Description

I believe T-Mobile (“TMUS”) is an attractive investment. While many investors fret about the likelihood of a Sprint/TMUS deal getting antitrust approval, I view the stock as tremendous value irrespective of whether the deal gets done. In a year, either the deal will be complete and you will have an attractive synergy-rich and “four to three” industry consolidation story, or the deal will be scuttled and you will have a capital-return story and an investment in the U.S. wireless industry’s secular share taker. In either situation the stock should be up meaningfully (50%+) over the next few years.

My key thesis points are:

1. TMUS is a structural winner in U.S. wireless over the medium-term given historical network investments and latent share gains
2. Broader wireless industry is better than investors appreciate
3. Great “story” no matter what happens with Sprint – either a synergy story or a capital return story
4. Great management who have created a lot of value through intelligent capital allocation
5. Under a base standalone case, very attractive EPS growth profile (~20%) at only a market multiple
6. Tons of asymmetric upside cases including a Sprint deal, relaxation of the competitive environment, and latent margin opportunity

Business Description:

TMUS is the 3rd largest U.S. wireless provider behind AT&T and Verizon, and ahead of Sprint. They have 75M customers nationwide, including 56M branded phone customers (35M postpaid, 21M prepaid), 5M other branded postpaid devices (largely tablets), and 15M unbranded wholesale customers which utilize Sprint’s network through M2M or MVNO relationships but are managed by other parties.

While historically the network has been perceived as lagging Verizon and AT&T’s (particularly in coverage), after significant investments in the last few years it is now comparable to Verizon’s (slightly better or slightly worse depending on who you ask), ahead of AT&T’s, and far ahead of Sprint’s.

They have differentiated themselves as being the “un-carrier” which in practice is an effective underdog strategy focused on introducing various customer-friendly initiatives to contrast themselves against AT&T and Verizon. As a result, they are often perceived as being a disruptive competitive influence on the industry, forcing others to copy their moves and negatively impacting the economics for other industry players over the last few years.

TMUS is controlled by Deutsche Telekom (“DTE”), which owns 64% of the outstanding shares.

Business History/Situation:

T-Mobile has operated as a subsidiary of DTE since 2001. Throughout the 2000s it was a pretty distant #4 player with <10% market share that struggled to compete effectively against the top three players. In September 2012 John Legere was hired as CEO. In October 2012 a transformative acquisition was announced with publicly-traded MetroPCS, giving TMUS additional scale and a large portfolio of prepaid customers. The deal closed in May 2013 through a reverse-merger, creating the structure that exists today.

In March 2013, the “un-carrier” strategy was introduced and since then TMUS has taken on a number of customer-focused initiatives as part of this strategy, including:

• Eliminating annual service contracts and providing plans without caps and overage charges
• Reducing international calling rates and roaming fees
• Reimbursement of early-termination fees for customers porting in from other carriers
• Automatic roll-over of high-speed data
• Free use of select audio and video streaming services without it counting against data caps
• Giving a share of stock to customers who refer a new active account
• Introduction of T-Mobile ONE, including unlimited calls, text and data
• All-in pricing including service fees and taxes

This strategy, along with significant investments in the network over the last 5 years, have made TMUS a large share gainer, with postpaid market share going from ~10% in 2012 to over  ~16% currently. In an industry where customers usually stick with their carrier for 5-10+ years, and market shares typically only move gradually as customers get traded between the carriers, this is a very large move, and was only achieved through TMUS reducing its monthly churn from 2%+ down to 1% and from capturing over 25% of gross adds since 2014.

A tie-up between Sprint and TMUS had been widely speculated for some time given the large cost benefits (only having to operate one network rather than two) and competitive benefits (moving from four providers to three) likely to be present in such a deal, as well as the presence of two controlling shareholders (Softbank for Sprint, TKE for TMUS) amenable to a deal. There were on-and-off discussions between the two companies throughout 2017 and 2018 culminating in the announcement of a deal in April 2018. In the week following the announcement, TMUS stock declined by 10% (note: it had run-up a bit earlier in the month on deal speculation), on some combination of investor skepticism the deal would pass anti-trust review, a more conservative pro forma outlook (driven in part by miscommunication on synergy timing and presence of one-time costs in forward estimates), and a turnover of the shareholder base as capital return-focused investors were replaced by event-driven investors. In the subsequent four months, the stock has recaptured this decline but much of this has been driven by market/industry moves rather than TMUS-specific re-rating.

This brings us to the current situation, where I believe many fundamental investors are writing TMUS off as a “deal-stock” with regulatory risk when in reality I believe it has a very attractive return profile regardless of what happens with the Sprint deal.

1. TMUS is a structural winner in U.S. wireless over the medium-term given historical network investments and latent share gains

I believe TMUS is a situation where perception is lagging reality. Before digging into the name, my view of TMUS was that of a disruptive challenger offering an inferior product at a discount price. I believe many investors and consumers share this view, however the reality is that TMUS has made material investments into their network over the past few years and is now offering a comparable product (if not better) than Verizon and AT&T, at a lower rate. This is evidenced by churn rates continuing to fall, and now approaching  VZ/T levels, as well as continued improvements in brand image (although this is also a lagging indicator), and industry-high NPS scores – they are not a hated company like VZ/T. They also recently invested in a substantial amount of low-frequency spectrum (ideal for densification and therefore, 5G capabilities), purchasing 45% of the 2017 600MHz spectrum auction for $8B. It is important to note that this is a latent benefit that is not currently showing up network quality, given that most phones are not yet configured to utilize this spectrum. However, over the next few years this will be an enabler of a step-change in network quality, setting TMUS up well for 5G deployment.

Even without a step-change in network quality, TMUS was already poised for continued share gains. The concept of “equilibrium” market share is important here. Basically, for a given churn level and share of gross adds taken, there will be an “equilibrium” level that will eventually be reached where TMUS’s share of total customer losses (through churn) will equal their share of total customer gains (through share of gross adds). Over the last 4-5 years, TMUS has consistently captured ~25% of gross postpaid adds while reducing annual postpaid churn from ~20%+ down to ~12-13%.

However, despite churn approaching industry-levels and taking ~25% of gross postpaid adds, TMUS still only has a 16-17% share of postpaid customers. Why is this? Basically, in an industry where the average customer life is 5-10+ years, it takes a very long time for market shares to converge to their “equilibrium” levels, however, over time it will happen.

I’ve calculated equilibrium market shares for TMUS at a variety of churn rates and gross add shares below. Basically, if churn rates were to remain constant and if gross add share was to slip a bit to 22.5%, there is still a very large latent share gain story as TMUS share will converge towards 21.5% over time from ~16-17% today – a +30% increase in customers and a larger increase in EBITDA given the largely fixed nature of network costs. With TMUS’s new spectrum coming online I believe it is reasonable that TMUS closes the churn gap with VZ/T completely and goes to 10%, and that share of gross adds increases slightly to 25-27.5% - in this scenario, the terminal market share is 27-30% or ~70-80% upside to customer counts over time (and again, more than that to EBITDA).

2. Broader wireless industry is better than investors appreciate

One pushback to owning TMUS is that the wireless industry is perceived to be a capital-intensive, fiercely competitive industry, with limited growth now that everyone already has a smartphone. I do not directly dispute any of this. However, looking at it in this way ignores the fact that the “fierce competition” has largely come from TMUS itself (making this a negative thesis point for an investment in VZ or T, not for TMUS) and that in the aggregate, the industry has consistently earned its cost of capital, with aggregate EBIT increasing commensurately with recent increases in invested capital (EBIT for the four majors + predecessors has increased 35% over the past five years vs. a 32% increase in tangible invested capital over the same time period). This is not a case of profitless prosperity, but rather a case of incumbents choosing to make value-creating investments and capturing a reasonable return on them. Furthermore, the above characterization ignores the many attractive attributes of the wireless industry – I could just as easily describe it as a highly recurring, non-cyclical, mission-critical, subscription model with insurmountable barriers to entry for anyone thinking of building out a new nationwide network. It would be a bit facetious to do so, but not inaccurate either – ultimately there are both positive and negative attributes for the broader industry.

3. Great “story” no matter what happens with Sprint – either a synergy story or a capital return story

While a true value investor may not care about a stock’s “story”, it certainly doesn’t hurt having one in order to get other investors interested as well. Currently I believe TMUS is between “stories” and this is partly to blame (thank?) for the current opportunity. Basically, because we have a deal that has a ~50/50 likelihood of going through and a regulatory process which is highly unpredictable, TMUS is in the throes of maximum uncertainty. In situations like this I believe most investors choose to “not play” and avoid uncertainty altogether rather than to risk-weight the various outcomes and construct a probability tree.

However, if we roll-forward a year, this uncertainty will be resolved one way or another, and there will be an attractive story for TMUS regardless of what happens with Sprint. If the deal closes, there is a massive synergy and pro forma earnings story. I will get into some light math in point #6 below, but suffice to say it is not hard to get to ~$100+ price targets a few years out. There are a variety of sell-side analysts who have attempted the pro forma math and I’ve seen ranges of $95-130 depending on how far out they go.

If the deal does not close, there is an extremely attractive capital return story which I believe is underappreciated. TMUS will not be a cash tax payer until 2024 given tax reform and existing NOLs, and at current net debt levels of 2.4x and significant EBITDA growth ahead, there is lots of room to take on additional debt. Recognizing this, the board has already authorized a $9B buyback over the next 3 years. While the Sprint deal is in process they cannot execute on this plan, but if it breaks they have committed to undertaking a large catch-up buyback to get back on pace. Given that DTE already owns 64% of TMUS shares, this $9B buyback represents 46% of the remaining float at current prices. Given the cash flow that TMUS will generate going forward, on a standalone basis it starts to look a lot like a creeping go-private.

4. Great management who have created a lot of value through intelligent capital allocation

The hiring of John Legere in September 2012 was a complete watershed moment for TMUS. He is a rockstar. Since he was brought onboard, the stock is up 4.6x (29% CAGR), dramatically outperforming VZ (6% CAGR), T (4% CAGR) and S (2% CAGR). He has created a truly distinct culture and identity for the firm, which even if “put on” has been remarkably effective at creating a cohesive culture and differentiation in the minds of customers. Glassdoor shows a 95% CEO approval rating and 4.1/5.0 rating for culture & values, compared with ratings of 40-68% and 3.0-3.4 for VZ/T/S.

So far management and the board appears to have been intelligent about capital allocation – they were not afraid to invest aggressively to drive profitable growth over the last few years, and now that they are starting to gush cash flow, are considering meaningful cash returns to shareholders, primarily in the form of buybacks.

Legere is very well-incented, with outright ownership of 1.6M shares (~$100M worth) and another ~$90M in potential awards that have not yet vested, and his short-term incentives are based on meaningful metrics that will create value for the company. He has never sold a single share on the open market other than to pay withholding taxes.

Management also has a history of guiding conservatively, and typically raise their guide 2-3x per year – they have never missed their guidance or taken down their ranges in 5 years. The sell-side community has caught on somewhat and holds them to a higher standard than their guide, but there is likely upside to company estimates.

5. Under a base standalone case, very attractive EPS growth profile (~20%) at only a market multiple

Since the beginning of 2017, TMUS stock has gone nowhere despite continued execution. EPS estimates have been revised up by ~25% and EBITDA estimates have remained steady, resulting in multiples contracting substantially, with 2-year forward P/Es coming in from ~23x to 14-15x. TMUS now trades at 16.3x FY19 EPS and 14.0x FY20 EPS – these are GAAP EPS estimates and I believe there is some upside to these numbers.

This valuation basically puts TMUS at or slightly-below the S&P market multiple, yet:

• S&P forward earnings growth is expected to be ~10% p.a. vs. ~20% near-term for TMUS (note: I believe this is easily achievable given consistent and persistent share gains)
• S&P earnings are highly cyclical, with a likely ~35% peak-to-trough contraction from a typical recession vs. a highly-defensive earnings stream from TMUS
• S&P earnings are adjusted vs. GAAP for TMUS

Basically, we have an opportunity to pay a market multiple for a near-term ~20% earnings algorithm, based only on conservative model drivers consistent with recent history. This does not require a transformational M&A event to occur. This does not require any improvement in the current competitive dynamics. It is merely the result of a very attractive, secular growth algorithm driven by sustainable trends that causes EBITDA to grow at MSD+ rates.

6. Tons of asymmetric upside cases including a Sprint deal, relaxation of the competitive environment, and latent margin opportunity

In addition to an attractive outlook on a standalone, “business as usual” basis which I believe will yield ~50% upside over a few years, I believe there are a lot of ways TMUS could generate “home run”-like levels of upside, including completing the Sprint deal, a relaxation in the competitive environment, and a latent margin/pricing opportunity.

Completion of Sprint Deal

Wireless M&A transactions are inherently very accretive given that the costs of building and operating a nationwide network are largely fixed in nature. In the Sprint deal presentation, TMUS called out $6B of synergies achievable over a 5 year basis, with an NPV of $43B. Note that this compares to a Sprint PF EV of ~$60B – in effect they are buying the Sprint standalone business net of synergies for only ~$17B. 60% of these synergies are from network consolidation which are very clear and tangible. While a transaction of this size is certainly unique, in their transformative MetroPCS deal, TMUS ended up being their synergy target by more than 40%. This deal also makes significant sense from a network optimization perspective - Sprint has a very large chunk of 2.5GHz spectrum (mid-band) which is essentially the “missing piece” for TMUS and is extremely complementary to TMUS’s own low-band and high-band spectrum.

While many are skeptical of the likelihood of a four-to-three deal being allowed given antitrust concerns, I believe there are some unique aspects in play here which increase the probability a bit. First, TMUS has been very careful about specifying that synergies here are not job-related. In fact, they claim that there will be direct jobs created, not lost, as a result of the merger. Whether this logic holds up to second-order thinking is unclear (there will certainly be some jobs lost at the tower or network equipment companies), but it’s a strong talking point. Second, the Sprint spectrum will allow them to turbo-charge 5G network development on a faster schedule than they otherwise would. This will spur broader industry-wide 5G development (light a fire under T/VZ) and will therefore create significant positive externalities outside of the wireless industry, the same way that 4G development laid the groundwork for many services we take for granted today. Finally, they claim that the deal will be pro-competitive and good for consumers – included in their pro forma numbers is the cost of migrating Sprint customers onto lower-cost TMUS plans. Again, it’s unclear whether in the long-run any four-to-three deal truly ends up being pro-consumer, but they are clearly taking the right actions to attempt to convince regulators that this situation is different.

You can get to some pretty home-run scenarios if management’s pro forma medium-term targets materialize (let alone the more aggressive long-term targets). Their targets imply $16.8B of EBITDA-Capex in 2021 with another $2B of synergies yet to be captured. At an 11x EBITDA-Capex multiple (equivalent to 7x EBITDA, consistent with historical industry multiples), this would result in a $200B EV company (adjusting for further integration costs) which equates to a ~$115-120/sh price target by the end of 2020, or ~$100 discounted back to today. As mentioned, this math is broadly consistent with sell-side pro forma analyses which get to $95-130 price targets depending on how far out they go.

Relaxation of the competitive environment

It is worth noting that the wireless industry has not always been as competitive as it is now. In fact, I believe we are pretty close to peak competitiveness driven by both TMUS and Sprint. The reason I believe we could be at peak competitiveness is that over the next few years, both TMUS and Sprint’s incentives will change. Re: TMUS, a smaller company is always more incentivized to be disruptive and competitive as compared to a larger company for two pretty obvious reasons: 1) they have less to lose (less customers means e.g. reducing prices has a lower $ effect) and 2) they have more to gain (the ability to gain share is more material if there are more potential customers to gain). While I don’t expect TMUS to dramatically change their positioning, I believe this means that on the margin they are more likely to “play nice” going forward as their share starts to approach the low 20s than they have been over the past few years when they were in the LDD share range and struggling to be relevant. Re: Sprint, I believe they have been particularly and uneconomically competitive in the past year (e.g. free service for a year if you port in from a competitor) in part to maximize their nuisance value for strategic reasons. In other words, if you can threaten to continue to make life hell for everyone else, you’re more likely to get someone to agree to acquire you on favorable terms to stop the pain. Even if the proposed TMUS/Sprint deal is disallowed, Sprint should lose some of the incentive to be as aggressive as they have recently been, as being an uneconomic nuisance is only of value if someone is able to pay you to stop it.

Latent Margin/Pricing Opportunity

A significant portion of TMUS’s SG&A is comprised of variable selling expenses that are directly associated with growth. As a result, TMUS operates with considerably lower EBITDA margins than peers (it’s tough to get apples-to-apples numbers but the gap appears to be in the range of 1000-2000bps on service revenues). Therefore, in a theoretical moderating growth environment, there is meaningful margin opportunity. Per comments made by management, “if our growth was to moderate, which it is not going to … you will see an explosion in margin with a significant reduction in SG&A”.

Somewhat tied to the above, there is also a latent pricing opportunity – as a pricing leader, TMUS maintains a pricing umbrella versus peers. If management ever decided to rebase pricing, perhaps as its network perception gets more in-line with peers and/or once market shares get into the 20s, there would be meaningful profit upside to capture.

Conclusion:

To conclude, I think TMUS presents a situation that is being structurally ignored given the hesitation of investors to get involved in situations which a high degree of uncertainty. As a result, a patent investor who is willing to look past the next year of deal-related headline noise can look forward to very attractive returns over a multi-year period. In the event the deal is consummated, there is significant pro forma upside driven by large network-related synergies and a potential relaxation in the overall competitive environment. In the event the deal is scuttled, TMUS turns into an attractive capital return story with 20% EPS growth over a multi-year period. Heads I win, tails I win. That’s some coin.

Risks:

• Increased competition from T given cross-promotion of TWX content
  • It’s possible, although I’m highly skeptical that the T/TWX combination will result in any meaningful cross-selling synergies – revenue synergies in large deals like this almost never pan out and the logic for this deal appeared to be driven primarily by T trying to diversify away from a poor competitive position in wireless rather than any inherent industrial logic
• Increased competition from cable companies eager to get into the space
  • This will certainly happen to a degree, cable companies are already capturing ~3% of gross adds. CMCSA is selling a solution through a Verizon MVNO and CHTR will be doing the same. Other than cross-selling/bundling it’s not clear what their value proposition is – they don’t’ have their own spectrum, and it’s unclear why customers will clamor to expand their relationship with a cable company they hate unless given a very good economic reason to do so. Longer-term if cable really wants to get into the space, TMUS/S likely becomes a consolidation candidate for CHTR or CMCSA given it’s the closest to a wireless pure-play and therefore more likely to pass anti-trust scrutiny
• Key man risk
  • Legere is 60 so I guess this is a risk, although I think he would miss trolling his competitors constantly

Catalyst

• Sprint deal being allowed
• Sprint deal being disallowed