I guess we are off today due to the Korean Gov't wacking carriers on subsidies again. This happens every year...ya da da...
Anlaysts continue to miss the boat on QCOM's business model. They tied it directly to CDMA subscribers in the beginning, and this was the right thing to do. But as the base has built, other factors have become much more relevant to QCOM's model, and very few (if any!) analysts recognize this, though it goes a long way in explaining QCOM's strength in chipset sales and backlog.
According to the CDG, there were 128,000,000 subs at the end of June. We know PCS lost subs in the quarter, VZW added, Korea, Japan, China on the plus side, etc. etc.
Probably netted 6,000,000 new subs in the quarter. Typically, the analysis stops here, and we'll hear about downgrades due to slowing sales, etc. etc, and QCOM's 20 Million chipsets shipping in the quarter must be going into inventory.
But what about churn? Negative to the provider, but positive to QCOM. Churn is measure of how many subs leave the provider (typically for another), and is the reason "net" adds get reported in lieu of "gross" adds. Churn runs just under 3% at most carriers, but this means 35% of the subscribers leave, but typically get replaced by new subs as well (churn from the competing carriers). So if the CDMA subscriber base (currently estimated at 134,000,000) stayed the same, QCOM would collect royalties on 46,900,000 new handsets over the next 12 months, a number almost double the net new subscribers projected over that period (@ 7 Million/quarter).
But there is much more! In addition, a report out today said that the average subscriber currently replaces their handset every 18 months. Apply this number to the 134,000,000 figure, and you sell an additional 88,440,000 CDMA phones over the next 12 months, a number more than 3X the projected new subscriber base.
How accurate is this analysis?
If we applied the same assumptions, to the same data sources as of December 31st, 2001, it would have projected 109,244,000 handsets to be sold during 2002.
I believe the actual for this period is landing in the 85-90 Million range (based on QCOM's production and backlog). So I guess we should apply a factor of 82% to the total of 163,340,000 (the sum of the new subs, the churn, and the replacements), for 12 month handset forecast of 133,938,000 (Oct-02 - Sept-03).
So while new subscriber growth may or may not slow over the next 12 months (which is all the analysts want to talk about), CDMA chipsets, phones, and QCOM's royalties could increase by 48% YoY.
However, we are down over a point today because Korea may gain 300,000 less new subscribers this quarter. Whatever... |
