I find it amazing that despite article after article of giving investors information about the auto sector, I still get emails asking me to clarify something that an analyst, or worse yet, a message board poster has said. It happened again this weekend, and once again I will attempt to outline the ins and outs of the auto channel as it relates to SiriusXM (NASDAQ:SIRI).
The latest deals with a message board poster, that seems to like to be very selective in the quotes and data they attribute to me. This poster, it would seem, has taken a stance that the auto manufacturers are "stuffing" the sales channel. Apparently his theory is that auto makers register cars and the reimburse dealers so that sales will look better than they are. There are a few problems with this theory. Once a car is registered it can no longer be sold as new. If this was happening to any substantial degree, it would quickly be vetted out. Truth be told, here in the U.S. this is a very, very small percentage. Setting that aside though, here is another thing this poster had to say:
"Auto sales came in weaker than expected but if you read Spencer this week, no one saw that coming. You probably missed another point in last week's post and that is that neither I nor anyone on the Street is using that 622K sub number when we discuss SIRI with our institutional clients. We all use models and we all buy proprietary info from Wards, Edmunds, and JD power. So let me explain the adjustment process. We can see the channel stuffing in GM and other car makers that count subs before the car sold. We can take the inventory build fro Q1 to Q2 and multiply by 70% penetration rate. That gives us the new parking lot subs and we take that away from 622k. That is the number my institutional clients get. It is too bad the guys at SA don't understand that computation."
This is very interesting. A week before the close of the month I wrote a piece that stated year over year sales would be up and that month over month would be down. I also outlined that the SAAR should come in at above 14 million. While I did see auto sales for the month being higher than what came in, the gist was that year over year would be good, and month over month would be bad. Even that is not the crux of what the issue with this poster is though. The real crux is this:
"We can see the channel stuffing in GM and other car makers that count subs before the car sold. We can take the inventory build fro Q1 to Q2 and multiply by 70% penetration rate. That gives us the new parking lot subs and we take that away from 622k. That is the number my institutional clients get. It is too bad the guys at SA don't understand that computation."
Wow! If this poster actually has clients, they are not getting very good information!
1. Sirius XM does not count GM cars before they are sold. GM is what I would term as a "Point-of-Sale" partner to Sirius XM. This means that when a car is sold it becomes a subscriber. This happens because GM does not pay Sirius XM any money until the car is sold.
2. Simply taking the inventory build and multiplying it by 70% accomplishes NOTHING near an accurate picture of SiriusXM's subscriber story. First of all the penetration rate is 65% not 70%. Now that that is out of the way let me spell out exactly why this posters calculations are fraught with error. SiriusXM garners subscribers in 3 distinct ways from the OEM channel. I developed the unique "Leading", "Point-Of-Sale", and "Trailing" categories to track this. The Leading category delivers subscribers at time of manufacture. This happens because SiriusXM receives payment at that point. This category encompasses about 38% of the market share. The Point-Of-Sale category delivers subscribers at the time of sale. Again, this is predicated on payment, which happens when a vehicle is sold. The Point-Of-Sale category has about a 31% share The Trailing category only supplies subscribers only after a free promotional period (which is not counted as a subscriber) and IF the consumer elects to become self paying. This category carries about a 31% market share in new cars, but also houses all of the used car promotions. As you can see, simply applying a 70% rate to inventory production will have you well off base in any given quarter. Especially when the manufacturers that supply "parking lot subscribers" only carry a 38% share of the market.
3. In order to best understand the subscriber picture of Sirius XM you need to look at various data points and timelines:
- Take about 65% of production slated for U.S. sale on Ford, Chrysler, Mercedes, BMW, Mitsubishi, Jaguar, Land Rover, Mazda, Volkswagen, Audi, Volvo, and Mini. These brands will supply counted subscribers on 100% of the satellite radio equipped cars produced for about 5 months on the European and Asian brands, 8 months on Ford brands, and 14 months on Chrysler brands (accounts for 2 months on the dealer lot).
- You need to look at about 65% of current sales on GM, Honda, Subaru, Suzuki, and Porsche. Each of these supplies a 3 month subscription on 100% of the satellite radio equipped cars sold.
- You then need to look at 65% of current sales on Toyota, Nissan, Hyundai, and Kia and understand that 45% of that 65% will become subscribers 4 months from now. This represents the conversion rate. Yes, Hyundai and Kia do virtually 100% satellite installs.
As you can see, the issue is much more complex than the data this poster allegedly sells to his clients. The Certified Pre-Owned and used car deals blur the lines even further, even with the knowledge that they are treated in the same manner as the Trailing category. Why is this such a big deal? When you look at the balance sheet the prepayments are booked as deferred revenue. That is a liability. Deferred revenue starts to become revenue when it is earned (i.e. a consumer listening to the radio). Deferred revenue is virtually unique to the "Leading" category. Further, you have costs. Each installed radio carries costs. With the Leading category you get cash back quickly from an installed radio. With the Trailing category you may never see cash back on 55% of the radios installed. As you can see there are positives and negatives beyond the subscriber number that need to be considered.
I could go on and on, but the essence of the way things are done is spelled out here. I have written many articles on these categories, SAAR, production and sales. They are available in the archives here.