PIVOTAL U.S.

Equity Research
Internet / Advertising
Pivotal Research Group

Snap Inc.: Cant Get There From Here. March 2, 2017

Initiating Coverage With SELL Rating and $10 Price Target

BOTTOM LINE: Snap is a promising early stage company with significant opportunity Brian Wieser, CFA
ahead of itself. Unfortunately, it is significantly overvalued given the likely scale of its 212-514-4682
long-term opportunity and the risks associated with executing against that opportunity. brian@pvtl.com
Significant ongoing dilution from share-based compensation will likely represent an
additional negative consideration for the stock. We value Snap at $10 per share on a Snap Inc. (SNAP)
YE2017 basis. As the stock priced well above this level in its IPO, we rate its shares RATING: Sell
Sell. (Previous: N/R)

Snap presents investors with the opportunity to invest in the company behind an Target Price: $10.00
innovative, large-scale, and distinctively young-skewing platform which is establishing (Previous: N/R)
itself as a magnet for business unit talent and content partners alike. Snap also offers Price (3/1/17): $17.00
investors a share of the significant economic potential that should follow from Snaps
ongoing business expansion.
MARKET DATA

At the same time, there are significant risks offsetting these opportunities. Investors in 52 Wk Hi Low N/A

Snap will be exposed to an upstart facing aggressive competition from much larger Market Cap. (MM) N/A

companies, with a core user base that is not growing by much and which is only Avg. Daily Vol (000) N/A

relatively elusive. It has a promising and innovative advertising offering, but so far it is
still mostly unproven and difficult to quantify its ultimate scale. Investors will also be EPS 2017E Prior 2018E Prior

exposed to what appears to be a sub-optimal corporate structure operated by a senior 1Q -$1.26 -$0.52
management team lacking experience transforming a successful new product into a
2Q -$0.46 -$0.53
successful company. High expenses and cash costs to run the company are negative
as well. And then there are other negatives for shareholders given the degree to which 3Q -$0.45 -$0.54

they will be diluted through aggressive share issuances to employees and through the 4Q -$0.43 -$0.53
lack of voting rights that they will possess.
FY -$2.60 -$2.12

While we consider ourselves cautious optimists on the business itself, our model feels
potentially stretched in even getting to $10 per share, or a $16bn valuation on a
Sales 2017E Prior 2018E Prior
YE2017 basis. As the stock priced well above this level in its IPO, we rate its shares
Sell. 1Q $173 $356

2Q $199 $406
Additional commentary is provided in the following initiation of coverage on Snap.
3Q $253 $444

4Q $337 $648

VALUATION. Our DCF-based valuation includes is largely driven by a long-term FY $963 $1855
discount rate of 12.0% and long-term growth of 6%. Sales in Millions

RISKS. Snaps risks include heightened uncertainty, the competitive environment,

slowing growth in the user base, a lack of a track record in building a successful BALANCE SHEET DATA (12/31/16)

business, high costs and ongoing dilution as well as sub-optimal organizational design Cash ($) $150

and corporate governance. Debt ($) $0

Source: Pivotal Research Group and Company Documents

Pivotal Research Group 286 Madison Ave., 16th Floor New York, NY 10017
Important Disclosures Are Located In The Appendix
Snap: Cant Get There From Here

Introduction

Snap presents investors with the opportunity to invest in the company behind an innovative, large-scale, and
distinctively young-skewing platform which is establishing itself as a magnet for business unit talent and content
partners alike. Snap also offers investors a share of the significant economic potential that should follow from
Snaps ongoing business expansion.

At the same time, there are significant risks offsetting these opportunities. Investors in Snap will be exposed to an
upstart facing aggressive competition from much larger companies, with a core user base that is not growing by
much and which is only relatively elusive. It has a promising and innovative advertising offering, but so far it is still
mostly unproven and difficult to quantify its ultimate scale. Investors will also be exposed to a novel corporate
structure operated by a senior management team lacking experience transforming a successful new product into
a successful company. High expenses and cash costs to run the company are negative as well. And then there
are other negatives for shareholders given the degree to which they will be diluted through aggressive share
issuances to employees and through the lack of voting rights that they will possess.

While we consider ourselves cautious optimists on the business itself, our model feels stretched in even getting
to $10 per share, or a $16bn valuation on a YE2017 basis. As the stock priced well above this level in its IPO, we
rate its shares Sell.

BUSINESS OVERVIEW

Flagship Platform Has Significant Absolute Usage and Reach Among Young Audiences

Snaps flagship Snapchat platform has grown from nothing to become one of the most significant digital media
platforms in the space of less than six years. With pioneering consumer and advertising propositions featuring
disappearing messages and vertical ads the company has rightly established a reputation as innovative.

Its user base and usage levels are also significant given the breadth and volume of usage of the platform,
especially among young audiences: according to its S1 filing indicating that on average, our usersspend 25 to
30 minutes on Snapchat every day and that it had 161 million Daily Active Users (DAUs) at the end of December
2016, including 69 million users in the broader North America region (including the US, Canada, Mexico and the
Caribbean). The company has also indicated that it has over 60 million DAUs in the United States and Canada.

Not provided by the company is their monthly active users (MAUs) base, which we think is more appropriate to
assess, both for purposes of comparing Snap to other companies and for looking at Snap as an advertising
platform. Most advertising campaigns require reach that accumulates over time, and a MAU is more likely the
relevant metric unless an advertiser runs a campaign that has a very short window of intensive marketing activity.
Consequently, we look for data from third parties, which are more likely to provide data that is consistently
measured across different media owners.

Our analysis of comScore data indicates that during December 2016, around 60% of adults 18-34 (more than 40
million people) in the US used Snap for an average of 7 hours each, not far from the lower bound of Snaps self-
provided global all-audience figure (for which the average figure would equate to 7.5 to 9 hours per month). This
was up from under 40% of people 18-34 and around 6 hours each during December 2015. Calculated in person-
hours, Snapchat was used for around 400 million person-hours among adults 18+, vs less than 200 million

-2- Brian Wieser 212-514-4682 Pivotal Research Group

person-hours in the year-ago period. The equivalent figure for 18-34 year-olds was closer to 300 million hours
during December 2016, also double vs. the December 2015.

The 400 million and 300 million figures above are slightly lower than equivalent figures for Instagram, which is
probably Snaps closest direct competitor for time and advertising money. Both are substantially smaller than
Facebook itself, with Snaps time metrics equal to around one third of the equivalent figure for Facebook among
18-34s, and slightly better than one tenth of the equivalent for Facebook among 18+ audiences.

Using Nielsen data to look at TV viewing for further reference, we can see that consumers 18+ spend around 35
billion person-hours with the medium in a month, or more like 40 billion person-hours including video consumption
on digital devices. Non-video consumption on digital devices probably adds another 20 billion person-hours on a
monthly basis. For adults 18-34, there are around 6 billion person-hours of traditional TV consumption in a typical
month, another 3 billion hours of total video consumption and probably another 6 billion hours of non-video
consumption on digital devices.

Consequently, we might compare the 400 million person-hours figure for Snapchat to Nielsens 60 billion person
hours of combined digital and TV consumption, or around two thirds of a percentage point of time-spent at
present. If we look at 18-34s only, we would calculate 300 million person-hours as a share of 15 billion, or around
2% of time spent with digital and video-based media.

Of course, Snap has made some pains to contrast its recent growth with the decline of consumption of traditional
TV. Specifically, its S1 notes that per Nielsen data, people between the ages of 18 and 24 spent 35% less time
watching traditional TV in an average month during the second quarter of 2016 compared to the second quarter of
2010. The -35% figure equates to a CAGR (compound annual growth rate) of around -7% over the six year time-
frame. However, the definition of TV on which this decline is based excludes most consumption of video content
through VOD, video game consoles, over-the-top devices, on laptops, desktops, tablets and other mobile devices
if they are not attached to television sets. Including these other platforms, viewing has likely risen slightly across
all age groups over the past six years. Excluding short-form content such as that which dominates YouTube,
consumption levels are probably closer to flat across most populations when we compare comprehensive data for
any given age group vs. that same age group in prior years.

A separate point worth considering when comparing longer-term media consumption trends for 18-24 year-olds is
that the two extremes can represent demographic groups with very different media consumption habits. 24 year-
olds consumed 3.7 hours of time with TV daily during 2016, vs. 2.7 hours per day for 18 year olds, a 38% gap.
While some of the gap is undoubtedly based on the emergence of new media technologies and services, life-
stage issues likely account for the bulk of the difference, which historically amounted to more like a 25% gap in
prior years. While it is certain that TV consumption trends are changing for young audiences, conventional TV is
still a massive medium, and as audiences age, the medium tends to become more important to those audiences
across a growing number of platforms.

Still, the comparison with traditional TV is not entirely inappropriate given the video-centric nature of Snap and its
efforts to build out Discover as an environment for premium video content. Towards these ends, another way to
look at Snapchats relative size is to compare its reach with individual TV properties. The big 4 broadcast
networks have the widest reach among almost every audience, including 18-34 year-olds. They typically reach
around 60% of adults 18-34 in a month, with 5-6 hours of viewing time per network for each viewer. This equates
to around 200mm person-hours per month. Top-ranking cable networks among young audiences have less
aggregated viewing, with the likes of MTV, VH1, Comedy Central and Cartoon Network / Adult Swim reaching
more like 20-30% of the population in a month. Adult Swim in particular scores high in view time (with around 7
hours of viewing per month per viewer) but other networks typically have less viewing than broadcast networks
do. Snap looks favorable vs. these metrics with more than 7 hours of activity during the month of December,
although we might guess that only a small share of activity at Snap occurred around its premium content.

-3- Brian Wieser 212-514-4682 Pivotal Research Group

Media Partners Are Providing Premium Content, Likely Supporting Usage Trends

We expect that consumption of premium content on Snap will grow over time, as Snap has emphasized licensing
this content in recent quarters. Specific data related to consumption of the Discover environment, were most of
the premium content lives, is not disclosed. Still, media partners are generally keen to establish and expand
relationships with Snap given a perceived opportunity to widen distribution and expose content brands more often
to Snaps audience, especially in Discover. We think that much of this activity could be characterized as
experimental and could yet evolve in different directions, but should generally lead to more usage of the platform
among a broader group of consumers. While early partners may have rights to sell ads directly, sharing the
revenues they collect with Snap, newer partners tend to have license agreements with Snap, who then shares
revenue with the partner. Incidentally, this dynamic distorts efforts to analyze the scale of current spending on and
around Snapchat at the present time.

Although specific arrangements are not typically disclosed, press reports have indicated that typical revenue
share arrangements with content providers provide for a 70/30 (content provider / Snap) split. On this basis, we
can estimate that advertisers spent ~$117mm through partners during 2016, of which Snap received $35mm (as
disclosed). At the same time, if the split is 50/50 when Snap licenses content and sells the ads, there would have
been $115.6mm of revenue associated with partner produced content recorded as revenue (based upon the
$57.8mm in revenue share payments to partners during 2016). If we added the aforementioned difference
between $117mm and $35mm to Snaps reported revenue, the company would have posted closer to $500mm in
activity last year. If the company is increasingly prioritizing its partner-sold model rather than a content producer-
direct-sold model, this could provide another near-term tailwind for revenue growth. On the other hand, we note
that the revenue share split will need to be continually optimized to ensure that content producers are
appropriately incented to license content. Advertisers will also need to be willing to prioritize audience reach
rather than content adjacencies with their Snap spending, as the Snap-sold model undoubtedly would result in
most sales running on a run-of-site model vs. partner-sold models which would allow partners to sell specific
content across platforms beyond Snap.

Snap Has Quickly Established Itself As A Magnet For Talent

The opportunity associated with this scale has also proven to be alluring for industry executives. Snapchat has
demonstrated that it is capable of attracting top tier talent to support its co-founders, for whom Snap is their first
post-college endeavor. Among the dozen individuals who (according to a The Information article from October)
report to its CEO are two senior former Mattel executives, Pandoras former CTO, a former senior News Corp
executive and former senior software engineers from Amazon and Motorola Mobility. Immediately below that level
other are other prominent hires, such as Viacoms former head of sales and marketing, the founder of Facebooks
audience network, Amazons former VP of Infrastructure (who was responsible for its global data centers), News
Corps former head of corporate development and President of the New York Post, the former President of The
Onion and a former producer and national political reporter from CNN.

With all of these elements, on paper, Snap has so far established itself as one of a handful of large scale digital
media platforms for younger audiences, and it has done so without the support of a corporate parent. The
company has expanded its revenue base faster than its peers have in relation to the inception of ad sales efforts.
It also has people in place who should be able to continuously develop Snap as a platform and drive significant
growth for many years to come.

-4- Brian Wieser 212-514-4682 Pivotal Research Group

Is Scaled Usage, Unique Content and Managerial Talent Enough For a Successful Business?

All of the aforementioned factors are not necessarily enough to establish a successful product into a successful
business over the long-run. Turning Snaps potential into reality on an ongoing basis will at minimum require
thoughtful choices related to overcoming the risks that the company faces, and some luck may help, too.

First among its risks is the competitive environment, which is made harsh by the dominant presence of Facebook
and Google among digital media companies. Because each company has a unique capacity to satisfy advertiser
goals better than any other digital media owner, the two duopolists gain access to a first look at most
advertisers wallets and position themselves to capture a disproportionate share of revenue, at least among digital
advertising budgets. We can observe something similar in television, where broadcast networks capture a
disproportionate share of revenue largely because they can satisfy any one advertisers goals more efficiently
than the next-best alternatives given their dominant audience-reach metrics, which have generally held up even
as total consumption of network TV has fallen.

Snap is doubly challenged because Facebooks Instagram is by itself larger than Snap in terms of users and
usage among 18-34 year-olds (at least in the United States). For many industry participants (advertisers and
content licensors, specifically), Instagram may very well be good-enough to prioritize working with at the
expense of Snapchat, given their needs to concurrently work with Facebook in some capacity. Snapchat
potentially offsets this risk because of its narrow audience focus today could result in better products. As well, it
may be better able to foster more innovation on an ongoing basis because of its smaller size vs. larger
competitors.

Audience Doesnt Need to Grow Necessarily, But Skew and Scale Impacts Opportunity

Whether because of Instagram or market saturation or other factors, Snap has indicated that user growth has
slowed significantly around the world in recent months, and that a disproportionate share of its growth is occurring
among audiences older than 25. To the extent that Snap does not expect to be ubiquitous and is not investing
against the business with any expectation that it will become so, nor conveying to advertisers that more growth is
at hand meaningful user growth from current levels is not essential for the company to build a durable business.

On the other hand, failure to sustain growth in consumption per user or minimize user losses will probably be
critical in the near term as it looks to encourage advertisers to invest time and effort with the platform. Perceptions
of momentum matter for companies which are positioned as bright shiny objects, at least until they are widely
viewed among their customers as making a meaningful and noticeable contribution to an advertisers marketing
efforts.

A bigger issue relates to whether or not the absolute size and unique audience skew it has will be sufficiently
meaningful to drive the business to meet expectations that the company and investors have for it. While young-
skewing audiences are viewed favorably by many advertisers, we think there are relatively few categories who
want to reach narrow and young audiences to the exclusion of others. We have seen research from the mid-
2000s indicating that only a single digit percentage of network TV buys were made with 18-34 year-old targets, for
example. Most advertisers who incorporate age and gender targets into their media buying want to run ads on
media properties with young skews for purposes of prioritizing their buys more than anything else.

-5- Brian Wieser 212-514-4682 Pivotal Research Group

Ad Products Need Time To Mature, Too

Another set of risks for Snap relate to its ad products. Among advertisers who are spending or looking to spend
with Snap, realizing the potential of a campaign is harder than if it were a TV network with TV ad inventory or a
pure digital media owner with digital ad inventory. The presence of innovative ad products and the capacity to
develop more of them on a continuing basis has pros and cons. In the long-run, possessing new and different ad
products at scale helps prevent a media owner from becoming a commodity. On the other hand, lack of
commoditization can make it difficult for an advertiser to justify working a media owner with unless that media
owner is so dominant that an advertiser cant run a campaign successfully without the media owner.

More specifically, a large brand marketer looking to realize Snapchats potential generally requires more
collaboration between internal and external creative and media teams than might otherwise occur because
Snaps advertising proposition is somewhat unique. If that advertiser can already reach Snaps audience
elsewhere with creative units and media buying processes they already have in place, the extra costs of time and
money that a marketer and its agencies will expect to incur in order to make a campaign on Snap successful may
not be viewed as worthwhile. This is especially true because advertisers have so many alternative places to
spend their money. Smaller advertisers with geographic constraints who may find geofilters to be valuable may
have fewer integration issues and may not require much work to develop and manage their ad units, but then
many of these advertisers won't have teams that are large enough to justify working with new and relatively small
(relative to Google or Facebook) media owners.

There are some kinds of advertisers movie studios in particular who seemingly cant spend enough with Snap.
It helps that Snap likely has so much usage from core movie-going audiences. It helps that Snaps daily usage
advantage vs. alternatives maps up well to how studios often want to buy media when they launch new
properties. It also helps that studios tend to manage their own creative processes internally, which makes the
integration between creative and media functions easier to manage. Other large advertisers can be grouped into
those which a) spent money early with Snap and couldnt justify the spend (or get sufficient help from Snap to
justify the spend) b) are spending money with Snap because it feels right, without knowing what appropriate key
success metrics are, especially in contrast to spending that runs on Facebook and Google or c) arent spending
money because they cant justify it yet.

One of the key lessons from Twitters challenges over the past couple of years is the observation that its sales
team was very effective in persuading marketers including CMOs, who might normally leave meetings with
media owners to their agencies to part with money on a trial basis. One consequence of this success, according
to various anecdotes, is that these kinds of mandates led to spending without much justification other than "gut"
feelings. Twitter subsequently struggled to justify the spending they captured and then further struggled in their
efforts to generate repeat business from many advertisers. Although Snap's S1 provides case studies indicating
that at least some marketers have seen success from their Snap campaigns, perceptions of success are far from
uniform among marketers.

In addition, we note that marketers we have spoken to about Snap have commonly indicated that they need
access to more third-party tools for their buying, management and assessments of campaigns with Snap. Snap
has made some recent announcements indicating support for some of these tools, including one with Nielsen that
would enable the use of Digital Ad Ratings. Snap has also made a recent announcement involving Oracles Data
Cloud formerly Datalogix - supporting targeting against Oracle audience segments. Snap has also indicated that
it is working with other partners such as 4C (an analytics provider), Moat (a viewability measurement vendor),
Merkle (for its audience targeting capabilities) and Videology (for its DSP). More partners and support for more
ways to use existing partners will be necessary to meet advertisers expectations. Snap will also likely have to
invest in brand safety-related capabilities on an ongoing basis as well.

-6- Brian Wieser 212-514-4682 Pivotal Research Group

Snaps Management Approach and Corporate Structure Appear Sub-Optimal

Managerial and corporate idiosyncracies also present risks to Snap investors despite the relatively favorable
outcomes which have followed from those idiosyncracies so far. Press reports have indicated that Snap is both a
very secretive company and not particularly collaborative internally. In terms of the physical layout of its offices in
the Venice, California area, Recode wrote early last year about its string of independent office buildings along
the...boardwalk that keeps teams separated and projects siloed. Even the company notes in its S1 that its
diffuse structure may prevent us from fostering positive employee morale and encouraging social interaction
among our employees and different business units. Echoing comments from the Recode article, Business Insider
reported in October that CEO Spiegel rarely addresses the company in all-hands meetings, in sharp contrast to
the weekly updates delivered by CEOs of many Silicon Valley tech companies. The same article went on to
describe Spiegels involvement in every decision, which hardly seems like a scalable approach for a rapidly
growing company, if this approach still holds true.

Although there are clearly individuals inside of Snaps orbit (on its board and at different levels within the
organization) who have the capacity to help Snap realize its potential, these and other reports make clear that the
companys direction lies in Spiegels hands, subject to his preferences. This is made even plainer by the absence
of voting rights associated with shares that are being sold in the IPO. Spiegel and CTO/co-founder Bobby Murphy
will maintain virtually all voting power of outstanding capital stock regardless of whether or not they remain
employed by the company. While there are, of course, examples of successful companies with dual or multiple
class shareholding structures, we view these structures unfavorably primarily because non-voting shareholders
incur risks that would not otherwise be present and because controlling shareholders are not sufficiently
incentivized to establish and communicate the business case behind their decisions. These points are probably
more important for a company which has relatively inexperienced controlling shareholder-managers, is early in its
evolution as a business and is known for its secrecy as described above.

-7- Brian Wieser 212-514-4682 Pivotal Research Group

FINANCIAL ANALYSIS

Our Approach To Forecasting Revenues

As a company which only began to generate revenue in the last three years, Snap is an upstart if there ever was
one, making forecasts particularly challenging. Forecasting revenue in particular is one of the most critical
variables we need to value the business, although the relative novelty of Snaps platform and the fact that Snaps
current sales leadership has not yet been in place for even a year means we have little history against which to
base our expectations.

Given its relative size, we think it is safe to say that advertisers pursuing young-skewing audiences will want to
have a presence on Snapchat. But will all inventory have value in the way that Facebook (or Instagram) or
television does today? Will targeting capabilities provide sufficiently granular audience segments? Or will granular
targeting be necessary? Or will only a limited amount of inventory be monetizable given the limited volumes of
premium content on Snaps platform today, and the likelihood that the bulk of usage will be user-generated? And
can Snapchat continuously improve its product enough to sustain or expand its user base and sustain or expand
usage? These and many other unanswerable questions will persist around Snap for some time, leaving the
estimates we establish below subject to significant uncertainty.

Focus On Share of Wallet Rather Than ARPUs. But Which Wallet?

For starters, one approach we think many investors will take - but should rule out - relates to forecasting Snaps
Average Revenue Per User (ARPU). ARPUs are not particularly meaningful for businesses which are dependent
on advertising for two key reasons. First, they ignore total addressable market-size limitations. Second, they
ignore that the relevant user who is spending money is the advertiser, not the consumer. Average revenues per
advertiser can be useful if we have insights into a media owners book of business. Unfortunately, relevant and
accurate data of this nature is difficult to come by, even for mature companies.

A better approach and one which will generally map to how budgets actually get allocated to Snap in the long-
run is to think about revenue in terms of the share of advertiser wallets. This roughly coincides with the ways
in which larger brand-oriented advertisers (such as those which Snap is focused on) tend to manage their
budgets. Estimating the relevant market and the share that Snap can capture still requires several additional
assumptions.

We think that Snap is likely better positioned on a relative basis (as a percentage of the companys future
revenue, that is) to capture budgets that would otherwise go to television than is the case for other digital media
owners. Advertisers looking to reach young audiences with ad units involving sight sound and motion have
appeared to us to be relatively more accepting of running ads alongside semi-professional or non-professionally
produced video content vs. other advertisers because of the relative scarcity of young skewing inventory. That
Snaps head of ad sales had the same role at Viacom Media Networks is helpful towards these ends.

Of course, many (and perhaps most?) advertisers will continue to bucket Snap as a digital media owner because
it looks and feels more comparable to other digital media properties, and fund spending there accordingly. This
will be especially true for advertisers who want to tightly integrate spending on Snap with other digital activity.
Revenue that might come from small or geographically contained businesses and revenue from performance-
based marketers is probably mostly coming from digital pools given the limited spending that many of these
advertisers have on television today.

For reference, Facebook and Twitter have for the most part generated revenue from budgets that were bucketed
as digital historically. Google has done much the same across its broader business, YouTube has become better

-8- Brian Wieser 212-514-4682 Pivotal Research Group

positioned to compete for TV budgets, and is actually doing so today. We would note that even today the share of
spending that YouTube collects which might have otherwise gone to traditional TV is relatively small certainly
hundreds of millions, but not likely more than a billion dollars. Still, it seems safe to say that spending on YouTube
which otherwise would have gone to other digital media owners is many multiples larger than the amount of
spending it collected which otherwise would have gone to TV.

Whichever the reference point for a total addressable market, we need to estimate what Snaps share could be.
Consumer time with a medium is a proxy for the potential to create ad inventory. Although there are real
limitations to this approach (as the relative value of ad formats can vary, the advertisers who find a platform to be
useful can vary and the ad loads can vary) it is at least a starting point.

Estimating Revenue For Snap: $6.7 Billion Globally By 2023

Because we are assuming that there should be some proportional relationship between tonnage of use and the
share of advertising budgets that Snap can collect, guesses around multi-year user and usage growth trends
within the United States and in countries around the world are among the most significant wild cards impacting
valuations. While we can imagine scenarios with significant upsides and significant downsides to recent trends,
for the present time we are assuming that the company can grow its usage (some mix of users and time per user)
in the United States by approximately half relative to its key peers over a multi-year period. We consider this an
optimistic assumption, but recognize that growth could well exceed or fall short of such an amount.

But taking that relative growth as a given, we would say that Snap may eventually generate 15% of the equivalent
time spent with Facebook. Following from that figure, we could estimate that 15% of Facebooks US revenue
could be viewed as the upper bound of whats possible for Snap. We note that while Snap can sell its ad inventory
units to advertisers at a premium today, if it sold a greater volume and percentage of inventory, its pricing
advantage probably goes away. This is because Facebook can target better than Snap and find higher paying
advertisers for many of its ad units, because it can sell a greater share of its ad units as video if that were an
optimal choice and because it is at least as likely as Snap to have a significant volume of premium content on its
platform against which premium priced ads may be sold. Further, as Snap grows it probably will need to trade
pricing premia down for greater volumes of buys.

On the other hand, Facebook (along with Google) has significant market power and will likely capture a
disproportionate share of industry revenue over long time horizons, much as network TV advertising does vs.
cable TV. Because they essentially reach everybody who accesses digital media, they are well-positioned to get
the first look at big brand wallets (again, similar to the dynamic between network and cable TV). This factor could
bring down Snap's upper share boundary to more like 10% of Facebook revenues.

Using the same 50% growth rate for Snap vs. TV and premium video, we can estimate that Snaps share of time
of video consumption would equal around 1.5% of total video time among adults 18-34, or perhaps closer to 3%
of time among adults 18-49. For Snap to capture an equivalent amount of TV spend, it would need to
demonstrate that the quality of its audience and the quality of its ad inventory is comparable to averages for TV. It
would also need to ensure that measurement, impact and executions are broadly similar to what it used for TV.
However, there should be upside to this percentage considering the heightened value for a younger audience
skew. On the other hand, there is downside if the UGC nature of most content reduced its relative value.

Using our digital benchmarks, we could calculate $3.4bn of spending on Snap by 2021. On the other hand, using
our TV / premium video benchmark we could calculate $2bn if we defined the market as only including traditional
TV.

-9- Brian Wieser 212-514-4682 Pivotal Research Group

With these boundaries in mind, at a mechanical level we have decided to model domestic revenues as a
percentage of national TV advertising paired with digital advertising excluding search, as these seem to be the
most likely buckets of advertising that Snap will compete for. On this basis, our model conveys that Snap
captured 0.43% of its addressable market in 2016, and assume that it will essentially double in 2017 before
gradually rising to 2.5% of this market by 2021.

We would separately add international revenues to these figures, which we assume will broadly follow US trends
with a multi-year lag. We note that at Facebook it was roughly five years following the launch of international sales
before revenues outside the US matched domestic ones. Meanwhile, at Twitter, six years following the
introduction of international sales, domestic revenues are still much larger than international ones despite a larger
international user base.

Mechanically, we are modelling international revenues as a percentage of TV and digital advertising ex-China,
Japan and Russia using Magna Globals estimates for the spending in those markets. We assume that Snap will
primarily fight for market share in developed markets but will not generally operate in major markets including
China, Japan and Russia given the presence of local competitors in those countries. On this basis, we calculate
that Snap captured 0.3% of its international revenue opportunity, although note that this small base is mostly
limited by Snaps absence of a sales force (or even the ability to sell) in most countries. Our model assumes that
they can capture 1.0% of this market by 2021, at which point international revenues would account for 43% of
worldwide revenues.

In total, our model estimates that Snap will generate $4.9 billion in global revenue from advertising by 2021, equal
to 1.6% of global digital advertising on Magna Globals estimates. The figure equates to 2.1% of global digital
advertising if we exclude China, Japan and Russia from the total. Alternately, we could describe Snaps revenues
as equaling 2.5% of global TV advertising by 2021 on Magnas figures, or 3.2% of global TV advertising excluding
China, Japan and Russia.

Outside of advertising, it seems inevitable that the company will make other efforts to generate revenues. The
Spectacles product is a current example, although it is difficult to anticipate whether the company will invest
enough into these initiatives to make a meaningful difference in the companys overall financial profile. For the
present time, we ignore other revenue streams.

Unfortunately for modeling purposes, it seems that Snap is unlikely to realize a steady-state margin profile
(described further below) for many years. Assuming that most costs rise in absolute terms (at minimum) as the
ecompany grows, we dont even see much opportunity for positive adjusted EBITDA for several years. Because
using a 2021 baseline for estimating long-term cashflow growth (as we do with every other company we cover)
would result in a growth rate that exceeds costs of capital, such a model would not work. Consequently, we need
to extend our model our further to point where the company can produce steady-state margins, which we assume
to be 2023. For 2022 and 2023, we assume that growth decelerates to teens levels, yielding yield estimates of
$6.7bn in global advertising in 2023. While $6.7bn is a significant amount of revenue and would place Snap
among the largest sellers of advertising globally, it would fall well short of Facebook and Googles sizes, as those
companies would probably have closer to $100bn and $200bn, respectively, in global advertising revenue around
that time.

Beyond that period, growth expectations for Snap match figures we use for Facebook and roughly converges
around a long-term CAGR of around +6%. Supporting these estimates is an assumption that Snap can continue
to innovate, and generally expand its scale of usage and audience reach vs. its two much larger competitors.
Such an outcome is far from certain.

- 10 - Brian Wieser 212-514-4682 Pivotal Research Group

Snaps Margins Should Be Relatively Low (At Least Lower Than Twitters), But CapEx Should Be
Relatively Light

Looking beyond revenue, we next consider what Snaps appropriate margins should be. We have very little in the
companys history to provide indications of what to expect. Although we can certainly look at cost-per-user-per-
quarter metrics, we dont have a sense of how these costs will change as the company grows. Costs are high,
certainly, likely driven by the video-centric nature of its products. There may be some advantages to scale which
lead to improving margins, although those advantages are mitigated by the variable-cost skew of the companys
financial profile following on its choice to outsource much of its infrastructure to Google and Amazon. Further, we
dont know how much of the companys user activity or revenue is due to professional content let alone how
license fees for professional content will trend. (Beyond any costs to host and serve the content, Snap either pays
a license fee for the content or collects only some of the related revenue if the partner retains responsibility for ad
sales.)

All of this points us to a long-term margin profile that is likely lower than what we see with other digital media
owners we cover. Snap is unlikely to ever have the scale that Facebook and Google have, and doesnt seem
likely to generate as substantial a portion of its revenues from the high-margin self-service advertisers that drive
significant portions of those two companies businesses.

If anything, it probably comes closer to Twitter in scale, and is similar in that both companies are probably equally
dependent on license fees for professional content over the long run, and neither one seems likely to have a large
volume of high margin self-service ad revenue. However, with this benchmark we think Snaps margins should
more likely be lower than Twitters because of the video skew of its product and because it outsources much more
of its infrastructure, trading off lower ongoing capital expenditures for lower operating margins. While Twitter has
established a 40-45% adjusted EBITDA target margin as a benchmark, we think a 30% adjusted margin feels
more plausible for Snap at maturity. Of course, it could also end up lower as Snap needs to continually invest in
its consumer experience, making the service an increasingly costly one to manage and execute against.

Despite its outsourcing arrangements with Google and Amazon, we think Snap will also likely incur relatively
significant capital expenditures, if much less than if they decided to own and operate all of their own infrastructure.
In 2016 the company incurred $66mm in capex, which equated to 8% of the companys operating expenditures
(excluding depreciation & amortization and excluding share-based compensation). For reference, during 2016
Twitters capital expenditure budget was equivalent to 12% of that companys operating budget on a comparable
basis. Arguably this was a constrained figure, as Twitters average over the period from 2010 to 2016 was 16%
(and during Twitters earlier period, it also outsourced much of its infrastructure, too). On balance, a low-single
digit percentage of adjusted expenses seems like a plausible assumption for capital expenditures over extended
time horizons (we assume $150mm per year in our models terminal yeary) so long as they continue to prioritize
outsourcing. We further assume that D&A generally offsets capital expenditures on our models DCF, as has
typically been the case for Twitter. We next assume that acquisition expenditures will remain modest, perhaps
growing slightly from 2016s level of $104mm each year, and roughly matching capital expenditures.

Significant Use of RSUs Attract Employees, But Dilute Shareholders

One of the more remarkable features of Snaps financial profile is the degree to which the company issues shares
to its employees. While the idea of rewarding the companys CEO and largest shareholder with 3% of the
companys equity for completing the IPO is noteworthy by itself, the number of shares issued by the company to
employees is even more significant. During 2016, the company issued 105 million new shares as restricted stock
units (RSUs), with a fair value of 1.7 billion as of the time they were issued. This equates to $1.4 million in stock
grants per average employee last year. Looking at Twitter for a point of reference, we note that during the most

- 11 - Brian Wieser 212-514-4682 Pivotal Research Group

comparable year 2013, as Twitter was itself going public - that company issued shares worth an average of
$612k. Twitters RSU grants per employee have generally fallen from that level, and amounted to $425k per
average employee in 2016. At Facebook the amount was $230k per average employee during 2016, and at
Google the comparable amount was $144k.

January and February suggest that we may see a similar pace of RSU issuances to employees this year, as Snap
indicated that it has issued 27mm RSUs in the first two months of 2017 alone. If this pace continues, the company
would end up offering around 160mm RSUs, or around 11% of outstanding shares over the course of the year.
While the nominal value is higher than the shares Twitter provided its employees with during its IPO year, Twitter
actually provided slightly more equity as a share of its total, with distributions of RSUs that were equivalent to
13% of outstanding shares at that time. In subsequent years, that percentage fell to 9%, 6% and 7%.

As another way of looking at how extensively the company is relying on RSUs is to look at them as a percentage
of operating expenses, excluding depreciation & amortization as well as share-based compensation. At Snap the
figure was 179% (i.e. nearly 2x the cash-like operating costs the company incurred) during 2016, up from 135% in
2015, while the equivalent figure at Twitter did not exceed 32% over the past five years (and it was only 23%
during its IPO year). The value of RSUs at Facebook has been somewhat higher than at Twittter, although only
amounted to 58% during the companys IPO year of 2012 and trended down towards 36% during 2016.

Importantly, we note that not all RSUs granted become actual common stock, unvested RSUs are typically
forfeited if employees leave the company. Still, we think that dilution from RSUs is a significant issue which
negatively impacts the companys value when comparing them to peers or when using conventional DCFs with
estimates of share counts that the company will report on their income statement. Our valuation of Snap is based
upon an estimate of the outstanding shares as of the end of 2017 as well as all RSUs issued during the year
(although as we noted previously, some percentage of those RSU shares will ultimately be forfeited). This
amounts to nearly 1.6bn shares as of year-end 2017.

While dilution from more mature companies is typically immaterial to investors, we think the practice is much more
important for investors to consider when looking at Snap. With an assumption that the share issuances diminish
over time (i.e. 120mm in 2018, 100mm in 2019, 80mm in 2020, etc.) and with a continuation of the companys
existing vesting schedule, we estimate the company will have close to 1.8bn fully diluted shares outstanding by
the end of 2021, up from close to 1.5bn shares outstanding at the end of 2017.

NOLs Are Another Source of Value

Yet another challenging valuation component relates to the assumptions we need to make in determining the
value of the Net Operating Loss (NOL) carry forwards that the company will produce, both from conventional
operating losses and from the operating losses that are enhanced by the generous use of restricted stock units.
Companies are generally allowed to offset taxable income in past or future years if they incur losses. Losses may
be carried back two years or carried forward for up to 20.

For points of reference, Facebook and Twitter both similarly made significant use of RSUs to create NOLs and
reduce cash taxes payable. Facebook which is highly profitable has managed to pay a cash tax rate of
between 3 and 4% during each of the past three years, despite a book tax rate that has ranged between 40 and
45%. Twitter has paid only a negligible amount of cash taxes in part because it has produced ongoing operating
losses. But it wont likely pay much tax any time soon: as of the end of 2016, the company had federal NOL
carryforwards worth $3.5bn and state NOL carryforwards worth $1.4bn.

The effect should be even more significant with Snap. In our model, we assume that the company doesnt even
produce a positive adjusted EBITDA result until 2022, let alone a GAAP profit, which probably wont occur for

- 12 - Brian Wieser 212-514-4682 Pivotal Research Group

many years after that. Excluding expenses associated with stock-based compensation, we estimate that the
company will produce more than $3bn of conventional losses and if our assumptions around ongoing share
issuances hold, with (among other simplifying assumptions) a stable price equal to the IPOs level we could see
another $6bn in deductible losses following from stock based-compensation. Applying a 25% tax rate (assuming
tax reform brings corporate taxes down to that level over the next decade) and discounting the NOLs back to the
present yields around $1.4bn of value, or approximately $1 per share. Our estimates around this source of value
will undoubtedly be refined as managements plans around RSU issuances or other formal guidance related to
this calculation become available.

Other Valuation Items Are Mostly Benchmarked (And Comparable To) Facebook, For Now

As we have indicated previously, our operating model for the company extends to 2023 rather than 2021 as is the
case for all other companies we cover. This is because our DCF-driven valuation needs something resembling a
steady-state baseline of profitability and cashflow against which long-term growth and costs of capital can be
applied. Using similar annual growth rates that we use for Facebook in a simplied long-term model beyond 2023
leaves us with a 6% long-term growth rate for Snap. Near-term discount rates are identical to the rates we use for
Facebook and Alphabet, but longer-term discount rates, which discount cashflows beyond 2023 are slightly higher
than the ones we use for Facebook and Google at 12%, reflecting the incremental uncertainty that we have about
the nature of their long-term business orientation. Arguably this discount rate is too low, especially considering
how uncertain investors can be about the future direction of the business and its likelihood of succeeding in
building out the business in the first place. However, we think that most investors will give Snap a benefit of the
doubt for a while, causing us to limit how high we think the appropriate discount rate should be.

Valuation Summary

In simplified terms, our model assumes that by 2023, Snap is generating $6.7bn in revenue with 30% adjusted
EBITDA margins, equal to $2.0bn. We assume modest cash outflows for capital expenditures and acquisitions,
equal to around $300mm annually. We assume no cash taxes are paid during our models time horizon, although
we do assume a 25% long-term cash tax rate. The NPV of the $23bn of value we calculate for a terminal value in
2023 is $12bn of YE2017 value at the growth and discount rates we have assumed, equal to $7 per share.
Modest cash generation and the present value of NOLs adds $1 per share, and the $2.8bn of cash we expect the
company to have on its balance sheet at the year of the year adds $2 per share of value. Overall, we calculate
$16bn of enterprise value, and 1.6bn shares in our YE2017 valuation, leading to a price target of $10 per share.
With the stock going public at $17 per share, we rate Snap Sell.

Risks

As Snap is essentially a venture stage company, investors face a host of risks that are driven by greater
uncertainty than might otherwise be presence in a digital media company.

In addition, Snap investors face the following company-specific risks:

Investors in Snap will be exposed to an upstart facing aggressive competition from much larger
companies

Its core user base that is not growing by much and is only relatively elusive, if still findable on other media

- 13 - Brian Wieser 212-514-4682 Pivotal Research Group

 The company has a sub-optimal corporate structure operated by a senior management team lacking
experience transforming a successful new product into a successful company.

High expenses and cash costs will likely persist for some time

Shareholders will be diluted through aggressive share issuances to employees

Management is effectively entrenched, and shareholders are entirely disenfranchised because Snaps
publicly traded shares lack any voting rights.

- 14 - Brian Wieser 212-514-4682 Pivotal Research Group

Snap Income Statement
Snap Income Statement
In $mm Except for Share Data FY15A FY16A 1Q17E 2Q17E 3Q17E 4Q17E FY17E 1Q18E 2Q18E 3Q18E 4Q18E FY18E FY19E FY20E FY21E FY22E FY23E

Total Revenue 58.7 404.5 173.0 199.0 253.3 337.4 962.8 356.1 406.1 444.4 647.9 1,854.6 2,899.7 3,862.7 4,889.1 5,866.9 6,746.9
Year-over-Year Growth ------ 589.5% 345.8% 177.2% 97.6% 103.7% 138.0% 105.9% 104.0% 75.4% 92.0% 92.6% 56.4% 33.2% 26.6% 20.0% 15.0%

Costs and Expenses

Cost of Revenue (182.3) (451.7) (182.9) (182.2) (222.4) (284.1) (871.7) (294.0) (313.8) (320.6) (429.2) (1,357.6) (1,487.0) (1,578.8) (1,617.6) (1,638.4) (1,707.6)
% of Revenue 310.8% 111.7% 105.7% 91.6% 87.8% 84.2% 90.5% 82.5% 77.3% 72.1% 66.2% 73.2% 51.3% 40.9% 33.1% 27.9% 25.3%
% of Revenue ex-Share Based-Comp 311.7% 111.8% 90.0% 87.5% 85.0% 82.5% 85.5% 80.0% 75.0% 70.0% 65.0% 71.3% 50.0% 40.0% 32.5% 27.5% 25.0%
Research and Development (82.2) (183.7) (1,190.7) (412.2) (394.5) (372.5) (2,370.0) (480.7) (504.3) (523.9) (526.2) (2,035.1) (2,258.7) (2,239.1) (2,160.6) (2,089.3) (2,008.1)
% of Revenue 140.2% 45.4% 688.4% 207.1% 155.7% 110.4% 246.2% 135.0% 124.2% 117.9% 81.2% 109.7% 77.9% 58.0% 44.2% 35.6% 29.8%
% of Revenue ex-Share Based-Comp 157.8% 50.8% 40.0% 40.0% 40.0% 40.0% 40.0% 30.0% 30.0% 30.0% 30.0% 30.0% 25.0% 22.0% 20.0% 18.0% 17.0%
Sales and Marketing (27.2) (124.4) (254.2) (119.7) (128.9) (144.1) (646.8) (165.4) (180.7) (192.7) (238.0) (776.7) (1,059.5) (1,274.2) (1,484.5) (1,682.4) (1,504.7)
% of Revenue 46.4% 30.7% 146.9% 60.1% 50.9% 42.7% 67.2% 46.4% 44.5% 43.3% 36.7% 41.9% 36.5% 33.0% 30.4% 28.7% 22.3%
% of Revenue ex-Share Based-Comp 52.4% 31.7% 30.0% 30.0% 30.0% 30.0% 30.0% 27.5% 27.5% 27.5% 27.5% 27.5% 27.0% 26.5% 26.0% 25.5% 20.0%
General and Administrative (148.6) (165.2) (329.0) (141.9) (148.4) (159.9) (779.2) (163.6) (175.7) (185.4) (211.6) (736.2) (813.9) (845.4) (878.9) (900.6) (887.4)
% of Revenue 253.3% 40.8% 190.2% 71.3% 58.6% 47.4% 80.9% 45.9% 43.3% 41.7% 32.7% 39.7% 28.1% 21.9% 18.0% 15.4% 13.2%
% of Revenue ex-Share Based-Comp 354.2% 42.2% 30.0% 30.0% 30.0% 30.0% 30.0% 20.0% 20.0% 20.0% 20.0% 20.0% 15.0% 13.0% 12.0% 11.0% 10.0%
Total Costs and Expenses (440.4) (924.9) (1,956.7) (856.0) (894.3) (960.6) (4,667.6) (1,103.6) (1,174.6) (1,222.5) (1,404.9) (4,905.6) (5,619.2) (5,937.4) (6,141.5) (6,310.8) (6,107.8)
% of Revenue 750.7% 228.7% 1131.3% 430.1% 353.0% 284.7% 484.8% 309.9% 289.2% 275.1% 216.8% 264.5% 193.8% 153.7% 125.6% 107.6% 90.5%
% of Revenue ex-Share Based-Comp 876.0% 236.5% 190.0% 187.5% 185.0% 182.5% 185.5% 157.5% 152.5% 147.5% 142.5% 148.8% 117.0% 101.5% 90.5% 82.0% 72.0%
Income from Operations (381.7) (520.4) (1,783.7) (657.0) (640.9) (623.2) (3,704.9) (747.4) (768.5) (778.1) (757.0) (3,051.0) (2,719.5) (2,074.7) (1,252.5) (444.0) 639.1
% Operating Margin -650.7% -128.7% -1031.3% -330.1% -253.0% -184.7% -384.8% -209.9% -189.2% -175.1% -116.8% -164.5% -93.8% -53.7% -25.6% -7.6% 9.5%
Interest Income (Expense), Net 1.4 3.2 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0
Other Income (Expense), Net (0.2) (4.6) 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0
Income Before Provision for Income Taxes (380.5) (521.7) (1,783.7) (657.0) (640.9) (623.2) (3,704.9) (747.4) (768.5) (778.1) (757.0) (3,051.0) (2,719.5) (2,074.7) (1,252.5) (444.0) 639.1
Provision (Benefit) for Income Taxes 7.6 7.1 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 (159.8)
% Effective Tax Rate 2.0% 1.4% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 25.0%
Net Income (372.9) (514.6) (1,783.7) (657.0) (640.9) (623.2) (3,704.9) (747.4) (768.5) (778.1) (757.0) (3,051.0) (2,719.5) (2,074.7) (1,252.5) (444.0) 479.3
% Net Margin -635.7% -127.2% -1031.3% -330.1% -253.0% -184.7% -384.8% -209.9% -189.2% -175.1% -116.8% -164.5% -93.8% -53.7% -25.6% -7.6% 7.1%

Adjusted EBITDA (439.9) (523.1) (144.7) (162.2) (202.3) (264.4) (773.5) (179.8) (188.2) (186.1) (250.4) (804.4) (382.9) 62.1 594.5 1,196.0 2,039.1
% Revenue -750.0% -129.3% -83.6% -81.5% -79.9% -78.4% -80.3% -50.5% -46.3% -41.9% -38.6% -43.4% -13.2% 1.6% 12.2% 20.4% 30.2%

Net Income (372.9) (514.6) (1,783.7) (657.0) (640.9) (623.2) (3,704.9) (747.4) (768.5) (778.1) (757.0) (3,051.0) (2,719.5) (2,074.7) (1,252.5) (444.0) 479.3
Stock Based-Comp (73.5) (31.8) 1,628.1 482.9 425.6 344.8 2,881.3 542.7 555.3 567.0 481.7 2,146.5 2,226.5 2,016.8 1,717.0 1,500.0 1,250.0
Depreciation and Amortization 15.3 29.1 11.0 12.0 13.0 14.0 50.0 25.0 25.0 25.0 25.0 100.0 110.0 120.0 130.0 140.0 150.0
Interest & Other Expense (1.4) (3.2) 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0
Other 0.2 4.6 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0
Taxes (7.6) (7.1) 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 159.8

Source: Pivotal Research, Company Reports

- 15 - Brian Wieser 212-514-4682 Pivotal Research Group

Snap Discounted Cashflow Model

SNAP DISCOUNTED CASH FLOW MODEL

In $mm Except for Share Data FY17E FY18E FY19E FY20E FY21E FY22E FY23E
Adjusted EBITDA (773.5) (804.4) (382.9) 62.1 594.5 1,196.0 2,039.1
Capital Expenditures (100.0) (100.0) (110.0) (120.0) (130.0) (140.0) (150.0)
Acquisitions (100.0) (100.0) (110.0) (120.0) (130.0) (140.0) (150.0)
Cash Raised in Equity Offerings 2,834.8 0.0 0.0 0.0 0.0 0.0 0.0
Cash Raised in Debt Offerings 0.0 0.0 0.0 0.0 0.0 0.0 0.0
Other Changes in Cash Balances 0.0 0.0 0.0 0.0 0.0 0.0 0.0
Changes in Cash Flows 1,861.2 (1,004.4) (602.9) (177.9) 334.5 916.0 1,739.1

NPV of Future Cash Flows (929.2) (516.0) (140.9) 244.9 620.6 1,089.9

Sum of Future Cash Flows 369.3

NPV of Terminal Value 11,622.7
Terminal Value (Based on 2023 FCF): 23,001.2
Value of Future Cashflows 11,992.0
Value of NOLs 1,237.3
Plus: 2017E Cash+Marketable Securities 2,848.6
Value of Cashflows, Cash and Investments 16,077.9
Less: Debt 0.0
2017E Common Equity Value 16,077.9

Shares Outstanding Plus Granted RSUs 2017E 1,576.9

Equity Value 2017E (Per Share) 10.00
Current Equity Value (Per Share) 17.00
2017E Equity Value Premium
Vs. Current Price -41.2%

KEY ASSUMPTIONS
Near-Term Discount Rate 8.1%
Long-Term Growth Rate 6.0%
Long-Term Discount Rate 12.0%

Source: Pivotal Research, Company Reports

- 16 - Brian Wieser 212-514-4682 Pivotal Research Group

Appendix: Important Disclosures
Analyst Certification
I, Brian W. Wieser, hereby certify that the views expressed in this research report accurately reflect my
personal views about the subject company and their securities. I further certify that I have not received
and will not receive direct or indirect compensation related to specific recommendations or views
contained in this research report.

Legal Disclaimers
Pivotal Research Group LLC is an independent equity research company and is neither a broker dealer
nor offers investment banking services. Pivotal Research Group LLC is not a market maker for any
securities, does not hold any securities positions, and does not seek compensation for investment
banking services. The analyst preparing this report does not own any securities of the subject company
and does not receive any compensation directly or indirectly from investment banking services.

Stock Ratings
Pivotal Research Group LLC assigns one of three ratings based on an expectation of absolute total return
(price change plus dividends) over a twelve month time frame. The ratings are based on the following
criteria:

BUY: The security is expected to have an absolute return in excess of 15%.

HOLD: The security is expected to have an absolute return of between plus and minus 15%.

SELL: The security is expected to have an absolute return less than minus 15%.

Ratings Distribution
Pivotal Research LLC currently provides research coverage of 50 companies, of which 48% are rated
BUY, 38% are rated HOLD, 12% are rated SELL, and 2% are not rated. Our company does not offer
investment banking services. This data is accurate as-of 3/2/17.

Price Chart and Target Price History

N/A

Other Disclaimers
Information contained in this report has been prepared from sources that are believed to be reliable and
accurate but are not guaranteed by us and do not represent a complete summary or statement of all
available data. Additional information is available upon request. Furthermore, information and opinions
expressed are subject to change without notice and we are under no obligation to inform you of such
change.

This report has been prepared solely for our institutional clients. Ratings and target prices do not take into
account the particular investment objectives, financial and/or tax situation, or needs of individual
investors. Investment decisions should take into account all available information, not just that which is
contained in this report. Furthermore, nothing contained in this report should be considered an offer or
solicitation by Pivotal Research Group LLC to buy or sell any securities or other financial instruments.
Past performance is not indicative of future performance and estimates of future performance contained
in this report are based on assumptions that may not be realized.

Material in this report, except that which is supplied by third parties, is Copyright 2017, by Pivotal
Research LLC. All rights reserved. No portion may be reproduced, sold, or redistributed in any form
without express written consent of Pivotal Research Group LLC.

Commission Sharing Arrangements

- 17 - Brian Wieser 212-514-4682 Pivotal Research Group

Pivotal Research Group LLC has commission sharing arrangements (CSA) with numerous broker-
dealers. Please contact Jeff Shelton at 212-514-4681 for further information.

Additional Information Available Upon Request

- 18 - Brian Wieser 212-514-4682 Pivotal Research Group