Apple settled with Qualcomm as few others had its technology edge; the move also settles the canards about excessive royalty
Given that Apple contributed around 12-13% of Qualcomm’s revenues at the peak, it is not surprising that the two ending their billions-of-dollar lawsuits sent the latter’s stock soaring 23% immediately. Apart from a six-year global patent licensing agreement, Apple is believed to have paid Qualcomm around $5-6bn of pending royalties. When the two snapped ties in 2017, and Apple started producing devices without Qualcomm chips, the latter’s stock crashed 17% between December 30, 2016 and January 27, 2017 which is few days after Apple filed its first lawsuit. And from $57.18 on April 15 this year, the stock soared to $79.08 on April 17, the day after the settlement.
The fact that Apple has got back with Qualcomm underscores the quality of the latter’s technology since, while Apple had started working more with Intel, it didn’t have the same edge. And with Qualcomm supplying 5G modems, Apple’s rivals like Samsung had a head start. Once Apple-Qualcomm patched up, Intel said it would quit the 5G modem market.
The Washington Post (wapo.st/2PvyEBb) quotes from a presentation Qualcomm (bit.ly/2L1nTrD) made during the trial that shows how, even while Apple argued Qualcomm’s royalties were too high, it acknowledged its technology was superior. Qualcomm argued (see graphics) that while Apple talked of price-gouging, a royalty of $7.5 per device was a small part of what Apple’s devices retailed at. In 2017, the presentation said, while Apple’s net sales were $229bn, Qualcomm’s revenues were just $22bn, and this includes sales from other clients as well. Qualcomm also showed that, while the speed of mobile device downlinks and uplinks was rising rapidly, the prices of chip fell equally fast.
More important, the Apple-Qualcomm rapprochement should help end the needless controversy over royalty-gouging in India by firms like Ericsson. It has been argued, by Apple amongst others, that the patent regimes followed by firms like Qualcomm and Ericsson could make royalties prohibitive . If each innovator wants 3-4% of the device sale price as royalty, some argued, overall royalties could rise to even 25-30% of a device’s price due to ‘patent stacking’. While this is a theoretical possibility, the reality is different. For one, only firms with several strong patents, like Qualcomm, can command such rates. Nor do all firms want a royalty, some trade their patents for a better deal while using other firms’ patents. IDC data confirm this since prices of smart phones fell from around $453 in 2007 to around $135 in 2016 while tablet prices have fallen from $701 in 2010 to $223 in 2016; none of this would be possible if ‘patent stacking’ was taking place in any meaningful manner.
Falling prices are directly related to the licensing model followed. In traditional patent models, such as in pharmaceuticals, a license to manufacture can be given to one or two firms, but this is exclusionary in the sense that they alone can manufacture. In the Qualcomm model, licences are given to any manufacturer, as long as it agrees to pay a reasonable royalty; it has to be this way since any phone could be using hundreds of patents at the same time. An industry standards body examines all patent applications for innovation etc and declares some of them to be Standard Essential Patents (SEPs); these SEPs are to be available to anyone on a Fair Reasonable And Non-Discriminatory (FRAND) basis. The technology that a Qualcomm supplies an Apple, for instance, is available to everyone; indeed, because of this, there has been an explosion in the number of device manufacturers, whether phones or tablets, and that is what has made prices crash.
Another argument is that royalty should be based on the value of the component being sold, say a chipset, and not on the phone’s price. While this sounds reasonable, the patent does not always reside at the chipset level, it can be anywhere in the network because the technology is for the overall system, not just for a component; if a patent helps utilise spectrum more efficiently,for instance, should the chipset-maker pay royalty, or the telco or the handset manufacturer? A Qualcomm can charge each phone-user a royalty and also charge each telco something, but charging this to just the device makes licensing simpler.
Another version of this argument is that an absolute value should be charged instead of pegging the royalty to the handset price. This also sounds reasonable since, if the same chip is used in a $100 handset and in a $1,000 one, it seems unfair that the patent-holder gets $3 in the first case and $30 in the other, assuming a 3% royalty rate. But since the $1,000 phone is better able to unlock the technology’s potential because of a bigger screen or better video capability, for instance, a higher royalty in absolute terms is reasonable. In any case, royalties don’t scale up in a linear fashion, there is a cap beyond a certain value.
While there are several ongoing disputes between Ericsson and some device manufacturing firms like Intex and Lava in India, and the Telecom Regulatory Authority of India (TRAI) is also examining the issue of how royalty rates should be charged, it is important to keep in mind that, with FRAND terms, technology suppliers are offering more and more capabilities to more and more device manufacturers, as a result of which the customer is benefitting; and more so since device prices also continue to crash even as they get more powerful. The Apple-Qualcomm patch up, above all, is testimony to that.