Friday, 19 September 2014

Stacking the Deck in Analysis of Smartphone Patent Licensing Costs

Estimates of patent licensing costs for smartphone manufacturers are greatly exaggerated. Allegations of excessive fees paid and resulting harm to manufacturer profits, incentives to invest and compete are faulty and unsupported by the facts -- which show much to the contrary.
Allegations of excessive royalties
 and harm pile high in smartphones
A "working paper" entitled The Smartphone Royalty Stack: Surveying Royalty Demands for the Components Within Modern Smartphones has been published by one in-house lawyer at Intel and two outside counsel from WilmerHale. Intel Vice President and Associate General Counsel Ann Armstrong and WilmerHale's Joseph Mueller and Timothy Syrett argue that aggregate patent licensing fees including SEPs and non-SEPs are excessive at around $120 per $400 smartphone. They conclude that “few suppliers are meeting the basic goal of selling devices for more than the costs incurred in supplying them,” imply that this is due to the alleged royalty stack, and state that “those costs may be undermining industry profitability—and, in turn, diminishing incentives to invest and compete.”

The paper’s economic and empirical analyses are deficient and defective. In contradiction to its findings, evidence shows that licensing fees:
  • Are not undermining profits and are not preventing manufacturers from covering more than their costs. According to Credit Suisse, handset manufacturer operating profits since 2007 have tripled to $51 billion on $326 billion revenues in 2013.
  • Are not excessive. There is no basis for arbitrary price caps on smartphone patent fees, or limits based on chip manufacturing costs. The latter are unrelated to patented technologies and the value they generate more broadly in the entire device, its use in mobile networks, or across the broader ecosystem including services and applications. Methods of determining charges follow well established principles and benchmarks in bilateral negotiation. Patent licensing fees are analogous to licensing fees for book, music, movie or software publishers, which typically exceed greatly the cost of the physical mediums on which they are published and distributed.
  • Are nowhere near $120 in aggregate; and there is copious evidence actual payments are much lower than purported. The Paper inexplicably and erroneously disregards fundamental offsets in cross-licensing which greatly reduce or eliminate fees paid to many patent owners. This figure is also systematically biased and inflated by including rates demanded by licensors, even where there is no evidence anybody—including those who have little or nothing to cross license —actually pays such rates. And, where there is, instead, copious evidence that rates actually paid, if at all, are substantially less—orders of magnitude less in some instances. For example, court-adjudicated rates were much lower than “demanded” rates in various cases, and yet the higher figures were used in calculating the above total.
  • Are helpful, not detrimental, to the highly competitive and flourishing smartphone ecosystem. By every measure the patent system and the risk-reward balance it strikes—spurring innovation, market entry and competition while not overburdening licensees—is unmistakably working very well. 
My full and detailed analysis, in a pdf document, of this working paper by Intel and Wilmer Hale includes copious evidence countering the latter's findings.

This follows a previous my previous IP Finance posting on alleged royalty stacking entitled Theories of Harm with SEP Licensing do not stack up in which I responded to papers co-authored by Mark A. Lemley and Carl Shapiro in 2006 and 2013, and my posting entitled Absurd (F)RAND licensing-rate determinations for SEPs that analyses some U.S. court judgments which have relied on these economists in their royalty rate determinations. 


 



No comments: