The title of this article is almost certain to inflame the sensibilities of the Sirius XM Holdings (NASDAQ:SIRI) bulls who will once again accuse me of being short or shilling for Liberty Media (NASDAQ:LMCA), the parent company of Sirius XM, and Liberty's chairman John Malone. The first accusation is false and the second is absurd. The bulls will quickly point to the fact that total paid subscribers rose 7% in 2013 to 25,559,310 or by 1,658,974 subscribers, from the 23,900,336 at the end of 2012. They will tell you this growth was despite the shift of one large OEM customer going from paid promotional subscribers to unpaid trials, and they would be correct. In short, they won't be concerned. I am.
The following chart looks at the growth of total subscribers and self-pay subscribers over the past 5 years, with a few other metrics included:
Average Self-pay Monthly Churn
All Subscriber Net Adds
Self-Pay Net Adds
Data Source: Conference calls, interviews, presentations and company 10Ks and 10Q.
As the economy improved and car sales increased, the churn rate held steady and even decreased slightly, but the conversion rate also began to decline. The result? The 2013 self-pay subscriber net adds were 149,989 (or just over 9%) less than they were in 2012.
Part of the decline in conversion rate can be traced to increasing new car penetration rates. On the Q4 conference call Sirius XM CEO Jim Meyer discussed the effect of increasing penetration into lower priced vehicles:
In both the new and used car markets, we are seeing increased numbers of sub-$75,000 households coming through our trial funnels.
...As we discussed on our last call, our new car subscribers are selling their cars more quickly than industry averages. Increasing vehicle turnover and income sensitivity in our subscriber base are contributing to churn.
The mix of our vehicle, owner, and subscription base will continue to move toward the demographics of the average driver in this country. I don't think it will surprise any of you that over the years, we have seen higher income households convert at higher rates and churn at lower rates than lower income households, but lower income subscribers remain a very profitable segment for us, because of our high contribution margin and low variable cost per acquisition.
He's right. I wouldn't be surprised to see that the increasing penetration in new and used cars would expose the company to lower income demographics and drive conversion rates lower and churn rates higher. In fact, we did see the conversion rate decline by one percentage point in 2013, but surprisingly, churn also declined by one-tenth percentage point. I would have expected to see the churn rate go up instead of declining or even holding constant. However, in the fourth quarter, as the penetration rate increased to 71%, the churn percentage ticked up to 1.9% (the same level as the prior three years) and the conversion rate for new vehicles declined to 42%. Is there a reason?
With the release of the third quarter results, the company lowered guidance on self-pay net adds from 1.6 million to 1.5 million. Meyer stated the reason for the reduced self-pay net add guidance:
With continued growth in new automobile sales and an increasing number of existing self-pay subscribers selling their cars and rotating back into our trial funnel, we are increasing our guidance for net subscriber additions and reducing our guidance for self pay subscriber additions by equal amounts.
During the conference call, he further explained the change in guidance:
We have seen that our existing new car subscribers are turning over their vehicles sooner than what would have been predicted based on historical industry trends. As a result, our subscribers are migrating from one radio to another in a newer car. This trend picked up in the third quarter and is ahead of our expectations and this trend of faster turnover is good for the business long-term as it leads to increased opportunities for used car subscriptions, but we are now estimating self-pay net adds of approximately 1.5 million for 2013.
This led to the following question by Brian Kraft of Evercore:
Could I just ask one other, just a clarification too? When someone is an existing subscriber and they got a new car and that's a free trial car, not a paid trial. Are they counted as self-pay churn?
And the answer by Sirius XM CFO David Frear:
Yes. Look, I think the 1.8% continues to be a up pretty good number, because your next logical question which I know you are going to ask. Is that well do you achieve 100% of those guys when they migrate back to the trial. Of course the answer is no, but when you are all done tracking it through, whether it's rotate through as a paid trial or an unpaid trial and you look at the net retention, we think 1.8% is a fair representation of what happens after the migration effect of people moving from one car to another, so it continues to be I think a good long-term planning number for you to use.
It seems to make sense. It also would have suggested that an increase in turnover by self-pay subscribers to new car trials would drive the churn rate higher. However, the 1.8% remained at the low end of the range over the past 5 years. Here's another interesting note from then 10K:
2013 vs. 2012: For the years ended December 31, 2013 and 2012, our average self-pay monthly churn rate was 1.8% and 1.9%, respectively. The decrease was due to a higher mix of existing subscribers migrating to paid trials in new vehicles which are not included in average self-pay churn.
Also from the 10K, the definition of churn:
Average self-pay monthly churn - is defined as the monthly average of self-pay deactivations for the period divided by the average number of self-pay subscribers for the period.
So, as far as self-pay monthly churn is concerned, when an existing self-pay subscriber cancels their subscription because they are buying a new car that comes with a trial that was paid for by the OEM, it IS NOT a deactivation and is not part of churn. BUT if the OEM is not paying for the trial, that deactivation IS part of churn.
That's certainly confusing, and maybe it has been that way all along, but the shift of that one OEM from paid-promotional trials to unpaid trials should drive an increase in reported churn figures. We may have already had a hint of this in Q4 when churn moved up to 1.9%. And, investors should be prepared for churn to climb higher in future years.
But that one OEM is not the only reason to expect an increase in churn. It's also the exposure to lower income subscribers. That exposure could also be having a significant effect on the new car conversion percentage. That percentage dropped to 42% in the fourth quarter. It is the lowest quarterly level I can recall seeing, and 42% is not only below the 44% for all of 2013, but also less than any annual new car conversion percentage since at least 2006 despite an economy that has continued to improve.
The company has stopped discussing the total trial funnel and has made it more difficult for analysts to make refined projections. On the last conference call Meyer revealed the following info:
...In 2013, new car buyers converted at about 44% and used car buyers converted at about 34%. We ran about 13 million total trials in 2013, and we expect this number to be over 15 million in 2014. Incremental penetration into more mid and lower end cars may cause a softening of conversion metrics, but with declining SAC per new vehicle, we are more than willing to take this tradeoff.
Consider the following:
- The company generated 1,658,974 total subscriber net additions and 1,511,543 self-pay net additions from 13 million total trials in 2013.
- The company has guided to 1.25 million subscriber net additions from 15 million total trials in 2013.
Regardless of whether those 1.25 million net subscriber additions are total net adds or self-pay net adds isn't all that relevant. The bottom line is that it will take a 15% increase in trials to generate 17%-25% fewer additional subscribers in 2014.
Average Revenue Per User ("ARPU")
One of the best performing metrics for Q4 was ARPU which rose sequentially by 1.8% from Q3 to $12.46 and was up by more than 2.2% for the year from $12.00 to $12.27. According to the 10K, the annual change was:
driven primarily by the contribution of the U.S. Music Royalty Fee, the impact of the increase in certain of our subscription rates beginning in January 2012, and an increase in subscriptions to premium services, partially offset by subscription discounts offered through customer acquisition and retention programs, and lifetime subscription plans that have reached full revenue recognition.
There were also other factors that would have a positive impact on Q4, including the incremental revenue from the connected vehicle services acquired from Agero and the loss of lower revenue subscribers due to the OEM shift from paid to unpaid promotional subscribers. It is also possible that the company was less aggressive in offering retention discounts during Q4, which could have contributed to the sharp drop in the conversion percentage in that quarter.
Going forward several factors will continue to affect ARPU. ARPU will increase as the revenue from the connected vehicle services will begin to contribute for full quarters, and that revenue is expected to double over the next few years. It should also increase as there will be a permanent reduction of lower revenue paid promotional subscribers due to the OEM shift. In addition, a base rate price increase of $6 per year starts to flow through the population as of January 1, 2014 and all subscribers can expect the MRF (music royalty recovery fee) to increase each year as the music royalty rate climbs 0.5% each of the next several years.
The ARPU growth could be constrained to some degree by lower priced subscription offers and more aggressive retention programs. Last fall the company announced $5.99 per month subscription for a suite of channels targeted at the Hispanic market. Part of the announcement included an extended free trial that will expire the end of this month. Depending on the success of the offer and the number of trial conversions to self-pay subscribers, the growth in ARPU could be constrained.
Since 2009 the company has been able to continue growing subscribers, aided by increasing car sales and by increasing the penetration rate with its OEM partners. It has also been creative in expanding its marketing efforts in order to attract new subscribers with Service Lane trials, its used car trial program and its Hispanic offering.
As Sirius XM continued to increase subscribers the share price had risen sharply, peaking at $4.18 in October of 2013. After the Q3 earnings were released on October 24th, along with a downward revision self-pay net add guidance, the share price began to fall.
The downward revision to guidance proved accurate as the full year subscriber growth at Sirius XM slowed down in 2013 and that growth is expected to continue slowing down in 2014. This is not just the rate of subscriber growth that is slowing down, but also the actual number of net adds that is expected to decline for the second consecutive year.
As Meyer pointed out, the company is moving into a market segment that has lower income and is expected to convert to self-pay at lower rates and cancel at higher rates. Investors would be well advised to closely follow churn, conversion rates and subscriber growth, and determine whether or not their growth expectations for the company are reasonable.
I am long Sirius XM, and I expect the company to be able to continue to grow revenue and subscribers over the next several years. However, I also expect those rates of growth will continue to decline as Sirius XM struggles to add subscribers.
Additional disclosure: I actively trade SIRI. In addition to my long positions in SIRI, I have January 2015 $4 covered calls written against several of these positions. I may initiate new covered call positions or close out or open new positions in SIRI at any time.