Description

LSXMA is far from a unique idea – one can look at the various write-ups the past 5 years highlighting the same opportunity.  While the LSXMA shares have increased over this time span, the discount to NAV has ballooned.  While down slightly during 2021, and considerably narrower relative to some of the crazy Reddit days in January, it still looks far too wide at just over 30 percent currently.  As has been widely reported, LSXMA will likely own over 80 percent sometime this year (roughly 77% as of 03/31/21).  Post the 80 percent threshold, SIRI will likely pay a larger dividend and LSXMA will likely use most of the proceeds to repurchase stock.  SIRI will continue repurchasing shares, and LSXMA hinted it may sell into the repurchases to retire LSXMA shares.  While a short-term merger is possible beyond 90 percent, arbitrage players specializing in appraisal rights could make a deal difficult.  Additionally, current and former SIRI executives own large amounts of SIRI stock and it could be awkward to cram current/former management.   Whatever the path, it seems highly that Liberty will close the discount over time.  Assuming 2022E free cash flow per share somewhere between $0.36-$.40, LSXMA is buying back its SIRI shares at an implied “look through” price of 10-11x free cash flow.  This seems remarkably cheap for a quality business.  I own SIRI unhedged, believing SIRI can continue to generate slow but steady top-line/free cash flow growth, that LSXMA will continue narrowing the underlying discount to SIRI shares.  Additionally, there may be some positive optionality on what the ultimately LSXMA/SIRI entity does with the free cash flow.  

From the time of the first iPhone through the current SPOT offering, there have been substantial questions about numerous services replacing the basic satellite radio service.  Yet, despite the rise of listening options/improved cell phone service/etc., SIRI has continued to grow subscribers, maintain conversion rates and reduce churn.  Additionally, with the recent addition of Toyota, and expanded agreements with GM, BMW and KIA, SIRI new car penetration has now risen to above 80 percent in 2021.  The new and expanded car deals do beg the question that if satellite radio is clearly in decline, why have OEMs chosen to expand their relationship with SIRI?  And despite a pandemic greatly limiting car commutes, SIRI still managed to grow self-pay additions during 2020.

From telephone books to video connections, there are admittedly several examples of stable telecom businesses that were disrupted by new technologies and it is still possible such a change occurs here.  Critics can also argue that a domestic only service will be unable to compete with a service that can spread content costs over a global footprint.  But, there just does not seem to be a lot of evidence this is occurring.  

In addition to the raw company KPIs, various customer service surveys suggest customer satisfaction results remain high.  Morgan Stanley does annual survey work (~2000 total participants, splits for gender, income levels, urban versus suburban, etc.).  While one can discuss sample size/other issues, I think the survey works generally paints a picture that:

• SIRI users are a satisfied group, spend a good chunk of their “listening time” using SIRI and use the service outside the car (~60%) more frequently than one might expect – encouraging, given the company’s efforts to expands its base of streaming customers.  Again, all of this has occurred despite the rise of SPOT over the past several years.  

• Lots of Apple, SPOT, and AMZN users seem to also use SIRI and the percentage has increased over the past couple of years.

• Surprise, surprise, the typical SIRI skews older than other streaming services

Certainly, one can argue the survey results are more anecdotal or that the market is rapidly evolving, and these results will look different in 5 years.  Maybe.  Another read is simply that the survey report seems to corroborate the actual subscription and churn data and simply reflects a sticky customer base who continues to value the subscription service.  It will be worth monitoring how these surveys evolve as 360L is rolled out throughout the base.    

And the 360L angle may be the most underappreciated part of the SIRI story.  SIRI has been competing against the scaling streaming services with one arm essentially tied behind its back given the slow rollout of 360L.  360L whose two-way data connections offer real-time customer data should offer a far better user experience, allow far more targeted advertising/other services and conceivably give more accurate (and cheaper) royalty data when the Sirius’ royalty rates come up in 2027.  The impressive ARPU/churn statistics have been achieved with rather pedestrian improvements in the actual product over the past ~15 years– 360L should represent a far more meaningful change.  From a data perspective, SIRI knows the make/model of vehicle, blue book value of the vehicle, whether the previous owner subscribed to SIRI and can make some basic guesses on the age/sex based on name, etc. – but this would likely be considered dark age stuff relative to the other streaming heavyweights.  With 360L, SIRI will know which songs the customer listens to (and conceivably a per song rate could be less expensive from a royalty perspective), how long they use the product, what time of day, etc.  It is difficult to think that the vastly improved data does not lead to some improvement in conversion rates, churn and possibly ARPU.  It should also open possibilities for adding ad supported channels for the ~60% of customers that choose not to subscribe to SIRI’s service – further expansion of Pandora stations will likely be part of these efforts as well.  The add-supported expansion concept could also be supported by freed up low band spectrum likely available in the next 5-6 years.  With only ~2 million 360L vehicles relative to a 130 million+ fleet, there is little side-by-side comparative data now and it will take several years to fully compare 360L vs. non-360L.  That said, the rollout should pickup considerable momentum with 360L anticipated to be installed in 25 percent of trial starts in 2021, rising to 80 percent by 2025.

From a royalty perspective, SIRI pays a mid-to-high teen rate for the various musical composition and sound recording royalties.  Pandora pays closer to 55 percent and is awaiting the results of the Web V review which could result in a stepped-up rate for its non-direct sound recording royalties.  As has been well discussed, Pandora continues to bleed daily active subscribers and there is considerable uncertainty on when/if these trends improve.  While Pandora has some of the same royalty challenges that investors have debated with SPOT, its focus on an advertising offering does result in a lower overall payout rate (obviously the latter is growing substantially, Pandora is shrinking).  Interestingly, SIRI believes they very well could be paying more on a per song basis versus SPOT – there is not exact data for the rates they pay.  They give lists of songs played and are charged relative to the total number of channels offered.  With a vastly expanded 360L base by 2027, it is possible SIRI might be able to base royalty rates on actual play rates which could be more economic – how much of this is overwhelmed by a far higher ask rate from the cartel remains to be seen.  

On the podcasting front, there is considerably debate how large the ~$1 billion market will grow, how many ultimately listen to podcasts, etc. etc.  SIRI’s Stitcher/Simplecast/99% Invisible/Other have provided a viable service – the same customer survey works appears to show SIRI customers like the podcast offering.  SIRI thinks SPOT has vastly overpaid for its exclusive content, but time will tell.  

It is finally worth noting that 2020 reattribution likely reintroduced concerns regarding assets moving around between tracking stocks.  This asset juggling has occurred multiple times in the past (LCAPA/LINTA, QRTEA/LVNTA, LVNTA/LBRDA among others) – this is part of the deal with Liberty names.  But looking back, most of these movements have tended to work out – and it certainly looks like this is the case for LSXMA/FWONK.  In May of 2020, it seemed far less clear that vaccines would be rolled out so quickly/be so effective and while Live Nation was clearly a special asset, I thought LSXMA could have driven a harder bargain versus foregoing obviously attractive repurchases and giving up the first 30 percent upside to ~$48 in exchange for a minority interest.  Oops.  LVY’s valuation clearly prices in a stronger franchise than existed pre-COVID, but this very well may materialize as the immodest Michael Rapino notes that demand for 2022 concerts is stronger than 2019 levels.  While the exact home for LYV post a future LSXMA/SIRI merger is less clear, LSXMA is stronger for owning the asset even if this meant foregoing attractive LSXMA repurchases during 2020.  Certainly, COVID could have evolved differently and the alternative door and the swap could have been far less attractive in that scenario.   But, it didn’t and LSXMA is stronger because of the swap.  

In summary, the risk/reward at LSXMA is attractive.  There is little evidence of franchise erosion at SIRI and the service could be even stronger as 360L is rolled out.  Meanwhile, LSXMA buybacks are likely to materially increase and the discount should narrow with the accelerated buybacks. 

Catalyst

-Accelerated buybacks at LSXMA

-360L rollout