Breakthrough Ideas (continued) [Although present state analyses rapidly become dated, this summary can serve as a benchmark for progress in the last six months.]
The State of CDMA Around the World
Qualcomm. In presenting the state of Qualcomm through the end of August 2002, I rely on the June Q3 press report, “q1000’s” Conference Call notes from July 25 2002, and slides from a Price-Waterhouse presentation in the last week of August.
qualcomm.com
siliconinvestor.com
At a time when Qualcomm’s direct competitors Nokia, Ericsson, and Motorola, reported flat to negative results, when these companies have been forced to reduce and delay estimated shipments of GPRS and WCDMA handsets and infrastructure, when many wireless carriers labor under massive and recently downgraded debt, as delays of WCDMA mount and rollouts extend into an indefinite future, when WCDMA equipment suppliers suffer from both delays and unsolved technical problems, when GPRS is poorly received by carriers and end-users and EDGE remains unproven, when 3GPP still makes numerous changes to Release 99 of a UMTS standard that was supposed to be commercial, when evidence emerges that Release 5 may require another forklift upgrade from Release 99, Qualcomm and its CDMA2000 prosper around the world.
On August 8, speaking with unusual candor, Vodaphone Germany’s Juergen von Kuczkowski summarized the state of European GMS/WDMA technology, “It is less a question of terminal quantity than quality standards, and we are simply not satisfied with various performance features. This is particularly so with handover problems. Our suppliers, Nokia and Motorola, are often not able to keep agreed timetables, and they delay again and again features and handset deliveries.”
In sharp contrast, Dr. Irwin Jacobs and his team reported that Q3 results exceeded expectations, both quantitatively (by 1 cent over high-end guidance) and qualitatively. With over 13 million subscribers, using over 100 3G devices on 16 networks in 7 countries, CDMA 1X is first-mover in 3G. Since Q3 ended, three more 1X networks rolled out: on July 22, Telecom New Zealand; on July 28, Smartcom PCS in Chile; and on August 12, Sprint PCS in the U.S. With gpsOne and a peak speed of 307 kbps, the MSM5100, the leading seller along with the two initial rollouts for MSM5500 1xEV-DO, with 2.4 mbps and gpsOne, produced record pro forma revenues and record shipments of both MSM phone chips and CSM-equivalent voice channels. In Q3, QTC shipped well over 16 million MSMs as compared to 14 mm last year and last quarter, for a cumulative total of 31mm 1X 3G chips. QCT shipped 1X and DO infrastructure chip that supported a record-setting 4 mm equivalent voice channels, compared to 3 mm a year ago and 2.5mm last quarter.
Jacobs concluded his remarks by noting that two improvements, the SMV vocoder, which will most often be used to supply today’s voice quality with a significant [about 20 to 30% in the 6xxx series] capacity increase over 1X, and the introduction of handset antennae diversity [in the 7xxx series], will double voice capacity over 1X.
In the CC, Jacobs delivered this Elevator Message, “Unlike GSM, a TDMA technology, CDMA immediately translates any reduction in average vocoder rate into a corresponding capacity increase, which is one of the reasons that all companies are basing their third-generation offerings on CDMA. When attempting capacity improvements in existing TDMA networks, improvement, for example with a GSM AMR vocoder, is limited since timeslot duration and the number of timeslots per second cannot be changed nor can more than one voice call be placed in a timeslot.” This continues a marketing theme of CDMA technology dominance: CDMA 1X, 1xEV-DO, and multimode 6xxx has it, so flaunt it.
At the same time, Jacobs continued his theme of cooperation, “Qualcomm will continue to actively support all 3G CDMA technologies with standards efforts, chip and software development and test platforms and test activity to enable each to reach market as early as possible, with high voice and data performance and desirable low-cost handsets.”
QWI”s Dr. Paul Jacobs reported that after only one and a half years, BREW has achieved many milestones: Three commercial launches, with more expected from ongoing trials, 1.3 mm users (approaching 2 million by the end of September) worldwide, who downloaded 8 million Brew apps over the air, with about 300 BREW applications on sale today, with digital signatures for over 2000 to go on sale tomorrow, running on 19 commercial BREW handsets from 7 manufacturers, with 50 more handsets in development. Paul Jacobs commented, “The fact that the BREW solution has been so readily accepted emphasizes the need for an end-to-end solution.”
Several of Paul Jacobs opening CC comments were noteworthy. According to the younger Dr. Jacobs,
“Voice is still the killer application in wireless. Even the most aggressive 3G operator, KT Freetel is predicting that 25% of their revenue will come from data in 2005, which means that voice will continue to generate 75% of the revenue. This does NOT mean that data is less important. In fact, what we see is that data applications today are driving consumer choices – market share and economics. …
· KTF has reported a 6-fold increase in data ARPU for its 1X and BREW subscribers.
· Verizon’s upcoming lineup of color screen phones illustrates their view that handset subsidies and CPGA [cost per gross add] can be reduced because high-end phones need lower subsidies because the consumer is willing to pay for additional value.
· In Japan, KDDI has built market share and reduced churn 20% through its camera phones and gpsOne applications, both of which depend on data services, not voice.”
Don Schrock reported that orders were strong for the December quarter, with plans to ship a record-setting 18 to 19mm chips in Q4, with 95% already booked. Schrock saw world demand favoring a strong transition to the MSM5100, currently the highest volume device because of its gpsOne and multimedia, and rising demand for the new, low-cost MSM6050, which sampled in March and has led already to14 development agreements. SnapTrack is winning E-911 business in Japan, USA, and Korea, with SK Telecom launching in June, Sprint and Verizon rolling out, and Alltel and Quest expected to buy E-911 services this year. 6xxx chipset development is progressing ahead of track, with successful interop testing in Europe and Japan, completion of WCDMA and GSM calls, and completely integrated GSM transceivers. Schrock said, “We have the smallest, most integrated GSM/WCDMA chipset solution in the industry.” In the September quarter, Shrock said they were on target to sample the MSM6100 with leading edge multimedia features like a Java hardware accelerator, 2D and 3D graphics accelerator, on-chip MPEG 4 decoder and encoder, on-chip camera interface, and Qualcomm’s new SMV vocoder. Also, Qualcomm will begin early customer sampling of the MSM6300, the industry’s first 1xGSM dual mode chipset and software solutions, “that can truly enable a phone that can operate anywhere in the world.”
These stellar achievements sharply contrast with European failures. In August, Nokia delayed its dual mode chipset into next year. So far, Nokia has never publicly shown or demonstrated a working GSM/WCDMA multimode handset. Its model under development is reported to use two standard systems residing in the same large plastic case, neither integrated as a single IC, nor able to handoff calls across modes and bands. On August 30, The AP reported that Finnish operator Sonera “…blamed the delay [in its planned September 26 launch, when Nokia was to unveil its new handset and chip] on delays in the development of the technology, called Universal Mobile Telephone System.” Qualcomm leads the race in GSM/UMTS multimode technology too, having demonstrated not only live voice calls, but also streaming media, 384 kbps packet data, and live wireless WCDMA to GSM calls in a small form handset at CTIA. Too bad, Europe continues to restrain trade to its “politically proprietary,” but not yet completed or fully proven WCDMA standard.
In the first 9 months of FY’02, pro forma revenues for the QCT segment were $1,107 mm and $604 mm for the QTL segment, 53% and 29% respectively. QTC pro forma EBT was $282 mm or 34%; QTL, $535 mm or 64%. This demonstrates the profitability of both segments, but especially demonstrates the high QTL margins of about 88% versus about 29% for QCT’s chips. Over 90% of the QCT/QTL revenues came from the sale of new CDMA handsets, showing that business is good and driven by new 1X and DO rollouts. Tony Thornley estimated that subscriber additions were in the 30% range and replacement/upgrades had about 70% share. [Analysts are missing the strength of the drivers for replacement sales: 1X throughput, color screens, camera phones, gpsOne, and BREW.]
Pro forma gross margin for Q3 was 67%, up 200 basis points due to improvement in QCT’s margins. At the end of June, Qualcomm has $2.6 billion in Cash & Marketable securities, with total assets of $6.3 billion and stockholder equity of $5.2 billion. Qualcomm gave positive guidance for Q4, expecting pro forma earnings per share in Q4 to be 26 to 27 cents and 93 to 94 cents for the year, based on shipping a record-setting 15 mm chips (actually expecting to sell 18-19 million).
All reported losses were unrealized losses, for example, $194 million in Leap. Goodwill amortization for SnapTrack ends in December, and the Telefonica Moviles purchase of Pegaso, which requires repayment of a loan within 60 day, will drop $200 mm as a loan repayment, plus other (Thornley recently said $400 to $500 mm total) cash back into Qualcomm’s coffers.
Qualcomm estimates its FY’02 cash flow at over $1 billion dollars. It should improve cash flow in FY’03 or ’04 as China and India sales ramp-up, as pro forma amortization and unrealized losses lessen, and as funding commitments end. Holding R&D and SG&A constant as chip sales increase in scale explodes margins and improves cash flow.
Thornley’s recent September summary was (emphasis added):
“Clearly, we have a very strong balance sheet – which is something that we value very highly. Our cash position, at $2.6 billion, is something that we think is extremely important – to have a liquid position, especially at this time in the market. We’re generating over $1 billion a year of operating cash flow. We are investing this year a little less than that in strategic investments. As you may have noticed on the chart, we had $800 million in commitment going forward; we will be funding these commitments over the next year or so. Looking to the future though, we expect to be very strongly free cash flow positive. And, so we are looking to the future as to what we can do with that cash. But, right now we think that having a strong cash position is the place to be.”
The good news is that up to 25 operators are expected to launch 1X services this year, and 7 more operators are expected to launch 1xEV-DO networks over the next six to twelve months. But most important, handset replacement sales are being driven by device improvements, position location, and BREW. Guidance for Q4 was positive: a Q2Q revenue increase of 10-13%, with pro forma earning of 26-27 cents, assuming record shipment of 18-19 million MSMs, with 15 million for 1X as of July. According to CDG, there were more than 127 million CDMA subscribers and 15 million 3G subscribers. Over the past year CDMA subscriber grew by 32%, which passed GSM’s slightly slower rate, while the CDMA2000 base increased nearly 21 times. CDMA subscribers are growing at a rate of 1.8 million per month because of these competitive advantages: (1) increased voice capacity; (2) Device Availability (over 120); (3) Unique Applications; (4) Spectrum Efficiency; (5) Frequency Flexibility; (6) Multimedia Services; and (7) Widespread Deployment. |
