of a fiber

deployment.

DSL technologies (vectoring in particular) have improved, which is helping

AT&T upgrade Uverse from 25 Mbps speeds to 45 (and eventually 100). VDSL

speeds can be close to 100Mbps within 2,000 feet and 200-400Mbps if bonded

with a few pairs of copper. There are vDSL2 products that use 8 copper pairs to

achieve 400Mbps speeds at 1,000 feet or 200Mbps at 2,000 feet. If fiber is

extended to 50% of homes passed, we believe vDSL would need to be extended

to 30/40% of subscribers, with improvements in wireless technologies being

suitable for the 10/20% of remaining households.

We estimate optical terminating units at the home have declined from $500 a

decade ago to $50 today. The optical industry is on the verge of a mass

migration from 10 gig wavelengths to 100 gig coherent (Finisar, Infinera, etc. are

driving this process). The components will be standardized on CSP2+, and this

is what has enabled a step-function decrease in prices. This has been occurring

in long haul, but has now become cheap enough to deploy in metro areas and

even datacenters. This is a once in 15-year migration enabling a 10x increase in

capacity. This capacity increase will make it much easier to deploy small cells

and fiber either to the home (we expect half the country) or closer to the home

(within 2,000 feet, VDSL can handle 100 Mb speeds).

For building out fiber, the regulatory environment has also greatly

improved (thanks to GOOG). Google has done the industry a large favor by

convincing the regulators to allow selective build-outs of plant with minimum

requirements.

GOOG is building out fiber to the home. Google is building out fiber in ~50

cities with a Gbps symmetrical broadband service at $70 a month (voice and

video are extra). In the cities it is in, it also provides free 5Mbps services with a

onetime install fee of $300.

Demand trends are not lost on cable companies. Cable companies have had an

advantage in wireline connections based on the COAX connections to each home, which

can handle about 1,000 Mhz of spectrum. Allocating 20% of this to broadband can enable

1Gpbs of capacity or ~6Mbps for each Mhz of capacity (using the standard 256-QAM).

DOCSIS 3.1 uses new modulation technologies that can increase speed per Mhz to

9Mbps and can also increase the number of channels used. CableLabs is coordinating the


15

new technologies defined in the converged cable access platform (CCAP), which

combines edge QAM and cable modem termination systems to drive video broadband and

IP convergence.

DOCSIS 3.1 is the fourth generation of DOCSIS standards developed by CableLabs, and

we believe this technology will enable cable networks to get to substantially higher

speeds, beginning by year-end. This technology enables broadband data transport over

the cable plant and has allowed cable companies to dominate broadband delivery where

deployed, usually gaining ~60% of household market share and ~85% of the bit share.

The 3.1 specifications were released on October 2013, with plans to support capacities of

at least 10Gbps downstream and 1Gbps upstream using 4096 QAM.

DOCSIS 3.1 is part of the entire upgrade to a Cable Converged Access Platform (CCAP),

and there are dozens of other technologies also needed to be deployed, both at the

physical layer and software layer (software-defined networking, SDN). However, the

advantage of DOCSIS 3.1 is that it is backward compatible and fits into existing operating

systems relatively easily. We expect 3.1 to have double the existing broadband speed

packages, but at premium prices with the ability to move to 1Gbps download speeds over

the very long term. We believe the rollout will take half a decade or longer as the cable

companies are not under competitive threat to act any faster to defend their wireline

advantage. In addition, much of the basic infrastructure will need to be upgraded to

support 3.1, the fiber POPs will need to be extended deeper into the network and/or have

more capacity added. More of the spectrum capacity on the COAX will have to migrate

from video usage to broadband usage. Practically speaking, getting 100Mbps speeds will

not be that difficult to achieve in two to three years.

Fixed wireless LTE could be the carriers’ answer to upgrading a lot of the last-mile

connectivity, particularly in rural areas or poor neighborhoods without having to extend

fiber all the way to the home to be competitive. However, in most areas at this point, fixed

wireless services do not reach over 25Mpbs and are relatively expensive with low data

caps compared to wireline offerings. However, as LTE progresses and technological

advances allow for more efficient use of spectrum, fixed wireless could become a viable

alternative to wireline services. We see fixed wireless LTE first penetrating rural markets

where connectivity is limited and capacity unused.

Overall, last-mile connectivity is the competitive differentiator. Fiber is the ultimate

last-mile connection, and we always believe the closer to the consumer, the better.

However, LTE is a truly revolutionary product enabling fast-speed mobile broadband, and

we believe will lead to some cannibalization of wireline broadband.

The bottom line is the incumbent network providers need to be focused on

defending their last-mile barrier. We believe Investing in wireless connectivity is the

right strategic move, and having fiber closer to the end user will help greatly with

speeds/capacity. The incumbents will need to be vigilant about how these technological

advances impact their revenue on the applications side and how quickly the competitive

landscape is moving. We also believe they will need to be players in the enterprise market

for cloud services and will need to transform their networks, utilizing SDN/NFV protocols

to offer a network that can appropriately leverage the capabilities of computing and

storage, something legacy networks cannot do efficiently.

OTT Video Is Finally a Substitute for Linear TV

We believe Internet video or OTT video will continue to evolve from a secondary, niche

service into a mainstream commercial offering capable of becoming a legitimate substitute


16

to linear television. Going forward, we expect 1-2% of households per year to drop or

reduce their paid video service. This compares to wireless cannibalization of wireline of

~3% per year. We see five key reasons this is happening.

Higher broadband speeds and penetration

Increased Internet content availability

Proliferation of smart mobile devices and applications

Stagnant disposable income among consumers

Video ARPU continues to rise above the rate of inflation. In this regard, paid TV

is close to $100 per month with a triple-play bundle at $160, with content costs of

just $40 of this amount.

We have seen several recent studies that support this hypothesis. To illustrate, a

recent study published by Nielsen indicated that the average time spent watching linear

TV in 4Q14 declined 13 minutes from last year to a total of 4 hours and 51 minutes per

day. While linear TV viewing remains substantial, the report goes on to note that 40% of

US homes now subscribe to streaming services (Netflix, Amazon Instant Video, Hulu, etc.)

vs. 36% in 4Q13. Another recent study by DigitalSmiths suggests that roughly 1.5M video

subs plan to cancel their cable subscription sometime in the near future and another 2.4M

subscribers plan to downgrade their service.

Cable TV pricing is likely to peak: Using Comcast as an example, video ARPU has

risen to $78/month from $49/month in 2004, for a CAGR of 4.6% vs. inflation of about

1.5%. When bundled with broadband, a typical cable bill approaches $160. Approximately

half the cost of television goes toward content versus a third back in 2004. Given that

consumer income and spending are relatively flat, we believe there is going to be pent-up

consumer demand to cut the video cord once OTT video has better content. We think we

are close to that point and that the next economic downturn will drive the cord cutting.

With companies like Google, Intel, Sony and even Apple all reportedly working on over-

the-top (OTT) services that deliver live TV broadcasts via an Internet connection, it’s clear

that the pay TV market is ripe for lower priced new entrants. The telcos will also be rolling

out 15-channel lineups using multicast over the wireless networks. The main challenge

historically is getting access to the content, but there are now enough independent

producers and middlemen such as Netflix that the legal and economic barriers to getting

content are declining.

We believe that part of the hold-up has been that the broadcast TV stations are now being

paid by the cable providers for retransmission fees. We guesstimate that these fees

totaled $3B in 2014 and will double in the next five years. We believe this has dis-

incentivized the content providers from offering their product to OTT players, as it would

dilute their video product, thereby making it less valuable to their current customers (the

MSOs and telcos).

Regulation may open the floodgates: We believe that FCC Chairman Wheeler wants

his legacy to be tied to the growth of OTT video, and may consider ways to force the

content providers to sell content to OTT video providers at more competitive rates and on

an a la carte basis. This would open the floodgates to potentially dozens of OTT providers,

although we would expect consolidation to leave us with two to four additional Internet

video players.

We believe this would have two major effects. The first would be a sharp decline in video

revenue and video subs, which would be partially offset by cable trying to increase the

price of standalone broadband, or potentially putting into place a broadband cap, which

would limit video consumption outside the linear TV model. We are skeptical, though, that


17

the cable providers will be able to retain the $120 per month they are currently receiving

net of video fees for a broadband customer.

Sony TV a harbinger of what’s to come? According to the WSJ, Sony will start

commercial operation of its PlayStation Vue service within two weeks in New York,

Chicago and Philadelphia. Sources indicate it will be tested by customers in those cities

on an invitation-only basis and go nationwide by year-end. Sony noted that it aims to offer

roughly 75 channels per market, and most importantly, it will include the major broadcast

networks, which will enable users to watch live sports programming. While pricing has yet

to be revealed, we believe the service would be competitive in the $50 area. We think

pricing will be critical, since if the bill exceeds $50, the cost will defeat the entire purpose

of the service. However, we believe a decent triple-play service could be priced around

$100 per month versus the $150 cable is receiving.

Cable stocks have priced in some of this expected impact. However, if revenue

growth slows to the 2% area from the 4% expected, it will be difficult to see further

margin expansion. For now we believe the cable companies can continue to grow

revenue around 4%, but this is likely peak growth. Luckily for Comcast, its NBC business

should do better in this transition, offsetting some of this slowdown in growth.

Network Transformation

As mentioned, we believe the demand for bandwidth is being driven by video, cloud

and mobility. Wireless broadband in turn is driving cloud computing adoption, and

adoption of cloud by enterprises and many legacy consumer applications (video, etc.) is

set to drive extremely strong network volume growth. This in turn is creating a revolution in

provisioning of services to more nimble, lower cost, general-purpose white-box servers.

The lower cost broadband and computing is propelling strong enterprise and consumer

volume growth driven by new applications. Most of the new applications and network

virtualization are being driven by Internet companies, which will force legacy telcos to

more aggressively upgrade their networks. The lower costs should drive more enterprise

adoption of cloud, video OTT and the Internet of Things. We believe the legacy telcos will

continue to rationalize assets, and this and the new technologies should enable much

lower cost of service.

Cloud architectures were born out of virtualization of computer servers which allowed for

low-cost sharing of resources. Internet companies such as GOOG and FB have already

virtualized much of their datacenters and networks, and we believe carriers will be forced

to adopt similar technologies to drive efficiencies, innovation, and to be able to compete

effectively. Business models of traditional communications companies will be transformed

over the next five years as we expect to see an increase in competition from both wireless

startups and large Internet companies, which will force a major restructuring of legacy

networks.

The incentive for the telcos/cable companies to aggressively upgrade their plant and

capabilities will be driven by Internet companies. The largest ones have been building

their own switch and optical technologies. Google is probably the most visible, and it has

been a clear innovator of cloud and datacenter interconnectivity. It has helped drive the

industry forward on virtualized computing and is looking to do the same with networking.

While we see GOOG (or other Internet companies) looking to move down market to

promote inexpensive mobile connectivity and clearly more video over-the-top to

consumers, we believe in the enterprise market, it (and other Internet companies) could

move to offer connectivity solutions to drive increased penetration of enterprise

applications. While we do not see this as an immediate risk, we do note that there is a


18

possibility companies with disruptive technologies move up market to offer connectivity to

drive further penetration of enterprise applications.

On the enterprise side, we believe carriers need to invest in converting legacy

networks to all-IP networks and use SDN/NFV protocols to drive down costs,

allowing for agile, flexible, and scalable networks, which we believe are imperative

to defend against new entrants and start-ups. We also believe that the incumbents will

look to greatly reduce operating expenses through this migration, essentially using cloud

computing themselves, which could save tens of billions in operating expenses.

Below, we go through some of the disruptive technologies (SDN/NFV) that can be used to

drive/defend enterprise market share. Along with competitive shifts highlighted above, we

believe this is set to be the most dramatic change to the communications industry since

the Internet started to become adopted by consumers 20 years ago.

Why Now?

IP traffic continues to grow at astounding rates, and mobile traffic is growing even

faster with the emergence of social media and streaming videos. The Internet of

Things only points to continued growth in IP traffic with a huge amount of data coming

from multiple devices per user. Cisco estimates by 2018 the average number of

connected devices in North America per user will be over eight.

Traditional networks were not built to handle the sheer volume of Internet traffic we are

witnessing today due to the demand for cloud services and mobility. Additionally,

traditional telecom networks were not built to handle the amount of traffic on both fixed

and wireless networks. We estimate that global IP traffic will grow ~30% annually through

2018, and growth will be led by mobile data traffic, which we estimate will grow over 65%

annually through 2018.

Exhibit 9: Global IP Traffic Growth Estimates

27%

28%

29%

30%

31%

32%

0

50,000

100,000

150,000

200,000

2013 2014 2015 2016 2017 2018

Pe

tab

yte

s p

er

mo

nth

Global IP Traffic Growth

Fixed Internet Managed IP Mobile Data YoY Growth

Source: Cisco VNI, Oppenheimer & Co. Inc.

Continued demand for bandwidth as well as competitive pressures to lower prices has led

to increased operating/CAPX expenses and diminishing returns for the incumbents.


19

Hardware requirements within networks require new services to be provisioned to specific

protocols and often require specific ancillary hardware purchases to ensure

interoperability with new services and applications. There are additional

maintenance/repair costs that go into legacy hardware throughout each central office a

telecom operator owns.

To bring networks up to speed with traffic growth and ensure scalable, agile, flexible

networks to compete in the current environment and meet demand, software-defined

networking and network function virtualization protocols must be used. Utilizing these

technologies will allow for significant cost savings and enable innovation leading to new

services, such as on-demand bandwidth provisioning. There are several operational shifts

that will take place as traditional physical networks shift to virtualized networks.

Exhibit 10: Expected Operational Shifts

From To

Hardware Centric Software Centric

Separate IT/data center & Network/Central Offices Common Technology & plant

Quarterly software releases Continuous software process

Geographically fixed, single purpose equipment Highly dynamic & configurable topology/roles

Tight coupling of Network Elements Separation of physical/logical components

Separation of service elements/support systems Integrated orchestration/automation/virtualization

Hardware monitoring appliances Software based monitoring

Special design and provisioning Configurable catalog/rule-driven delivery frameworks

Optimized provider network/ops process Optimized customer experience

Highly constrained, independent control planes Highly integrated/automated control planes

Limited service dimensions Multifaceted service dimensioning

Highly constrained data translation/synchronization Shared management technology between network/systems

Slow tooling changes requiring coding Rapid tooling changes using policies/rules

Network management Customer experience management

Long lead provisioning times Real-time provisioning

Static billing and charging Granular and dynamic usage-based charging/billing

Source: AT&T, Oppenheimer & Co. Inc.

While we acknowledge that software-defined networking is nothing new, we believe

the incumbent carriers have been slow to adopt new technologies used by Internet

companies and used in cloud datacenters. New computing/networking architectures

forced datacenters to become more of a virtualized resource. To take advantage of

compute and storage virtualization, it became evident that even further efficiencies could

be gained through the movement of a software-defined datacenter. The software-defined

datacenter builds on server and storage virtualization and utilizes SDN protocols to further

automate the flow of data and the movement of workloads and is the foundation for hybrid

cloud architectures.

The primary chip vendor for the white-box hardware is Broadcom which enables the

Internet companies to write the software code that essentially controls the hardware from

a remote datacenter. We believe that by doing so, the hardware/operating costs at these

companies are one-fifth that of a telecom operator’s.


20

We believe that by embracing open standards and developing/utilizing SDN/NFV

protocols, Internet-facing companies, especially GOOG, could move up market to begin

offering wireless and wireline solutions to consumers and enterprises; given GOOG’s cost

and technological advantage, this is a major threat to incumbent carriers.

Disruptive technologies are usually adopted and developed by new market entrants and

often resisted by incumbents because they usually predicate a move down to a lower

value network (cost structure). As SDN/NFV climbs up the S-curve of adoption,

companies that are deploying this technology often find new use cases, which we believe

in this case will be the ability to offer high-speed on-demand/flexible services and

connectivity demanded by customers and enterprises. In “The Innovator’s Dilemma” by

Clayton M. Christensen, he explains how disruptive technology usually gets its start in

emerging value networks (in this case, cloud computing) before invading established

networks. He further explains that after the technology progresses in its first use, it can

progress to the point where it can satisfy the level and nature of performance demanded

in another value network. The disruptive technology can then invade, knocking out the

established technology and its established practitioners. The exhibit below depicts a

disruptive technology S-Curve where SDN/NFV (technology 2) has been developed and

used inside datacenters and inside networks of Internet companies (mainly between their

datacenters). We believe these disruptive technologies have progressed enough to

become applicable to carrier networks (Application A below).

Exhibit 11: Disruptive Technology S-Curve

Source: Clayton Christensen, The Innovator’s Dilemma

We note that given the complexity in building networks as well as the massive scale of the

incumbents’ networks, it would take a herculean effort from GOOG or another Internet

company to rapidly take market share.

There is a natural marriage between Internet-facing applications, mobile

connectivity and cloud computing, and the carriers can take advantage of it to

reshape the industry for their benefit if they are successful in upgrading their

networks. The carriers can positively adjust their cost structure and have more flexibility

and resources to deploy in the face of increased competition from Internet and cable

companies if a shift to a multipurpose all-IP network occurs, which will be facilitated by

SDN/NFV. Among the major implications of this will be an explosion in innovation and

usage, helped by lower cost end-user devices and computing (in the datacenter). In

essence, to be competitive in the enterprise market, carriers will need to offer a network


21

that appropriately leverages the flexibility, agility and scalability of compute and storage

resources (cloud computing) to facilitate the utilization of innovative applications that drive

new products and services, and SDN/NFV provides the framework to do so.

Exhibit 12: Reasons for Change

Reason Description

Optimizes OPEX Move from telco OPEX model, to lower cost IT model

Lowers CAPX Move from dedicated appliances to virtualization

Accelerates time to market

Deploy new software/services quickly from months to minutes

Open platforms accelerates innovation

Broadens access to partners who can innovate

Rapidly scale up/down apps

Modify QoS

Deliver New Services Faster

Speed

Innovation

Agility

Source: Lightreading, HP, Oppenheimer & Co.

By utilizing SDN and NFV protocols the carriers can redefine their networks to be more

flexible, agile, and cost-effective, allowing them to innovate and offer new

products/services as well as reinvest cost savings to bring faster speeds closer to the

premise. By utilizing new technology and bringing connectivity closer to the premise not

only to increase speeds and reliability but to optimize networks, carriers will benefit from a

more optimal cost structure, lower capital requirements, speed of deployment, innovation,

and agility.

Industry Implications

Exhibit 13: High-Level Cloud Networking Architecture

Source: Cyan


22

By utilizing SDN/NFV. Service providers are being forced to move to an all-IP network to

leverage the benefits of SDN which effectively allows previously specialized hardware to

be replaced by unintelligent, white-box hardware. This will allow the carriers not only to

buy hardware for less, but leverage open standards to design protocols that will be

compatible with multiple vendors and suppliers. It will also allow the hardware to be used

for multiple tasks and applications. Combining network function virtualization on top of

software-defined networking will enable industry participants to leverage commodity

hardware to quickly develop, test and launch new services and utilize resource and

applications only when needed.

Specific cost savings are difficult to estimate given the transformational process. As

organizations compete to hire top IT talent well versed in relatively new technological

advances, some operating expense cost savings from being able to centrally manage a

network and not having to maintain hardware at each central location may be offset by

increased hiring and R&D costs in the near term. AT&T estimates that it will be able to

save ~$3B by transitioning its network to all-IP (and utilizing SDN/NFV) by 2020. In the

near term, we believe savings on CAPX is more likely, given the pure cost of white-box

hardware compared to legacy hardware.

Further, we believe that SDN and NFV will also allow carriers to reduce the number of

central offices they own by moving switching away from local networking into the cloud in

tandem with a move to all-IP networks. Outsourcing these functions to an external cloud

datacenter makes sense, with most of the functions residing in software, rather than

hardware, and allowing the management of those functions to be centrally located, rather

than in each of the central offices.

We believe that the most successful organizations are the most focused ones, and

outsourcing and/or selling off datacenters and legacy wireline assets will allow the

incumbent carriers to focus on defending their main competitive advantage, which is last-

mile connectivity. Becoming more horizontally integrated would benefit the organizations

to drive innovation and allow them to compete effectively in each area of focus. We

believe the most successful organization in the race for enterprise dollars will be the one

that decides to utilize disruptive technology to its advantage, redefining its networks and

leveraging cloud computing to offer enterprises the connectivity, flexibility, agility, and

scalability they need to bring a superior experience to their end users.

Positively, the need to utilize SDN and NFV is not lost on carriers, and many have begun

testing, including:

CTL (utilizing Cyan’s platform)

T (Domain 2.0)

VZ

Deutsche Telkom AG

Orange

KDDI Corp.

NTT Group

Telefonica


23

APPENDIX

What Is Software-Defined Networking?

Software-defined networking separates the control plane (which deals with the overall

network map of how to deal with different types of data packets/how they should be

routed) from the data plane (which deals with how data and the overall network topology

are managed), allowing for a centralization of network intelligence/state and separating

the underlying infrastructure from network applications and features. This is a shift from

traditional network devices (switches, routers, etc.) that have built-in control, data,

management and service planes that rely on conventional protocols to route traffic and

manage the network.

At the heart of software-defined networking is an SDN controller, which abstracts network

elements and allows for end-to-end control of a network, replacing element-based

configuration. This abstraction allows for applications, through APIs, to be programmed to

manipulate the abstracted network elements. Essentially, SDN moves a large portion of

what a physical switch had done, and virtualizes the capabilities to allow the data to be

dealt with on an abstracted level. This allows physical ports to be virtualized and allows

data flows to be more controlled, which allows for fewer hops and less latency.

Exhibit 14: Software-Defined Networking Architecture

Source: Gartner

What Happened To Enable SDN?

Traditional networking approaches have been based on closed, proprietary technology

that has become exceedingly complex to handle various applications. Networks have

been built using switches, routers, and other devices that implement distributed protocols

to support various functionality and applications. However, it has becoming increasingly

difficult for Internet companies to manage their traffic growth and has limited some of the

features and functionality they had wanted to deploy. Traditional networks are difficult to

optimize and costs continue to grow, making it difficult to introduce new revenue-

generating services. Also, networks continue to have problems with security, robustness,

manageability, and mobility.


24

As traffic continued to grow, it became expensive to continue to add hardware that went

underutilized into datacenters that were inefficient. That is when computer servers

became virtualized, and shortly after storage followed.

With virtualization, compute and storage became pools of resources that were provisioned

on an as-needed basis, which became the basis of the private cloud for on-premise

workloads. AWS took this one step further, introducing a business model that provided

shared infrastructure off-site, in a pay-as-you-go model, which is the essence of the public

cloud. These new architectures forced datacenters to become more of a virtualized

resource. To take advantage of compute and storage virtualization it became evident that

even further efficiencies could be gained through the movement of a software-defined

datacenter. The software-defined datacenter builds on server and storage virtualization

and utilizes SDN protocols to further automate the flow of data and the movement of

workloads and is the foundation for hybrid cloud architectures.

Using virtualization and SDN protocols, enterprises can now shift workloads between

private clouds and public clouds and can utilize the features and resources of both the

private and public clouds, or traditional on premise client-server architectures in a hybrid

environment.

We believe the next logical step in the virtualization chain is the network. We believe that

in the next five years we will have progressed from compute/storage virtualization to

complete network virtualization.

Inside Google’s Network To support all of the product and services GOOG offers (web search, GOOG+, Gmail,

YouTube, Maps, etc.), it needed a network that could appropriate/leverage the scale that

advancements in compute and storage brought to the table. To support its mission “to

organize the world’s information and make it universally accessible and useful,” it required

a global wide area network (WAN) with economies of scale that could deliver cost

efficiency, higher performance, better fault tolerance and manageability, operating the

WAN as a fabric, not just boxes. The vendors could not support Google’s ambition.

GOOG’s network is organized as two backbones, an Internet-facing network and an

internal network that carriers traffic between datacenters. Within the latter, GOOG has

deployed an OpenFlow-powered SDN solution. With Google’s scale it had to build its own

network switch utilizing outsourced chips and open-source routing stacks. The company

also built a centralized traffic engineering service, allowing it to collect real-time utilization

metrics/topology data from the underlying network as well as bandwidth demand from

apps/services. The data then allowed for the computation of the path assignments for

traffic flows, which then programmed the paths to the switches using OpenFlow. The

service can recompute the path assignments and reprogram switches on a real-time

basis. Traditional (telecom) networks do not have this ability given the intelligence is inside

the hardware, not the software, and hardware applications cannot be re-used for different

applications and services.

GOOG’s datacenter-to-datacenter WAN successfully runs on an SDN/OpenFlow-enabled

network and has greatly improved the manageability, performance, utilization and cost-

efficiency of the WAN. Network utilization, according to GOOG, is up to 95%, which is

unheard of in the industry. GOOG began deploying this network in 2010, and it has been

fully functioning since 2012. We believe telecom operators are five to ten years behind

Google’s network capabilities. Interestingly, as of 2013, more than 60% of all web-

enabled devices exchange traffic with GOOG’s servers.


25

Exhibit 15: GOOG WAN Network Timeline

Source: Google

We believe that Google has had the scale to leverage advancements in compute and

storage driving SDN that enables a network that is flexible, agile, and can handle

enormous volume growth. We believe that the incumbent carriers need to emulate

GOOG’s network to drive an end-user experience that GOOG and other Internet

companies are attempting to provide; if not, they face serious threats from Internet

companies being forced to move down market to drive customer lock-in and a superior

experience.

A Look at Facebook’s Open Compute Project

Large-scale Internet companies have been using virtualization and software-defined

technologies, developing open-source technologies to support their core business. Instead

of relying on technologies that are proprietary to suppliers, these companies avoid vendor

lock-in and inflexible hardware protocols by developing open-source software to be used

with generic white-box hardware.

As an example, over the last four years Facebook has contributed to the Open Compute

Project (OCP) to disaggregate core components of the traditional datacenters to build new

systems that are more flexible, scalable and efficient. In 2013 the Open Compute Project

began focusing on networking, which according to opencompute.org, is a project to build

“a disaggregated switch to enable a faster pace of innovation in the development of

networking hardware; help software defined networking continue to evolve and flourish;

and ultimately provide consumers of these technologies with the freedom they need to

build infrastructures that are flexible, scalable and efficient across the entire stack.”

Since 2014 Facebook has introduced a number of advancements in achieving OCP’s

goals including “Wedge,” “FBOSS,” and a few weeks ago “6-pack,” a scalable switch.

Facebook’s goal in developing and introducing these products was to make its network

look, feel and operate like the OCP computer servers it has already deployed in its

datacenters. While we recognize these technologies are largely deployed within and

between datacenters, we believe they can be applied to carrier networks as well.


26

Exhibit 16: Facebook OCP Progression

Source: Facebook

Wedge – A top-of-the-rack switch that utilizes an open-source server module to allow for

provisioning with Facebook’s Linux-based operating environment. By doing so, it allows

for the deployment, monitoring and controlling of the systems alongside servers and

storage resources, allowing engineers to focus more on innovation than management of

existing systems.

FBOSS – A Linux-based operating system that is designed to leverage software

libraries/systems Facebook currently uses to manage its server fleet. By controlling the

programming of the (wedge) switch hardware, FB can implement forwarding software

much faster. The service layer allows both distributed/centralized control and allows

flexibility for where the control logic resides. By using a central control, it allows the

company to find the optimal network path for data at the edges of the network, and by

doing so, Facebook claims it has managed to boost utilization of edge network resources

to more than 90%. FBOSS also allows the company to leverage existing tools for

environmental monitoring.

6-Pack – Uses wedge as the basis and includes 12 independent switching elements.

Each element runs its own operating system and is completely independent. This allows

for modification of any part of the system, with no system-level impact on either the

software or the hardware. All of Facebook switch designs use off-the-shelf, white-box

hardware. The open, module, technology-agnostic design of the 6-Pack allows for

customized solutions and centralized control of switches using an open-source operating

system such as FBOSS.

If carrier networks can make the transition to all-IP, utilizing open-sourced technology like

those highlighted above would be transformative.


27

What Is Network Function Virtualization?

Network function virtualization (NFV) is the concept of traditional network functions,

previously implemented in specialized hardware, being run in software on industry

standard, or white-box hardware. For example, traditional proprietary hardware-based

network applications or functions, such as routing, packet inspection, firewalls, load

balances and mobile packet gateways are removed from the proprietary hardware and

embedded in software. Each of these virtual network functions (VNFs) allows network

operators to spend less on hardware on centrally controlling their networks. This can be

deployed for individual network elements. SDN looks to tie all of the elements together.

Exhibit 17: Network Function Virtualization

Source: ETSI

Vendors

Large companies from CSCO to IBM are involved with the SDN/NFV supply chain. Below

we highlight some interesting companies that are involved in the supply chain.

Cyan – Cyan provides the SDN platform, packet-optical transport platform, as well as a

hyper-scale platform. Cyan’s Blue Planet is a carrier-grade SDN platform built for network

operators. It integrates the capabilities of a multi-domain service orchestrator, multi-layer

SDN controller, and multi-vendor management system. Applications include:

- Planet Orchestrate - A multi-domain service orchestration app. for

allocating/automating virtual and physical resources across carrier cloud and

software-defined WANs.

- Planet Operate – A multi-layer SDN controller/management app. for WANs

- Planet View – A multi-vendor performance monitoring SLA assurance app.

- Planet Inventory – An inventory reporting app.


28

- Planet Design – A multi-layer network planning app.

The platform centralizes control and management of physical network elements and

virtual SDN/NFV resources to simplify service orchestration across carrier data centers

and the WAN. It supports open northbound and southbound APIs to enable integration

with existing systems/accelerate service and app. development. Its open architecture

design enables a fully software-defined, open, and virtualized network that gives network

operators the flexibility to select multi-vendor solutions.

Its Z-Series Packet Optical Platform is a foundation for transformation to more

manageable/scalable multi-layer networks. Metro/regional networks consist of multiple

layers─WDM, OTN, and Ethernet─and what the Z-series has done is introduce a scalable,

COTS-based platform that leverages the latest advances in commercial off-the-shelf

(COTS) technology, including merchant silicon and photonic components, to combine

carrier-grade Carrier Ethernet, DWDM, OTN and ROADM functionality within a single,

open platform. It supports a broad array of module options that provide native switching

aggregation, and transport functions for a mix of Ethernet, wavelength, OTN, and

SONET/SDH services.

The N-Series Hyper scale provides a fundamentally new model for scaling datacenter

interconnect (DCI) networks that leverage best-in-breed COTS silicon and photonic

components, combined with a disaggregated Linux operating system. This has the effect

of eliminating vendor lock-in associated with proprietary, vertically integrated solutions,

and significantly reduces the cost of high-capacity optical transport. It also allows

datacenter operators to enable, control and program the DCI network using the

management and automation systems already in place for the compute/switching

infrastructure.

Orion VM – Orion VM has built a proprietary wholesale low-cost cloud service, which uses

white-box hardware combined with its own software. The company will only wholesale it

primarily through systems integrators and telco service providers. It primarily uses

Cumulus networks (highlighted below) and Linux networking stack. It also takes

advantage of modular switching Facebook has provided (highlighted above). The

company uses a chipset from Broadcom called Trident II, and the switches themselves

are manufactured by ODMs out of Taiwan (Quanta, Acton). Utilizing open/industry

standard hardware designs and writing its own software, Orion VM is able to provide cost-

effective white-label cloud services to system integrators and telco operators.

Cumulus Networks – Cumulus positions its Cumulus Linux product as the first, true Linux

OS for hardware networking. The product boasts a broad application ecosystem working

with existing open source and commercial Linux apps that run natively on industry standard

switches. New applications can be developed and integrated rapidly, and customers can

write applications to run directly on switches and implement any software agent they would

like. The company develops agents for network orchestration tools such as OpenStack and

VMware NSX. Cumulus OS works with a number of open networking switches and utilizes

Broadcom chipsets.


29

Exhibit 18: Cumulus Linux Hardware Compatibility List

Manufacturer Switch Silicon

DELL Broadcom Trident II

Edge Core Broadcom Trident II

Penguin Computing Broadcom Trident II

Quanta Broadcom Trident II

Agema Broadcom Trident

DELL Broadcom Trident

Edge Core Broadcom Trident +

Penguin Computing Broadcom Trident +/II

Quanta Broadcom Trident +

Agema Broadcom Triumph2

Edge Core Broadcom Apollo2

Penguin Computing Broadcom Triumph2

Quanta Broadcom Firebolt3

40 GB Portfolio

10 GB Portfolio

1 GB Portfolio

Source: Cumulus Networks, Oppenheimer & Co. Inc.

Big Switch Networks – According to the company’s website, Big Switch was founded

with roots in the original Stanford research team that invented SDN. Beginning in 2013,

the company began packaging technology components into Bare Metal SDN Fabric

Solutions. The company combines commodity Ethernet switches with sophisticate SDN

control software, developing modern networking fabrics. The company is currently focused

on datacenter networks and monitoring networks, with switching and routing being a core

area of focus.

Broadcom – Broadcom, in compliance with the Open Compute Project’s Networking

Project, has developed an open switch specification, delivering the foundation for efficient,

high-performance, flexible network architectures. The switching specification is based on

the Trident Switch architecture, which sports a large ecosystem of networking operating

systems and apps. According to the company, its Trident II brings the most advanced and

comprehensive feature set into the open-switch ecosystem. In conjunction with open

standards and what the SDN/NFV solutions promise, the adoption of the open-switch

specification will help enable faster innovation in the market, while providing more choice

to datacenter operators and telecom service providers. The Trident II is a single-chip

solution for common fixed top-of-the-rack aggregation and line card switching applications

and can be found in many of the commodity hardware devices that run SDN/NFV

protocols. Customers can use both the hardware and software functions offered by

Broadcom, or choose to write their own software given the open source characteristics of

Broadcom’s portfolio. In essence, Broadcom’s product portfolio provides reference

designs and hardware/software building blocks that enable advanced SDN

implementations, targeting datacenter and carrier/service provider applications.


30

Exhibit 19: SDN/NFV Vendors by Network Layer

Source: Gartner.

Stock prices of other companies mentioned in this report (as of 03/12/15):

Dish Network Corp. (DISH – NASDAQ, $73.82, Not Covered)

Deutsche Telecom (DTEGY:US-OTC, $18.14, Not Covered)

Cablevision Systems Corp. (CVC – NYSE, $17.86, Not Covered)

Sony Corp. (SNE – NYSE, $26.56, Not Covered)

Cyan , Inc. (CYNI – NYSE, $3.99, Not Covered)

Orange (ORAN - NYSE, $16.19, Not Covered)

KDDI Corp. (KDDIF - OTC, $65.75, Not Covered)

NTT DOCOMO, Inc. (DCM - NYSE, $18.17, Not Covered)

Telefonica, S.A. (TEF - NYSE, $13.97, Not Covered)


31

Exhibit 20: Annual Communications Industry Cash Flow Model Last Updated: 3/13/2015

2009 2010 2011 2012 2013 2014 2015E 2016E

Revenues*:

AT&T (T) 117,018 119,702 122,980 126,329 128,713 132,417 134,180 136,541

Verizon (VZ) ** 108,211 107,106 110,836 115,648 120,246 126,075 130,119 133,623

Sprint Nextel (S) 33,415 33,637 34,627 36,215 36,304 35,725 34,797 34,855

T-Mobile (TMUS) 21,531 21,347 20,618 19,719 26,130 29,564 32,354 33,727

Comcast (CMCSA) 33,742 35,363 37,226 39,604 41,836 44,140 46,047 48,151

Time Warner Cable (TWC) 17,868 18,868 19,675 21,386 22,120 22,812 23,681 24,970

Charter Communications (CHTR) 8,017 8,419 9,108 9,608 10,108

CenturyLink (CTL) 20,841 19,431 18,692 18,376 18,095 18,031 18,058 18,289

Windstream (WIN) 6,259 6,239 6,140 5,988 5,851 5,838 5,887

Level 3 (LVLT) 6,231 6,140 6,318 6,376 7,850 8,168 8,349 8,763

LEAP (LEAP) 2,481 2,697 3,071 3,142 2,898

MetroPCS (Acquired by TMUS) 3,481 4,071 4,846 5,101

Clearwire (CLWR) *** 274 558 1,259 1,265

Total 365,093 375,179 386,387 407,317 418,599 431,890 443,031 454,914

Growth YoY % 1.2% 2.8% 3.0% 5.4% 2.8% 3.2% 2.6% 2.7%

Capital Expenditures:

AT&T (T) 17,335 20,302 20,277 19,728 21,228 21,433 18,207 18,536

Verizon (VZ) ** 16,044 15,707 15,372 15,199 15,654 16,265 16,852 17,131

Sprint Nextel (S) 1,419 1,675 2,574 5,126 7,153 5,458 6,270 6,298

T-Mobile (TMUS) 3,692 2,819 2,729 2,901 4,241 4,317 4,702 4,781

Comcast (CMCSA) 5,037 4,853 4,806 4,921 5,403 6,154 6,711 7,077



CenturyLink (CTL) 2,450 2,555 3,151 2,858 3,001 3,026 3,034 3,073

Windstream (WIN) 647 892 1,049 812 764 817 883

Level 3 (LVLT) 487 603 628 743 1,262 1,256 1,269 1,315

LEAP (LEAP) 700 399 441 434 133

MetroPCS (Acquired by TMUS) 831 790 890 889

Clearwire (CLWR) *** 1,540 2,644 246 183

Total 52,766 55,924 54,943 58,942 63,938 64,990 64,381 65,811

Growth YoY % (7.7%) 6.0% (1.8%) 7.3% 8.5% 1.6% (0.9%) 2.2%

Cap-ex as % of Revenues 14.5% 14.9% 14.2% 14.5% 15.3% 15.0% 14.5% 14.5%

2009 2010 2011 2012 2013 2014 2015E 2016E

EBITDA

AT&T (T) 41,035 41,948 40,713 42,002 42,567 41,582 42,500 43,368

Verizon (VZ) 35,638 36,167 35,906 38,531 42,899 44,136 46,091 47,896

Sprint Nextel (S) 6,419 5,621 5,067 4,796 5,442 6,119 6,621 6,963

T-Mobile (TMUS) 5,919 5,478 5,310 4,886 5,317 5,636 7,137 8,234

Comcast (CMCSA) 13,767 14,561 15,288 16,255 17,205 18,112 18,854 19,806



CenturyLink (CTL) 8,378 8,303 7,799 7,656 7,361 7,043 6,919 6,976

Windstream (WIN) 2,406 2,430 2,384 2,318 2,144 2,137 2,151

Level 3 (LVLT) 1,258 1,239 1,297 1,432 2,218 2,443 2,630 2,936

LEAP (LEAP) 485 525 562 597 408

MetroPCS (Acquired by TMUS) 956 1,176 1,331 1,512

Clearwire (CLWR) *** (871) (1,364) (381) (157)

Total 119,478 122,936 122,548 130,582 136,662 138,633 144,805 151,032

EBITDA as % of Revenues 32.7% 32.8% 31.7% 32.1% 32.6% 32.1% 32.7% 33.2%

2.9% (0.3%) 6.6% 4.7% 1.4% 4.5% 4.3%

Unlevered FCF

AT&T (T) 23,700 21,646 20,436 22,274 21,339 20,149 24,292 24,831

Verizon (VZ) 19,594 20,460 20,534 23,332 27,245 27,871 29,239 30,766

Sprint Nextel (S) 5,000 3,946 2,493 (330) (1,711) 661 351 666

T-Mobile (TMUS) 2,227 2,659 2,581 1,985 1,076 1,319 2,435 3,454

Comcast (CMCSA) 8,730 9,708 10,482 11,334 11,802 11,958 12,143 12,728


Charter Communications (CHTR) 1,048 1,094 969 969 969

CenturyLink (CTL) 5,928 5,749 4,648 4,798 4,360 4,017 3,885 3,904

Windstream (WIN) 1,759 1,538 1,335 1,506 1,381 1,320 1,268

Level 3 (LVLT) 771 636 669 689 956 1,187 1,361 1,621

LEAP (LEAP) (214) 126 121 162 275


Clearwire (CLWR) *** (2,411) (4,007) (626) (340)

Total 66,712 67,012 67,606 71,640 72,724 73,643 80,424 85,221

UFCF as % of Revenues 18.3% 17.9% 17.5% 17.6% 17.4% 17.1% 18.2% 18.7%

Source: Company Reports & OPCO Research Estimates.

Notes:

* Includes wireless equipment sales.

** Proforma for wireless/wireline acquisitions/dispositions

*** Part of Sprint as of 3Q13


32

Exhibit 21: Annual Wireline Industry Cash Flow Model

Last Updated: 3/13/2015

2009 2010 2011 2012 2013 2014 2015E 2016E

Revenues*:

AT&T (T) 63,514 61,202 59,765 59,566 58,814 58,425 58,117 58,210

Verizon (VZ) 46,080 42,919 40,682 39,780 39,223 38,429 38,232 38,401

Sprint (S) 5,629 5,040 4,326 3,881 3,537 2,916 2,627 2,499

CenturyLink (CTL) 20,841 19,431 18,692 18,376 18,095 18,031 18,058 18,289

Windstream (WIN) 6,259 6,239 6,140 5,988 5,851 5,838 5,887

Level 3 (LVLT-TWTC) 6,231 6,140 6,318 6,376 7,850 8,168 8,349 8,763

Comcast (CMCSA) 33,742 35,363 37,226 39,604 41,836 44,140 46,047 48,151


Charter Communications (CHTR) 7,032 7,208 8,017 8,419 9,108 9,608 10,108

Total 206,216 202,254 200,131 203,126 205,882 207,879 210,557 215,278

Growth YoY % (2.3%) (1.9%) (1.0%) 1.5% 1.4% 1.0% 1.3% 2.2%


AT&T (T) 11,407 11,166 10,513 8,932 10,037 10,050 7,558 7,570

Verizon (VZ) 8,892 7,473 6,399 6,342 6,229 5,750 5,664 5,607

Sprint (S) 258 231 158 242 320 282 158 150

CenturyLink (CTL) 2,450 2,555 3,151 2,858 3,001 3,026 3,034 3,073

Windstream (WIN) 386 647 892 1,049 812 764 817 883

Level 3 487 603 628 743 1,262 1,256 1,269 1,315

Comcast (CMCSA) 5,037 4,853 4,806 4,921 5,403 6,154 6,711 7,077



Total 32,148 31,667 30,795 29,998 32,116 33,599 31,729 32,393

Growth YoY % (13.5%) (1.5%) (2.8%) (2.6%) 7.1% 4.6% (5.6%) 2.1%

Cap-ex as % of Revenues 15.6% 15.7% 15.4% 14.8% 15.6% 16.2% 15.1% 15.0%

2009 2010 2011 2012 2013 2014 2015E 2016E

EBITDA

AT&T (T) 21,162 20,194 18,780 18,407 17,176 15,954 15,692 15,717

Verizon (VZ) 11,103 9,707 9,417 8,803 8,700 8,917 9,137 9,332

Sprint (S) 1,221 1,090 800 649 494 85 79 75

CenturyLink (CTL) 8,378 8,303 7,799 7,656 7,361 7,043 6,919 6,976

Windstream (WIN) 2,406 2,430 2,384 2,318 2,144 2,137 2,151

Level 3 1,258 1,239 1,297 1,432 2,218 2,443 2,630 2,936

Comcast (CMCSA) 13,767 14,561 15,288 16,255 17,205 18,112 18,854 19,806



Total 67,796 66,968 65,713 66,274 66,399 66,116 67,364 69,693


Unlevered FCF

AT&T (T) 9,755 9,028 8,267 9,475 7,139 5,904 8,133 8,147

Verizon (VZ) 2,211 2,234 3,018 2,461 2,471 3,167 3,474 3,724

Sprint (S) 963 859 642 407 174 (197) (79) (75)

CenturyLink (CTL) 5,928 5,749 4,648 4,798 4,360 4,017 3,885 3,904

Windstream (WIN) 1,759 1,538 1,335 1,506 1,381 1,320 1,268

Level 3 771 636 669 689 894 1,187 1,361 1,621

Comcast (CMCSA) 8,730 9,708 10,482 11,334 11,802 11,958 12,143 12,728


Charter Communications (CHTR) 1,384 3,701 1,048 1,094 969 969 969

Total 34,625 35,301 37,254 36,276 34,222 32,517 35,635 37,300




33

Exhibit 22: Annual Wireless Industry Cash Flow Model

Last Updated: 3/13/2015

2009 2010 2011 2012 2013 2014 2015E 2016E

Revenues*:

AT&T Mobility (T) 53,504 58,500 63,215 66,763 69,899 73,992 76,063 78,332

Verizon Wireless (VZ) 62,131 64,187 70,154 75,868 81,023 87,646 91,887 95,222

Sprint Nextel (S) 27,786 28,597 30,301 32,334 32,767 32,809 32,171 32,356

T-Mobile (TMUS) 21,531 21,347 20,618 19,719 26,130 29,564 32,354 33,727

LEAP (LEAP) 2,481 2,697 3,071 3,142 2,898

MetroPCS (Acquired by TMUS) 3,481 4,071 4,846 5,101

Clearwire (CLWR) ** 274 558 1,259 1,265

Total 171,188 179,957 193,464 204,192 212,717 224,011 232,475 239,636

Growth YoY % 4.9% 5.1% 7.5% 5.5% 4.2% 5.3% 3.8% 3.1%

Wireless Revenue as % of Total 47% 48% 50% 50% 51% 52% 52% 53%


AT&T Mobility (T) 5,928 9,136 9,764 10,796 11,191 11,383 10,649 10,966


Sprint Nextel (S) 1,161 1,444 2,416 4,884 6,833 5,176 6,112 6,148

T-Mobile (TMUS) 3,692 2,819 2,729 2,901 4,241 4,317 4,702 4,781

LEAP (LEAP) 700 399 441 434 133


Clearwire (CLWR) ** 1,540 2,644 246 183

Total 21,004 25,466 25,459 28,944 31,823 31,391 32,652 33,418

Growth YoY % 2.9% 21.2% (0.0%) 13.7% 9.9% (1.4%) 4.0% 2.3%

Wireless Cap-ex as % of Total 39.8% 45.5% 46.3% 49.1% 49.8% 48.3% 50.7% 50.8%

Cap-ex as % of wireless Revenues 12.3% 14.2% 13.2% 14.2% 15.0% 14.0% 14.0% 13.9%

2009 2010 2011 2012 2013 2014 2015E 2016E

EBITDA

AT&T Mobility (T) 19,873 21,754 21,933 23,595 25,391 25,628 26,808 27,651


Sprint Nextel (S) 5,198 4,531 4,267 4,147 4,948 6,034 6,542 6,888

T-Mobile (TMUS) 5,919 5,478 5,310 4,886 5,317 5,636 7,137 8,234

LEAP (LEAP) 485 525 562 597 408

MetroPCS (Acquired by TMUS) 956 1,176 1,331 1,512

Clearwire (CLWR) ** (871) (1,364) (381) (157)

Total 56,096 58,560 59,512 64,308 70,263 72,517 77,440 81,339


Unlevered FCF

AT&T Mobility (T) 13,945 12,618 12,169 12,799 14,200 14,245 16,159 16,685


Sprint Nextel (S) 4,037 3,087 1,851 (737) (1,885) 858 430 741

T-Mobile (TMUS) 2,227 2,659 2,581 1,985 1,076 1,319 2,435 3,454

LEAP (LEAP) (214) 126 121 162 275


Clearwire (CLWR) ** (2,411) (4,007) (626) (340)

Total 35,093 33,095 34,053 35,364 38,440 41,126 44,789 47,920



Notes:

* Includes equipment sales.

** Part of Sprint as of 3Q13


34

Exhibit 23: Annual Wireless Statistics Model

Last Updated: 2/20/2015 Year Over Year % Change CAGR

Subscribers (000s): 2009 2010 2011 2012 2013 2014 2015E 2016E 2009 2010 2011 2012 2013 2014 2015E 2016E '11-16

Net-Adds

Verizon Wireless 5,552 3,673 6,421 7,663 4,908 5,783 5,200 4,600 (2.6%) (33.8%) 74.8% 19.3% (36.0%) 17.8% (10.1%) (11.5%) (6.5%)

AT&T 7,278 8,853 7,699 3,764 2,721 5,608 5,100 4,500 8.6% 21.6% (13.0%) (51.1%) (27.7%) 106.1% (9.1%) (11.8%) (10.2%)

Sprint Nextel* (1,132) 1,777 5,111 605 (2,221) 575 2,372 2,722 NM NM NM NM NM NM NM NM (12%)

T-Mobile USA 1,033 (55) (549) 203 4,408 8,334 5,000 3,800 (64.9%) NM NM NM NM NM NM NM NM

MetroPCS 1,273 1,516 1,191 (459) (9.4%) 19.1% (21.4%) NM

Leap Wireless 1,110 241 417 (637) (745) 17.8% (78.3%) 73.0% (252.8%) 17.0%

Others (1,760) (239) (339) (1,356) (309) (1,291) (1,395) (1,395)

Total U.S. - Net-Adds 13,353 15,766 19,951 9,783 8,762 19,009 16,277 14,227 (5.2%) 18.1% 26.5% (51.0%) (10.4%) 116.9% (14.4%) (12.6%) (6.5%)

Incremental Market Share

Verizon Wireless 41.6% 23.3% 32.2% 78.3% 56.0% 30.4% 31.9% 32.3%

AT&T 54.5% 56.2% 38.6% 38.5% 31.1% 29.5% 31.3% 31.6%

Sprint Nextel (8.5%) 11.3% 25.6% 6.2% (25.3%) 3.0% 14.6% 19.1%

T-Mobile USA 7.7% (0.3%) (2.8%) 2.1% 50.3% 43.8% 30.7% 26.7%

MetroPCS 9.5% 9.6% 6.0% (4.7%) 0.0%

Leap Wireless 8.3% 1.5% 2.1% (6.5%) (8.5%)

Others (13.2%) (1.5%) (1.7%) (13.9%) (3.5%) (6.8%) (8.6%) (9.8%)

Ending Subs

Verizon Wireless 98,573 102,246 108,667 116,330 121,238 127,021 132,221 136,821 36.8% 3.7% 6.3% 7.1% 4.2% 4.8% 4.1% 3.5% 4.7%

AT&T 85,120 95,536 103,247 106,957 110,376 120,554 125,654 130,154 10.5% 12.2% 8.1% 3.6% 3.2% 9.2% 4.2% 3.6% 4.7%

Sprint Nextel 48,133 49,910 55,021 55,626 53,405 53,980 56,352 59,074 19.5% 3.7% 10.2% 1.1% (4.0%) 1.1% 4.4% 4.8% 1.4%

T-Mobile USA 33,791 33,734 33,185 33,389 46,684 55,018 60,018 63,818 3.2% (0.2%) (1.6%) 0.6% 39.8% 17.9% 9.1% 6.3% 14.0%

MetroPCS 6,640 8,155 9,347 8,888 23.7% 22.8% 14.6% (4.9%)

Leap Wireless 4,954 5,518 5,934 5,297 4,552 28.8% 11.4% 7.5% (10.7%) (14.1%)

Others 17,636 17,971 17,633 16,277 15,968 14,677 13,282 11,887 (54.7%) 1.9% (1.9%) (7.7%) (1.9%) (8.1%) (9.5%) (10.5%) (7.6%)

Total U.S. - Ending Subs 294,846 313,070 333,034 342,764 352,223 371,250 387,527 401,754 9.1% 6.2% 6.4% 2.9% 2.8% 5.4% 4.4% 3.7% 3.8%

Penetration 95.6% 100.7% 106.3% 108.4% 110.3% 115.2% 119.2% 122.4%

Market Share


AT&T 28.9% 30.5% 31.0% 31.2% 31.3% 32.5% 32.4% 32.4%

Sprint Nextel 16.3% 15.9% 16.5% 16.2% 15.2% 14.5% 14.5% 14.7%

T-Mobile USA 11.5% 10.8% 10.0% 9.7% 13.3% 14.8% 15.5% 15.9%

MetroPCS 2.3% 2.6% 2.8% 2.6% 0.0% 0.0% 0.0% 0.0%

Leap Wireless 1.7% 1.8% 1.8% 1.5% 1.3% 0.0% 0.0% 0.0%

Others 6.0% 5.7% 5.3% 4.7% 4.5% 4.0% 3.4% 3.0%

Total 100.0% 100.0% 100.0% 100.0% 100.0% 100.0% 100.0% 100.0%

Churn


AT&T 1.5% 1.3% 1.4% 1.4% 1.4% 1.5% 1.4% 1.3%

Sprint Nextel 2.7% 2.8% 2.5% 2.5% 2.8% 2.8% 2.7% 2.4%

T-Mobile USA 3.2% 3.4% 3.5% 3.2% 3.3% 3.1% 3.2% 3.2%

MetroPCS 5.5% 3.6% 3.8% 3.5%

Leap Wireless 4.5% 4.8% 3.8% 4.3% 3.9%

Total ARPU

Verizon Wireless $46.65 $46.72 $46.56 $47.42 $48.56 $49.01 $47.91 $46.90 (9.5%) 0.2% (0.3%) 1.8% 2.4% 0.9% (2.3%) (2.1%) 0.1%

AT&T $50.68 $49.68 $47.77 $46.99 $47.41 $43.79 $40.54 $40.09 0.2% (2.0%) (3.8%) (1.6%) 0.9% (7.6%) (7.4%) (1.1%) (3.4%)

Sprint Nextel $44.18 $44.48 $43.69 $43.33 $44.55 $42.35 $38.15 $36.35 (3.6%) 0.7% (1.8%) (0.8%) 2.8% (4.9%) (9.9%) (4.7%) (3.6%)

T-Mobile USA $47.31 $46.31 $45.84 $43.09 $38.81 $36.69 $35.86 $35.75 (7.7%) (2.1%) (1.0%) (6.0%) (9.9%) (5.5%) (2.3%) (0.3%) (4.9%)

MetroPCS* $42.45 $40.79 $41.32 $41.14 (3.3%) (3.9%) 1.3% (0.4%)

Leap Wireless $42.14 $39.65 $41.19 $42.36 $44.88 (3.3%) (5.9%) 3.9% 2.8% 6.0%

Data ARPU

Verizon Wireless $15.66 $18.74 $21.69 $24.74 $27.21 $29.10 $31.44 $31.44 21.8% 19.6% 15.8% 14.1% 10.0% 7.0% 8.0% 0.0% 7.7%

% of Total ARPU 33.6% 40.1% 46.6% 52.2% 56.0% 59.4% 65.6% 67.0%

AT&T $14.71 $16.86 $18.52 $20.59 $22.44 $24.18 $25.69 $25.69 22.1% 14.7% 9.8% 11.2% 9.0% 7.7% 6.3% 0.0% 6.8%

% of Total ARPU 29.0% 33.9% 38.8% 43.8% 47.3% 55.2% 63.4% 64.1%

Sprint Nextel $15.89 $18.28 $20.11 $21.92 $23.89 $25.70 $26.33 $26.33 23.4% 15.0% 10.0% 9.0% 9.0% 7.6% 2.5% 0.0% 5.5%

% of Total ARPU 36.0% 41.1% 46.0% 50.6% 53.6% 60.7% 69.0% 72.4%

T-Mobile USA $9.81 $11.93 $15.56 $17.35 $19.08 $20.75 $22.67 $22.67 11.2% 21.6% 30.4% 11.5% 10.0% 8.8% 9.3% 0.0% 7.8%

% of Total ARPU 20.7% 25.8% 33.9% 40.3% 49.2% 56.6% 63.2% 63.4%

Source: Company reports, Oppenheimer & Co. Inc. estimates.


35

Exhibit 24: Annual Wireless Statistics Granular Model Last Updated: 2/20/2015 Year Over Year % Change CAGR

Subscribers (000s): 2009 2010 2011 2012 2013 2014 2015E 2016E 2009 2010 2011 2012 2013 2014 2015E 2016E '11-16

Ending Postpaid Subs

Verizon 82,194 83,125 87,382 92,530 96,771 102,079 106,779 110,779 22.7% 1.1% 5.1% 5.9% 4.6% 5.5% 4.6% 3.7% 4.9%

AT&T 64,627 68,041 69,309 70,497 72,638 75,931 78,131 80,131 8.3% 5.3% 1.9% 1.7% 3.0% 4.5% 2.9% 2.6% 2.9%

Sprint Postpaid 33,967 33,112 33,014 31,877 30,837 29,904 30,554 31,554 (7.4%) (2.5%) (0.3%) (3.4%) (3.3%) (3.0%) 2.2% 3.3% (0.9%)

Sprint Affiliates 276 296 316 336 376 498 620 742 (70.2%) 7.2% 6.8% 6.3% 11.9% 32.4% 24.5% 19.7% 18.6%

T-Mobile 26,765 24,574 22,367 20,293 22,299 27,185 29,885 31,885 (0.2%) (8.2%) (9.0%) (9.3%) 9.9% 21.9% 9.9% 6.7% 7.3%

US Cellular 5,482 5,416 5,299 5,134 4,267 4,175 4,075 3,975 6.8% (1.2%) (2.2%) (3.1%) (16.9%) (2.2%) (2.4%) (2.5%) (5.6%)

Alltel 0 0 0

Clearwire 642 1,529 2,563 2,500 2,958 2,958 2,958 2,958 35.4% 138.1% 67.7% (2.5%) 18.3%

Others 2,115 2,623 2,223 1,823 1,423 1,023 623 223 (38.4%) 24.0% (15.2%) (18.0%) (21.9%) (28.1%) (39.1%) (64.2%) (36.8%)

Total 216,068 218,716 222,474 224,990 231,569 243,753 253,625 262,247 1.8% 1.2% 1.7% 1.1% 2.9% 5.3% 4.1% 3.4% 3.3%

% total 73% 70% 67% 66% 66% 66% 65% 65%

Ending Prepaid/Wholesale Subs

Verizon 16,379 19,121 21,285 23,800 24,467 24,942 25,442 26,042 NM 16.7% 11.3% 11.8% 2.8% 1.9% 2.0% 2.4% 4.1%

AT&T 20,493 27,495 33,938 36,460 37,738 44,623 47,523 50,023 18.1% 34.2% 23.4% 7.4% 3.5% 18.2% 6.5% 5.3% 8.1%

Sprint 13,890 16,502 21,691 23,413 24,141 25,527 27,127 28,727 19.1% 18.8% 31.4% 7.9% 3.1% 5.7% 6.3% 5.9% 5.8%

T-Mobile 7,026 7,287 8,389 13,096 24,385 27,833 30,133 31,933 18.0% 3.7% 15.1% 56.1% 86.2% 14.1% 8.3% 6.0% 30.6%

Leap Wireless 4,954 5,518 5,934 5,297 4,552 0 0 0 28.8% 11.4% 7.5% (10.7%) (14.1%)

MetroPCS 6,640 8,155 9,347 8,888 23.7% 22.8% 14.6% (4.9%)

US Cellular 945 656 592 664 1,185 1,186 1,091 996 (11.0%) (30.6%) (9.8%) 12.2% 78.5% 0.1% (8.0%) (8.7%) 11.0%

Alltel 0 0 0

Others 8,737 9,620 9,384 6,155 4,186 3,386 2,586 1,786

Total 79,064 94,354 110,560 117,773 120,654 127,497 133,902 139,507 41.2% 19.3% 17.2% 6.5% 2.4% 5.7% 5.0% 4.2% 4.8%

% total 27% 30% 33% 34% 34% 34% 35% 35%

Total Gross Additions

Verizon Wireless 22,357 23,291 23,691 25,035 24,380 26,905 27,410 27,028 38.5% 4.2% 1.7% 5.7% (2.6%) 10.4% 1.9% (1.4%) 2.7%

AT&T 21,540 22,986 23,925 20,810 20,628 25,963 25,781 24,453 1.7% 6.7% 4.1% (13.0%) (0.9%) 25.9% (0.7%) (5.2%) 0.4%

Sprint Nextel 14,085 18,206 20,817 17,230 15,998 18,647 20,070 19,678 51.3% 29.3% 14.3% (17.2%) (7.1%) 16.6% 7.6% (2.0%) (1.1%)

T-Mobile USA 13,939 13,599 13,660 13,087 21,333 27,579 27,119 27,370 0.6% (2.4%) 0.5% (4.2%) 63.0% 29.3% (1.7%) 0.9% 14.9%

MetroPCS 5,324 4,756 5,291 3,353 33.9% (10.7%) 11.3% (36.6%)

Leap 3,523 3,234 3,016 2,347 1,521 42.6% (8.2%) (6.8%) (22.2%) (35.2%)

Other 7,591 6,065 5,493 4,112 4,569 2,779 2,256 1,742

Total 79,512 84,147 87,586 80,274 86,908 101,873 102,637 100,271 6.1% 5.8% 4.1% (8.3%) 8.3% 17.2% 0.8% (2.3%) 2.7%

Total Postpaid Gross Additions

Verizon Wireless 14,363 12,645 13,910 14,829 15,068 17,796 17,858 17,053 12.7% (12.0%) 10.0% 6.6% 1.6% 18.1% 0.3% (4.5%) 4.2%

AT&T 12,565 10,880 11,057 10,512 10,861 12,520 12,830 12,445 (0.0%) (13.4%) 1.6% (4.9%) 3.3% 15.3% 2.5% (3.0%) 2.4%

Sprint Nextel 5,391 6,927 7,267 6,725 6,124 6,973 8,086 8,080 (11.7%) 28.5% 4.9% (7.5%) (8.9%) 13.9% 16.0% (0.1%) 2.1%

T-Mobile USA 7,548 6,211 5,400 3,938 6,317 9,607 8,257 7,930 (10.0%) (17.7%) (13.1%) (27.1%) 60.4% 52.1% (14.1%) (4.0%) 8.0%

Other 1,571 991 850 642 (221) 587 471 376

Total 41,438 37,655 38,484 36,645 38,149 47,483 47,503 45,885 (6.1%) (9.1%) 2.2% (4.8%) 4.1% 24.5% 0.0% (3.4%) 3.6%

% of Total 52% 45% 44% 46% 44% 47% 46% 46%

Total Prepaid/Wholesale Gross Additions

Verizon Wireless 7,994 10,646 9,782 10,206 9,312 9,109 9,552 9,974 135.0% 33.2% (8.1%) 4.3% (8.8%) (2.2%) 4.9% 4.4% 0.4%

AT&T 8,975 12,106 12,868 10,298 9,767 13,443 12,951 12,008 4.2% 34.9% 6.3% (20.0%) (5.2%) 37.6% (3.7%) (7.3%) (1.4%)

Sprint Nextel 8,694 11,279 13,550 10,505 9,874 11,673 11,984 11,597 171.4% 29.7% 20.1% (22.5%) (6.0%) 18.2% 2.7% (3.2%) (3.1%)

T-Mobile USA 6,391 7,387 8,260 9,149 15,016 17,972 18,863 19,440 16.9% 15.6% 11.8% 10.8% 64.1% 19.7% 5.0% 3.1% 18.7%

MetroPCS 5,324 4,756 5,291 3,353 33.9% (10.7%) 11.3% (36.6%)

Leap 3,523 3,234 3,016 2,347 1,521 42.6% (8.2%) (6.8%) (22.2%) (35.2%) (100.0%)

Other 6,020 5,074 4,643 3,470 4,790 2,192 1,785 1,366

Total 46,921 54,481 57,409 49,328 50,280 54,390 55,135 54,385 25.9% 16.1% 5.4% (14.1%) 1.9% 8.2% 1.4% (1.4%) (1.1%)

% of Total Wireless Industry Subscribers59% 65% 66% 61% 58% 53% 54% 54%

Postpaid Net Additions

Verizon Wireless 4,135 2,550 4,252 5,024 4,118 5,482 4,700 4,000 (22.8%) (38.3%) 66.7% 18.2% (18.0%) 33.1% (14.3%) (14.9%) (1.2%)

AT&T 4,199 2,153 1,429 1,438 1,776 3,290 2,200 2,000 (7.2%) (48.7%) (33.6%) 0.6% 23.5% 85.2% (33.1%) (9.1%) 7.0%

Sprint Nextel (Direct Postpaid) (3,546) (855) (98) (1,137) (2,209) (933) 650 1,000 (12.9%) (75.9%) NM NM NM (57.8%) NM NM NM

T-Mobile USA (41) (1,068) (2,205) (2,074) 2,006 4,886 2,700 2,000 NM NM NM NM (196.7%) 143.6% (44.7%) (25.9%) NM

Other (352) (433) (497) (545) (1,227) (302) (378) (378)

Total 4,395 2,347 2,881 2,706 4,464 12,423 9,872 8,622 (45.0%) (46.6%) 22.8% (6.1%) 65.0% 178.3% (20.5%) (12.7%) 24.5%

% of Total Wireless Additions 32.9% 14.9% 14.4% 27.7% 50.9% 65.4% 60.6% 60.6%

Wholesale/Prepaid Net Additions

Verizon Wireless 1,417 1,123 2,169 2,639 790 301 500 600 NM (20.7%) 93.1% 21.7% (70.1%) (61.9%) 66.1% 20.0% (22.7%)

AT&T 3,079 6,700 6,270 2,326 945 2,318 2,900 2,500 41.5% 117.6% (6.4%) (62.9%) (59.4%) 145.3% 25.1% (13.8%) (16.8%)

Sprint Nextel (Direct Prepaid) 2,552 1,638 2,512 798 (145) (683) 400 500 NM (35.8%) 53.4% (68.2%) (118.2%) NM (158.6%) 25.0% (27.6%)

T-Mobile USA 1,074 1,013 1,656 2,277 2,402 3,448 2,300 1,800 (4.1%) NM 63.5% 37.5% 5.5% 43.5% (33.3%) (21.7%) 1.7%

MetroPCS 1,273 1,516 1,191 (459) (9.4%) 19.1% (21.4%) NM

Leap 1,110 241 417 (637) (745) 17.8% (78.3%) 73.0% (252.8%) 17.0% (100.0%)

Other (1,408) 1,188 2,855 133 1,052 1,202 305 205

Total 9,096 13,419 17,070 7,077 4,298 6,586 6,405 5,605 49.3% 47.5% 27.2% (58.5%) (39.3%) 53.2% (2.7%) (12.5%) (20.0%)

% of Total 68.1% 85.1% 85.6% 72.3% 49.1% 34.6% 39.4% 39.4%



36

Exhibit 25: Annual Video Subscriber Model

Last Updated: 03/13/2015 CAGR

($ in millions, except subscribers) 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015E 2016E 10-15

U.S. Households 113.9 115.3 116.7 117.3 117.9 117.9 117.9 117.9 119.0 120.2 121.4

MSO Video Subscribers 65.6 64.9 63.5 62.2 60.4 59.1 57.8 56.0 55.9 54.9 54.0 (1.9%)

YOY Growth 0.31% -1.07% -2.09% -2.08% -2.94% -2.11% -2.20% -3.16% -0.11% -1.83% -1.59%

Satellite Video Subs 29.1 30.6 31.3 32.7 33.4 33.9 34.1 34.3 34.3 34.4 34.5 0.6%

YOY Growth 6.94% 5.34% 2.25% 4.35% 2.13% 1.49% 0.85% 0.50% 0.06% 0.23% 0.23%

Telco Video Subs 0.2 1.2 3.0 4.9 6.5 8.0 9.3 10.7 11.6 12.6 13.7 14.3%

YOY Growth 459.0% 152.4% 66.3% 31.1% 23.3% 16.3% 15.8% 8.1% 8.8% 8.7%

MSO Market Share 69.3% 68.0% 65.0% 62.3% 60.3% 58.6% 57.1% 55.4% 54.9% 53.9% 52.8%

Total U.S. Video Subs 94.9 96.7 97.8 99.8 100.2 100.9 101.2 101.0 101.8 101.9 102.2 0.3%

YOY Growth 2.5% 1.9% 1.2% 2.0% 0.4% 0.7% 0.3% (0.2%) 0.8% 0.1% 0.3%

Penetration 83% 84% 84% 85% 85% 86% 86% 86% 86% 85% 84%

Cable Video Revenue

Video ARPU $57.00 $61.00 $64.00 $67.20 $69.89 $72.68 $74.86 $76.36 $77.89 $79.45 $81.04 2.6%

Total Video Revenue $44,870 $47,507 $48,800 $50,173 $50,646 $51,559 $51,940 $51,303 $52,271 $52,339 $52,537 0.7%

EBITDA Margin 37.5% 38.3% 38.5% 38.3% 38.0% 37.5% 36.0% 35.0% 34.0% 33.0% 32.0% (2.8%)

EBITDA $16,826 $18,195 $18,788 $19,216 $19,245 $19,335 $18,698 $17,956 $17,772 $17,272 $16,812 (2.1%)

Broadband Revenue

Broadband Subs 30.1 34.0 37.1 39.4 41.9 44.2 46.5 47.9 48.8 48.8 48.8 3.1%

ARPU $42.96 $43.10 $42.24 $41.82 $42.86 $44.15 $45.69 $47.29 $48.95 $50.66 $52.43 3.4%

Total Broadband Revenue $15,501 $17,579 $18,807 $19,752 $21,546 $23,401 $25,501 $27,172 $28,677 $29,680 $30,719 6.6%

EBITDA Margin 50.0% 55.0% 55.0% 55.0% 55.0% 56.0% 57.0% 58.0% 59.0% 60.0% 61.0%

EBITDA $7,751 $9,668 $10,344 $10,864 $11,850 $13,105 $14,535 $15,760 $16,919 $17,808 $18,739 8.5%

Telephony Revenue

Telephony Subs 6.6 12.9 18.0 20.7 21.9 23.2 24.5 25.4 26.1 26.3 26.6 3.7%

ARPU $43.00 $42.14 $41.30 $40.47 $39.26 $38.47 $38.47 $38.47 $38.47 $38.47 $38.47 (0.4%)

Total Telephony Revenue $3,404 $6,539 $8,929 $10,060 $10,326 $10,722 $11,294 $11,709 $12,039 $12,142 $12,280 3.3%

EBITDA Margin 8.0% 12.0% 18.0% 25.0% 30.0% 33.0% 33.0% 34.0% 35.0% 35.0% 35.0% 3.1%

EBITDA $272 $785 $1,607 $2,515 $3,098 $3,538 $3,727 $3,981 $4,214 $4,250 $4,298 6.5%

Other Revenue(1)$2,244 $2,375 $2,440 $3,010 $3,545 $4,254 $5,105 $6,126 $7,351 $8,822 $10,586 20.0%

EBITDA Margin 40.0% 40.0% 40.0% 40.0% 40.0% 40.0% 40.0% 40.0% 40.0% 40.0% 40.0%

EBITDA $897 $950 $976 $1,204 $1,418 $1,702 $2,042 $2,450 $2,941 $3,529 $4,234 20.0%

Total MSO Revenue $66,019 $74,000 $78,976 $82,996 $86,063 $89,936 $93,839 $96,309 $100,338 $102,983 $106,122 3.7%

YOY Growth 13.8% 12.1% 6.7% 5.1% 3.7% 4.5% 4.3% 2.6% 4.2% 2.6% 3.0%

Total ARPU $83.87 $95.02 $103.57 $111.16 $118.76 $126.79 $135.26 $143.35 $149.51 $156.32 $163.69 5.6%

YOY Growth 13.5% 13.3% 9.0% 7.3% 6.8% 6.8% 6.7% 6.0% 4.3% 4.6% 4.7%

(1) Includes Advertising, Installation, Content, Wireless etc.

Cash Flow Analysis

EBITDA $25,747 $29,598 $31,715 $33,799 $35,612 $37,679 $39,003 $40,147 $41,846 $42,858 $44,083 3.8%

EBITDA Margin 39.0% 40.0% 40.2% 40.7% 41.4% 41.9% 41.6% 41.7% 41.7% 41.6% 41.5%

EBITDA Growth 16.6% 15.0% 7.2% 6.6% 5.4% 5.8% 3.5% 2.9% 4.2% 2.4% 2.9%

CAPEX $11,700 $15,000 $13,426 $13,694 $13,770 $13,490 $13,137 $12,520 $13,044 $13,388 $13,796 (0.6%)

As % of revenue 17.7% 20.3% 17.0% 16.5% 16.0% 15.0% 14.0% 13.0% 13.0% 13.0% 13.0%

Cash Flow (Unlevered, Pre Tax) $14,047 $14,598 $18,289 $20,105 $21,841 $24,189 $25,865 $27,627 $28,802 $29,471 $30,287 6.2%

YOY Growth 26.8% 3.9% 25.3% 9.9% 8.6% 10.7% 6.9% 6.8% 4.3% 2.3% 2.8%

Total Debt $77,191 $74,271 $70,613 $66,603 $62,235 $57,397 $52,224 $46,699 $40,938 $35,044 $28,987 (10.9%)

Debt to EBITDA 3.0X 2.5X 2.2X 2.0X 1.7X 1.5X 1.3X 1.2X 1.0X 0.8X 0.7X

Source: Company Reports & Opco Research Estimates.


37

Exhibit 26: Annual Video Subscriber Model Bottom-Up Last Updated: 03/13/2015 YoY Percentage Change (%) CAGR

2009 2010 2011 2012 2013 2014 2015E 2016E 2009 2010 2011 2012 2013 2014 2015E 2016E '10-'15

Total Pay-TV Subscribers (mil)

Cable 62.218 60.389 59.113 57.816 55.987 55.925 54.900 54.027 -2.1% -2.9% -2.1% -2.2% -3.2% -0.1% -1.8% -1.6% -1.9%

DBS 32.660 33.356 33.852 34.140 34.310 34.330 34.410 34.490 4.3% 2.1% 1.5% 0.9% 0.5% 0.1% 0.2% 0.2% 0.6%

Telco Facilities-Based 4.926 6.990 8.522 9.715 11.282 12.512 13.392 14.552 66.3% 41.9% 21.9% 14.0% 16.1% 10.9% 7.0% 8.7% 13.9%

Total 99.804 100.736 101.487 101.670 101.579 102.767 102.702 103.069 2.0% 0.9% 0.7% 0.2% -0.1% 1.2% -0.1% 0.4% 0.4%

Total Pay-TV Subscribers (market share %)

Cable 62.3% 59.9% 58.2% 56.9% 55.1% 54.4% 53.5% 52.4%

DBS 32.7% 33.1% 33.4% 33.6% 33.8% 33.4% 33.5% 33.5%

Telco Facilities-Based 4.9% 6.9% 8.4% 9.6% 11.1% 12.2% 13.0% 14.1%

Total 100.0% 100.0% 100.0% 100.0% 100.0% 100.0% 100.0% 100.0%

Total Pay-TV Subscriber Additions (mil)

Cable (1.324) (1.829) (1.357) (1.298) (1.829) (0.062) (1.025) (0.873) -2.5% 38.2% -25.8% -4.4% 41.0% -96.6% nm -14.9%

DBS 1.361 0.696 0.496 0.288 0.170 0.020 0.080 0.080 97.8% -48.9% -28.7% -41.9% -41.0% -88.2% nm 0.0%

Telco Facilities-Based 1.963 2.064 1.531 1.193 1.568 1.230 0.880 1.160 9.7% 5.2% -25.8% -22.1% 31.4% -21.6% -28.4% 31.8%

Total 2.000 0.931 0.670 0.183 (0.091) 1.188 (0.065) 0.367 78.7% -53.4% -28.0% -72.6% nm nm nm nm

Total Pay-TV Subscriber Additions (market share %)

Cable -66.2% -196.4% -202.4% -707.2% 1999.5% -5.2% 1573.8% -237.5%

DBS 68.0% 74.7% 74.0% 157.0% -185.8% 1.7% -122.8% 21.8%

Telco Facilities-Based 98.1% 221.7% 228.4% 650.3% -1713.7% 103.5% -1351.0% 315.7%

Total 100.0% 100.0% 100.0% 100.0% 100.0% 100.0% 100.0% 100.0%

Telco Facilities-based Video Subscribers (mil)

AT&T 2.065 2.987 3.791 4.536 5.460 5.943 6.593 7.343 97.6% 44.6% 26.9% 19.7% 20.4% 8.8% 10.9% 11.4% 17.2%

Verizon 2.861 3.472 4.173 4.726 5.262 5.649 6.019 6.369 49.2% 21.4% 20.2% 13.3% 11.3% 7.4% 6.5% 5.8% 11.6%

CenturyLink 0.000 0.000 0.000 0.106 0.175 0.242 0.270 0.300 na na na na 65.1% 38.3% 11.6% 11.1%

Cincinatti Bell (Fioptic Video) 0.040 0.055 0.074 0.091 0.100 0.120 23.2% 9.4% 20.0%

Frontier Comm. 0.000 0.531 0.558 0.347 0.385 0.587 0.410 0.420 na na 4.9% -37.8% 11.2% 52.2% -30.1% 2.4%

Total 4.926 6.990 8.561 9.770 11.357 12.512 13.392 14.552 66.3% 41.9% 22.5% 14.1% 16.2% 10.2% 7.0% 8.7% 13.9%

Telco Facilities-based Video Subscriber Additions (mil)

AT&T 1.020 0.922 0.804 0.745 0.924 0.483 0.650 0.750 25.3% -9.6% -12.8% -7.3% 24.0% -47.7% 34.6% 15.4%

Verizon 0.943 0.611 0.701 0.553 0.536 0.387 0.370 0.350 -3.3% -35.2% 14.7% -21.1% -3.1% -27.8% -4.4% -5.4%

CenturyLink 0.000 0.000 0.000 0.000 0.069 0.067 0.028 0.030 na na na na na -2.9% -58.2% 7.1%

Cincinatti Bell (Fioptic Video) 0.016 0.019 0.017 0.009 0.020 -9.9% -50.0% 132.6%

Frontier Comm. 0.000 0.538 0.026 (0.211) 0.039 0.201 (0.177) 0.010 na na -95.1% nm nm nm nm nm

Total 1.963 2.071 1.531 1.103 1.587 1.155 0.880 1.160 9.7% 5.5% -26.1% -28.0% 43.9% -27.2% -23.8% 31.8%

Cable Basic Subscribers (mil)

Comcast 23.559 22.790 22.331 21.994 21.689 22.382 22.247 22.197 -2.6% -3.3% -2.0% -1.5% -1.4% 3.2% -0.6% -0.2% -0.5%

Time Warner 12.859 12.422 12.061 12.218 11.393 10.992 10.473 10.013 -1.6% -3.4% -2.9% 1.3% -6.8% -3.5% -4.7% -4.4% -3.4%

Cox 5.282 5.176 5.099 5.048 4.997 4.947 4.898 4.849 -1.5% -2.0% -1.5% -1.0% -1.0% -1.0% -1.0% -1.0% -1.1%

Charter 4.824 4.520 4.440 4.286 4.177 4.160 4.035 3.914 -4.4% -6.3% -1.8% -3.5% -2.5% -0.4% -3.0% -3.0% -2.2%

Cablevision 3.063 3.314 3.250 3.197 2.813 2.681 2.627 2.575 -1.4% 8.2% -1.9% -1.6% -12.0% -4.7% -2.0% -2.0% -4.5%

Mediacom 1.238 1.193 1.069 1.000 0.945 0.890 0.846 0.803 -4.3% -3.6% -10.4% -6.5% -5.5% -5.8% -5.0% -5.0% -6.7%

Other 11.393 10.973 10.864 10.073 9.972 9.873 9.774 9.676 -0.8% -3.7% -1.0% -7.3% -1.0% -1.0% -1.0% -1.0% -2.3%

Total 62.218 60.389 59.113 57.816 55.987 55.925 54.900 54.027 -2.1% -2.9% -2.1% -2.2% -3.2% -0.1% -1.8% -1.6% -1.9%

Cable Basic Subscriber Additions (mil)

Comcast (0.623) (0.769) (0.459) (0.337) (0.305) 0.693 (0.135) (0.050) nm 23.4% -40.3% -26.6% -9.5% nm nm -63.0%

Time Warner (0.210) (0.437) (0.442) 0.157 (0.825) (0.401) (0.519) (0.460) 15.4% 108.1% 1.1% nm nm -51.4% 29.4% -11.4%

Cox (0.080) (0.106) (0.078) (0.051) (0.050) (0.050) (0.049) (0.049) 95.7% 32.1% -26.5% -34.3% -1.0% -1.0% -1.0% -1.0%

Charter (0.222) (0.304) (0.080) (0.154) (0.109) (0.017) (0.125) (0.121) 28.2% 36.9% -73.5% 91.5% -29.2% -84.4% nm -3.0%

Cablevision (0.045) 0.251 (0.064) (0.053) (0.384) (0.132) (0.054) (0.053) 200.0% nm nm -17.2% nm -65.6% -59.4% -2.0%

Mediacom (0.055) (0.045) (0.124) (0.069) (0.055) (0.055) (0.045) (0.042) 77.4% -18.2% 175.6% -44.4% -20.3% 0.0% -19.1% -5.0%

Other (0.089) (0.420) (0.110) (0.791) (0.101) (0.100) (0.099) (0.098) -91.4% nm -73.9% nm -87.3% -1.0% -1.0% -1.0%

Total (1.324) (1.829) (1.357) (1.298) (1.829) (0.062) (1.025) (0.873) -2.5% 38.2% -25.8% -4.4% 41.0% -96.6% nm -14.9%

DBS Video Subscribers (mil)

DirecTV 18.560 19.223 19.885 20.084 20.253 20.352 20.392 20.422 5.3% 3.6% 3.4% 1.0% 0.8% 0.5% 0.2% 0.1% 1.2%

DISH 14.100 14.133 13.967 14.056 14.057 13.978 14.018 14.068 3.1% 0.2% -1.2% 0.6% 0.0% -0.6% 0.3% 0.4% -0.2%

Total 32.660 33.356 33.852 34.140 34.310 34.330 34.410 34.490 4.3% 2.1% 1.5% 0.9% 0.5% 0.1% 0.2% 0.2% 0.6%

DBS Video Subscriber Additions (mil)

DirecTV 0.939 0.663 0.662 0.199 0.169 0.099 0.040 0.030 18.9% -29.4% -0.2% -69.9% -15.1% -41.4% -59.6% -25.0%

DISH 0.422 0.033 (0.166) 0.089 0.001 (0.079) 0.040 0.050 nm -92.2% nm nm -98.9% nm nm 25.0%

Total 1.361 0.696 0.496 0.288 0.170 0.020 0.080 0.080 97.8% -48.9% -28.7% -41.9% -41.0% -88.2% nm 0.0%

Source: Company Reports; Oppenheimer & Co, Inc. estimates


38

Exhibit 27: Annual Broadband Subscriber Model Last Updated: 03/13/2015 CAGR

Subscribers (000s): 2009 2010 2011 2012 2013 2014 2015E 2016E 2009 2010 2011 2012 2013 2014 2015E 2016E '11-'16

Cumulative Subscribers

RBOCs 27,821 30,010 30,651 31,033 31,431 31,315 31,780 32,580 5.4% 7.9% 2.1% 1.2% 1.3% (0.4%) 1.5% 2.5% 1.2%

Others 6,143 5,000 4,849 4,901 5,012 5,477 5,525 5,580 7.2% (18.6%) (3.0%) 1.1% 2.3% 9.3% 0.9% 1.0% 2.9%

Total Telco (DSL/Fiber) 33,964 35,010 35,500 35,934 36,443 36,791 37,305 38,160 5.7% 3.1% 1.4% 1.2% 1.4% 1.0% 1.4% 2.3% 1.5%

Total Cable 39,364 41,891 44,431 46,800 48,290 50,772 52,178 53,509 6.1% 6.4% 6.1% 5.3% 3.2% 5.1% 2.8% 2.6% 3.8%

Total Broadband Subscribers 73,328 76,901 79,930 82,734 84,734 87,563 89,483 91,669 5.9% 4.9% 3.9% 3.5% 2.4% 3.3% 2.2% 2.4% 2.8%

Incremental Additions

RBOCs 1,419 3,248 641 382 398 (116) 465 800 1.4% 128.9% (80.3%) (40.4%) 4.2% (129.2%) (499.5%) 72.0% 4.5%

Others 411 (2,202) (151) 52 111 464 49 55 (35.5%) (635.5%) (93.1%) NM NM NM NM NM #NUM!

Total Telco 1,830 1,046 490 434 509 348 514 855 (10.2%) (42.9%) (53.2%) (11.3%) 17.3% (31.7%) 47.6% 66.5% 11.8%

Market Share - Incremental Additions 44.8% 29.3% 16.2% 15.5% 25.5% 12.3% 26.7% 39.1%

Total Cable 2,259 2,527 2,540 2,369 1,490 2,482 1,407 1,331 (27.5%) 11.8% 0.5% (6.7%) (37.1%) 66.5% (43.3%) (5.4%) (12.1%)

Market Share - Incremental Additions 55.2% 70.7% 83.8% 84.5% 74.5% 87.7% 73.3% 60.9%

Total Broadband Net Adds 4,090 3,573 3,030 2,803 2,000 2,830 1,920 2,186 (20.7%) (12.6%) (15.2%) (7.5%) (28.7%) 41.5% (32.1%) 13.8% (6.3%)

Broadband Market Share

RBOCs 37.9% 39.0% 38.3% 37.5% 37.1% 35.8% 35.5% 35.5%

Others 8.4% 6.5% 6.1% 5.9% 5.9% 6.3% 6.2% 6.1%

Total Telco 46.3% 45.5% 44.4% 43.4% 43.0% 42.0% 41.7% 41.6%

Total Cable 53.7% 54.5% 55.6% 56.6% 57.0% 58.0% 58.3% 58.4%

Broadband Penetration - HHs 62.5% 65.2% 67.3% 69.2% 70.4% 72.2% 73.4% 74.9%

YOY Change 3.2% 2.7% 2.1% 1.9% 1.2% 1.8% 1.2% 1.4%

Broadband Penetration - Online HHs 81.5% 83.0% 84.1% 85.4% 86.8% 89.2% 90.6% 92.3%

YOY Change 1.9% 1.5% 1.2% 1.2% 1.4% 2.5% 1.4% 1.7%

Total Narrowband 16,651 15,778 15,065 14,162 12,937 10,596 9,265 7,672 (5.9%) (5.2%) (4.5%) (6.0%) (8.6%) (18.1%) (12.6%) (17.2%) (12.6%)

Narrowband Penetration - HHs 14.2% 13.4% 12.7% 11.8% 10.7% 8.7% 7.6% 6.3%

Narrowband Penetration - Online HHs 18.5% 17.0% 15.9% 14.6% 13.2% 10.8% 9.4% 7.7%

Narrowband Subs Migrating To Broadband 26% 24% 24% 32% 61% 83% 69% 73%

Total U.S. Households (mil.) 117.3 117.9 118.7 119.6 120.4 121.2 121.8 122.5 0.5% 0.5% 0.7% 0.7% 0.7% 1.0% 0.5% 0.5% 0.6%

Total U.S. Computer Households (mil.) 90.0 92.7 95.0 96.9 97.7 98.2 98.7 99.3 3.5% 3.0% 2.5% 2.0% 0.8% 0.5% 0.6% 0.6% 0.9%

As % of HHs 76.7% 78.6% 80.0% 81.0% 81.1% 81.0% 81.0% 81.1%



39

Exhibit 28: Acess Line Analysis Model Last Updated: 03/13/2015 CAGR

2008 2009 2010 2011 2012 2013 2014E 2015E 2016E 11-16E

Total U. S. Access Lines4 162.8 152.9 149.7 143.3 138.1 133.2 130.6 128.6 126.7 (2.4%)

YOY Growth 2.7% (6.0%) (2.2%) (4.2%) (3.7%) (3.5%) (2.0%) (1.5%) (1.5%)

Primary Access Lines 152.3 142.8 139.7 133.5 128.4 123.7 121.1 119.2 117.4 (2.5%)

YOY Growth 3.3% (6.2%) (2.2%) (4.4%) (3.8%) (3.7%) (2.1%) (1.5%) (1.5%)

Secondary Access Lines 10.5 10.1 9.9 9.8 9.7 9.6 9.5 9.4 9.3 (1.0%)

YOY Growth (5.0%) (3.0%) (2.0%) (1.5%) (1.3%) (1.0%) (0.9%) (0.9%) (0.9%)

By End Market

Total Residential 97.8 91.4 87.0 82.8 78.9 75.3 74.5 74.6 75.4

YOY Growth 4.2% (6.6%) (4.8%) (4.9%) (4.7%) (4.6%) (1.0%) 0.1% 1.1%

Switched Access Line 78.2 68.6 60.0 51.9 44.6 37.6 33.1 29.4 26.2

ILEC 72.5 63.6 55.7 48.4 41.5 35.0 30.8 27.4 24.4

Competitor 5.6 5.0 4.3 3.7 3.1 2.6 2.3 2.0 1.8

Interconnected VoIP 19.7 22.8 27.0 30.9 34.3 37.7 41.5 45.2 49.2

ILEC 0.3 1.0 2.5 4.2 6.1 8.0 10.0 12.0 14.4

Competitor 19.0 21.4 24.6 26.7 28.2 29.7 31.4 33.2 34.8

Total Business 65.0 61.5 62.6 60.5 59.2 58.0 59.2 61.4 65.1

YOY Growth 0.6% (5.3%) 1.7% (3.4%) (2.2%) (2.0%) 2.1% 3.7% 6.0%

Switched Access Line 62.9 58.3 57.9 54.7 51.5 47.7 45.3 43.1 40.9

ILEC 45.4 42.2 39.0 36.4 33.7 31.0 29.2 27.4 25.8

Competitor 17.4 16.3 17.9 18.4 17.8 16.7 16.1 15.6 15.1

Interconnected VoIP Business 2.1 3.2 4.7 5.8 7.6 10.3 13.9 18.3 24.2

ILEC 0.3 0.6 0.4 0.5 0.7 1.0 1.6 2.3 3.5

Competitor 1.7 3.0 4.3 5.2 6.9 9.2 12.3 16.0 20.7

VOIP Market Share 5.2% 7.6% 9.5% 12.9% 17.7% 23.4% 29.8% 37.1%

VOIP

Total ILEC VoIP 0.5 1.6 2.9 4.7 6.9 9.0 11.6 14.3 17.9

201.7% 79.3% 64.5% 46.0% 31.8% 27.9% 24.0% 24.9%

Total Competitive VoIP 20.7 24.4 28.8 32.0 35.1 38.9 43.7 49.1 55.5

17.7% 18.2% 10.9% 9.8% 10.9% 12.4% 12.3% 12.9%

Total VOIP 21.3 26.0 31.7 36.7 41.9 48.0 55.3 63.5 73.4

22.2% 21.9% 15.7% 14.4% 14.3% 15.4% 14.8% 15.6%

Coax Cable 20.1 23.2 25.9 27.8 29.3 30.9 32.4 34.0 35.7

139.8% 15.2% 11.7% 7.3% 5.6% 5.3% 5.0% 5.0% 5.0%

SaaS VoIP (Residential + Business) 0.6 1.2 2.9 4.2 5.8 8.0 11.3 15.1 19.7

96.9% 141.6% 42.4% 37.6% 39.3% 41.1% 33.3% 30.8%

Market Share

ILEC Residential Market Share 74.4% 70.7% 66.9% 63.3% 60.4% 57.1% 54.8% 52.8% 51.4%

ILEC Business Market Share 70.4% 68.9% 62.8% 61.0% 58.3% 55.3% 51.9% 48.5% 45.0%

Incumbent Access Lines 118.5 107.0 97.5 89.4 81.6 75.1 70.3 66.8 63.5 (6.6%)

YOY Growth (8.6%) (9.7%) (8.9%) (8.3%) (8.8%) (8.0%) (6.3%) (5.0%) (5.0%)

Absolute Declines (11.20) (11.48) (9.52) (8.07) (7.84) (6.51) (4.75) (3.55) (3.33)

Competitive Line Analysis

Total Lines 44.3 45.9 52.2 53.9 56.5 58.2 60.2 61.8 63.2 3.2%

YOY Growth 54.1% 3.7% 13.6% 3.3% 4.8% 3.0% 3.6% 2.6% 2.3%

Cable (mostly VoIP1) 19.1 21.3 23.1 24.3 25.3 26.0 27.1 27.7 28.1 2.9%

YOY Growth 31.8% 11.5% 8.2% 5.4% 4.2% 2.7% 4.3% 2.2% 1.4%

Consumer VoIP1 5.3 5.3 5.7 5.9 6.0 6.2 6.2 6.2 6.2 1.2%

YOY Growth 36.8% 0.5% 6.9% 3.0% 1.4% 4.8% (0.4%) 0.1% 0.1%

CLEC and Other2 19.8 19.3 23.4 23.7 25.2 25.9 26.9 27.9 28.9 4.0%

YOY Growth 92.0% (2.8%) 21.3% 1.3% 6.3% 2.8% 3.9% 3.7% 3.6%

Estimated Substitute Lines

Total Lines 58.9 67.2 75.5 83.6 90.7 95.7 98.5 101.3 104.1 6.1%

YOY Growth 13.1% 14.1% 12.4% 10.7% 8.5% 5.5% 2.9% 2.8% 2.8%

Wireless 40.0 48.0 56.0 64.0 71.0 76.0 79.0 82.0 85.0 5.8%

YOY Growth 17.6% 20.0% 16.7% 14.3% 10.9% 7.0% 3.9% 3.8% 3.7%

Broadband (2nd line substitution) 18.9 19.2 19.5 19.6 19.7 19.7 19.5 19.3 19.1 (0.5%)

YOY Growth 4.4% 1.6% 1.6% 0.5% 0.5% 0.0% (1.0%) (1.0%) (1.0%)

Normalized U.S. Access Lines3 221.7 220.1 225.2 226.9 228.8 228.9 229.1 229.9 230.8 0.3%

YOY Growth 5.3% (0.7%) 2.3% 0.8% 0.8% 0.1% 0.1% 0.4% 0.4%

Source: Company reports and Oppenheimer & Co. Inc. estimates.

Notes:

1. VoIP includes both consumer and business lines.

2. Other represents primarily facilities based CLEC mostly provided by the integrated long distance carriers.

3. Normalized lines represent actual lines plus estimated substitute lines.

4. U.S. reported lines peaked at 192 million in 2000.


40

Important Disclosures and CertificationsAnalyst Certification - The author certifies that this research report accurately states his/her personal views about thesubject securities, which are reflected in the ratings as well as in the substance of this report. The author certifies that no partof his/her compensation was, is, or will be directly or indirectly related to the specific recommendations or views containedin this research report.Potential Conflicts of Interest:Equity research analysts employed by Oppenheimer & Co. Inc. are compensated from revenues generated by the firmincluding the Oppenheimer & Co. Inc. Investment Banking Department. Research analysts do not receive compensationbased upon revenues from specific investment banking transactions. Oppenheimer & Co. Inc. generally prohibits any researchanalyst and any member of his or her household from executing trades in the securities of a company that such researchanalyst covers. Additionally, Oppenheimer & Co. Inc. generally prohibits any research analyst from serving as an officer,director or advisory board member of a company that such analyst covers. In addition to 1% ownership positions in coveredcompanies that are required to be specifically disclosed in this report, Oppenheimer & Co. Inc. may have a long positionof less than 1% or a short position or deal as principal in the securities discussed herein, related securities or in options,futures or other derivative instruments based thereon. Recipients of this report are advised that any or all of the foregoingarrangements, as well as more specific disclosures set forth below, may at times give rise to potential conflicts of interest.

Important Disclosure Footnotes for Companies Mentioned in this Report that Are Covered byOppenheimer & Co. Inc:Stock Prices as of March 17, 2015Apple Inc. (AAPL - NASDAQ, $124.95, OUTPERFORM)Akamai Technologies (AKAM - NASDAQ, $72.08, OUTPERFORM)American Tower Corp. (AMT - NYSE, $94.61, OUTPERFORM)Amazon.Com, Inc. (AMZN - NASDAQ, $373.35, OUTPERFORM)Broadcom Corporation (BRCM - NASDAQ, $44.94, OUTPERFORM)Crown Castle International (CCI - NYSE, $85.80, PERFORM)Cogent Communications (CCOI - NASDAQ, $35.24, OUTPERFORM)Comcast (CMCSA - NASDAQ, $59.75, OUTPERFORM)Salesforce.com (CRM - NYSE, $66.43, OUTPERFORM)Cisco Systems (CSCO - NASDAQ, $28.30, OUTPERFORM)CenturyLink (CTL - NYSE, $34.50, PERFORM)Equinix Inc. (EQIX - NASDAQ, $236.14, OUTPERFORM)Facebook, Inc. (FB - NASDAQ, $78.07, OUTPERFORM)Google, Inc. (GOOG - NASDAQ, $554.51, PERFORM)Intel Corp. (INTC - NASDAQ, $30.83, PERFORM)InterXion Holding N.V. (INXN - NYSE, $30.05, NOT RATED)Level 3 Communications, Inc. (LVLT - NASDAQ, $54.55, PERFORM)Microsoft Corporation (MSFT - NASDAQ, $41.56, OUTPERFORM)Netflix, Inc. (NFLX - NASDAQ, $421.97, OUTPERFORM)Nokia Corporation (NOK - NYSE, $7.78, OUTPERFORM)QUALCOMM Incorporated (QCOM - NASDAQ, $70.00, OUTPERFORM)Rackspace Hosting, Inc. (RAX - NYSE, $53.06, OUTPERFORM)Sprint Nextel (S - NYSE, $5.15, UNDERPERFORM)AT&T, Inc. (T - NYSE, $33.06, OUTPERFORM)T-Mobile (TMUS - NYSE, $32.47, OUTPERFORM)Twitter, Inc. (TWTR - NYSE, $46.43, PERFORM)Verizon (VZ - NYSE, $49.27, OUTPERFORM)


41

Windstream Corporation (WIN - NYSE, $7.59, PERFORM)Yahoo! Inc. (YHOO - NASDAQ, $43.51, OUTPERFORM)

All price targets displayed in the chart above are for a 12- to- 18-month period. Prior to March 30, 2004, Oppenheimer & Co.Inc. used 6-, 12-, 12- to 18-, and 12- to 24-month price targets and ranges. For more information about target price histories,please write to Oppenheimer & Co. Inc., 85 Broad Street, New York, NY 10004, Attention: Equity Research Department,Business Manager.

Oppenheimer & Co. Inc. Rating System as of January 14th, 2008:

Outperform(O) - Stock expected to outperform the S&P 500 within the next 12-18 months.

Perform (P) - Stock expected to perform in line with the S&P 500 within the next 12-18 months.

Underperform (U) - Stock expected to underperform the S&P 500 within the next 12-18 months.

Not Rated (NR) - Oppenheimer & Co. Inc. does not maintain coverage of the stock or is restricted from doing so due to a potential conflictof interest.

Oppenheimer & Co. Inc. Rating System prior to January 14th, 2008:

Buy - anticipates appreciation of 10% or more within the next 12 months, and/or a total return of 10% including dividend payments, and/orthe ability of the shares to perform better than the leading stock market averages or stocks within its particular industry sector.

Neutral - anticipates that the shares will trade at or near their current price and generally in line with the leading market averages due to aperceived absence of strong dynamics that would cause volatility either to the upside or downside, and/or will perform less well than higherrated companies within its peer group. Our readers should be aware that when a rating change occurs to Neutral from Buy, aggressivetrading accounts might decide to liquidate their positions to employ the funds elsewhere.

Sell - anticipates that the shares will depreciate 10% or more in price within the next 12 months, due to fundamental weakness perceivedin the company or for valuation reasons, or are expected to perform significantly worse than equities within the peer group.

Distribution of Ratings/IB Services Firmwide

IB Serv/Past 12 Mos.

Rating Count Percent Count Percent

OUTPERFORM [O] 320 55.08 144 45.00

PERFORM [P] 251 43.20 93 37.05

UNDERPERFORM [U] 10 1.72 2 20.00

Although the investment recommendations within the three-tiered, relative stock rating system utilized by Oppenheimer & Co. Inc. do notcorrelate to buy, hold and sell recommendations, for the purposes of complying with FINRA rules, Oppenheimer & Co. Inc. has assignedbuy ratings to securities rated Outperform, hold ratings to securities rated Perform, and sell ratings to securities rated Underperform.

Company Specific Disclosures

Oppenheimer & Co. Inc. makes a market in the securities of AAPL, AKAM, AMZN, BRCM, CCOI, CMCSA, CSCO, DCM,DISH, EQIX, FB, GOOG, INTC, MSFT, NFLX, QCOM, SLA, T, TMUS and YHOO.

Oppenheimer & Co. Inc. expects to receive or intends to seek compensation for investment banking services in the next 3months from CMCSA, FB, INXN, LVLT and WIN.

Additional Information Available


42

Please log on to http://www.opco.com or write to Oppenheimer & Co. Inc., 85 Broad Street, New York, NY 10004, Attention: EquityResearch Department, Business Manager.

Other DisclosuresThis report is issued and approved for distribution by Oppenheimer & Co. Inc. Oppenheimer & Co. Inc. transacts business on all principalexchanges and is a member of SIPC. This report is provided, for informational purposes only, to institutional and retail investor clients ofOppenheimer & Co. Inc. and does not constitute an offer or solicitation to buy or sell any securities discussed herein in any jurisdictionwhere such offer or solicitation would be prohibited. The securities mentioned in this report may not be suitable for all types of investors.This report does not take into account the investment objectives, financial situation or specific needs of any particular client of Oppenheimer& Co. Inc. Recipients should consider this report as only a single factor in making an investment decision and should not rely solelyon investment recommendations contained herein, if any, as a substitution for the exercise of independent judgment of the merits andrisks of investments. The analyst writing the report is not a person or company with actual, implied or apparent authority to act on behalfof any issuer mentioned in the report. Before making an investment decision with respect to any security recommended in this report,the recipient should consider whether such recommendation is appropriate given the recipient's particular investment needs, objectivesand financial circumstances. We recommend that investors independently evaluate particular investments and strategies, and encourageinvestors to seek the advice of a financial advisor. Oppenheimer & Co. Inc. will not treat non-client recipients as its clients solely by virtueof their receiving this report. Past performance is not a guarantee of future results, and no representation or warranty, express or implied,is made regarding future performance of any security mentioned in this report. The price of the securities mentioned in this report andthe income they produce may fluctuate and/or be adversely affected by exchange rates, and investors may realize losses on investmentsin such securities, including the loss of investment principal. Oppenheimer & Co. Inc. accepts no liability for any loss arising from theuse of information contained in this report, except to the extent that liability may arise under specific statutes or regulations applicable toOppenheimer & Co. Inc. All information, opinions and statistical data contained in this report were obtained or derived from public sourcesbelieved to be reliable, but Oppenheimer & Co. Inc. does not represent that any such information, opinion or statistical data is accurate orcomplete (with the exception of information contained in the Important Disclosures section of this report provided by Oppenheimer & Co.Inc. or individual research analysts), and they should not be relied upon as such. All estimates, opinions and recommendations expressedherein constitute judgments as of the date of this report and are subject to change without notice. Nothing in this report constitutes legal,accounting or tax advice. Since the levels and bases of taxation can change, any reference in this report to the impact of taxation shouldnot be construed as offering tax advice on the tax consequences of investments. As with any investment having potential tax implications,clients should consult with their own independent tax adviser. This report may provide addresses of, or contain hyperlinks to, Internet websites. Oppenheimer & Co. Inc. has not reviewed the linked Internet web site of any third party and takes no responsibility for the contentsthereof. Each such address or hyperlink is provided solely for the recipient's convenience and information, and the content of linked thirdparty web sites is not in any way incorporated into this document. Recipients who choose to access such third-party web sites or followsuch hyperlinks do so at their own risk.This research is distributed in the UK and elsewhere throughout Europe, as third party research by Oppenheimer Europe Ltd, which isauthorized and regulated by the Financial Conduct Authority (FCA). This research is for information purposes only and is not to be construedas a solicitation or an offer to purchase or sell investments or related financial instruments. This research is for distribution only to personswho are eligible counterparties or professional clients and is exempt from the general restrictions in section 21 of the Financial Servicesand Markets Act 2000 on the communication of invitations or inducements to engage in investment activity on the grounds that it is beingdistributed in the UK only to persons of a kind described in Article 19(5) (Investment Professionals) and 49(2) High Net Worth companies,unincorporated associations etc.) of the Financial Services and Markets Act 2000 (Financial Promotion) Order 2005 (as amended). It is notintended to be distributed or passed on, directly or indirectly, to any other class of persons. In particular, this material is not for distribution to,and should not be relied upon by, retail clients, as defined under the rules of the FCA. Neither the FCA’s protection rules nor compensationscheme may be applied.Distribution in Hong Kong: This report is prepared for professional investors and is being distributed in Hong Kong by OppenheimerInvestments Asia Limited (OIAL) to persons whose business involves the acquisition, disposal or holding of securities, whether as principalor agent. OIAL, an affiliate of Oppenheimer & Co. Inc., is regulated by the Securities and Futures Commission for the conduct ofdealing in securities, advising on securities, and advising on Corporate Finance. For professional investors in Hong Kong, please [email protected] for all matters and queries relating to this report. This report or any portion hereof may not be reprinted, sold, orredistributed without the written consent of Oppenheimer & Co. Inc.

This report or any portion hereof may not be reprinted, sold, or redistributed without the written consent of Oppenheimer & Co. Inc. Copyright© Oppenheimer & Co. Inc. 2015.


http://www.opco.com

mailto:[email protected]

cloud convergence - zacks investment research

Documents