Description

I know Charter is a familiar name to most people on this site but given the recent stock performance and the noise being made around fiber and fixed wireless, I figured it would be worth writing up again and restarting the conversation around the name and industry. 

This thesis can be broken down into a few key points/assumptions. If you agree with these, I think the stock can return ~17% CAGR over the next 5 years:

1. Fiber and fixed wireless are changing the industry (no more monopoly), but Charter can continue adding internet subscribers, even if at a slower pace. Statements by fiber and 5G operators about taking cable share just aren’t showing up in the numbers and converting legacy DSL subscribers to fiber/5G is flattering gross adds.
2. Switching video subscribers for wireless increases the value of the bundle and the stickiness of the customer. 
3. Charter’s EBITDA margins are only going in one direction (up) as video subscribers drop, wireless scales, and customer service costs continues to decrease on a cost per customer basis.

Background

Charter is the 2^nd largest internet provider in the U.S. (behind Comcast), and also provides video, voice, wireless, and business services.  In 2016 it obtained significant scale with the acquisition of TWC/BHN, and while it took a few years to digest the acquisitions, the business has grown passings, revenue, customer relationships, and EBITDA over the last 5 years.  They carry a significant amount of debt (~$93bln net), with a target leverage ratio of ~4-4.5x EBITDA.  To stay at this leverage goal, they repurchase shares using free cash flow and issue incremental debt as necessary.  Summary of historical financials shown below.

1. Cable vs Fiber and fixed wireless

Cable

The cable thesis hinges on future competition from fiber and fixed wireless.  This is new to cable, who have operated in a near monopoly for much of the last decade (DSL was an inferior technology).  Fiber already covers ~35% of Charter’s footprint and with well-funded overbuilders, this is expected to increase to ~55% by 2025 (research my MoffettNathanson).  Fiber is a superior technology, but I think the market is overreacting to slowing internet adds as tailwinds from the pandemic and internet adoption subside. 

Let's start with the customer; what do they care about?  They want an internet provider that offers a product that keeps up with their requirements (multiple connected devices, streaming, gaming etc.) and they want it at a reasonable price.  Some consumers (e.g., serious gamers, content creators, and general tech nerds) will covet fiber given its superior upstream capability, but overall, symmetrical speeds aren’t needed for the average home; coax can easily provide enough upstream bandwidth for casual online gaming, Zoom meetings etc.  Cable does, however, have a cost-efficient path to 10gb downstream and 6gb upstream with DOCSIS 4.0.  While this is a year or so away, testing has been positive with Comcast recently showcasing these speeds in lab tests.  They are also making incremental improvements to the network with existing technology (DOCSIS 3.1); this includes upgrades such as high splits (that allow for more spectrum to be dedicated to upstream performance) and wiring fiber deeper into the network. Simply put, coax is flexible enough that they can continue improving performance, keep ahead of customer needs, and avoid digging up yards to lay FTTH. 

The key is they are still comfortably ahead of consumer requirements, there’s no widespread activity that requires fiber; both cable and fiber usually offer the same top tier download speed ~1gbps.  Charter can also deploy FTTH in targeted areas where they see churn picking up (a hypothetical gaming community for example).  This does increase execution risk, but Charter has the advantage of a large footprint and can assess overbuilders neighborhood by neighborhood, learning lessons from each one. Management has proven themselves more than capable operators over the years and can be trusted to execute.

Cable is also the incumbent.  The median length of home ownership in the U.S. is ~16 years.  There are ~80mm owner-occupied homes, so ~60% of households are living in their house for 16 years+.  I’m generalizing here but if the average household is getting the performance they need and switching won’t save them a significant amount of money, there’s no major incentive to switch.  Having someone come to your house and dig up your yard for a similar product thats costs the same just isn’t that appealing.  Despising a cable provider could be a reason a customer is willing to go through this; I’ll address that in the video section after one comment here.  People don’t hate their cable provider because of ~$60 a month internet that everything else in their home relies on.

Fiber

So, what are fiber operators saying about their overbuild plans?  Pretty much every quarterly call with a fiber overbuilder has the same sequence of questions:

Analyst: “where are you taking share with fiber, are you taking share from cable?”

Company: “we are taking share from cable, fiber is better, we expect to continue taking share”

It’s true fiber is adding connections at solid rates, but taking share from cable?  Both Charter and Comcast have increased penetration and passings while fiber has been deployed.  It’s possible that fiber overbuilders are taking share from small cable operators and taking a chunk of the adds from new builds, but the most obvious place they’re getting customers is from their legacy copper subscribers.  Some Q1 numbers:

Looking at these numbers with no context, there’s an obvious question.  Why is AT&T adding ~5x as many fiber subs as Verizon?  They both have similar fiber passings and penetration (AT&T ~17mm and Verizon ~16mm).  Adding the context… AT&T has ~8mm non-fiber internet connections they can chase as they overbuild, with only a few hundred thousand for Verizon.  Upgrading these subscribers to fiber is still good for AT&T, the avg. rev per customer is ~40% higher for a fiber subscriber and the market values a fiber passing ~5x higher (using Frontier as a proxy).  It also stops copper subs being taken by cable and, given they already have the customer in their ecosystem, they have a high chance of conversion. 

But there's a big difference between converting a current DSL customer to fiber and taking a cable customer. The cable customer likely has all the performance they need, the price of fiber probably won’t be cheaper, and they’re potentially bundled in with other products (voice, video, wireless).  Acheiving 40% penetration in areas where you already have customers using your DSL product is one thing, trying to acheive the same without that tailwind will be another. I think the attractiveness of overbuilding will shrink as the low hanging fruit (copper customers) dwindle down.  There's ~15mm DSL and VDSL connections left in the U.S. (per a recent LightReading article). Assuming ~30% of these are in Charter’s footprint, that’s ~5mm customers (~9% of Charter’s passings) that overbuilders will likely be targeting before they have to truly take significant share from cable.

I think Verizon is probably a good proxy for true fiber share gains as they have a very small amount of legacy copper subs.  In 2021, they had ~150k net adds.  Assuming those are all from cable, on ~16mm passings that’s ~1% share gain (assumes 0 added passings).  Add in the fact ~100k of these came from the footprint where they compete with Altice, and I think the narrative around fiber quickly eating up cable’s core base is misleading.  Fiber can edge in slowly over the years and a 50/50 split between fiber and cable is a realistic end point, but for fiber to take significant share quickly, I think coax has to become obsolete.  I haven't seen any evidence this is going to happen any time soon. 

Fixed wireless

5G allows for wire type speeds without the need for a wire in the home.  I see this as a threat around the edges for cable, potentially reducing the economics of their edge outs but not a huge threat to their core base.  The census is that the C-band is the sweet spot for fixed wireless, as it can cover miles without much interference, produces acceptable speeds (~300mbps), and can be rolled out quickly.  This has led to some impressive add numbers; T-Mobile added the most broadband connections of any internet provider in Q4 2021.  The same scenario as described above plays out on earnings calls though, with claims of taking share from cable; from Q1 2022 T-Mobile earnings:

“Demand just continues to build from dissatisfied suburban cable customers to underserved customers in smaller markets and rural areas”. 

Research from MoffettNathanson’s (based on Comlinkdata) tells a different story:

“T-Mobile's FWA subs significantly over-index in rural areas, and in areas where DSL is the only wired option… Although T-Mobile has indicated that its 5G Home Internet product is taking share from cable operators, Comlinkdata found that T-Mobile's FWA subs substantially over-index in ILEC-only areas, and meaningfully under-index in markets where both cable and fiber-fueled services from telcos are available.”

Taking share from cable is a narrative telcos want to push (they think the market wants to hear it), and while it’s true they’re adding in some cable areas, the number of true cable customers they’re taking, i.e. someone who was signed up to Comcast or Charter, is significantly smaller. 

Price, on the other hand, is a real risk that could put downward pressure on cable pricing; ~$50 is an attractive offer and while I don’t believe the comments from T-Mobile about $50 forever, it’s a smart marketing ploy (acquiring an internet sub is the hard part).  Capacity restraints should help cable, with T-Mobile being selective about where they offer 5G.  Some of their comments hint at potential capacity issues:

“Average users are using 300 to 400 gigs a month. We have mid-single digits using more than a terabyte… Their (cable) averages are only higher because they have some 10%, 20% of people that use multiple terabytes… we can support some of that”.

Data usage is rising rapidly and dropping your cable tv package sends the average data usage up to ~700gb a month. If you’re taking share from cable, they’re probably dropping their cable tv package as well… Average wireless data usage is ~30gb a month, but this includes data consumed using WiFi.  If ~10gb of this is on cellular networks, that means each FWA customer is worth ~1/50th of a wireless customer (on a per gig basis).  While they have excess capacity that's fine, but I don't see this as the threat that will take down cable.  Having a coax wire to the home (that's already there) is just a better and more efficient way to deliver internet.      

Charter did make some comments on 5G during their Q1 2022 call, claiming they saw 5G customers as a "parking lot": 

"I talked a little bit about to the extent that fixed wireless access is building up potentially in the areas that we're moving into from a rural perspective, that's in the future DSL pile or parking lot of subscribers that we can go acquire. So it's DSL today, it's VDSL and it's inferior broadband product relationships that are getting created today that provides that fuel for growth for us in the future from a mix standpoint."

I'm not sure I agree.  While this might be true long-term, 5G is a significant upgrade to DSL/VDSL and those subscribers will be more difficult to pry away once they've switched to 5G (assuming adequate performance). So, for the same reasons I like cable as the incumbent, I don't think dismissing 5G as a "parking lot" is correct.   

I see T-Mobile having a large first mover advantage in this space and collecting DSL subscribers quickly in areas where cable and fiber aren’t.  There’s potential their low prices could move some customers away from cable, but having to balance other priorities like capacity and their mobile offering should allow cable to keep its market leading position. 

2. Video and wireless

Another risk to cable is that customers despise them so much after years of price hikes, they’re willing to deal with the hassle of switching now there's other options.  Altice is an example of this; in the footprint where they compete with Fios they have been bleeding subscribers.  Since the Drahi acquisition they’ve pushed price, cut expenses, and outsourced customer service.  They’re now overbuilding themselves with fiber (hail Mary if you ask me), but if your customer hates you enough and another option comes along, they might be willing to deal with the switching costs. 

I don’t think Altice is a good proxy for the rest of cable industry, especially Charter; Charter has U.S. based customer service and hasn’t pushed price to the same degree as others.  Additionally, this is where I think losing video subscribers might actually help.  Video pricing is where customers get upset with cable providers.  Historically, a customer would pay more every year, had limited other options, and the product was getting less valuable.  Now there's other options like virtual MVPDs, skinny bundle, or no live tv at all.  Options help reduce customer resentment towards cable providers and this portion of the business is low EBITDA margin anyway.  As customers cut the cord, I think customer satisfaction increases; removing a potential fiber selling point ("we're not cable"). 

There is a potential replacement for video in Charter’s bundle, and it’s currently their fastest growing revenue stream. The launch of a wireless offering has been a huge success; they’ve signed ~4mm lines since 2018 and with an average of 1½ -2 lines a household, this is still only ~5% household penetration.  Operating as an MVNO (on Verizon's network) has meant they can scale quickly; the cable wireless model has managed to remove the biggest hurdle for previous MVNOs, negative brand perception.  Customers may complain about their cable provider, but the speed at which they’ve signed up to Charter and Comcast’s offerings show they value the brands.  There’s a lot of customers who wouldn’t go near Mint, Cricket, Boost etc., but offer them a discounted bundle price attached to a quality brand like Spectrum or Xfinity and they’re willing to make the switch.  They haven't disclosed much in the way of details about the agreement with Verizon, but the timing of the updated agreement, at a time when they were also negotiating with AT&T would hint the economics skew in Charter's favor.  The length of the agreement also appears to be long-term, with Verizon EVP Ronan Dunne claiming: "we've created the conditions to ensure we're their (Charter's) primary provider of connectivity for the foreseeable future".

A family of four is going to pay ~$200 a month for 4 lines and 400mbps internet with a Spectrum bundle (post internet promotion and BYOD).  A telco just can’t offer that value. Cable only has to maintain their cable network, whereas telcos have to maintain a wireless network, roll out 5G, and build out fiber.  It’s true cable has to pay for data by the gig, but ~70% of phone data is consumed via WiFi (pre-pandemic), and CBRS spectrum gives Charter optionality to offload traffic onto their own network if the economics make sense.

The wireless offering is allowing Charter to switch out relationship straining video for value add mobile. Management has also talked about how mobile is lowering churn. I expect this is especially helpful if someone is moving apartments, but staying in Charter's footprint (around 1/2 of moves in the U.S. are crosstown and 75% are intra state).   

One note on the recent joint venture announcement with Comcast.  I'm not sure how it will go. I think management are good stewards of capital, so I'm assuming the math makes sense, but I don't think it's a requirement that the joint venture be a success for this thesis to workout.  

3. EBITDA margins

This section can be kept relatively short.  EBITDA margins will continue to grow faster than revenue for the following reasons:

1. Low margin video subscribers continuing to cut the cord.  Cost per video subscriber will continue increasing but gross programming costs should stay relatively flat over the next 5 years. 
2. Costs to service customers will continue to decrease (even if at a slower rate that previously). Gross cost to service customers has gone down over the last 5 years, despite Charter adding ~5.5mm internet connections.  The current year will see a slight increase due to heightened bad debt expense and an increase in the starting wage to $20, but should revert back to decreasing on a per customer basis.   
3. Wireless is approaching scale and margins should turn positive in the next year or so.

Even if Charter doesn’t push price, this should still get EBITDA margins to >40% over the next few years. 

Valuation

Key assumptions shown below.  Passings growth slows as edge outs become less economic due to 5G, offset by RDOF passings (~1mm locations by 2029).  Internet penetration stays roughly flat, and internet adds avg. ~500k a year throughout the forecast (~2.5mm total).  Revenue per Internet customer increases as wireless grows, offset by video revenue declines (I’m using internet customers as a proxy for cust. relationships).  Commercial revenue I’ve projected to continue growing at ~4%.  EBITDA margin increases due to the reasons discussed above and share count gets reduced to 127mm by 2026 (I’ve assumed ~18% increase in buyback price throughout the forecast, consistent with forecasted price), including 1% annual dilution for stock comp.  Net debt at the end of the forecast is ~$102bln.  I’ve also added a more detailed model in the appendix. 

These assumptions mean Charter would generate ~$26bln of EBITDA in 2026 and ~$10.4bln of free cash flow (assuming 40% EBITDA to free cash flow conversion).

Using a 9x EBITDA multiple gets us to ~$235bln EV and ~$130mm equity value.  This would be ~8% fcf yield and ~$1k share price in 2026, up from ~$475 today and an IRR of ~17% through 2026.  There's potential upside on the multiple here if cable shows it can hold its own in the new oligopoly broadband market (~5% fcf yield potentially on the upside).

Appendix

Links

Home ownership

https://ipropertymanagement.com/research/average-length-of-homeownership#:~:text=The%20average%20length%20of%20homeownership,for%2040%20years%20or%20more.

Owner occupied homes

https://policyadvice.net/insurance/insights/home-ownership-statistics/#:~:text=21.,occupied%20homes%20in%20the%20US.&text=Owner%2Doccupied%20housing%20in%20the,million%20of%20them%20back%20then.

DSL customer

https://www.lightreading.com/broadband-tech/thoughts-on-evolving-battle-among-us-fiber-cable-and-5g-providers/a/d-id/777372?

Comcast DOCSIS 4.0 test

https://www.lightreading.com/cable-tech/how-comcast-is-paving-road-to-10g/d/d-id/776103

MoffettNathanson fiber overbuild

https://www.telecompetitor.com/will-cable-broadband-market-share-decline-as-telcos-deploy-fiber/

MoffettNathanson 5G

https://www.lightreading.com/5g/t-mobiles-fixed-wireless-sub-base-skews-heavily-rural---study/d/d-id/776745

Wifi usage

https://www.lightreading.com/cable/cable-wi-fi/an-inside-look-at-cables-mvno-business-model/d/d-id/752938

Charter RDOF funding

https://dgtlinfra.com/charter-communications-invests-5bn-broadband-rdof/

Moving trends

https://www.move.org/moving-stats-facts/ 

Catalyst

Buybacks

DSL customers start to disappear and fiber gross adds slow

Continued EBITDA growth