categories

HOT TOPICS

Subscribe to our Feed

QualComm: The Nokia War

Posted on Friday, Oct 12th 2007

By Vijay Nagarajan, Guest Author

Over the past few articles, I have taken a look at QualComm’s potential revenue flow over the next few years if it got its way. Before I proceed to look at the effect of the lawsuits and the industry consolidation on the company’s revenues and long term interests, I wish to diverge and look at the company’s recent bickerings with Nokia (NOK).

Let me begin by re-iterating the unstated fact, ratified by my previous analysis: It is all about who gets those percentage margins per phone sold. Nokia, with close to 40% of the market share, has been a prominent QualComm licensee in the past. With the migration to WCDMA/HSDPA picking steam, the Finland company will pay royalties to QCOM on most phones it sells in the near future.


The center of all the lawsuits between QCOM and Nokia is the cross-licensing agreement that expired this summer. The parties were in discussion to renew the cross-licenses that span CDMA, GSM and now UMTS (WCDMA/HSDPA). Nokia made a strategic move to walk away from CDMA. There are atleast 3 reasons that came out during my exchange with industry veterans and confirmed by the analysis so far –

  • Nokia is betting away from CDMA2000 and EV-DO. The Finnish giant may stand vindicated with the handset shipments of these standards likely to stagnate over the next few years. With its weight away from CDMA2000, I believe it has been instrumental in some of the migration away from CDMA2000 in emerging markets giving the QualComm-standard a perception of high cost as compared to the more ubiquitous GSM. QualComm’s past aggression to capture market share in the CDMA2000 space has exacerbated this situation with the lack of competition.
  • Nokia wants a more even position at the licensing tables. Neither party disputes that QCOM should be paid royalties. It is a question of how much. If it continued to stay in the CDMA2000 business, Nokia would have had to negotiate with the patent value heavily skewed in favor of San Diego. On the other hand, for 3GPP, at least in terms of the number of patents, the two giants are on equal terms (based on publicly done studies). I do think that the QCOM patents are still the heart of the phone and by that logic perhaps more valuable. So, Nokia will probably be happy with around 2% of ASP or slightly higher as the fee to QCOM. (This is a number that has popped up in the EC anti-trust case.) QCOM might also be willing to budge but for the ripple effect with other vendors and continents resulting in its licensing model breaking down. The upside of such an agreement is perhaps a higher chipset market share for QCOM.
  • Royalty as a percentage ASP may no longer be acceptable. This is an important point that I had overlooked thus far. As illustrated earlier, QCOM gets its royalty as a percentage of the handset ASP. So, while the chipset itself costs only about $24, QCOM makes the same amount of money on the higher-end phones just on royalty. On the other hand, in laptops, QCOM makes royalty money as a percentage of the CDMA components alone implying that the value is closer to $2 or so. With data-centric HSDPA phones taking center-stage in the future, the phone manufacturers would like to get the licensing model closer to the laptop model and not the handset model. As a friend pointed out, this makes sense in a device which predominantly connects to the net through WiFi or WiMax with EV-DO or HSDPA being used sparsely and on the move.

The move away from CDMA2000 is only one sign that all is not well with this almost imperative industry relationship. With the license extension discussions moving to a deadlock, the two companies have engaged in lawsuits and counter-lawsuits since 2005 with no end at sight. These lawsuits span 3G and QualComm’s essential IP in GSM/EDGE (Yes, GSM!). For a concise history of the lawsuits, please refer to this zdnet article. It appears that QCOM allowed Nokia to use its GSM patents until now in return for CDMA royalties. With the discussions falling apart, QCOM has been filing for infringement in GSM so it can impact Nokia’s immediate productization thereby forcing an early resolution in its favor. Nokia, on its part, has been seeking ‘fairer’ terms of licensing. For now, the posturing has taken a formal structure with the lawsuits and the outcome of the EC antitrust probe will be quite decisive for the industry with the 3GPP community centered in Europe.

While I am not willing to take sides, I definitely think that Nokia has made calculated efforts to lower QCOM’s patent play led by its CDMA2000 pull-out. What gives the handset vendor more leverage and support is the pervasiveness and competition in the UMTS/GSM standard. So, I will keep a close watch on the EC case. Its outcome and QCOM’s strategic options at that stage will be crucial to its future.

This segment is a part in the series : QualComm

Hacker News
() Comments

Featured Videos

Comments

Nokia has not walked away from CDMA2000, or pulled out of that market, although they have deemphasized it. Their plans announced last year after the dissolution of the planned CDMA JV with Sanyo are to selectively participate in key CDMA markets, with special focus on North America. They had already reduced R&D on CDMA in San Diego and this spring they completed the ramp down of all manufacturing of CDMA2000 devices in their own facilities which utilized 1xRTT ASICs of their own design fabricated for them by TI. Nokia has begun the transition to an ODM to OEM model for all CDMA2000 devices with those ODMs responsible for sourcing components and paying applicable. Since the beginning of this year all Nokia branded 1xRTT and 1xEV-DO CDMA2000 devices have been manufactured for them by ODMs in Taiwan and/or China. The entry level Nokia 1255 was the 1st of these devices and it 1st shipped to India’s Reliance in April 2006 before shipping into Latin America and it utilized a QUALCOMM 1xRTT chipset. Both China Tech Faith and Foxconn have been named as the ODM. Nokia has since shipped several rebadged 1xRTT and 1xEV-DO Pantech models and China’s BYD and Taiwan’s Compal are now rumored to be ODM partners. We are likely to here more about Nokia’s CDMA2000 plans at the overlapping Amsterdam events that commence on December 4 (Capital Markets and Nokia World).

Cheers,

  • Eric –
Eric L. Friday, October 12, 2007 at 1:34 PM PT

Vijay, a few comments on your “QCOM- the Nokia War.”

1) Re: “… I definitely think that Nokia has made calculated efforts to lower QCOM’s patent play led by its CDMA2000 pull-out.” As was pointed out, NOK has not “pulled-out” of CDMA2000, effectively nullifying
some of your rationale for their lower royalty rate demands

2) Re: your QCOM revenue estimates- + “Royalty as a percentage ASP may no longer be acceptable,”

  • “ the chipset itself costs only about $24, “
    ……….QCOM’s chipset ASP is closer to $20, or slightly under. And, the “chipset (MSM)” includes several chips (baseband, RFR, TFR, PM) except in their newer single chip products (QSC).

  • “QCOM makes the same amount of money on the higher-end phones just on royalty”.
    ………QCOM’s average revenue per handset is about $8 (4% of $200 ASP). Higher-end phones would yield a higher royalty, but there is a maximum limit on which the standard royalty is computed ($400 -$500???)

  • (royalty from laptops / aircards)- “value is closer to $2 or so.”)
    ……..QCOM’s royalty from laptop sales is about $6 -$8, and based upon the aircard ASP which is $150 to $200.

    Re: NOK v QCOM and “While I am not willing to take sides…”

I don’t imagine any company looks favorably on paying royalties to another, but I believe you’re buying Nokia’s (and its cronies) story “hook, line, and sinker” without much understanding of Qualcomm’s business model and the wireless industries IPR history.

Don’t you think it’s rather hypocritical of the dominant industry incumbents Nokia (#1 handsets- 40% share / #2 infrastructure), Texas Instruments (#1 chipsets), and Ericsson (#1 infrastructure/ #1 WCDMA handsets) to be accusing Qualcomm of anticompetitive behavior and reneging on commitments to license their WCDMA IPR following FRAND? Are your aware>>>

  • that all three negotiated WCDMA licensees with Qualcomm?

  • that Nokia’s most recent WCDMA license with Qualcomm was negotiated in mid-2001, after the WDCMA standard was approved and under those same FRAND provisions ?

  • that the dominant incumbent GSM companies unsuccessfully attempted to design around QCOM’s patents, and in so doing actually complicated the technology (while increasing their patents), and caused significant delays in commercial WCDMA launches.

  • that QCOM was instrumental in correcting those flaws in the WCDMA standard and provided vital assistance to the carriers in deploying their WCDMA networks. That would not have been possible without the QCOM business model which employs royalty revenue to fund those R&D and SG&A efforts?

  • that Nokia let that license expire earlier this year, and now is attempting to negotiate better terms (reneging)?

  • Qualcomm’s average royalty rate is about 4% of the handset ASP, and QCOM’s rate to Nokia’s is believed to be somewhat less?

  • the incumbent GSM handset manufactures (NOK, ERIK, MOT) via their royalty free cross license agreements were able to thwart competition form the Asians with cumulative GSM royalty rates “roughly between 8% and 28%” according to the ITSUG?

  • that QCOM’s royalty of about 4% of the handset ASP may well be a bargain, considering BRCM’s extraction of $6 per handset from Verizon for its 6 flimsy purchased patents developed in 1995 by an obscure (now defunct) company not involved in cellular wireless. It should be noted that QCOM’s 4% royalty (~ $8 per handset) is for its total portfolio of over 5,000 patents mostly developed by its own engineers (large portion of its 11,000+ employees) over the past 20 years.

  • that Qualcomm has now enabled these Asian companies to compete in the wireless handset market place by leveling the 3G playing field with a standard royalty rate structure and access to state of the art WCDMA chipset solutions?

  • that there would not be 3G as we know it today without Qualcomm pioneering CDMA technology?

jim mullens Tuesday, October 16, 2007 at 8:35 AM PT

Jim,

Excellent discussion above. Your last point is probably one worth acknowledging the most: Qualcomm’s pioneering role in CDMA and in the wireless industry overall. There is absolutely no question about that, and Qualcomm has handsomely benefited financially and in terms of getting the industry’s respect.

I think, one of the issues that is driving the negotiations on behalf of the licensees is the margin squeeze that everyone is operating under. Handsets, by and large, is a very low margin business, except in the very high end, and I think, the dynamics you are seeing in the other vendors is raw survival instincts.

Bottomline, Qualcomm eats too much of the handset players’ margins. This is just as true about the Asian vendors, isn’t it? It’s not history, it is plain and simple current reality.

Sramana

Sramana Mitra Tuesday, October 16, 2007 at 10:11 AM PT

Jim,

Thanks for your detailed note. Please find my answers below –

  1. I was aware of the shift to the OEM model by Nokia. In fact, there are more updates on that as of today. But the point is that Nokia is betting away from this entire standard. If I understand it right, the licenses will be borne by the ODMs with whom Nokia will have tie-ups. This means that CDMA2000 still does not have to figure in the licensing tables during the re-negotiation. So, my reasoning still stands.
  2. I am aware that chipset implies all components and not just the base-band chip. Its value estimate depends on what the assumptions are. In my case, I have culled out numbers from QCOM’s 2006 annual report and calculated this number based on revenues and total chipsets shipped. It does to a certain extent aggregate other fringe products that QCOM may sell but forms a good basis for the revenue forecasts.
  3. Again, if we use the numbers from the annual report, we will get close to 4.8% as the net licensing fee percentage. But, I have used some skewing techniques to obtain higher percentage numbers for WCDMA vs CDMA. This is in line with the fact that the ASP for WCDMA is much higher than the CDMA handsets. The $200 ($215 in the annual report) is an aggregate of both CDMA2000 and WCDMA handsets. In fact, the higher-tier handsets are priced at $400 or so on an average.
  4. Your $6-8 number for laptops is not necessarily true. Internet for laptops through these cards is more an add-on, not the life-blood as in the mobile. All the more if we look at embedded base-band processors for laptops and even combination cards (e.g. WiFi plus WCDMA etc). In such cases, the CDMA components are used for royalty determination and not the total price of the unit (e.g. laptop itself, or the combination card).
  5. Let me reiterate that I am not trying to determine who should win or lose. But as an industry observer, I base some of my opinions on what I have seen happen in first person. I do not think it is hypercritical of the parties stated to act that way just as I dont think QCOM is acting naive by portraying itself as the technology aggregator for the industry. It is a dog-eat-dog world and the companies will do what is best for them. For QCOM, it needs to vigorously protect its margins, and for the others, it is a question of getting those margins back.
  6. Lets see..Am I aware if –

all three negotiated WCDMA licenses with Qualcomm?

  • Yes, I do.

that Nokia’s most recent WCDMA license with Qualcomm was negotiated in mid-2001, after the WDCMA standard was approved and under those same FRAND provisions ?

  • Sure. But remember that Nokia was very much in CDMA2000 development then. So, I guess Nokia would dispute your ‘same FRAND provisions’ remark.

that the dominant incumbent GSM companies unsuccessfully attempted to design around QCOM’s patents, and in so doing actually complicated the technology (while increasing their patents), and caused significant delays in commercial WCDMA launches.

  • I dont think the companies ever intended to have an air interface different from CDMA. In that sense, QCOM’s contributions were going to be there. What was attempted was to minimize the impact of not just QCOM but to try and remove the need for GPS-based synchronization. Some of these work-arounds are in fact successfully implemented in the standards and contribute to its robustness. On the other hand, certain QCOM patents like power-control, soft-hand-over, each of which are elementary to the CDMA air-interface had to be adapted for want of simpler, optimal schemes.

that QCOM was instrumental in correcting those flaws in the WCDMA standard and provided vital assistance to the carriers in deploying their WCDMA networks. That would not have been possible without the QCOM business model which employs royalty revenue to fund those R&D and SG&A efforts?

  • As I said, it is rather imperative for a CDMA air-interface to have some basic QCOM patents. And by virtue of having set-up CDMA networks, QCOM was in a position to insert itself to help set-up the WCDMA infra-structure. And it is precisely for this reason that QCOM as a company has great relations with carriers. It understands the pulse of the carriers better than most others in the industry. I do not dispute that the R&D effort is not possible without the royalty-based business model. In fact, that is the entire rationale behind my margins analysis.

that Nokia let that license expire earlier this year, and now is attempting to negotiate better terms (reneging)?

  • Yes. I am aware of the mid-summer expiry. I had pointed it out in the article. I do not think Nokia is necessarily reneging. Again, with CDMA2000 no longer directly impacting the licensing tables, Nokia thinks that it has to now re-negotiate given that the patent balance has changed from 20-80 to 50-50 or so. The argument now is whether QCOM is really licensing based on FRAND. That is subjective and is for the courts to decide. For the moment, I do think that Nokia at least has a case to argue.

Qualcomm’s average royalty rate is about 4% of the handset ASP, and QCOM’s rate to Nokia’s is believed to be somewhat less?

  • ‘Somewhat less’…true! But Nokia seems to want the ‘somewhat’ to be changed to ‘much lower’. The problem is that if it becomes ‘much lower’, the others will want lower too, and then the ‘much lower’ will become ‘somewhat’ again. And therein lies the problem. As I have stated in the article, QCOM will consent to Nokia’s wishes but for this ripple effect.

the incumbent GSM handset manufactures (NOK, ERIK, MOT) via their royalty free cross license agreements were able to thwart competition form the Asians with cumulative GSM royalty rates “roughly between 8% and 28%” according to the ITSUG?

  • Perhaps. Now, if they burnt the Asians, they know what would happen if they get victimized this time around. So, they are only being pro-active.

that QCOM’s royalty of about 4% of the handset ASP may well be a bargain, considering BRCM’s extraction of $6 per handset from Verizon for its 6 flimsy purchased patents developed in 1995 by an obscure (now defunct) company not involved in cellular wireless. It should be noted that QCOM’s 4% royalty (~ $8 per handset) is for its total portfolio of over 5,000 patents mostly developed by its own engineers (large portion of its 11,000+ employees) over the past 20 years.

  • $6 is posturing. And as you had pointed out in an earlier comment, if Verizon sells emough phones, that is just an eye-wash. The number is out there to take the sting out of QCOM’s patents, thats all. The others believe it is higher and it not a bargain by any means.

that Qualcomm has now enabled these Asian companies to compete in the wireless handset market place by leveling the 3G playing field with a standard royalty rate structure and access to state of the art WCDMA chipset solutions?

  • -True that LG and Samsung are now competitive in the market. But I think they have gone with QCOM among so many chipset vendors more for the engineering excellence than for the royalty structure itself. Remember that QCOM will get its royalty from these players, no matter what.
  • >

that there would not be 3G as we know it today without Qualcomm pioneering CDMA technology?

  • True. And that is why this is also the golden age for QualComm. Come 4G and OFDMA, QCOM will have to re-invent itself given the flatter patent-play. Right?
Vijay Tuesday, October 16, 2007 at 11:49 PM PT

Sramana-

Re: On the behalf of “licensees” / “margin squeeze that everyone is operating under” /” Bottomline, Qualcomm eats too much of the handset players’ margins”

I believe Vijay’s essay was about NOK’s “war” with QCOM, and NOK’s handset margins are quite good and by far the highest in the industry. No doubt part of that is because the bulk of their sales are still GSM which “currently” bares no QCOM royalty. However, with close to 40% world wide handset market share, (50% in GSM, and probably 60% or more in Euro GSM), their volume significantly disadvantages their Asian competitors much more than QCOM’s royalty, and that volume advantage is the direct result of NOK’s real anticompetitive practices.

Now, NOK is hypicritically claiming QCOM is anticompetitive, yet NOK’s volume advantage was built upon real anticompetitive actions that for many years stifled competition from the Asians as detailed in my initial comments.

In 3G (WCDMA and CDMA2000), Qualcomm’s business model levels the playing field by subjecting all to a standard royalty rate (with adjustments likely for volume / IPR offsets). No doubt we all want a free ride whenever we can get it, but without a way to monitize essential IPR there is no incentive for Qualcomm or anyone to innovate.

Further, if NOK and all the other handset makers are paying roughly the same royalty rate (~4%) to Qualcomm, none are disadvantaged because of QCOM’s royalty line within their BOM. Thus they will a pay Qualcomm about $2 on a $50 handset, $4 on a $100 handset, and $8 on a $200 handset.

When one really thinks about the the total cost involved in 3G, the royalty that Qualcomm has and is asking for pioneering / enabling this enormously innovative and life changing technology is very misiscule in the whole scheme of things.

………………………………Total………..QCOM royalty
+ handset cost- …………….$ 200…………$8
+ Voice contract- 2 years .…1,440
+ Data services- 2 year ??……….480
+ Devise accessories……………50
+ Devise insurance- 2 years……144
sub total…………..…….$2,314…………$8.. or 0.34%

  • Spectrum Licenses- ??? $ 100+ billion in Europe alone
  • Infrastructure build out- $ 100+ billion ???????
  • Carrier operations- $ 100+ billion ??????
    …….recurring cost

Although Verizon is not saying so publically, their willingness to quickly hand over $6 to Broadcom for its 6 flimsy purchased patents developed in 1995 by an obscure (now defunct) company not involved in cellular, suggests the real value of Qualcomm IPR as they are willing to pay up for whatever IPR no matter how insignificant to continue to sell new phone models (developed with real QCOM IPR).

Funny, I haven’t heard you, Vijay, or anyone complain about “Broadcom eats too much of the handset players’ margins” (of course the handset mfg’s are not paying it, but Verizon is and that cost must be eventually captured within their rate structure and eventually passed on to the consumer)

jim mullens Wednesday, October 17, 2007 at 7:26 AM PT

Vijay-

1) “But the point is that Nokia is (g) etting away from this entire standard. If I understand it right, the licenses will be borne by the ODMs with whom Nokia will have tie-ups. This means that CDMA2000 still does not have to figure in the licensing tables during the re-negotiation. So, my reasoning still stands.”

Yes, with NOK using ODM’s rather than in-house manufacture of CDMA handsets, it does avoid needing a QCOM CDMA license itself as the ODM pays the royalty to Qualcomm. However, you earlier wrote that was a primary reason for “walking away from CDMA” (“Nokia made a strategic move to walk away from CDMA”. I would submit that NOK did not “walk away from CDMA’, but only from the design and manufacturing busines, and that was primarily because they and their chipset fab partner (TXN) were several generations behind in the technology aspects of CDMA2000 and finally realized they could never compete technology-wise with Qualcomm. Thus Nokia capitulated to Qualcomm in CDMA, and “saved-face” by using ODMs rather than directly purchase Qualcomm’s chipsets themselves.

2) “Let me reiterate that I am not trying to determine who should win or lose. But as an industry observer, I base some of my opinions on what I have seen happen in first person. I do not think it is hypercritical of the parties stated to act that way just as I dont think QCOM is acting naive by portraying itself as the technology aggregator for the industry. It is a dog-eat-dog world and the companies will do what is best for them. For QCOM, it needs to vigorously protect its margins, and for the others, it is a question of getting those margins back.”

…that Nokia’s most recent WCDMA license with Qualcomm was negotiated in mid-2001, after the WDCMA standard was approved and under those same FRAND provisions ?

…Sure. But remember that Nokia was very much in CDMA2000 development then. So, I guess Nokia would dispute your ’same FRAND provisions’ remark. “

Lets review some facts.>>

  • Nokia originally signed a royalty (~ 4%) bearing CDMA license with Qualcomm in the early 1990’s.

  • In mid 2001, NOK renewed that license with QCOM and extended it to include WCDMA subscriber units at the same royalty rate as was in the original license.

http://www.qualcomm.com/press/releases/2001/press77_print.html

“As part of the agreement, Nokia will continue to pay royalties to Qualcomm for subscriber equipment at the same rates established by the terms of the original cross-license agreement in 1992, irrespective of the CDMA standard implemented.”

  • The WCDMA standard was completed by mid 2001, and Nokia at that time knew its own IPR position, Qualcomm’s and all the others that contributed to that WCDMA standard.

  • The WCDMA standard was under the same FRAND provisions as are applicable now, and NOK (and the others) signed the Qualcomm license extension in 2001 fully aware of QCOM’s IPR contributions, royalty rate, and other provisions.

  • Now Nokia (and their cohorts/ other dominant incumbents) are claiming Qualcomm is not living up the their FRAND commitment. In fact, when NOK writes and speaks publicly about this, they “interestingly” refer to all the changes that have taken place since their original license in 1992, conveniently omitting the fact that they renewed their license in 2001 knowing full well the status of the relative WCDMA IPR contributions and in so doing accepted Qualcomm’s royalty rate and other terms as being FRAND.

3) “the incumbent GSM handset manufactures (NOK, ERIK, MOT) via their royalty free cross license agreements were able to thwart competition form the Asians with cumulative GSM royalty rates “roughly between 8% and 28%” according to the ITSUG?

Perhaps. Now, if they burnt the Asians, they know what would happen if they get victimized this time around. So, they are only being pro-active. “

Yes, Nokia “burnt” the Asians by their exclusionary practices (anticompetitive) and gained a destinct competitive advantage (~40% market share) by doing such which has resulted huge volume and scale. With Qualcomm’s business model the royalty playing field is essentially leveled for 3G and all will be charged reasonable, FRANDly negotiated rates. However, despite Qualcomm’s positive industry contributions, the Asians are still at a considerable disadvantage due to NOK’s past anticompetitive actions which still yield significant cost concessions from their suppliers due to scale.

4) re: our differences in royalty rate / chipset prices, my numbers are for Qualcomm current FY thru Q3FY07.

jim mullens Wednesday, October 17, 2007 at 6:16 PM PT

texas term life insurance rates…

Definitely needed info. thanks!…

texas term life insurance rates Tuesday, October 30, 2007 at 1:58 AM PT

[…] Qualcomm With its legal troubles with Nokia coming to an end, top wireless semiconductor company Qualcomm (QCOM) is on a roll. It is also a […]

Top 10 Semiconductor Stocks - Sramana Mitra on Strategy Wednesday, August 13, 2008 at 8:45 AM PT