A Close Look At The 2016 Sirius XM Subscriber Guidance

| About: Sirius XM (SIRI)


Sirius management has issued guidance for 2016 subscriber growth 40% below 2015.

A look at the key drivers of new car sales, free trials and cancellations.

Increasing penetration resulting in declining conversion rates.

After adding 2.3 million subscribers in 2015 a number of investors were left scratching their heads when Sirius XM Holdings (NASDAQ:SIRI) issued guidance for only 1.4 million net additions for 2016. Even though Sirius management has long professed its proclivity to issue conservative guidance, the decline of 40% seems excessive at first glance. Perhaps a closer look at the drivers that affect the net adds may provide some additional insights.

Most of the Sirius subscribers still come from marketing its service to buyers of new cars. Currently 75% of the cars sold in the United States come with a Sirius satellite radio installed in the dashboard, and buyers of those new vehicles get a free trial to the service that typically lasts from three months to a year (although in some instances, it's a even longer). As the end of the trial approaches, Sirius amps up the marketing campaign, trying to convince that new car buyer to convert to a self pay subscriber.

The rate that those trials convert to self-pay subscribers has declined over the years as the penetration rate into car dashboards has risen. As of the end of the third quarter, the penetration rate had risen to 75%, and because of the lag with trials being converted to self-pay, we have probably not yet seen the conversion rate fully impacted by the latest increase. Here are the historical penetration rates and conversions as taken from various company sources, including annual meeting presentations, conference calls ,and the 10-k's and 10-Q.

















Conversion Percentage







Average Self-pay Monthly Churn







Once converted to self-pay subscribers, Sirius loses those self-pay subscribers at a rate that has been 1.9% per month for five out of the past six years.

Projecting Subscriber Growth

Last year there were a record 17.5 million vehicles sold in the US, exceeding the previous record of 17.4 million set in 2000. Forecasts for 2016 indicate we could see another record, with some figures as high as 18 million units. Assuming the figure is 18 million, and 75% of those come equipped with satellite radio, and 41% convert to self-pay subscribers, as many as 5,535,000 new self-pay subscribers would be added from new car sales. At the same time, there would be a loss of self-pay subscribers due to self-pay monthly churn of 1.9%, or nearly 23% of the average number of self pay subscribers for 2016.

New OEM Self-pay Sub Gains and Churn losses

To determine the approximate average number of self pay subscribers during 2016, we can start with the year end 2015 figure. The recent press release by Sirius that included preliminary 2015 results stated:

Self-pay net subscriber additions in 2015 were 1.77 million, exceeding the company's increased 2015 guidance of 1.6 million and resulting in self-pay subscriptions of over 24 million at year end.

How much over 24 million did Sirius achieve? Well, since it finished 2014 with 22.52 million self-pay subs, and added 1.77 million in 2015, we get 24.29 million self-pay subs at the start of the 2016. And, if we use the 1.4 million net add guidance for 2016, we arrive at a year end number of 25.69 million and an average number of self-pay subscribers for the year equal to 24.99 million. Applying that churn statistic of 1.9%/month to the 24.99 million average tells us that Sirius will experience ~5.70 million self-pay subscriber losses. These losses would be 0.16 million higher than the 5.54 million self-pay subscriber additions from the new car channel.

Effect of Increased Penetration

To simplify the analysis, it is best to ignore the impact of the used car business for the moment. The penetration rate increased from 71% in 2014 to a 75% rate in Q3 of 2015. During the Q3 earnings call, CEO JIm Meyer noted:

Our penetration rate reached a record of about 75%, up nearly 4 points from the third quarter of 2014, resulting from gains at virtually all of our OEMs.

This would indicate that the increase affected OEMs with both paid and unpaid trials, and could be partially responsible for the number of paid promotional subscribers rising from 4.79 million at the end of 2014 to 5.31 million subscribers by the end of 2015, a rise of .52 million, or nearly 11%. This is a number that should rise by a much smaller amount in 2016, even if the expected new car sales increase to 18 million. That 0.5 million increase in new car sales, or 2.9%, along with the increased penetration rate to 75% for a full year, would indicate that the OEM paid promotional trials would increase by, at most, ~0.3 million.

So far, we can show a 0.3 million increase in paid promotional trials offset by a net loss of 0.16 million self-pay subs from a combination of new OEM self-pay net adds and the loss of self-pay subs due to churn.

Used Car Activations

So, if there is a net gain of just 0.14 million net adds from the above combination, it means that Sirius needs to acquire 1.26 million self-pay subs from the used car channel to get to 1.4 million net adds. That number would certainly appear to be one that could be exceeded based on prior comments by management about used car additions. In early 2014 Meyer discussed that market as follows:

We are very excited about this opportunity and well-positioned to capitalize on it. We are on track to grow self-pay additions in this segment from approximately 1.5 million last year [2013] to close to 2 million this year [2014].

There hasn't been much guidance since then, but based on this set of assumptions and the used car business trends, the 1.4 million would appear to be very conservative.

Other Factors Governing Net Adds

There are, however, other factors that could negatively impact the calculations that go into the forecasts. For instance, we know that Sirius tends to use average or consensus estimates of new vehicle sales rather than some of the higher numbers, such as the 18 million that I chose. If that occurs, there will be fewer trials and fewer conversions.

As an example, two assumptions that turned out to be significant under-estimations and affected the 2015 net add forecast were discussed early in 2015 at a Morgan Stanley conference:

We are assuming kind of like the market about 16.7 million new car sales, roughly 70% penetration rate.

When new car sales came in 700,000 higher and the penetration rate rose to 75% percent, the company significantly beat the original forecast. Sirius was able to run well over 1 million more trials than expected and would see higher than anticipated paid promotional trials and self-pay net adds by the end of the year. Company management is likely to be asked similar questions about its as the year progresses, and investors will get additional clarification about the assumptions used in the 1.4 million unit guidance.

An important consideration should be the correct choice for the new and used car conversion rates. As the company increased the penetration rates on new cars over the past several years, we saw a decline in the conversion rate. There is every reason to expect that we will see another decline based on the penetration rate rising from 71% to 75%. In Q1 of 2015 the conversion rate was 40%, so it is possible it could drop to 40%, or even 39% in 2016 based on increased penetration for the full year.

Management has discussed the link between incomes and conversion rates and churn on multiple occasions. Here is one such comment made by CFO David Frear in the middle of 2014 when discussing pricing and income levels:

People ask us a lot about income sensitivity and there is definitely a difference. So people who have the highest incomes convert higher and churn less and you can see a drop by a couple of points in every $25,000 of income. Zero to $50,000 a year households comprise believe it or not, a significant portion of new car sales in the United States and a very significant portion of used car sales.

And so these are households that we are sensitive to as well. The average household income in the United States is roughly $51,000 or $52,000 a year and we want to make sure that we are pricing in a way to make the product relevant to everybody.

I believe it is reasonable to assume that as penetration increases, there will be more cars at lower price points with satellite radios, and those lower priced cars will be targeted to lower income buyers. The result should be lower conversion rates. But it's not just new car buyers with lower incomes that will be an increasing part of the trial funnel. The same situation should also carry over to used car buyers.

Look at the chart that showed the penetration rate in 2010 was 62%. Not shown is the penetration rate in 2009 that was just 55%. The used cars entering the market in 2016 are likely to be coming from an installed base that has a declining average initial MSRP, reflecting the OEM and Sirius trend of installing more radios in lower priced vehicles during the past six years. To the extent that those lower priced used vehicles are more likely to be purchased by lower income families, the likelihood of a used car converting should also decline. And if trying to place a precise number on the conversion figure wasn't difficult enough, the company can affect this by choosing how much of an acquisition discount to offer a potential subscriber.

Another factor that will affect the net adds is the self-pay monthly churn. While the quarterly figures in recent years have ranged from 1.6%-2.1%, by the end of the year we have seen the figure typically come in at 1.9%. Should it drop to 1.8% or rise to 2.0%, the impact could be significant. This is a number that can be affected by how aggressively the company chooses to pursue self pay subscribers and how far they are willing to go with retention discounts.


Is the company being ultra conservative in its forecast? Perhaps, or perhaps it is simply being somewhat conservative. We really can only make estimates depending on how accurately we can answer the following questions.

What is the correct forecast for new car sales in 2016? Will the company see a drop in the new car conversion rate or an increase in churn because it is marketing to buyers of new cars from a lower income demographic? Or, will it ratchet up acquisition and retention discount offers in order grow self-pay subscribers? Which companies will win market share -- those that generate paid promotional trials where the trial counts as a subscriber, or those that are unpaid trials? Will more companies sign up for longer paid promotional trials increasing the number of paid promotional trial subscribers at year end, or will some of the current OEMs with extended trials migrate to shorter trial plans, reducing the paid promotional trial number?

How many used car trials will the company run? Will the conversion rate of used cars be maintained in the low 30% range or will it decline as the average price of used cars with OEM installed satellite radios follows the decline in the average price of new cars with OEM installed satellite radios? Will the used car churn rates suffer as well?

The answer to each of these questions can have an impact on how accurate the net add guidance turns out to be. Directionally, I expect to see a decline in the full year conversion rate and I will be surprised to see new vehicle sales reach 18 million, both factors which will contribute to fewer net adds. I also see significantly slower growth in paid promotional trials in 2016.

Will that be enough to result in fewer net adds in 2016 than we had in 2015? I think so, although probably not as low as 1.4 million. I also recognize that the company has the ability to manage subscriber growth with its discount programs.

Bernard Baruch said, "The main purpose of the stock market is to make fools of as many men as possible." The main purpose of forecasting would seem to be the same thing.

Disclosure: I am/we are long SIRI.

I wrote this article myself, and it expresses my own opinions. I am not receiving compensation for it (other than from Seeking Alpha). I have no business relationship with any company whose stock is mentioned in this article.

Additional disclosure: I have written $4 and $4.50 covered calls on Sirius that will be expiring at the end of the week. I could let the shares be assigned, close my positions on the options, roll the options forward, sell Sirius and/or buy Liberty at any time.