Streamsong, Florida

December 2017

@Find_Me_Value (twitter)
 Goals for Investing
• Earn above-average RoR, typically 10-20% in this environment,
depending on how well I know the company
• Invest only in companies I know well (no excuses)
• Minimize errors
• Have fairly concentrated portfolio of 8-15 holdings
• Search for after-tax ROR, which lends itself to longer term holdings, as
well as if an investment performs well, relative hurdle rate for
replacement investment becomes higher due to tax reasons
• Long-only

Streamsong 2017
 Why 8-15 Holdings?
• Less than, and I feel uncomfortable, reckless to an extent
• More, and hard to find enough attractive ideas, earn solid RoR
• Also - one man band, not efficient to try to own too many, only so
much time in the day
• Many companies don’t qualify for what I am looking for, based on my
personality and return goals

Streamsong 2017
 What Shapes My Process?
• Realization that there is no certainty
• Things can change, including competitive advantages
• As a passive shareholder, management becomes crucial, as they can
destroy capital, or compound capital effectively, even with a solid
underlying business
• A good management team cannot overcome a business with poor
economics, poor prospects, in a tough industry (horse vs. jockey)
• Learn from past mistakes is an excellent way to improve, but can also
mislead you if you have biases
• Willingness to wait and buy a company I truly want to own
• Belief I can compound money by minimizing errors more than picking huge
winners

Streamsong 2017
 Process
• Start with the business / industry I am interested in
• Learn about the businesses, the industry any way I can
• Do I think there are businesses in this industry that I would be attracted to,
excluding valuation?
• Primary focus is on understanding the business inside-and-out
• Why the customers purchase from them, and is it sustainable
• What sort of competitive advantage protects them from easy entry of new competitors
• How sticky is the customer relationship
• What protects their ability to earn attractive returns on incremental capital deployed
• Decide on what I think is a fair range of valuations, and an what price would I
become interested
• What is the bear case? What could go wrong?

Streamsong 2017
 What Type of Companies
• Reasonable, not heroic, revenue growth (important)
• Example: AZO (AutoZone) – 24x bagger from 2000-2013, yet revenue only increased by 6%
per year (5% from store increases)
• Revenue growth driven in some part, or most, by volume (shows
continual demand for product, i.e. runway)
• Pricing power, or potential pricing power, but within reasonable, and
is sustainable without crippling their customers
• Still in a growth mode, but not early stage explosive growth
• Strong incremental returns on capital is highly important

Streamsong 2017
 What Type of Companies
• Legacy ROI can be very unattractive, yet company can be scaling and contribution
margins be exceptional and capital efficient; or, legacy ROI can be very attractive
but limited ability to grow going forward (like a bond / fixed annuity stream)
• Best for me: (a) runway for reinvestment, i.e. demand is there and visible (b)
company is capable of capturing some of that demand (c) can capture at strong
incremental returns on capital (d) typically implies management is very
competent
• Ideally the business can continually reinvest all FCF and would earn very high
rates of return; however, some companies, once they scale, earn excess FCF and
are capital light, and have more flexibility (i.e. share repurchases)
• Share repurchases as a use of FCF is excellent as long as growth runway is in tact,
adequately funding to monetize and capture this demand, and company is highly
cash flow generative

Streamsong 2017
 Examples:

SiriusXM (SIRI)
 SiriusXM (SIRI)
• Legacy ROIC was terrible - company almost went bankrupt post-XM
merger in 2009
• Very high cost structure, high fixed costs, high SAC (>$120 per
installation pre-2006)
• Contribution margins 70%+, EBITDA margins now close to 40%
• Runway for reinvestment was: increased penetration in new vehicle
market, increased conversion, churn through product improvement,
and now used car market
• Current strength is content provider, not just music (a commodity)
• Exceptional management added returns through higher FCFPS growth
> FCF

Streamsong 2017
 SiriusXM (SIRI)
Pre-Merger:
• XM had $3.19bn in operating
losses from 2000 – 2007
• SIRI had $4.14bn in
operating losses from 2001 –
2008
Pre-Merger

Post-Merger:
• From Q1 2009 – Q3 2017,
combined SiriusXM has had
$8.25bn in operating income

Streamsong 2017
 SiriusXM (SIRI)
• SiriusXM had/has:
• Solid product
• Monopoly of satellite radio, news, talk, sports, and other content
• Combined entity went from poor ROIC to very high ROIIC once scaled
• Now >70% contributions margins (revenue – royalty on music – customer service – billing –
COGS equipment)

Leveraged high subscriber acq. costs = Higher margins due to scale efficiencies (high = Excess FCF, and now ~$9bn share repurchases
ROIIC) as subs grow since 2013

Streamsong 2017
 SiriusXM (SIRI)

Went from < $0.50 stock price to $5.44 (11/26), which is > 1,100%
Even last 5 years = +87%

Stock price largely a reflection of the strength of the business model.

Streamsong 2017
 Examples:

Visa (V) and MasterCard

(MA)
 Visa / MasterCard
• IPO’d late 2000s each
• Very strong brands, globally recognized
• Benefited from strong network effects
• The more people that had cards, the more appealing it was to merchants to accept it, the
more banks wanted to partner with them due to brand recognition, ease of use “anywhere”
• More merchants accepted, the more people would use, which would drive additional
payment / processing volume over the card network “rails” but also benefits each party
involved
• Both were able to leverage their fixed cost base and capital / opex-light
incremental transaction to scale margins
• Visa EBIT margins went from 34% in 2007 to 66% last quarter
• MasterCard EBIT margins went from 34% in 2007 to 55% recently

Streamsong 2017
 Visa: Costs per Payment Transaction

Streamsong 2017
 Visa / MasterCard
• The network infrastructure can handle multiples more than current volume
• There is no additional capital needed to process the next incremental transaction
• Capital is spent on data centers, IT, R&D, security / fraud prevention
• Visa tangible invested capital is ($4bn) and will have around $13bn+ in EBIT next
year

2007 – Current

Streamsong 2017
 MasterCard
• ($1bn) in tangible invested capital and will earn about $6bn+ next year
• EPS / FCFE-per-share almost 2x revenue growth due to running additional
transactions over a low variable cost / steady fixed cost base

Streamsong 2017
 Current Examples:

Comcast (CMCSA) and Charter

Communications (CHTR)
 The Cable Business
• Not nearly as capital light as Visa, MasterCard, or SiriusXM; however, cable can earn very
high returns on incremental capital invested and has a decent runway for
reinvestment….
• Very few competitors due to legacy monopoly contracts with municipalities, led to ~ no
overbuild. Very little overbuild of one cable company over another (current example:
“WOW”)
• Very high capital costs to build out the infrastructure to provide products
• Equipment / cables on utility poles
• Digging trenches, laying cables
• Hundred of thousands of miles of cable (HFC coaxial, some fiber)
• Equipment costs – set-top-box, modems
• Other: trucks, testing equipment, customer service representatives, headends, amplifiers, nodes
• Initially very high capital expenditure to build out with no promise of any ROI
• In most areas today: one legacy telco / one cable operator / DISH and DirecTV
Streamsong 2017
 The Cable Business
• On a legacy ROIC standpoint, the business appears unattractive
• High capex needs
• High incremental capital needs
• Initially = no customers
• In mid-to-late 2000s, cable companies were struggling
• Satellite had national reach, thus had video scale in negotiating with programmers
• Satellite had superior video as cable was all analog, one-way (no PPV, DVR, etc.)
• DirecTV had NFL Sunday Ticket, which was costing them hundred of millions but attracted
subscribers
• Home broadband needs and expectations were far lower
• Each technological change called for another roll-out of new set-top-box equipment to
support it, which was very expensive, very disruptive to the customers
• Cable companies were non-competitive with satellite companies

Streamsong 2017
 The Cable Business
• The times have changed…..
• Cable companies went from analog to digital, opening up bandwidth for more
channels, more HD offerings, DVR, PPV, etc. and thus had a comparable video
product to satellite
• Furthermore, the cable coax-fiber infrastructure was advantaged
• Satellite: unable to provide adequate strength and capacity to support desirable broadband
speeds
• Telco: old twisted copper infrastructure built only for land-lines, and now DSL, are inferior
versus HFC cable
• Last 2-3 years: cable has taken substantial market share in video and home broadband from
telco and satellite

Streamsong 2017
 The Cable Business
• The greatest opportunity are the homes that cable infrastructure is already
connected to (legacy spent capex) and yet there is not a subscriber
• In other words: increasing penetration on the large capital base leads to very high
incremental returns on capital
• Costs per product
• Video: Can provide this product at a ~35-40% gross margin (but declining – this is a fact)
• Broadband: minimal “COGS”, and thus 90-95% gross margin
• Phone: similar to broadband, just not as lucrative
• Differences in strategy as to how to maximize the historical large capital spent on
infrastructure between cable operators

Streamsong 2017
 Cable Companies: Difference in Strategy
• Cable ONE (CABO)
• Focused on broadband-first, video is not a necessity
• Strategy has increased EBITDA margins to best-in-industry, as broadband requires far less capital
(no STB, truck rolls, less testing equip.) and less operating costs (less call center reps)
• Some of this likely due to their size and minimal gross margin on video
• Altice (ATUS)
• Focusing mostly on cutting costs (potential underinvestment in people, customer service)
• Expanding FTTH buildout to >5m HP (majority of footprint, ~100% of Cablevision) instead of
upgrading to DOCSIS 3.1
• Charter Communications (CHTR)
• Will finish all-digital in TWC and BHN, then DOCSIS 3.1 to get up to 1GB speeds, then 2-3 years
DOCSIS 3.1 Fully Duplex (to get up to 10GB symmetrical speeds)
• Comcast (CMCSA)
• To have DOCSIS 3.1 (1Gb) completed by E2017, likely rollout FD3.1 by 2018-2019
• Incremental cost to get to 10Gb is minimal, and far less than building out a network (VZ and small
cells, for ex.)

Streamsong 2017
 The CHTR and CMCSA thesis
• Legacy capex spent is now being utilized far more efficiently due to the improved
product set and technological advantages over their competitors
• This combination is leading to high incremental returns on investment as they
increase the penetration of their footprint
• ~75-80% of the capital spent by CHTR and CMCSA is “success based”, which most
investors don’t understand
• Many investors want to see a cutback in capital spending to free up FCF for dividends or
share repurchases
• However, the more capital spent (on a customer relationship and homes-passed basis, not as
a % of revenue basis as many refer to), the faster the company would grow
• CABO spends about half as much per customer on CPE (customer premise equipment) but
are also growing subscribers at a fraction of CHTR and CMCSA
• This is ultimate concept of hoping a company can reinvest large sums of capital and earn high
incremental returns on that capital

Streamsong 2017
 Cable Capital Spent
• Maintenance capital needed:
• There are some maintenance costs per customer, and on the infrastructure
• Repairing headends, fixing coax due to weather, erosion, etc.
• Replacing set-top-boxes or modems that break, or remote controls for the TV
• High incremental returns on capital:
• The large costs are already spent regardless
• The other capital spent is due to a “home” transitioning into a customer
• New set-top-box for the house
• Remotes, broadband modems
• The technician installing (truck roll) the service, connecting the cables, etc.
• More customers in an area leads to some bandwidth congestion = invest in more nodes, amplifiers, or
extending the fiber
• At this point, the best thing that can happen is the customer remains a customer for as long
as possible (lower churn), which maximizes the lifetime value of the subscriber and capital
spent

Streamsong 2017
 The Opportunity
• CHTR has 27m relationships on a total of 50m homes passed; thus, there are 23
million homes that have had large capital spent and would be very high
incremental returns on that capital if they became a customer of CHTR
• CMCSA has 29m relationships on 57m passings; i.e. 28m passings that are “could
be subscribers”
• The real growth engine over the last 2-3 years, and will continue to be:
• Broadband subscriber growth
• Business / enterprise subscriber growth

Both of these lines of business are very margin accretive!

Streamsong 2017
 The Opportunity:

Streamsong 2017
 Brief Views on Each – CHTR and CMCSA
• Neither are dirt cheap, or punchcard investments, at these prices
• There is some still uncertainty and headwinds, and potential likelihood of much
lower rates of return than I hope
• Both fall in to: think both can earn 10% ROR, with potential for higher returns
(CHTR more upside, CMCSA more downside protection)
• CHTR is a pure play cable operator, CMCSA earns about 75% of EBITDA from
cable. In my view, CMCSA is trading far more based on the cable prospects than
anything else.
• Being concerned about minimizing errors, I may be proven to be conservative on
my ~10% CAGR over the next 3 years on these companies
• Would not be surprised if both achieved 12-15% CAGR if they operated as they
can

Streamsong 2017
 Charter Communications (CHTR)
• Stock price of $330 versus a high of $408
• Around 10x EBITDA (LTM)
• Opportunity exists due to:
• Concern over video subscriber losses finally becoming similar to other non-cable distributors,
due to an acceleration of cord cutting
• 5G uncertainty
• Uncertainty as to whether cable HFC infrastructure can be supportive for 5G fixed wireless
• Leverage ~4x debt/EBITDA
• Potential M&A by “four suitors” now off the table, it seems
• Legacy TWC churn issues and the timing of going all-digital when it appears video landscape
is worsening

Streamsong 2017
 Comcast (CMCSA)
• $180bn cap / $231bn EV (stock price $36/share)
• 2.0x net debt / EBITDA
• Cable business: $52bn revenue / $21bn EBITDA / $13bn EBITDA – CapEx
• NBCUniversal: $32bn revenue / $8bn EBITDA / $5bn EBITDA – Capex/ Software
• Opportunity exists:
• Cable companies got hit beginning with the early September comments by CMCSA about
losing 100k – 150k subs in Q3 due to hurricanes and “highly competitive marketplace”
• Inferred that DirecTV / AT&T was expanding and being highly promotional (possible to a point
of no margin on some products)
• Comcast has highest video penetration of large cable companies at ~39% and is very focused
on growing video / bundling still (i.e. maybe they have the most subscribers to lose in the
transitioning of large bundles to a la carte products and OTT video)

Streamsong 2017
 Comcast (CMCSA)
• Thoughts on 2018:
• Programming costs will be far more normal
• Capex is elevated currently due to rolling out X1 STB (mid 50% penetration, peak out around
75%) and then CPE costs should decline per customer quite a bit
• MVNO with Verizon is not even considered much yet (250k subs from May – Sept)
• Impact of MVNO on churn, additional product bundling (Hulu, NFLX) and broadband is not
discussed, but should be.
• Share repurchases of $5bn
• $10.5bn - $11.5bn in FCFE and 4.638b S/O = $2.25 - $2.50 FCFE/PS
• Range of ~ 14x – 16x 2018 FCFE/sh.

Streamsong 2017
 CMCSA Valuation Summary
• Paying ~15x – 16x FCFE is reasonable for a business that:
• Is run by highly capable management – look at NBCUniversal acquisition, for example
• Underleveraged compared to industry at 2x net debt / EBITDA versus 4-6x
• Is investing heavily in their long-term core products
• Broadband: XFi app to manage broadband completely, XB6 wireless gateway, bundling wireless (unlimited
$45/mo) with broadband, adding DOCSIS 3.1 to all passings by E2017 to get up to 1Gb speeds, extending some
fiber to the homes, adding Hulu, SlingTV, Netflix to Xfinity user interface
• Still growing broadband subs 4-5% with upside pricing power for broadband-only subscribers
• Total cable revenue growing 5-7%, yet most of the growth is from (a) video pricing vs volume,
which is margin accretive (b) broadband and business services, which is margin accretive
• Potential to see EBITDA growth of high single digits over next 3 years given product mix,
MVNO churn potential, less programming burden
• EBITDA – CapEx will grow faster than EBITDA with CapEx now peaking and will be coming
down.
• Possibly >10% CAGR on FCFE basis, with business diversification, well managed,
underleveraged
Streamsong 2017
 Comcast (CMCSA)
• Cable networks and broadcasting strength is underappreciated due to headlines
of cord cutting, concern/uncertainty around skinny bundle inclusion and
economics
• Theme parks are not discussed enough, but is a very strong asset, and growing
• Bought remaining stake in Universal Japan
• Added new water park mid-2017 in Orlando – Volcano Bay
• Superior product currently (in my experience) over Disney due to cost, IP, crowds, quality of
the rides
• Comparable in cost to Disney on a daily-pass basis, but heavily under-monetized on an
annual pass basis
• Fresh new content potential (?): James Bond ride, Harry Potter Fantastic Beasts, Jason
Bourne, replacing the Fear Factor theater, Super Mario world
• Add 1-2 new rides per year to drive attendance
• Reality is “US” has been taken share from Disney each year, which is why Disney is finally
reinvesting back into their Orlando parks (see: Avatar, Star Wars world)

Streamsong 2017
 What Is the Downside Concern?
Risks:
• Cord cutting accelerating, impact on video margin, programming costs, and
broadband-only churn
• Slowing down recently of broadband subscribers
• 5G residential wireless as a replacement of cable fixed broadband
• Antitrust lawsuits currently with AT&T / Time Warner, and what that implies
about larger scale M&A in the cable space
• Leverage

Streamsong 2017
 Risk: Cord-Cutting
Concern:
• Cord cutting shifts customers away from the bundle
• Cable loses scale in terms of negotiations with cable networks, thus programming costs-
per-sub increase
• Elevated churn, what happens to relationship with other products – broadband?
What will likely happen:
• Video gross margins already have been declining, yet EBITDA margins have improved due
to faster growing broadband and business services
• Less video subs = lower CPE costs (look at CABO CPE per CR costs) and less opex spend
on customer service (look at CABO headcount trend)
• Potential to offer skinnier packages, or bundle OTT products (similar to wireless) but
increase broadband pricing
• Video will be a loss-product in 2-3 years as is
• Potential to increase pricing on broadband for broadband-only subscriber, also through
increasing speeds, upgrading customers to faster tiers (higher ARPU)

Streamsong 2017
 Video Margin vs. EBITDA Margin

Source: LBRDA Investor Day 2017

Streamsong 2017
 Risk: Cord-Cutting
Potential Upside:
• Video GM has been in decline. Faster higher-margin mix may increase
profitability per subscriber
• CHTR could still grow subscribers, or be flat due to:
• 100% of losses in Q3 were low value, low quality basic subs
• Yet, actually had growth in high tier, expanded basic video subscribers
• Only 41% of TWC and BHN customers have updated pricing and packaging
• Household growth and footprint expansion of 1-2% per year could lead to ~flat video sub
growth, but lower video penetration levels

Streamsong 2017
 Risk: Broadband Subs Slowing
Concern:
• Broadband growth will start to taper off
• Maybe due to competitiveness of AT&T, some unlimited wireless shifting on the fringes,
and reaching closer to full penetration in US of broadband
• Lower broadband sub growth means they will rely more on pricing going forward, which
may create regulatory concerns, or additional competitiveness in larger markets with 5G
/ fiber build out
What I think:
• Still ample room for CHTR and CMCSA to grow broadband, despite AT&T 14m HP
buildout plans by July 2019 (from 6m currently)
• At ~45-47% penetration, still could get to ~55%
• Additional tailwind of expanding footprint
• Cable highly advantaged as lowest cost for them to increase speeds to 1Gb and 10Gb
versus VZ, AT&T
• DSL and vDSL still will constitute largest overlap in their footprint, which cannot compete

Streamsong 2017
 Risk: 5G
• Much talked about, especially with Verizon’s recent announcements
• Reality:
• Is at least 3-5 years out before any impact
• Possibly uneconomical in most places
• VZ only targets 20% -30% penetration on 30m homes; i.e. only certain markets “make sense”,
not most
• 5G is still somewhat capital intensive (equipment, truck roll, installation, what about inside
the home?)
• Still too uncertainty to guess true economic impact, much more of a scary “what if” as of
now that is too early to know; potential minimal impact

Streamsong 2017
 How does CHTR / CMCSA fit in to Process
• Both have above average management, have show ability to operate well, be S/H
friendly
• Businesses are above average in terms of quality, due to stickiness of customer,
necessity of products, limited true alternatives
• Competitive positions are strong, not to say they are unable to erode, just that
current positions and foreseeable future – in a good position
• Fair pricing, could earn decent returns under modes assumptions; but if
companies executive ‘as they believe they can’, potential for mid-teens ROR very
capable
• Runway for reinvestment – broadband, business services, untapped pricing in
some markets on broadband, low cost provider for higher Mbps vs. telco (DSL,
copper) – and majority of capex is success-based = higher return on incremental
capex spent
• Current uncertainties due to some of the headwinds offers more margin of safety
Streamsong 2017
 Comcast (CMCSA)
• Similar secular tailwinds as Charter (CHTR)
• Increasing demand for data (video, 4K, streaming, multiple devices, higher quality, IoT, etc)
• ~85% of data consumption is fixed, not wireless
• Broadband and business services growth will offset video sub declines; however, video
margins have already been in a free fall, yet cable margins have been improving despite due
to efficiencies and business mix – this should continue
• Broadband still has room for increased penetration from mid-40% to mid-50% despite AT&T
fiber expansion, extra growth coming from footprint expansion of 1% - 2% / year
• Broadband pricing has pent-up potential – but more so in some areas (less competition) than
others
• 5G may be a competitive threat, but not in entire footprint, and not anytime soon
• Cable will bundle things differently to drive customer growth and/or increase satisfaction
(lower churn), such as wireless + broadband + OTT product (which reinforces value of strong
data speeds)
• Capital intensity on a per-customer relationship, per-home passed will come down due to
efficiencies, set-top-box strategy and costs, cloud-based, and business mix (broadband lower
capital intensity – no truck roll, self-install, modem only, minimal customer service req.)

Streamsong 2017
 Comcast (CMCSA) and Charter (CHTR)
• Fair valuation – not super cheap but not expensive
• Multiple is far from expensive, potential for it to hold (or expand) over next 3 years
• Underlying FCF/sh. Growth is satisfactory enough to provide double digits RoR
• Multiple expansion potential due to continued competitive strengths, less concerns about other items (doubt 3 years
from now “cord-cutting” will be as prevalent a theme as there will be more disruption, strategy shifts by cable)
• Lower capital intensity a secular trend and based on more efficient use of customer premise equipment – investors
should cheer lower capital intensity, have less fear in historical large cap-ex ramp ups outside current levels of
spending

Streamsong 2017
Thank you.