Apollo,
This post is from the Motley Fool of Jan 1 re: Q. Has this been posted/responded to here? (I was in Europe through the holidays and while having read through 2000 posts here I wouldn't bet that I didn't miss one!)
Jeff
_____________________________________________________________
(Jan. 1) Motley Fool site www.fool.com The post can also be
referenced at boards.fool.com <http://boards.fool.com/message.asp?id=1220008001535003&sort=postdate>
This is post 5319 by Frank Dean of the QCOM thread at Motley Fool
*****************************
I agree with your assessment on Qualcomm's growth
in the next 10 years (~20% year). However, I don't
think it will be as clean and easy as everyone believes. There is a fair amount of risk.
Keep in mind there is a huge difference between
CDMA (IS-95) and W-CDMA (the next generation
system). There is zero compatibility between the
two systems. Your TV uses FM to carry sound, but
that doesn't mean your car radio can be used as a
TV. That said, Qualcomm does own some significant
pieces of IP that will be needed to make W-CDMA
(aka 3G) work. However, everyone is pushing to
make 3G a open, world standard, much like GSM. The
idea behind open world standards is that all IP
must be licensed on a fair basis. And (here is the
real kicker) the cost of phones is falling so
quickly that the days of Qualcomm extracting $40
(estimate for chip and IP) from each phone that
ships are coming to a close. A whole GSM phone
only costs $55!
Here's my quick work up on this. It was very
educational.
The world market for cellphones in 1999 was about
160 million units with about 400M subscribers
worldwide. Nokia estimates that by the
end of 2002 we'll have 1B subs worldwide. To reach
that goal, and also satisfy the replacement market
(folks that want a new model every 2-3 years), the
industry will have to pump out 300B handsets in
2000, 500B in 2001 and about 550B in 2002. Growth
will slow slightly after that, but assuming 15%
growth and a desire to replace handsets once every
two years, we can expect 2B handsets to be made in
2010 with 3B users worldwide. Yes, a 45% world
penetration rate. Far fetched? Perhaps. But
consider some countries will never run copper wire
and will largely rely on cellular phones as their
telecom network. Also consider that families that
could never afford a full blown PC will **have**
to rely on a $100 phone as their "computer" for
the information age. And we're already seeing >50%
penetration rates in the Nordic countries. Finland
has estimate >100% penetration rate as consumers
have several models. At $100, why not?
So, suffice to say there are going to be a
sickening number of handsets made each year.
Currently, there are just two world cellular
standards that really have a future: GSM and CDMA.
GSM holds about 55-60% of the world market and
CDMA has about 25-30%, driven largely by Asian
markets right now. Both standards will continue to
maintain those market shares as the total market
grows until 2002. At that point, CDMA will start
to decline, I think, due to Japan's adoption of
W-CDMA and aggressive GSM pushing in China
(Qualcomm is trying to get into China too, but
Nokia, Ericsson, Motorola are cutting sweet deals
to make sure the country is mostly GSM). This will
slow the growth of CDMA and eventually lead to a
decline in 2006 for both CDMA and GSM as 3G's
WCDMA comes on line. In 2006 WCDMA will surpass
CDMA, and in 2008/2009 WCDMA will surpass GSM.
Now, the interesting point to keep in mind here is
that while Qualcomm is a major force in CDMA, they
have nothing in GSM. The GSM chipset and IP market
is interesting to look at because the major 3G
players (Nokia, Motorola, Ericsson) are all
interested in seeing the same model replicated for
3G: That is, an open global standard that favors
their experience and know-how.
Today, the latest Qualcomm Annual Report shows the
silicon chipset group had revenues of $1.13B and
sold about 39M chipsets, averaging about $29 per
chipset. Great money! However, the mature GSM
market (with chipset makers such as TI, Motorola,
VLSI, Analog Device, Conexant, etc) is only
getting about $10 per chipset today, and probably half that in 2005. That is due to the
fierce nature of the chipset business. Until last
year, Qualcomm was the only CDMA chipset maker. In
the last year, Nokia, Motorola, DSPC and VLSI have
introduced chipsets. When they stood alone,
Qualcomm was assured of a 90% chipset market
share. With these other titans in the ring (many
who are also pushing handsets) expect the prices
(and Qualcomm's market share) to fall fast. If you
are Kyocera trying to sell to Sprint, how long
does it take Sprint to say "your phone is $60 more
than Motorola and Nokia's, and we're no longer interested" (the $60 figure comes from usual markup on $20 chipset premium).
Thus, in a mature market with several competitors,
Qualcomm will be forced to slash
silicon prices. In addition, they'll lose market
share simply because Motorola and Nokia will be
shipping Motorola and Nokia silicon. So, I think
we'll see Qualcomm's chipset marketshare erode
from 90% today to 50% over the next few years.
IP revenue is a huge source of income for
Qualcomm, and with 88% net margins on IP, they are
laughing all the way to the bank. Looking again at
their annual report, my amatuer eyes see they rake
in about $450M on (I'd guess) 50M worldwide CDMA
licenses. That works out to about $9 per license.
Pure gravy.
But again, keep in mind that when the total price
of a phone is $50, there isn't room for this much
IP. The price pressures are too fantastic.
Learning from GSM, this $9 figure will have to
fall drastially. How low? Probably not any more
than 4% of the total unit price. For a $100 W-CDMA
phone in 2006, this is $4.
The good news is that even though Qualcomm will
see half as much IP revenue, they'll make up for
it in 3G volume.
Grinding all this data together puts Qualcomm's
chipset and IP business earning $1.3B on $2.7B
sales in 2000 (48% net margin), growing steadily
to about $6B on $14B in 2010 (42%). At 60X
earnings, I think that puts the 2000 stock price
around $485, the 2005 price at $1,079 and the 2010
target at about $2,565.
I have no clue how a 2001 price target of $1000
can be realistically reached. |
