|Shares Out. (in M):||262||P/E||0||0|
|Market Cap (in $M):||78,600||P/FCF||0||0|
|Net Debt (in $M):||70,621||EBIT||0||0|
DISCLAIMER: The author of this posting and related persons or entities ("Author") currently holds a long position in this security. The Author makes no representation that it will continue to hold positions in the securities of the issuer. The Author is likely to buy or sell long or short securities of this issuer and makes no representation or undertaking that Author will inform the reader or anyone else prior to or after making such transactions. While the Author has tried to present facts it believes are accurate, the Author makes no representation as to the accuracy or completeness of any information contained in this note. The views expressed in this note are the only the opinion of the Author. The reader agrees not to invest based on this note and to perform his or her own due diligence and research before taking a position in securities of this issuer. Reader agrees to hold Author harmless and hereby waives any causes of action against Author related to the below note.
We understand this is not the most original idea, we do hope you get some good/new information from the write-up.
Charter Communications Inc (CHTR) is the second largest cable operator in the United States providing video, internet, and voice services to approximately 27.5mm residential and business customers. The company provides, through a subscription based service, a set of products that are considered essential in today’s day and age.
While CHTR provides video and broadband services to 21% of US households, its market dominance at the local level is much stronger, running local monopolies and duopolies through most of its footprint. Broadband penetration of homes passed, is currently at 49% and increasing. The company is the market share leader in 22 out of the company’s top 25 DMAs with an average market share of 61% in the markets where it is the incumbent. The average market share of the number two in those 22 DMAs is 22%. DMA level data actually overstates the competitive environment given that competition is at a municipal level.
We believe CHTR will keep gaining market share in its footprint at attractive economics. The reason for that is that iIn most of its footprint the company competes with telephone companies (or telcos) and small overbuilders who either have a lower quality product or are subscale and poorly run. CHTR’s geographic overlap with other cable operators such as Comcast, Cox, or Altice is actually very limited. The lower quality internet product from the telcos of <20 mbps – 70 mbps will become increasingly uncompetitive with CHTR’s 200+ mbps as data needs continue to increase. This is a disadvantage that the telcos/overbuilders will have issues overcoming given the uneconomical prospects of upgrading from their copper network to a hybrid coax or pure fiber network.
CHTR’s management is widely regarded as the best management team in the industry. Management, along with controlling shareholder John Malone, has developed an operating model that has rewarded and is likely to keep rewarding shareholders. Through shrewd execution and operating leverage, we believe management can grow EBITDA around 7% a year on normalized margins, which through its levered equity model, should translate into mid-teens FCF per share growth for the next 5+ years.
Despite the attractive competitive and financial position of the company, its stock is currently trading at 9x our 2020 levered FCF estimates (13x EV/EBITDA-Capex). The reason for the low valuation is (1) fear of new entrants, and (2) execution questions around the TWC/BHN acquisitions that have temporarily hindered margins. We believe the new entrants will have a difficult time taking share from CHTR while making an economic return on their investment. In terms of margins and integration, we are nearing the tail end of the integration process which should lead to an inflection in growth and cash flow.
We believe CHTR is worth around $453 per share or about 50% above the current price. Our view is predicated in the following four points:
CHTR will maintain its superior competitive position and product offering given uneconomical prospects for competitors attempting to upgrade their networks to compete on even footing.
Strong tailwinds in the broadband market will make CHTR’s infrastructure and pipe into the home more valuable despite a secular decline in video customers.
Decreased duplicate costs and integration expenses will create an inflection in cash flow generation with margins and capital intensity to revert to levels in-line with industry.
Best in class management team and a strong controlling shareholder to maximize the value of the asset and shareholder returns.
As mentioned, CHTR is the second largest cable operator in the US providing video, internet, and voice services to approximately 27.6mm residential and business customers through a hybrid fiber coax network composed of a national backbone, regional/metro networks and a “last-mile” network. Below a table comparing CHTR to the biggest telecommunications providers as well as a map of CHTR’s footprint.
As of 2017 74% of Charter’s footprint was all-digital and is expected to get to 99% by year-end. Through the digitization of the footprint and implementation of DOCSIS 3.1 the company is expected to provide 1 gbps offering as well as HD channels across its footprint. Approximately 59% of CHTR’s customers subscribe to a bundle of services.
The company has one of the lowest residential ARPUs in the industry at $110 (vs Comcast at $151 and Altice at $140). Importantly, this differential is not due to any inferiority of CHTR’s network, product or customer base, but rather a product of CHTR’s market share strategy. CHTR’s pricing power provides meaningful upside to our forecast, which assumes only 3% pricing increases mostly derived from TWC pricing roll-offs.
On the cost side, the network is a high fixed cost infrastructure with marginal costs to wire a new customer within its footprint. There are, however, a number of variable costs such as programing costs ($53 per video customer), servicing a customer (between $12-$30 per customer per year), marketing costs ($110-$440 to acquire a new customer), and truck rolls ($85 per truck roll). Capex is also a combination of fixed and variable depending on the plants and upgrade needs, current capex/sales is at 19% which we believe is temporarily elevated and should move closer to 15% in the next two to three years.
1. CHTR will maintain its superior competitive position and product offering given uneconomical prospects for competitors attempting to upgrade their networks to compete on even footing.
Cable operators, who have largely comparable offerings, have divided the country amongst themselves and do not compete with each other. CHTR’s main competition is with the telcos (especially AT&T and Frontier) and to a lesser extent with smaller overbuilders who vary regionally. Cable operators made heavy investments in the 1990s and 2000s to lay out the basic infrastructure that is currently used. During that time the telcos went to DSL by using the copper wires they had set up on poles for landline usage, which required a smaller capital investment but is an inherently worse medium to transport data. As a result, telcos struggle to provide broadband speeds and are currently capped around 70 mbps, while CHTR is expected to offer 1 gbps speeds across its footprint by year end.
This has allowed the cable operators to take market share from the telcos. As thesis point #2 will show, as data needs increase, the telco’s disadvantage will become increasingly relevant.
CHTR’s superior offering has allowed it to increase its internet penetration at a faster rate than other cable operators, going from 34% in 2012 to 48% in 2017. As we will touch upon later, CHTR management team has a unique operating model that has allowed it and should continue to allow it to gain share at a faster rate than the industry. In fact, CHTR currently represents over 50% of the total broadband industry net additions adding 1.3mm new subs in 2017.
Providing telecommunication services to consumers is a capital intensive, high fixed cost business, that benefits from high penetration levels. Providing a pipe into a home takes those dynamics and makes it relevant at the local level. This is an important distinction that makes cable a more attractive business model than other capital intensive high fixed cost businesses such as cellular phone, which competes at a national level. While the four telephone companies fight for every possible phone customer in the US, cable providers compete over a town. Building a cable system is an expensive endeavor, towns that already have one or two providers give little room for a third entrant to sign enough subscribers to make it economically viable. As a result, over a third of households in the US have only one broadband provider and the vast majority of households have no more than two.
In any given DMA (most DMA and zipcode data we have is from SNL), the second biggest player has been able to achieve a market share between 14%-22%, creating an ideal environment for monopolies/duopolies. In CHTR’s top 25 DMAs 88% of those markets have no more than two competitors with more than 15% market share. The weighted average market share of the incumbent (in almost all cases, CHTR) is 46%. The high fixed cost of building the infrastructure with low probabilities of getting customers to switch leads to unsatisfactory ROIs in markets with an incumbent. Historically, the costs and technologies have changed, but the overarching dynamics are similar. The table below to the left shows the aggregated market share rankings for charter in its top 25 DMAs, the table to the right shows the main competitor in the 22 markets where CHTR is the incumbent.
Our research suggests that the cost of laying fiber in a new city vary with a number of factors but generally range between $2k-$5k+ per passing in cities where an incumbent is lobbying to block the buildout. The biggest cost is getting the fiber to the curb. The biggest cost component is labor (which is not scalable) at 40%-90% of the overall cost. Some other major components are $30k-$50k a mile in aerial planning, machinery, equipment, and regulatory. Last mile costs can be $200-$500 excluding consumer premise equipment (“CPE”). The costs vary most with population density, lot sizes, labor costs, aerial vs underground builds, and planning requirements. The cost of the actual cable is small at $2 a mile.
In order to understand whether there is a return to be made on the $2k-$5k cost per passing let’s see what an overbuilder would stand to make. Penetration will be the key determinant in the success of the buildout. Penetration levels for the second biggest operator in a market hovers in the high teens to low twenties. In a duopoly a third entrant should expect to get less than that, but for the purpose of the exercise we can assume 14%-22% market share. This brings the cost per subscriber to between $9k and $36k. In order to get a 7% return on that investment the new competitor would have to get a $52-$210 monthly margin per customer (net of all operating costs and depreciation).
CHTR is a scaled competitor with ARPUs at $110 and 35% EBITDA margins. That means that CHTR makes about $38 in monthly EBITDA margin per subscriber making the $52-$210 required margin for a new entrant unachievable (CHTR’s numbers exclude maintenance capex which would be another $10+). Furthermore, the new entrant knows that as soon as they try to enter, the incumbent will do everything in its power to defend its market. There is a wide array of tools, such as lowering prices, locking people in multi-year contracts, and increasing marketing spend, to name a few. The incumbent has the established footprint and paying customers to weather a new entrant whereas the challenger has to come out of pocket. This has led to low returns on new build, for example, as it has been pointed out on VIC before, Verizon’s (VZ) wireline segment pre-tax return on capital are 1%.
In addition, while $2k-$5k might not sound like many dollars, the numbers can become quite large in the context of the number of passings required to wire a market. A mid-sized market with 500k households would cost $1bn-$2.5bn to wire and that is excluding all the up-front operating costs to get people signed up and using the service. Wiring 10mm households would cost $20bn+. The high degree of uncertainty, big up-front costs, and poor penetration figures make overbuilding an irrational proposition.
There are ways to expand at the margin and improve existing technologies, which is what the cable operators do. The telcos have some pockets where their pre-existing presence gives them the ability to upgrade their network at a lower cost per passing. However, these are few and far between, and can cannibalize their current offering.
Given the large geographical overlay with AT&T’s U-verse, Frontier, and other regional overbuilders, we believe there is room for CHTR to keep gaining share and increasing their penetration metrics to drive FCF growth.
2. Strong tailwinds in the broadband market will make CHTR’s infrastructure and pipe into the home more valuable despite a secular decline in video customers.
Eighteen years ago 5% of US households had a residential fixed broadband connection, today that number is closer to 81%. The strength of that tailwind and its defensiveness is such that in the financial crisis the cable industry’s revenue and subscriber behavior grew.
While the runway for overall penetration is smaller than ten years ago, there is still room to get a few percentage points a year. Broadband penetration in CHTR’s markets is comparable to that of the US. Growing the footprint at ~2% a year would bring overall penetration to 89% in 5 years (assuming no population growth). Most predict the total addressable market around ~95% of households.
What the increased penetration figures do not account for is society’s increased dependence on broadband and the fact that data consumption is increasing at a fast pace. Given CHTR’s superior offering compared to the telcos and overbuilders, the more that data needs increase, the more attractive CHTR’s offering will be on a relative basis.
Data traffic in the US has grown and should continue to grow at an exponential rate. As the below graph shows data traffic has increased 16x in the last ten year. As households keep increasing the number of connected devices, watch more videos online, buy 4k televisions, and VR gaming becomes widely available, the continued increase in throughput will be of great importance to consumers. Cisco predicts traffic to grow by 65% between 2018 and 2021 (a projection that might prove overly conservative).
As it currently stands, the best way to provide ever higher speeds to the consumer is through an all fiber network or a hybrid fiber coax network. There are few all fiber networks given the high cost to deploy them, that leaves most of the US dependent on the hybrid fiber coax network, such as CHTR’s, to bring the needed speeds. There is a clear roadmap for the cable operators to get to 10 gbps symmetrical through Full Duplex DOCSIS 3.1 with limited capital investments. The 10 gbps symmetrical is expected to be relevant for the next 10+ years. It is widely believed that legacy copper wires cannot reach these speeds without a makeover of most of the infrastructure, which would be costly. Small overbuilders with a fiber or hybrid network face the issue of a small installed base and limited resources creating a scale disadvantage.
With increased traffic needs cable’s superior product will become increasingly more attractive, giving them the opportunity to keep taking share from the telcos and other overbuilders.
3. Decreased duplicate costs and integration expenses will create an inflection in cash flow generation with margins and capital intensity to revert to levels in-line with industry.
CHTR is currently under-earning with EBITDA margins and capex/sales temporarily depressed as it finishes integrating the TWC/BHN acquisitions. As the below table shows, CHTR is currently running the worst EBITDA margins and capex/sales metrics in the industry as the company runs duplicative costs through the business and upgrades TWC’s infrastructure. These excess costs should run off sometime between 2020 and 2021. On the below table, CHTR PF is pro-forma for our estimates of normalized margins for CHTR.
We believe CHTR’s margins are depressed by 3% on EBITDA and 4% on capex vs what would be normalized margins.
Current CHTR is a combination of former CHTR, TWC, and BHN. If we look at each one of those companies prior to Rutledge’s involvement the weighted average EBITDA margin for those is 35%, which is right in line with today’s margins. That seems counterintuitive because since those margins were measured (2015 and prior depending on availability), CHTR has added new subscribers and realized cost synergies from the acquisitions. Customer growth within the footprint and cost synergies should flow through to the bottom line.
We believe the difference between the expected margin expansion and today’s margins is explained by duplicate and integration costs that are nonrecurring in nature, such as training a new sales and field force, running duplicate customer service centers while building a new insourced one, going from 13 billing systems to one, and dealing with excess inbound customer calls from legacy bill step-ups, amongst others. We can run through a few assumptions to get to an estimate of where normalized margins ought to be:
From 2015 to 2017 CHTR has added a net 2.1mm new subscriber, mainly through increased penetration. Each of those new subscribers should add at least $30 a month in excess margin above current EBITDA margins.
At the time of the TWC/BHN transaction CHTR disclosed $800mm of cost synergies. We assume that two thirds of those synergies (or $533) are on opex such as procurement, vendor costs, overhead, etc.
As the below table shows, those two numbers alone drive an incremental 3.1% of EBITDA margins.
Some of the excess costs might also be related to customer acquisition activity which is NPV positive (as backed by management’s track record and conversations with former employees).
CHTR’s cumulative capex has increased by 25% since the acquisitions. Conversely, EBITDA increased 17% in that two year period while customer relationships increased 8.5%.
The above numbers do not include a change in accrual on capex which was $820mm in 2017.
There is an abnormally high level of capital expenses in four areas of the business: CPE, digitizing the TWC footprint, transition, and above average newbuilds. We can make assumptions in three out of those four categories to help us get to what a normalized level of capex should look like:
CPE: If we assume that in a steady state the cost of spending on CPE should have increased by 8.5% (in line with sub growth), the spending would have been $2.87bn (instead of $3.38bn) freeing ~$510mm in capital. That would decrease capital intensity by 1.2% of sales.
Transition: The company breaks out transition expenses for us at $489mm. If we assume that 25% of those are either CPE or costs that will stay in the business, the residual $367mm would improve capital intensity by 0.9%.
Newbuilds: For the FCC to approve CHTR’s acquisitions, CHTR committed to deploy and offer “at least 2 million additional mass market customer locations” within five years ending in 2021. Originally the commission required one million of the new buildouts to be overbuilds but have now allowed it to be in rural areas. If we assume that the 1mm of rural builds is excess capex and that it costs $1,500 per build over five years, then that is $300mm of run rate capex coming off in 2021. That improves capital intensity by 0.7%
Sub-total: With the first three points we can achieve 2.8% of capex/sales expansion. To that number we can add the one third of cost synergies from the acquisitions allocated to capex (or 0.6% of sales) to get to 3.4%. That means that we would need to get 0.6% or so from digitizing the TWC footprint to get to the 4% improvement in capex/sales (from 15% to 19%) that we estimate in our pro forma.
In addition to the above, in 2015 CHTR management presented its best estimates projection of the pro forma company’s cash flows to the board. The projections were then made public in a proxy. MarAzul’s November 2017 write-up includes it and we shamelessly copy his table below. We can infer that the projections were management’s best estimates given that they went to the board.
The table has management’s projections for 2017-2019. The projections are most relevant for the capex expectations because the company controls how much its spends, has good line of sight into the next 4-5 years, knows the three companies better than anyone, and accurately predicted its 2017 projections. The company has 2019 capex dropping by $2bn from 2017 to 2019. The acceleration of DOCSIS 3.1 deployment to fend against fixed wireless might prove them off by 1%-1.5% of sales, but to be wrong by more than 3% of sales on such a predictable part of the business is unlikely.
MarAzul’s table taken from the company’s proxy
4. Best in class management team and a strong controlling shareholder to maximize the value of the asset and shareholder returns.
CHTR is run by CEO Tom Rutledge and COO John Bickham who are long time cable operators and are widely considered the best team in the industry. We actually became interested in CHTR through a recommendation from an industry executive (not by looking at 13Fs!). From there, every primary research contact we spoke to has spoken highly of CHTR’s management team. That positive perception has been backed by a history of strong financial results.
Rutledge was hired in 2012 from Cablevision, where he was COO, to help CHTR resume its growth trajectory, which had stagnated. At CHTR the amount of value creation has been impressive. The below table shows CHTR’s compounded levered and unlevered FCF/share growth from 2011 to 2017. That is despite the fact that CHTR is currently undergoing the large integration of TWC and BHN which is depressing current cash flow generation.
Rutledge’s management philosophy is different from most of the industry. In an industry historically known for poor customer experience in service and pricing, his approach is to build a better product and price it competitively to gain share. He also invests substantially in customer service. These initiatives initially increase costs, and are therefore difficult to implement in a public company. However, in the longer term can create a profitable mousetrap by reducing churn and customer service interactions, which improves the customers’ lifetime value .
Importantly we believe the strategy brings an additional layer of security to our investment. Happy captive customers can go a long way on the revenue line.
CHTR also benefits from its controlling shareholder John Malone. Malone is mimicking with CHTR the levered equity approach he has successfully implemented in different parts of his career. In a stable and growing asset such as CHTR, if shares are repurchased below intrinsic value, the accretion from the buyback is meaningful. According to our estimates, buybacks can add between 5-10 points of annual FCF per share growth before we would sell our shares.
As we go through the below, keep in mind that there is a long precedent of competition trying to take business away from cable unsuccessfully. Firms with great brands and massive balance sheets such as Google have tried and failed. Most overbuilders have been able to gain some market share, however their gains have largely been at the expense of telcos and not cable operators.
5G Fixed wireless
In November 2017, VZ announced a plan to aggressively pursue the densification of some of its footprint in order to offer a fixed wireless product powered by 5G technology. The product would reach broadband speeds of up to 1 gbps and will be marketed to consumers as a substitute to broadband internet. VZ is targeting 30mm households in dense urban areas and is hoping to achieve penetration levels between 20% and 30%.
Assuming the technology works reliably, which is still a major question, the capabilities will be comparable to CHTR’s and most of the other cable operators’ current offerings. In other words, VZ’s plan can be seen as a normal overbuild. We have spent a long time studying VZ’s plan and believe the company has a strong uphill battle achieving its stated goals. We question the strategy as the company will be targeting dense urbanized areas, which already have two competitors fighting for that customer. As we see it, VZ is going to spend $30+bn to be the third provider into the home with (in the best case scenario where the new fixed wireless technology has no limitations compared to cable) no differentiation on capabilities or price.
We think there are two questions that will determine the plan’s success:
Is it economically feasible from a cost perspective?
Can they reach the levels of penetration that they are targeting?
There is little available information on 5G making it impossible to truly understand the cost economics of VZ’s plan. There is also a lot of uncertainty as VZ’s public statements regarding fixed wireless largely run contrary to what the entire broadband industry believes about the technology. There are a number of analyst estimates, all of which show that fixed wireless is uneconomical, however, it is difficult to rely on them. What we can do is use a few data points that would help us take a view.
5G uses millimeter wave spectrum that is high throughput/low latency but that does not travel well or far. As a result, in any given footprint, a 5G provider needs to employ small cell sites (with certain technology) close to one another in order to be able to offer those speeds. The small cells then use fiber backhaul in order to transmit the data to the internet.
Given that the small cells need to be near each other, at no more than a 1 km radius, the nature of those locations are similar to the nodes present in today’s cable infrastructure. That means that in order to achieve 5G fixed wireless, VZ needs to buy, build, or rent the most expensive portion of the cable infrastructure. In other words, from the node inward (away from the home), the infrastructure has to look quite similar to a cable system. VZ has some access to fiber through the acquisition of XO communications and some organic buildouts, but their footprint is not robust. That means that on the most expensive part of the network VZ will be at a cost disadvantage.
As it relates to the other part of the network, from the node outward (into the home), which is mainly composed of the small cell (also known as ‘the drop’) and the CPE, it seems that a 5G product would likely also be at a cost disadvantage to cable. We have cost figures from a sell side report on a small wireless broadband provider called Starry. Their antenna can reach about 1,000 homes (albeit in the densest cities, likely only relevant to the densest ~20% of households) and costs $30k with installation costs at $30k. That translates to $60 per passing for the antenna in a dense area. There is also a receiver that connects the household to the small cell which costs $200 and $100 to install. Lastly the router, which distributes the wifi signal around the home, costs $140. All those costs add up to $500 per passing assuming 100% penetration for the antenna. By comparison, it costs CHTR between $300-$600 per subscriber to wire a house from the node (including CPE). These costs exclude regulatory costs, which can be meaningful depending on the municipality.
As shown above, the costs for setting up a fixed wireless infrastructure are likely higher than building out fiber. The economic prospects at a large scale look dubious.
The next question deals with whether subscribers will switch to VZ’s fixed broadband offering. VZ’s penetration assumption of 20%-30% is aggressive for the third offering into the home in the already competitive markets that it is trying to enter. VZ is a well-known brand and receiving broadband through the air might appeal to some people given the “newness” factor of fixed wireless. On the other hand, historically, customers have only changed if (1) the new product offering is superior, (2) is better priced, or (3) they dislike their current offering. In the case of VZ’s fixed wireless, the product is, best case, comparable to cable, will likely be priced in line with the current offerings in the market, and there is no sign that the customer service is suffering at CHTR. Because of these we think VZ will have a difficult time poaching subscribers from CHTR. Again, all of this assumes VZ’s fixed wireless technology is rolled out seamlessly, which we are skeptical of.
If we assume that VZ overcomes the logistical and penetration challenges, we can try to determine CHTR’s exposure under an adverse scenario. VZ has not yet disclosed all the cities and many of the selection characteristics are unknowable. However, a conservative calculation suggests that no more than ~2mm subs from CHTR should be at risk and it would be over a 4+ year period. Happy to provide the detail of our assumptions in the Q&A.
Two million subs represents about 7.4% of CHTR’s total subs. In the last two years CHTR has gained an average of 922k residential relationships per year. If successful the VZ overbuild would cut CHTR’s growth rate by half over the next 5 years.
AT&T fiber build-out
One of the conditions for the approval of AT&T’s purchase of DirecTV was for AT&T to upgrade to 12.5mm fiber locations. AT&T currently offers fiber to 7mm locations (5.6mm homes) is expected to buildout an additional 7mm locations to get to 14mm (instead of the 12.5mm required). The fiber buildout is expected to deliver 1 gbps.
AT&T has flexibility on how to provide those speeds. They can be built, leased, or acquired. They have only had limited success on previous fiber build outs, so they will likely not compete against the large cable incumbents head on.
The cities where AT&T has a copper offering are an opportunity for CHTR to take share. The cities where AT&T wants to offer FTTP are a threat. JPM estimates that AT&T has 75mm customer locations (55mm-60mm homes passed) in 21 states. According to industry sources, in 2015 AT&T had a 29% penetration on its FTTP offering, whereas total penetration is ~43%. The high penetration DMAs are concentrated in rural areas where there is no competition.
Out of CHTR’s top 25 DMAs, AT&T has at least a partial fiber footprint in 18 DMAs. As the table below shows, CHTR dominates markets that have AT&T fiber, even in the areas that have high coverage of AT&T fiber (i.e. greater than 15%). The two exceptions are Dallas-Ft Worth and Atlanta. In Dallas-Ft Worth, the two companies are competing head to head. In Atlanta, CHTR barely has a presence with 7% coverage. The below data suggests that CHTR has not lost meaningful share to AT&T fiber.