Breakthrough Ideas (continued)
Value Capture.
Every company is entitled to recapture a portion of the value it creates for its customers. To survive, it must remain within the profit zone. Qualcomm recaptures value in several ways. First, it licenses its CDMA intellectual property. These license agreements are legal contracts in which the licensee agrees to accept Qualcomm’s rights to its intellectual property in exchange for the right to use their codified-knowledge of CDMA in their products. Qualcomm has done the R&D that these customers license and use, saving them time, money, and hassles.
Second, in the licensing agreement, the manufactures agree to a pay per-use royalty for this IP based upon the manufacturer’s average selling price (ASP) of its CDMA-based products. As you learned earlier, network operators pass on legal liability to manufacturers for all patent claims through a special clause in their purchase contracts. So, Qualcomm’s royalties recapture a small portion of the value that most benefits end-users and network operators from manufacturers who install network infrastructure or sell mobile terminals.
These license agreements segment this market by specific product, for instance, fabricating chips versus manufacturing handsets, line cards, or PDAs, by use, whether internal use or merchant market, by geography, say, national, regional, or world, and by generations, specific standards within generations, and the like. The more specific the segmentation, the better it is for Qualcomm, since this means that licenses for new segments become required for companies with expanding ambitions. For instance, this segmental specificity meant that 2G licenses did not cover 3G products, nor does a WCDMA license cover cdma2000.
On July 16 2002, Qualcomm and Alcatel entered into a CDMA infrastructure patent license agreement. Two sentences from the press release illustrate representative features, which are highlighted: “Under the terms of the royalty-bearing agreement, Qualcomm has granted Alcatel a worldwide license under Qualcomm’s CDMA patent portfolio to develop, manufacture, and sell third-generation (3G) WCDMA (UMTS) and TD-SCDMA infrastructure equipment. As part of the agreement, Alcatel has granted Qualcomm a royalty-free, worldwide license under Alcatel’s patent portfolio to develop, manufacture, and sell semiconductor components for incorporation into wireless CDMA and multimode CDMA equipment.” The first sentence covers the features of the royalty-bearing license that has Alcatel paying Qualcomm. The second sentence cross-licenses any of Alcatel’s IP for UMTS (or GSM if that were not already covered in their 2G license) to Qualcomm royalty-free for use in multimode CDMA equipment. That is, Qualcomm cannot directly compete by generating GSM and UMTS chips as a single mode, but can, as part of their radioOne multimode approach, develop a world phone without worrying about patent suits from Alcatel. Licenses with Ericsson, Nokia, and others were similarly structured.
On August 26 2002, the last major European holdout, Siemens restructured its existing 2G-license agreement to include infrastructure equipment for all CDMA standards and subscriber units for UMTS. The multi-million dollar expansion granted Siemens a royalty-bearing license and granted Qualcomm royalty-free rights under Siemens’ patents to market and sell CDMA components, including multimode chipsets. In a September 13 interview with Louis Lupin, Qualcomm Senior VP and general counsel, Lupin said (emphasis added),
“We sort of look on it as a milestone event in the sense that Siemens was the last of the major telecom equipment makers who was not licensed for 3G or for infrastructure equipment. So, it like the last piece of the puzzle …I think Qualcomm at least believed, at the time that there was development of the W-CDMA standard, that there had been some effort to design around some of our patent portfolio. …There was an effort to try and design something independently, but when it came time to solve some of the fundamental problems—like power control, like how to structure the wave form, and do things like orthogonal channelization—it just wasn’t possible to avoid the innovations that we had already come up with, and that were subject to our patent portfolio. …We certainly believed for a long time that whatever flavor of it ultimately got adopted, that CDMA would be the backbone for the next generation, the third-generation systems. …History has confirmed that. CDMA as a technology just is inherently more efficient in terms of its use of spectrum and its ability to deliver higher data rates in an efficient way. That’s what 3G is all about.”
The signing of the GSM Big Five to 3G licenses was a milestone event for Qualcomm. It demonstrated conclusively that Qualcomm met the gorilla criterion of possessing a proprietary, but selectively open, architecture. Although the Big Five said it would never work; then, that they had invented it first; and finally, that their WCDMA architecture had proprietary GSM-features. Gorilla power was demonstrated when each paid for a license and accepted the necessity of royalty payments from the Big-Five to Qualcomm, who cross licensed their IP without paying fees or royalties. This meets one prong of the definition of managing its value chain, a key indicator of gorilla power.
Each license covers all of the intellectual property to date, with the documentation that is necessary to be successful, including post-sale support. Occasionally, Qualcomm accepts equity in startup companies in lieu of a portion of this fee to encourage and support a new member of its value chain who cannot afford this licensing expense. Just as any violation of any claim in any patent is actionable under patent law, the licensing contract itself, more significantly, becomes legally actionable under contract law. All of this seems to be standard best business practice.
All of the major telecommunication companies in the world, somewhere over 100, have now taken licenses from Qualcomm. Before doing so, each scrutinized Qualcomm’s patent portfolio to be sure this was necessary. If it were not, they would hardly pay Qualcomm a license fee or royalties. This scrutiny and the signed legal contacts, in addition to Qualcomm’s legal victories in many patent disputes, confirm the value of the patent portfolio’s intellectual property, which, indeed, has been carefully developed and managed.
The royalties on manufacturer’s ASPs are negotiated as a percentage that fairly covers the value of the IP. Thus, CDMA IP is more essential to a CDMA handset than it is to a CDMA card in a laptop that also performs general purpose computing. Just because a laptop has a higher manufacturer’s ASP, it does not automatically generate a proportionally higher price-based royalty to Qualcomm. Value added is what counts.
A major problem facing the UMTS manufacturers will be that about 100 companies claim essential patents. Whereas, Qualcomm offers a single source of protection against competing CDMA claims against its own intellectual property. This could entail a lengthy series of patent negotiations, and a combined royalty rate of, some say, 20%, some say, up to 100%, on top of Qualcomm’s 5-6%. This is not a competitive advantage for UMTS.
Qualcomm has settled minor spread spectrum IP claims in the past to cap a potential source of future problems. No licensed company has asked Qualcomm to set aside money to cover any potential competing claims to its intellectual property because Qualcomm is indeed the first-mover, the, innovator, commercializer, and continuing technology-leader of mobile wireless spread spectrum. Despite the noisy chatter of expectation’s management, everyone recognizes Qualcomm’s actual lock on IP. They simply deny it loudly until their business interest requires that they enter a commercial market. That is, non-commercial trial networks don’t count; patent violations require the selling of another’s patent-protected intellectual property as your own. In the end, everyone quietly signs a licensing contract.
Qualcomm’s royalties are well deserved, but also potentially very remunerative as the market scales up, the degree and breadth of penetration increases, upgrades cycle rapidly, and replacement rates rise to take advantage of newer handset features and technologies. All forms of CDMA must pay royalty to Qualcomm. Thus, Qualcomm has a financial stake in the rapid rollout and success of UMTS.
These license and royalty agreements are confidential. Hence, accurate information is hard to come by, but rates are sometimes quoted by journalists or estimated by analysts. For instance, Korean manufacturers are reported to pay 5.25% for domestic sales of CDMA and 5.75% for international sales. Whereas, given their customer power, China MII was reported to negotiate a royalty of 2.65 % for domestic and 7.0 % for exports. The contracts may also contain terms that specify a percentage of MSMs that must be purchased from Qualcomm, which was 80% for China. Modal estimates for royalties are 4 or 5 per cent. This royalty rate applies to all 3G mobile wireless manufacturers’ ASP. This royalty rate does not decline because others also claim to have essential IP.
Third, Qualcomm sells MSMs to manufacturers. These manufacturers can deduct the portion of royalties reflected in the percentage of the average sales price, a discount that presumably advantages Qualcomm’s MSMs by making them fractionally less expensive. As Qualcomm builds in more intellectual property into its MSMs, the value and ASP of handsets increase. Volume discounts are also customary.
Fourth, Qualcomm sells CSMs to infrastructure providers who were selected by carriers to install their networks. Infrastructure CSMs usually represent about 10-to15% of chip revenues, with a higher margin than MSMs.
The infrastructure providers also design, install, and service the RAN. That is, the infrastructure companies assemble and test a system to an operator’s specifications for network mode, but prefer using and selling equipment that they design or manufacture whenever feasible. Base stations require antennae. Control stations will include switches and servers in the Core network. From the Alcatel licensing agreement discussed above, Alcatel claims, “We are now, more than ever, well positioned to provide mobile operators with field-proven 3G solutions with everything in place, for UTRAN, Core network, ATM switches to Open service Platform, and Software Solutions.” The UMTS radio infrastructure (UTRAN) is produced in a joint venture with Fujitsu, which Alcatel claims leverages its experience in GSM, GPRS, EDGE, as well as in ATM and IP protocol.
On August 20 2002, the CDMA Development Group (CDG), which is a CDMA promotional association of value web members, announced that eight wireless vendors had signed a MOU committing them to delivery of CDMA2000 infrastructure and terminals in the 2.1 GHZ frequency band, which KDDI will use. I have rearranged the alphabetical list of 8 companies according to my impression of prominence within the value chain: Qualcomm, Lucent, Nortel, Samsung, LG Electronics, Motorola, Ericsson, and ZTE. This reflects my view that Motorola and Ericsson are losing prominence. In terms of future potential, the Korean company Samsung, which I believe to be the emerging leader in infrastructure and terminals, along with LG Electronics, SK Telecom (which was not on CDG’s MOU list of eight), and China’s ZTE, has immense potential to rollout CDMA2000 handsets or infrastructure.
Lucent and Nortel both have extensive and successful experience installing CDMA2000 networks. Lucent decided to concentrate exclusively CDMA2000 (or UMTS) rather than GSM, GPRS, or EDGE. Subsequently, Lucent signed a contract with Cingular, which raised a few eyebrows and some SI speculation about the firmness of Cingular’s commitment to GPRS and EDGE; Cingular has no available spectrum for UMTS. The Japanese, having such mastery of consumer electronics and computing, cannot be counted out. Kyocera, who acquired Qualcomm’s handset division, along with Toshiba, and Sony, who is in a not-too-successful joint venture with Ericsson for handsets, are all valuable web members. Sharp, with its demonstrated excellence in LCDs over the years, is a significant contributor of handsets and components.
A new source of revenues is BREW. Because it is such a new entrant, it is difficult to estimate its potential contribution, which some speculate could be as high as $4 to $5 per user a month because of increased ARPUs. On September 4 2002, at the SSB Tech Industry Conference, Tony Thornley said (emphasis added): “The most important part of our business, in fact, is how quickly we can drive the replacement of devices, which drives our royalties revenues, our chip revenues, and, in the future, I think, our services. Because in the future, services will abound as devices abound. The services themselves are the third leg of the stool that will start contributing more of our revenue in the future, particularly the BREW platform, which has had considerable success.” Thornley continued by adding that BREW was driving ARPUs -- a selling point in rolling out this new platform. |
