What will Make 3G Grow?

Format for Printing

Format for printing

Request Reprints


By BRational
May 17, 2005

Posts selected for this feature rarely stand alone. They are usually a part of an ongoing thread, and are out of context when presented here. The material should be read in that light. How are these posts selected? Click here to find out and nominate a post yourself!

[Ed's Note: Post of the Day will take a brief hiatus, and return on Monday.]

3G Growth and Adoption Models

Much of the discussion of performance of key wireless telecom players has hinged on the rate of uptake of 3G, especially in its WCDMA (UMTS) flavor. From handset makers like Nokia, to silicon purveyors like TI, to IP royalty toll collectors like Qualcomm, growth in revenues and earnings, and hence the stock price, appear to be increasingly tied to the fate and rate of WCDMA adoption by the legions of consumers in GSM land.

Of course, for Qualcomm, the WCDMA picture is especially interesting because of its potential to contribute to both the chipset side of the business as well as to the royalty collection side. We have seen in the last quarter's numbers ( just what kind of impact it is already having on the royalty stream (the company's QTL business segment). We have yet to see the impact on the chipset segment (QCT), other than the contribution to lower margins of the investment in engineers and R&D specifically to bolster the company's ability to serve WCDMA customers.

1. Some simple arithmetic: WCDMA and QTL royalties. The projected 45~50M WCDMA handsets for the current year translate into 27~30% increase in total CDMA handsets relative to the 04 numbers. Assuming a conservative 25% higher ASP for WCDMA relative to the average CDMA ASP, the impact on QTL revenues would be an increase in the range of 34~38% (relative to 04)�all converting to earnings at a robust efficiency rate of better than 90% (I would also anticipate margins to increase to the 92~94% level, as the marginal cost of royalties is close to nil).

With near 100M WCDMA handsets forecast for 06, the growth rate relative to the 05 total CDMA base would be about 45%; adjusted for 25% higher relative ASP, brings us to about 57% growth in QTL revenues (all assuming uniform royalty rate, which is not entirely correct given preferential rate to Korean and Chinese manufacturers). This is the kind of tornado effect that many long term investors had hoped for; it seems to be a near certainty at this stage of the game.

2. WCDMA and QCT. On the chipset side, we wish the picture would be nearly as rosy�it is not yet, though even modest WCDMA market share would help the bottom line considerably. The math gets more complicated because of the time lags between when a chipset is ordered, shipped, booked and eventually incorporated in a handset shipped by the manufacturer and purchased by a consumer. An optimistic estimate would be for QCT to ship 5M WCDMA chipsets in the remainder of the company's fiscal 05 (i.e. by September 05). If this number is all concentrated in the September quarter, that would be about 15% growth in chipset volume for the quarter. Assuming that WCDMA ASP's would be about 50% higher than the average chipset ASP (which as we have seen has been trending lower), that would mean about 23% growth in revenues due to WCDMA. The company has been rather unforthcoming about the extent of WCDMA presence in its projected chipset volume in the next couple of quarters-- I am assuming none has been factored in current guidance, and I consider the 5M chipsets in this fiscal year to be very optimistic.

3. Looking ahead at the next year though, and assuming that QCT can achieve a 20% market share, we'd be looking at about 20M WCDMA chipsets for 06�and I would consider that number to be fairly conservative. With the 50% ASP premium, WCDMA chipsets might contribute to nearly 20% growth in annual QCT revenues. Assuming that Sanjay achieves even 25% margin for his operation, the impact on earnings growth would be substantial; coupled with surging royalties on handsets, annual growth rate in company earnings in 06 looks to be in the 40~45% range. This is only the WCDMA effect, and assumes that all other CDMA revenues stay at essentially the same level (i.e. this does not factor in EVDO take-up in the US and around the world, 3G in China, and other CDMA in low-end markets).

Of course, one can redo the calculations with more optimistic or pessimistic assumptions�the fact remains that even modest take-up of WCDMA among the huge base of GSM users would translate into substantial growth rates in QCOM revenues and earnings.

4. WCDMA and GSM handset makers. The picture is different for the handset makers that have dominated the GSM world for the past few years, especially Nokia, and to some extent Motorola. WCDMA take-up essentially means replacement at the high end of the GSM handset spectrum. It does not mean major new markets, or expansion of the GSM base, but rather would fit in the dynamic of change within the GSM market. Both handset makers and carriers who have deployed the expensive infrastructure would like to see accelerating adoption of WCDMA 3G at this time. For the former it means higher-ASP handsets, while for the latter it means potential for higher data ARPU. Just how fast can they expect such take-up to happen?

5. What affects WCDMA take-up? The answer depends on many factors�most obvious among those are: (1) handset availability, features and prices; (2) service availability, coverage, quality, awareness and pricing; and (3) availability of compelling applications and services exclusively to 3G users. We have had many discussions about these, and the popular press is showering us with the promise of all kinds of goodies through our mobile phones�TV, music, games, sex, and so on�a true Swiss Army knife combining communication, entertainment and utility (why we want all this on Lilliputian screens when laptop makers are finally recognizing our desire for 17" screens is a matter for another conversation). The main point I want to revisit has more to do with how 3G is branded and promoted than the specific services offered (or not).

6. Is history a good guide? To predict WCDMA adoption, one might turn to previous technological waves in the wireless telecom realm, and use that as a basis for inference. Specifically using Qualcomm's experience, the introduction of CDMA 2000 1X turbocharged both handset replacement and base growth in the CDMA world. While that was technically 3G, we know that in practice it was somewhere between 2.5 and 2.75G�but that point is not directly relevant to WCDMA introduction today. There appears to be a fundamental difference between the way in which 1X was introduced and the way WCDMA is being promoted. The former was introduced by the carriers essentially with little fanfare; its introduction coincided with a new generation of handset capabilities�color screens, polyphonic sounds, silver clamshells, and a little later camera phones. People wanted the phones, and did not particularly know nor care about the faster data rates. They did not sign on to 1X; they simply got new phones, or signed up for service. All CDMA replacement became 1X, building up the claim of 3G adoption regardless of whether the users knew the difference. Call this model passive adoption. Eventually, once they had the phones, people progressively tried the data services�a stock quote here, a snapshot there, and so on...

In contrast, GSM carriers did not automatically convert everyone to GPRS-capable handsets once GPRS was deployed. It remained an option, which was adopted only timidly in the first couple of years. There was some economic justification for that: GPRS did not add capacity to the GSM network, but rather used part of the available spectrum, crowding out voice during peak periods, whereas 1X nearly doubled the combined voice and data capacity, so carriers had an incentive to switch users as quickly as possible to 1X handsets once the network was up and running.

7. 3G adoption models. With WCDMA, the conversion from GSM is not automatic. Call it active adoption -- a more challenging approach to gaining rapid traction, as it requires convincing skeptical customers initially of the medium's "wow" factor (for the techno adopters), and eventually of its value proposition (for the broader market, albeit still its higher end). WCDMA is a premium service, for which carriers expect to make premium profits. Consumers must actively opt into the world of 3G services, by acquiring more expensive handsets, and signing on to the premium service. In Japan, FOMA was introduced as a new service, initially in addition to NTT Docomo's voice and i-mode services; the take-up was slow, albeit partly because of clumsy handsets, limited coverage and high price. The new improved FOMA came back with sleek handsets and new branding as a full-fledged service to replace basic service, and there has been no looking back. It is in (mostly western) Europe today that the principal WCDMA take up battle is being waged; and it is in Europe that the initial results have been less than overwhelming.

8. Winning over the Euro skeptics. First, it really is too soon to draw any meaningful conclusions for the future. The range of handset offerings at most major wireless carriers in Europe is still very limited, and underwhelming in my opinion, and the new services that go with it appear to be none too compelling (I have been to different places in Europe three times since February, and have surveyed for this note several of the major wireless carriers' web site offerings�I will report in more detail on that in a separate post). There seems to be little to wow or be wowed about for the enthusiasts, beyond what you could get already on a high end and better designed smartphone, and many quality-skeptics will undoubtedly sit out this round until the handset makers bring out a more sophisticated second iteration. Most carriers seem to be displaying 3G models alongside the other offerings�in most cases without much visible fanfare and without clear differentiation other than along a single dimension (e.g. Vodaphone UK pushes 3G Live as a video call service). Under these conditions, there is little reason to expect WCDMA take up to occur any faster than the "natural" adoption rate of high end models�perhaps even a little slower as the new plans required for the new services, uncertainty about the new technology and unfamiliarity may act as deterrents.

Under the active adoption approach, faster take up will occur when the high end models slide towards the middle end of the pricing spectrum, and the better features are available at what consumer perceive as a reasonable premium over plain vanilla. Handset makers have become quite adept at exploiting this sweet spot of the handset features adoption curve�especially Nokia that has carefully tuned the repeated cycles of the feature/price relationship for its portfolio of models, in attempt to balance squeezing more profit of older but fully-discounted models vs. losing ground in the prestige positioning game. A major difference in the case of WCDMA is that carriers now have a greater interest in accelerating this adoption than they do for regular high-end models�in order to sell premium data and multimedia services (and start seeing some return on the huge outlays for licenses and infrastructure).

9. Relevant segment analogies. From Nokia's standpoint, it appears that its positioning in WCDMA is not too different from its positioning in the previous few rounds of high-end features: it is not the first to introduce higher end features, but generally delivers these, after a time lag, bundled with a comfort factor that reaches the broader realms of the consumer market (in many ways, Nokia seems to be the Toyota or Honda of the handset world� but that's another discussion altogether). The main point is that I see no reason to expect any eye-popping lift to Nokia's handset revenues or earnings from WCDMA�it is a more crowded segment of the high end market, and so far only a moderate driver of replacement decisions in Western Europe. The impact on the infrastructure division is different�we have already seen considerable improvement in that division's performance partly due to WCDMA deployment by carriers.

10. Changing the adoption model. The hesitation and seeming lack of sharp focus of European carriers in selling 3G may be typical of the early stages of the technology cycle, where services may not be quite up to par with the technology's capabilities, or the latter may not yet quite have the features to deliver the eventual sparkle and sizzle of live high-resolution interactive human communication. Like previous technology enhancements, the features will get better and so will the services, though growth expectations for makers of wireless chipsets and handsets, and for content designers and producers, seem to outpace the natural rate of improvement and adoption. When the perfect storm of compelling services, handset features and pricing come together, we may see a faster adoption rate than suggested by the normal high end replacement rate model. To further boost that, and until it happens, carriers may want to induce a certain degree of passive adoption�by progressively making 3G the norm on all models instead of the high end exception, and aligning the pricing of 3G models with that of the regular models. The stream of data-driven revenues in a society that has fully embraced mobile access would undoubtedly provide returns that exceed many times whatever subsidy investment might be necessary by the carriers.

For Qualcomm, as we saw earlier, growth at even the natural rate of high end replacement would translate into hefty increases in revenues and earnings. For the GSM-centric industry, WCDMA (and more generally 3G) seems to be treated as part of the natural cycle of evolution of a maturing wireless telecom industry. In the calculus of quarterly results of large corporations, this is an appropriate perspective. However, a more relevant perspective might be to view it as an embryonic stage of the mobile Internet. Just as the fixed Internet has created new forms of value through electronic commerce, virtual supply chains, interactive gaming and all sorts of community activities, its mobile incarnation carries the promise (threat?) of still unfolding services and applications that will add value across a still weakly-linked chain of agents. As wireless carriers seem fixated on the profit potential of their role as gateway keepers, their insistence or perhaps inability to sell 3G as anything other than the next phone feature may well be stunting the growth and delaying the potential of the next frontier in the virtual wild west.

WCDMA adoption is undoubtedly a complex phenomenon. Additional comments, opinions, and thoughts would be highly appreciated.


Become a Complete Fool
Join the best community on the web! Becoming a full member of the Fool Community is easy, takes just a minute, and is very inexpensive.