SPOT - Spotify

[Liberty]

https://traviswiedower.com/2019/07/11/spotifys-moats-management-and-unit-economics/

[chesko182]

this is problably good for the industry as it may generate more cooperation between the labels and streaming platforms (Tencent owns 58% of Tencent music, which in turn did an equity swap with SPOT where they each own ~10% of each other) There's a lot of headlines around labels seeing distributors as their enemies but I don't think that's actually the case.

 

It may bring more discipline to the industry and I think it makes sense for them to do this.

 

https://www.wsj.com/articles/tencent-in-talks-to-buy-stake-in-universal-music-group-11565078272?mod=hp_lead_pos5

[Spekulatius]

I looked at Aswath Damodaran's SPOT valuation sheet from last year and decided to update it using this years numbers. One of the issues is that the growth extends further into the future, so I decided to reduce the growth rate from Y1-Y5 a bit from 25% into 22% (to account for the year that passed). FWIW, SPOT has grown >30% during the last 12 month.

 

Surprisingly, I got a valuation that exceed the current market cap quite a bit - the value/share I got was $185 vs a current share price of $125. A few words of caution - I don't get how Aswath came up with a 21% gross margin last year, I put in a 25% gross margins for the premium subscriber part, which is actually lower than what SPOT achieved in Q2 2019 (26%).

 

Most of the value is driven by the value of new subscribers that have yet to be acquired. I just took those from Aswath's model, minus my adjustment for the Y1-Y5 growth rate. The result is somewhat surprising, but maybe it isn't and just takes into account the 30% growth from last year, while the shares haven't moved. it seems that there might be a good value to be had. I also played with some numbers and found the valuation to cover the share prices as long as they can get a 13% growth rates for the next 5 years.

 

 

I will attach the spreadsheet for those that are interested.

SpotifyModelQ2-2019.xlsx

[griezeman23]

Has anyone read the WFC report yet? My impression was that WFC is constrained by the 12m PT that they must give (and admit so multiple times). I also believe that their argument for ST nearer GMs to not improve materially is fair but that they discount too much that the labels will go nuclear - 25% of the labels revenues are now from Spotify (according to WFC). It is hard to imagine they'd go nuclear on that. Anyways, as long as management can keep the labels on the proverbial "hook", then it should allow Spotify to continue to focus on customer acquisition and growth in the ST and further entrench SPOT as the audio platform. If one believes in audio streaming (and Spotify by extension), then once at appropriate scale, one would hope for a hockey stick inflection in earnings as management renegotiates contracts with the labels and podcast producers.

 

My initial thesis is that this point is not too far off into the future (less than five years) but I have not finished my research yet.

 

Spek: any idea where Damodaran got his churn numbers??? I am seeing churn closer to 4%/month rather than 5% for the year... If so, that drastically changes the picture on the current valuation using Damodaran's DCF.

[Spekulatius]

I checked the churn numbers and you are correct - the updated monthly churn should be ~ 4.5%, which translated into 42.5% yearly. So here goes the thesis, because the customer value isomer $16 rather than over $100.

FWIW, got stopped out of SPOT at $117. I rarely do stop losses, but for those growth/ momentum stock I do as when they’re moving against you, they really can move a lot. And then, the thesis is broken ( sort ) of anyways now.

 

Working on updating  Aswaths spreadsheet for Amazon from last year as well and found some smaller issues.

 

The churn issue with SPOT is huge, so perhaps you should let Aswath know?

[griezeman23]

I think the difficulty with the churn number is: is this all premium subscribers? I.e. does the 4% monthly number include non-payers who are on-trial, or is the number purely on premium subscribers? Furthermore, the company does state that 40% of its premium subscribers who churn, rejoin within 3 months, 45% within 6 months and 50% within 12 months. So is 4% to 5% monthly churn the right number? I am not sure and hope someone in the community has some more insight than I do. 

 

I think the stop/loss for growth companies is an interesting way to play these names, as it does prevent too many losses on the downside (can be massive) for a long, long story. However, how/when do you decide to buy back in?

[Spekulatius]

I instituted the stop loss because I had little confidence in the valuation to begin with.  I usually don’t use stop losses. In terms of how to use them, I would think that on would generally look at the price momentum and get back in when the selling subsides.

 

The bigger question however in this case is that I don’t have a reason to get back into SPOT , because the thesis is broken.

 

I do agree that looking at the churn in a more granular would give more insights, but short of having these numbers, we need to use what we have and based in that it doesn’t seem like SPOT is a value right now, not even close.

 

FWIW, I think most of the churn comes from users signing up for a promotion and then cancelling when SPOT  goes to full price after the first 3 month or so.

[griezeman23]

Spek,

 

I'd agree with you on the churn with regards to free subscribers.

 

Why do you say the thesis is "broken"? Sure, in the short-term it is (no leverage with contracts with Big 3) but I'd think SPOT's SG&A is getting to scale given the last three quarters (TTM R&D expense grew 13% y/y down). I think it also shows discipline from management and a confirmation that 2019 contracts will not be renewed at more favorable rates like 2017.

 

It seems like the main point of contention is whether or not SPOT will be able to pressure the labels enough to reduce their take rate. Was 2017 a blip b/c of the impending IPO so Warner and Sony could sell their stakes? Likely but I still think it's hard to argue that labels take rates (in the LT) will remain at 70 cents on the dollar.

 

[spartan]

I've been thinking about Spotify almost to the point of insanity for the past week and can't get around their gross margin problem. Ben Thompson wrote about it a while ago: https://stratechery.com/2018/lessons-from-spotify/

 

Spotify has a lot going for itself:

1) Product is way better than anything out there (I'm a music guy..all of my millenial friends use it too)

2) User switching costs are high (they have pricing power, as evidenced by their price increases in Scandinavia)

3) There's a lot of room for growth (steady increase in user base and pricing, product tinkering will lead to better customer experience, etc.)

 

I can't get to a point where I see them holding more bargaining power than the major record labels.

 

Initially, I thought that artists would want to emancipate themselves from the labels (which would lead to bargaining power erosion and supplier fragmentation). But they don't.

 

Even worse, Spotify removed the 'direct upload' option for artists. Why? Probably to satisfy the demands of labels before the latest negotiating round. At this stage, it is almost impossible to go around the record labels. They still hold all the cards because they own the music..

 

One could argue that none of this matters. Labels and streaming services are symbiotic, label management are too dependent on cash coming from streaming services and suffer from institutional problems (if I'm a record label manager, I want to make my bonus and avoid getting fired). But labels still hold all of the cards. They own the music. Spotify will never break free from them.

 

The only bull case I see is this: steroid-induced revenue gains with steady gross margins (25%) while maintaining sane operating expenses. This can still lead to very healthy profits. But again, in my opinion, the labels will not tolerate that and can squeeze them. They sold off their equity stakes in the company, so interests are not aligned.

 

And who's to say that Spotify can be trusted with the serious task of capital allocation? Will they be kind to shareholders? Time will tell.

 

It is entirely possible that I am entirely wrong. Spotify can pull this off. (And I hope they do...record labels have caused a lot of pain to artists.)

 

But then I'm reminded of the following Buffett quote: "I don't try to jump over 7-foot hurdles: I look for 1-foot hurdles that I can step over."

[griezeman23]

Great points spartan. I have done much of the same.

 

I think Spotify is not an easy long call to make; the two main points of the bear thesis: label power and big tech competition are relatively hard to argue against.

 

Label power: I have a hard time, however, seeing either Spotify or the labels really exacting power over each other. If Spotify maintains its market leadership (Apple Music can't really expand beyond the iOS ecosystem and YouTube Music & Amazon Music both have user bases 10% the size of Spotify's; you also mention the stickiness of the product), then its hard to see the labels squeezing Spotify more. Vice versa, I also have a hard time believing that Spotify will squeeze the labels much, because their business is distribution! Thus, I'd pushback on your GMs get squeezed comment (plus they are moving into podcasts which should theoretically begin to help in the 3 to 5 year range).

 

What I have been struggling with is what does Spotify look like in 10 years? Let's assume podcasts fail to change the margin profile, we stay at a constant 25% GM and that they have an 90/10 split b/w premium and ads. JPM estimated that SPOT could have 1bn premium users based on further geo expansion by 2023 (I think this is aggressive as it implies a CAGR of close to 30% for the next five years, extending this out to 10 years leaves it closer to 15% CAGR; more reasonable given that the marginal user is now coming more often from EM not DM). At an ARPU of $4/month, this yields $48bn in premium revenues. Using our previous split, that would yield $5bn in ad-supported revenues. At 25% GMs and 5% to 10% net margin, this would yield $2.5bn at the low end and $5bn at the high end. Discounting back to today, that gets us $1bn to $2bn in current peak earnings. Thus, SPOT trades at 10x to 20x peak earnings, today. Thus, the market assuming that Spotify reaches this scale with 50%+ probability . 

 

Ultimately, this analysis is flawed, as it is simple and does not take into the account the impact of podcasts. If one does believe that SPOT can get to 35% GMs, then net margins could go closer to 15%, which would yield SPOT trading at closer to 6x peak earnings (a more interesting proposition, as the likelihood of success is now closer to 30%). Again, we need stellar management (I think Ek and McCarthy are) and belief in the company strategy.

 

Just some more thoughts and FWIW I think the Buffet quote is apt here.

 

[spartan]

Great points spartan. I have done much of the same.

 

I think Spotify is not an easy long call to make; the two main points of the bear thesis: label power and big tech competition are relatively hard to argue against.

 

Label power: I have a hard time, however, seeing either Spotify or the labels really exacting power over each other. If Spotify maintains its market leadership (Apple Music can't really expand beyond the iOS ecosystem and YouTube Music & Amazon Music both have user bases 10% the size of Spotify's; you also mention the stickiness of the product), then its hard to see the labels squeezing Spotify more. Vice versa, I also have a hard time believing that Spotify will squeeze the labels much, because their business is distribution! Thus, I'd pushback on your GMs get squeezed comment (plus they are moving into podcasts which should theoretically begin to help in the 3 to 5 year range).

 

What I have been struggling with is what does Spotify look like in 10 years? Let's assume podcasts fail to change the margin profile, we stay at a constant 25% GM and that they have an 90/10 split b/w premium and ads. JPM estimated that SPOT could have 1bn premium users based on further geo expansion by 2023 (I think this is aggressive as it implies a CAGR of close to 30% for the next five years, extending this out to 10 years leaves it closer to 15% CAGR; more reasonable given that the marginal user is now coming more often from EM not DM). At an ARPU of $4/month, this yields $48bn in premium revenues. Using our previous split, that would yield $5bn in ad-supported revenues. At 25% GMs and 5% to 10% net margin, this would yield $2.5bn at the low end and $5bn at the high end. Discounting back to today, that gets us $1bn to $2bn in current peak earnings. Thus, SPOT trades at 10x to 20x peak earnings, today. Thus, the market assuming that Spotify reaches this scale with 50%+ probability . 

 

Ultimately, this analysis is flawed, as it is simple and does not take into the account the impact of podcasts. If one does believe that SPOT can get to 35% GMs, then net margins could go closer to 15%, which would yield SPOT trading at closer to 6x peak earnings (a more interesting proposition, as the likelihood of success is now closer to 30%). Again, we need stellar management (I think Ek and McCarthy are) and belief in the company strategy.

 

Just some more thoughts and FWIW I think the Buffet quote is apt here.

 

Great points. I think that using a 10 year horizon is the right approach. That will give them enough time to build a sizable user base which, given their superior product, is not a hard argument to make.

 

Re: competition, large tech's products remind me of Rory Sutherland's TED Talk when he talks about "TV / DVD players". Large tech is good at other things, not music. (And your point about Spotify dominating non-iPhone is bang on, so Apple is at a natural disadvantage.) In my opinion, this is an important psychological distinction in the minds of consumers. Spotify does music, and does it well. Period.

 

In any event, even if Spotify wildly grows its user base (I think JPM is being optimistic with 2023 assumptions..), the cost structure will stay the same (at best) due to record label strength, which eventually dampens the profitability of user growth. Not a genius insight...

 

If Spotify makes 2 Billion FCF by 2030 we get a discounted EV of $22 Billion (15 multiple, 3.0% discount rate, 10 years).

If Spotify makes 3 Billion FCF by 2030 we get a discounted EV of $33 Billion (15 multiple, 3.0% discount rate, 10 years).

If Spotify makes 5 Billion FCF by 2030 we get a discounted EV of $56 Billion (15 multiple, 3.0% discount rate, 10 years).

 

If they quadruple revenue by 2030 ($6 Billion -> $25 Billion) and get GM to 30%, that still leaves only $7.5 Billion of GM. After operating costs, you're not left with much.

 

With current EV of roughly $17 Billion (net cash of $3 Billion), this is not a slam dunk. Come to think of it, maybe I should short it...haha

[griezeman23]

3% discount rate? Would 10% not be more appropriate?

 

I also think 15x may be a bit a low in today's environment, given that SPOT should grow faster than global population growth in the LT (mostly targeted towards younger cohorts, so death rate should be lower, increasing cohort TAM growth). We can squabble on multiples but SPOT clearly does not have the best business model relative to other consumer internet players.

 

Why I have been interested: management bought back shares in the $120s range, there have been no insider sales since the co. went public via direct listing and I like management. Agree it is not a slam dunk.

[chesko182]

You also need to account for the cash build over the next 10 years, you can't just ignore that... This is FCF positive company even though it loses money on a GAAP basis, because of the neg working capital dynamics (subscribers pay upfront, royalties go out much later). So there's certainly going to be an interesting cash generation over the next 10 years along with some operating leverage in R&D and SG&A.

 

a 7.5bn gross profit would certainly leave something after operating costs, I wouldn't describe this as "not much".. how much will they be spending in R&D and SG&A by then? prob much less on a % of revenue basis than they are now.

[griezeman23]

That's a great point and not factored in to this very basic analysis. The ARPU number is also relatively conservative, imo, given SPOT most likely has some pricing power in certain regions (Scandinavia they do!). It requires some more work on my end, but I think SPOT does look interesting here. Again, I am not sure that it is a a slam dunk given the growth (and at scale) that they need to see. But I may change my mind as I do more work.

[chesko182]

Can you think about other examples where a few suppliers choke off their biggest & fastest growing customer to the point where it becomes permanently unprofitable? I can't. I agree about the point that the label's product (the back catalog) is basically irreplaceable and a must-have as there are no alternatives but switch this point of view around -invert always invert!- and ask yourself: why would any of the three labels ever think about removing their catalog from spot? spot accounts for 25% of their revenues now, sometime in the future, this may be between 40-50%. Will they really risk losing that share of revenue (and give more revenue to their competitors) and have to deal with all the artists complaining that they stopped getting paid because they decided to remove the catalog? for what for exactly, to threaten them and get a few % more in royalties? My base case is they figure something out that works for both of them.

 

A good comparison may be the Cable Co's/Content Providers historical relationship in the 90s or so (this is broken now). Cable companies needed to have the content and content providers needed them for distribution (spot essentially owns the distribution pipes for music right now). The relationship was extremely profitable for both parties. Yes, in the case of labels there's only three controlling most of the pie, but I think it's still valid to look at it through this lens. With TV content, there were still a handful of channels that you really had to carry if you wanted to compete, and these were controlled by a few content companies (think HBO, CNN etc). So I think this will end up in a similar place when things normalize.

 

This will def be an interesting dynamic, but my bet would be they figure something out that is beneficial to both parties. One needs the other one to survive and this should eventually be reflected in their economics.

 

There are multiple paths for higher gross margins, either by lowering label payments (high bar and will take time) or other monetization methods, spot has already said its 2-sided marketplace strategy will do just this: using all the data they have and sell this to labels & artists, which would go straight to the bottom line (confirmed in their last conf call by the CFO).

 

Also think about podcasts (original content), advertising (huge in podcasts, under earning right now), concerts/ticketing, merchandise etc. Spot pretty much owns the customer relationship and when you're on top of the funnel you usually monetize better.

 

Don't think this is a 1 foot hurdle, but at the current price there's pretty much no optionality being reflected considering the multiple paths to higher gross margins they have. And whether they get to 1bn users in 5 or 10 years don't think matters that much, at that level of users I don't think any company is a pushover, less so in a business were scale is so important. If you have any confidence that they can get anywhere close to that number, I think this makes sense as a long term investment as the rest should take care of itself.

[dwy000]

"Can you think about other examples where a few suppliers choke off their biggest & fastest growing customer to the point where it becomes permanently unprofitable?"

 

I think you described the perfect example - cable companies and content.  ESPN, local sports, local retransmission, high profile cable channels - they all got so expensive that the cable companies started dropping them or moving them to higher priced tiers (where customers then dropped them).  A couple of cable companies have gotten out of cable entirely (eg CABO) claiming that it was no longer profitable given they were being choked by the content providers.  Cable is no longer a very profitable business and it is largely just retained to maintain the ownership of the customer.

 

I suspect the labels will go wherever they can get the most $ for their product.  And it's not exclusive so anywhere and everywhere as long as they are willing to pay.  Spotify is another customer for the labels (a big one but still just a customer).  The labels are life and death to Spotify.  They will bleed Spotify, if not to death, to the point of indifference if they can.

[Spekulatius]

Spek,

 

I'd agree with you on the churn with regards to free subscribers.

 

Why do you say the thesis is "broken"? Sure, in the short-term it is (no leverage with contracts with Big 3) but I'd think SPOT's SG&A is getting to scale given the last three quarters (TTM R&D expense grew 13% y/y down). I think it also shows discipline from management and a confirmation that 2019 contracts will not be renewed at more favorable rates like 2017.

 

It seems like the main point of contention is whether or not SPOT will be able to pressure the labels enough to reduce their take rate. Was 2017 a blip b/c of the impending IPO so Warner and Sony could sell their stakes? Likely but I still think it's hard to argue that labels take rates (in the LT) will remain at 70 cents on the dollar.

 

The thesis is broken because my valuation models gives me NPV numbers far below EV when I put in higher churn numbers. The business model may work, but I think the valuation doesn’t. One pretty much needs to assume that they generate growth besides the music vertical to justify the current valuation, or higher profitability or growth rates than the generous assumptions in Aswath spreadsheet. I like the management, the branding, but it seems that their profitability is restrained by the cut that labels allow it. the music ecosystem is interesting and attractive, so maybe I just need to find another way (buy a company owning the labels like Sony or Bollore) to make it work within my valuation framework.

 

Or perhaps Aswath and by extension my valuation is wrong, but I haven’t seen anything they suggest otherwise.

[blainehodder]

I just don't see how the labels retain take rate indefinitely when they are an almost completely useless middleman in music today. They are basically the patent trolls of music. The bargaining power they have today seems certain to erode. How that plays out I do not know, but useless entities tend to get cut out of profits in time. Labels ownership over living artists music seems certain to trend towards zero. T-Swift is 100% right about the direction, and new pop music absolutely dominates streams.

[chesko182]

Strong Q3 results:

-MAUs up 30%

-Premium subs up 31%

-FCF positive, Op Income positive

-Churn improved 19bps YoY

-Podcast usage up 39%

-Unfortunately CFO Barry McCarthy is retiring

-Started to announce Two-sided MarketPlace strategy:

  Sponsored Recommendations - Available to select major and independent label partners as part of our recently announced paid beta in the U.S., this is Spotfy’s first cost

per click ad product which leverages our listener graph of music tastes to promote new releases to free and paying users.

 

As always, a pleasure to read the letter:

 

https://s22.q4cdn.com/540910603/files/doc_financials/2019/q3/Shareholder-Letter-Q3-2019-[Final].pdf

[griezeman23]

Overall, a really great quarter. I did like how Spotify called out their performance relative to expectations since the direct listing and their comments about competition but disliked how McCarthy was particularly strong-worded against competition in an FT article. https://www.ft.com/content/fd0802e0-f96e-11e9-a354-36acbbb0d9b6

 

It does get me a bit concerned; however, as some of the comments seem a bit petty. On the flip side, if people keep questioning your viability as a business for, then I can imagine it does stir up some reasonably strong emotions.

 

Last thing was the commentary on podcasts was pretty poignant. I think it can be read two ways: a pump vs. a strong data point. My read is that Ek/McCarthy are great leaders, so I'd lean towards the latter argument but I have not followed Spotify (at a very in depth level) for a long enough to be 100% certain.

[Liberty]

Good podcast interview with Daniel Ek:

 

http://investorfieldguide.com/ek/

[chesko182]

A very interesting take https://venturedesktop.substack.com/p/spotify-the-ambient-media-company

[chesko182]

Q1 out.

 

"Despite the global uncertainty around COVID-19 in Q1, our business met or exceeded our forecast for all major metrics. For Q2 and the remainder of the year, our outlook for most of our key performance indicators has remained unchanged with the exception of revenue where a slowdown in advertising and significant changes in currency rates are having an impact."

 

"Q1 2020 was the third consecutive quarter of year on year growth above 30%. In all four of our regions, MAU grew faster in Q1 2020 than it did Q1 2019. Growth in North America accelerated for the 2nd straight quarter led by outperformance in the US, as did growth in our largest region, Europe. Our Latin America and Rest of World regions continue to see the fastest growth, with those segments growing 36% and 65% Y/Y, respectively."

 

https://s22.q4cdn.com/540910603/files/doc_financials/2020/q1/Shareholder-Letter-Q1-2020-Final.pdf

[chesko182]

Up almost 11% today, and rightfully so. Huge move today with signing Joe Rogan exclusively for Spotify INCLUDING videos...(a first for spotify)

 

Pretty insane

 

https://newsroom.spotify.com/2020-05-19/the-joe-rogan-experience-launches-exclusive-partnership-with-spotify/

 

https://www.theverge.com/2020/5/19/21263927/joe-rogan-spotify-experience-exclusive-content-episodes-youtube

[Liberty]

Sucks for the podcast ecosystem generally, but good for Spot.

 

Too bad their podcast player kind of sucks. Can't believe the speed settings aren't more granular (goes from 1.5 to 2) and there's no dynamic silence trimming... I also hear it sometimes loses where you are and starts over. Sounds like a RealPlayer experience... Surprised they haven't been better about that since they obviously are focusing on podcasting a lot.