By Amine Mansour*
China’s competition authorities have the reputation of being aggressive. Qualcomm’s case confirms such reputation. On 10 February 2015, China’s National Development and Reform Commission (NDRC) imposed a fine of 6.088 billion Yuan (975 million dollars) on Qualcomm calculated on the basis of 8% of the company’s 2013 sales in China. This adds to several royalty concessions offered by Qualcomm to local manufacturers. Notwithstanding the significance of the amount of the fine, the highest antimonopoly fine ever exacted in China, Qualcomm decided not to appeal the decision.
According to the NDRC, the company held a dominant position in the CDMA, WCDMA and LTE chipset markets. The official statement (here in Chinese) comes out with very succinct information and characterizes the abuse in extremely broad terms. Below you will find a brief presentation of the case followed by some comments.
First, the decision points out that Qualcomm was charging excessive royalty fees. In this respect, the company refused to provide its licensees with a list of its available patents from where they could choose and forced them to accept a package that includes out of date patents. Taking this into account, the NDRC held that the price charged was unreasonable. Second, the press statement highlights a tying mechanism in Qualcomm’s patent licensing activity. In particular, the NDRC points out the fact that Qualcomm tied, with no justification, licenses for wireless and non-wireless technology forcing Chinese customers to pay for unwanted licenses. Third, it is stated that Qualcomm abused its dominant position by adding unreasonable conditions on the sale of baseband chips. As a result the NDRC states that “Qualcomm’s acts to eliminate or restrict market competition, hinder and inhibit technological innovation and development and harm the interests of consumers violating China’s anti-monopoly law”. To address those issues, Qualcomm offered a corrective measures package. This package consists of the following actions (details in the Qualcomm press release).
Patent and products covered by Qualcomm’s concessions. One has to bear in mind that what is affected in the present case is a small fraction of Qualcomm’s patented technology. In particular, the decision focuses on the pricing of Qualcomm’s patented technology adopted as a standard for 3G and 4G devices (phones, tablets, laptops…).
Concessions offered. Qualcomm will be offering a separate and less broad license that includes only its 3G and 4G essential Chinese patents. Other essential and non-essential patents will be offered through a separate license. In addition, Qualcomm commits itself to make a list of its patents available to licensees during the negotiation process. Additional concessions focused, first, on grating a fair compensation to Chinese licensees in case of a cross-license agreement and, second, on removing unreasonable terms from license agreements including restrictions on licensees’ ability to challenge terms of their agreements. Interestingly, patents covered by the new commitments will be offered at new terms. These new conditions induce a reduction of the royalty base from 100% of the selling price to 65%. The applied rate will be « 5% for 3G devices (including multimode 3G/4G devices) and 3.5% for 4G devices (including 3-mode LTE-TDD devices) that do not implement CDMA or WCDMA ». The above-mentioned terms apply only to products to be sold for use in China. This means that a licensee’s sales of devices used in the US, EU, Japan or other major market will not, in principle, be impacted by the new terms. Of course, no one doubts that they will be offered to both existing and new licensees that may elect or refuse to take them. However, in the industry, licensees may also act as a contract manufacturer or an original design manufacturer (ODM). It then becomes less clear whether the new terms extend, for instance, to products designed and manufactured for another company. The rectification plans leaves open several implementation issues some of which are discussed in the following comments.
The decision opens more questions than it solves. From both the NRDC and Qualcomm’s statements it is not clear whether the agreed upon figures constitute the exact new terms to be offered to all existing licensees or whether they constitute a simple starting point for negotiation (royalty cap). Both interpretations give the idea that the NRDC is engaging in a sort of price regulation of intellectual property assets as it condemns the high level of royalties as such. Recently, Qualcomm itself reported that it is being targeted by EU and US regulators, and has been involved in private litigation concerning his licensing practices of its standard-essential patents in the US. In my personal view, it seems unlikely that the FTC will adopt an approach similar to the one put forward by the NRDC.
Regarding the EU, it is not completely clear how things will evolve. The EU Commission was extremely careful when targeting excessive prices especially in the case of immaterial assets. In this respect, the Commission investigations in the Qualcomm case provide a concrete example on how difficult it is to demonstrate that royalties are excessive and thus exploitative within the meaning of Article 102TFEU. On the other hand, there are many similarities between the settlement Rambus negotiated in 2009 in a case also about SEP’s and the corrective measures in the present case, which is an indicator that the Commission favors terms close to the ones in the Qualcomm case in China.
However, there may be an argument that by lowering royalties to be collected for devices sold for use in China, Qualcomm discriminates against branded devices to be exported to the EU and others countries and thus harming consumers in those parts of the world. The signal it may then send to the Commission is that it is charging excessive/unfair royalties for its essential European patents. This may prompt the Commission to intervene in order to bring royalties collected in the case of products sent to the European market to a level similar to the one applied for branded devices sold in China. Again, this seems very unlikely given that Qualcomm’s new terms in the Chinese case were extracted under the threat of a judicial proceeding.
In any case, price regulation (as done by the NRDC) differs from an eventual intervention from competition authorities in the EU or the US. This is to say that any move from the Commission or the FTC may not be as focused on the royalties amount in itself but rather on assessing whether the behavior of the dominant firm is designed as to limit existing or potential competition and as such to maintain the extraction of a supra competitive profits. However the only problem in this case is that there exists no competition and there will be no entry (unless the industry moves from the existing standards). Given the involvement of patents, excessive prices are not going to be challenged by competitors. This is what makes the issue of fairness and reasonableness a pressing one in the present case (the debate over the meaning of the terms “fair” and “reasonable” contained in the FRAND commitments is of critical importance here).
Should the Commission or the FTC intervene as “rate-setting” institutions and thus solve the issue relating to the meaning of the words “fair” and “reasonable”? The real answer should ensure that the likely impact on dynamic gains is fully taken into account. More clearly, profit is the main driver of R&D expenditures. However it appears that NRDC’s ruling added to the ex-ante uncertainty about profit with an ex-post cap. This effect can be better understood if we take the case of the industry’s eventual evolution toward the 5G technology. Firms will be less keen to invest in R&D for the 5G technology knowing that their returns are not only uncertain but will most likely be caped in one of the biggest markets in the world. In the name of fairness and reasonableness, price regulation of intellectual property assets can prove highly counterproductive having a detrimental impact on the evolution of technology.
Another striking aspect of the case is that Qualcomm’s licensing activity may be considered as abusive or not depending on the end user of the affected product. Devices sold for use in China will be subject to a specific regime in which royalties are still collected on the basis of the selling price but at a specific rate. In this way, the decision introduces what can be called the « Chinese consumption » criteria. Royalties applied for devices sold outside China are not considered as excessive (and so, not abusive) simply because they harm foreign consumers. Given that most of the devices are manufactured in China, the resulting mechanism has something that resembles to the idea of indifference toward an export cartel in that the involved practice affects mainly consumers outside the jurisdiction in which the conduct was initiated.
Vis-a-vis to Qualcomm’s licensees, the decision terms regarding royalties will most likely extend to other jurisdictions. It is hardly conceivable that Qualcomm will convince its licensees, at least the new ones, of any rate and base higher than what is adopted for devices sold for use in China. This domino effect is even more likely given that license agreements usually contain the « Most favored royalty rate provision » clause which means that the new terms need to be offered not only to licensees for sales of branded devices for use in China but also for the sales in other locations. In this respect, the NDRC’s decision may initiate a snowball effect that is hard to stop.
*Co-editor, Developing World Antitrust