Tag Archive for "AT&T" tag

As an analyst in the wireless industry, I was often asked to compare KPIs (Key Performance Indices - subscriber counts, ARPU, Churn, etc.) of various carriers.  Unfortunately, most carriers earnings reports do not provide breakout of KPIs for their prepaid and/or reseller (aka. MVNO) businesses.  Unless they are from a pure-play prepaid provider like Virgin Mobile (now less true with their acquisition of Helio) or Tracfone, prepaid KPIs are not easy to come.

In this article, I will share the method I use to estimate prepaid subscriber count, prepaid ARPU, and prepaid churn for AT&t using its latest 4Q 2008 earnings report.

First, you will have to be fairly familiar with AT&T’s Statement of Segment Income — GAAP for the Wireless segment [master.xls].  Under the tab, Wireless Segment, AT&T provides total Subscriber count, Net adds, Postpaid sub count, churn and its ARPU.  This is great especially AT&T is doing quite well in its postpaid business.  But if you want to get to Prepaid/Reseller ARPU and churn, you’ll have to do some math.   Let’s calculate AT&T’s Prepaid + Reseller :

• Subscribers (row 40, 41)
• ARPU (row 59, 60)
• Churn (row 51, 52)

You can download the work file here in Excel format (see the red Wireless Segment tab).

Prepaid Subscribers

All you need is ONE reasonable assumption, the number of Reseller subscriber in 4Q 2008 to get to the estimated prepaid subscriber count.  Fortunately, we can make one fairly easily since most of their reseller subscribers come from the Tracfone, the largest MVNO in the US.  As of 3Q 2008, Tracfone reported 10.449M subs.  If in 4Q 2008, Tracfone added the same number of subs in as in 4Q 2007, then 10.499M + 711k  = 11.16M (see cell I41).  AT&T has 77M total subs, 60M postpaid subs, and 11.16M reseller subs (est).  So AT&T prepaid subscriber count is 77M - 60M - 11.16M = 5.7M.  We will use this estimate for the rest of the calculation.

Prepaid/Reseller Churn

To estimate the churn, you first calculate the number of disconnects for total subscribers and postpaid subscribers using the reported churn rate (see row 53-55).  As you can see, it looks like the disconnects for Prepaid + Reseller segement would have to be 1.479M for AT&T to have a consolidated total churn of 1.6%.  Hence, the Prepaid + Reseller churn is about 3.1%.  [Gut check: Tracfone's last reported churn rate was 3.6% in 4Q 2007 and 3.8% in 3Q 2008.]  If AT&T’s composite Prepaid + Reseller churn is 3.1% and the Reseller churn (Tracfone’s churn) is ~3.8%, the true AT&T Prepaid churn rate is likely lower than 3%.  This would be quite good for US Prepaid providers.

Prepaid + Reseller ARPU

Estimating ARPU can be done in a similar fashion.   The only thing you need to do is read the Investor Briefing (the pdf file they released).  It states that the Postpaid ARPU was $59.59 in 4Q 2008.  With that plus the total service revenue, you can back out the postpaid only service revenue which you can use to calculate the remainder of the service revenue for the prepaid and reseller segments.  The Prepaid + Reseller service revenue was 955M so using the Prepaid + Reseller subscriber count for 3Q and 4Q we calculated earlier, you can estimate the ARPU.  It turns out the Prepaid + Reseller ARPU was about $19.27.  The true Prepaid-only ARPU for AT&T is likely to be much higher.  Tracfone (reseller) reports ARPU of $11 in 3Q 2008.  Don’t forget that this is at the retail level.  As a wholesaler (AT&T in this case), the ARPU AT&T takes is likely in the $4-6 range depending on the price of the wholesale minute which can range from $0.03-0.08 per minute based on a Tracfone average of 78 MOU (minutes of use).

I hope this has been helpful in estimating AT&T’s prepaid KPIs .  All of these estimates are done using assumptions and should be use for “back-of-the-envelope” calculations only.

The recent 4Q 2008 earnings simply reinforced the facts.  Verizon, T-Mobile, and Sprint are not losing market share to AT&T.

Take a look at their respective net adds (show below).  Clearly, no one expected net adds will grow in a saturating mobile market.  As can be seen in the yearly numbers, net adds across the carriers dropped YOY.   But AT&T dropped the least among the three here.

What is more impressive is when you look closer at the postpaid net adds.  AT&T increased the number of postpaid net adds from 4Q 2007 while Verizon and T-Mobile dropped significantly.  The effect of adding more new subs into your base also helps reduce churn which can be seen at AT&T while again Verizon and T-Mobile showed higher churn than the previous year.

If this continues and Verizon is not able to introduce new devices that will slow the iPhone love affair, I might want to entice consumers with more attractive offers on their rate plans.  They are currently the most expensive among the four national carriers (given others offer bonuses like 7pm nights, rollovers, etc.).

Some have criticized AT&T’s increase in acquisition cost in 2008 especially when it picked up fewer net subscribers for dollars spent than it did in 2007.  However, 2007 was the year when it acquired massive number of low quality subscribers from prepaid.  With higher churn, lower ARPU, and intense competition, these prepaid subscribers can have a lifetime value 3-5x lower than an average postpaid subscriber.

Let’s look at the numbers more closely…..

AT&T spent $13B in 2007 for Selling, General, and Administrative expense vs. $14B.  So roughly the same.  However, in 2007 it picked up roughly 5.3M net additional subscribers while only 4.7M in 2008.  But of the 5.3M in 2007, one out of four were prepaid subscribers whereas in 2008 nearly all net adds were postpaid.

Not only were they postpaid subscribers, they were mostly iPhone postpaid subscribers that have a reportedly 1.6x ARPU than the average postpaid subscriber.  Extracting the numbers from the Investor Briefing, more than 50% of all postpaid gross adds where iPhones (1.9M out of the 3.5M total in postpaid) coupled with 60% of the net adds being smartphone subscribers……I would say the increase in acquisition cost was money well spent.  (I don’t buy the lower churn rate of iPhone subs yet since these subs have not been around long enough.)

But don’t iPhone subs cost more to acquire and it is hurting AT&T’s CPGA and bottom line as can be seen in the recent earnings??

Not true.  Despite the heavy acquisition cost for a iPhone sub as can be seen by the earnings shortfall at AT&T, in the long run, the residual value from higher iPhone ARPU and lower churn should justify the investment.  Assume that the iPhone is subsidized an additional $200 in CPGA. If ARPU was truly 1.6x higher than average postpaid ARPU, $60x.6 = $36.  $36 x 40% margin = $14.40 in additional gross contribution per month.  Hence, the breakeven point for the additional $200 in CPGA is a little more than 1 year ($200/14.40 = 14 months). This seems well worth the investment given that the subs are under a 2-year contract.