Case 3:11-cv RBD-TEM Document 305 Filed 07/25/13 Page 1 of 27 PageID 11914

Transcription

1 Case 3:11-cv RBD-TEM Document 305 Filed 07/25/13 Page 1 of 27 PageID PARKERVISION, INC., Plaintiff, UNITED STATES DISTRICT COURT MIDDLE DISTRICT OF FLORIDA JACKSONVILLE DIVISION v. Case No. 3:11-cv-719-J-37-TEM QUALCOMM INCORPORATED, Defendant. PARKERVISION S RESPONSE TO QUALCOMM S DAUBERT MOTION TO PRECLUDE PAUL BENOIT DAMAGES TESTIMONY

2 Case 3:11-cv RBD-TEM Document 305 Filed 07/25/13 Page 2 of 27 PageID TABLE OF CONTENTS I. INTRODUCTION...1 II. BACKGROUND...1 III. ARGUMENT...3 McKool A. Benoit s Use of Qualcomm s Own Estimates to Derive the Benefit or Portion of Operating Profit Attributable to ParkerVision s Invention is Proper Benoit s Methodology is Sound, As Confirmed by Powell v. Home Depot Benoit Relies on Reliable Evidence He Relies on the Best Evidence of How Qualcomm Would Value ParkerVision s Technology Benoit s Evaluation of the Prior Negotiations Between Qualcomm and ParkerVision is Not a Comparison of Past Patent Licenses Analysis Differences in the Market Between 1999 and 2006, Considered Within Benoit s Analysis, Present Fact Issues for the Jury...8 B. Benoit s Reasoned, Fact-Specific Application of the Nash Bargaining Solution Cannot be Characterized as an Improper Rule of Thumb Benoit Reliably Applies the Nash Bargaining Solution by Conducting a Careful Analysis Closely Tied to the Facts of this Case A 50/50 Split Of Incremental Profit Under the NBS Does Not Constitute An Impermissible Rule of Thumb...15 C. Unlike the VIA Agreement, the Technology at Issue was the Motivation for the Symbol Agreement D. Benoit Presents An Economically Sound And Legally Sufficient Model for Convoyed Sales...18 E. Benoit s Opinion on Indirect Infringement Damages Applies to Damages for Both Induced and Contributory Infringement...20 i

3 Case 3:11-cv RBD-TEM Document 305 Filed 07/25/13 Page 3 of 27 PageID TABLE OF AUTHORITIES CASES Page(s) Amakua Dev. LLC v. Warner, No. 05-CV-3082, 2007 U.S. Dist. LEXIS (N.D. Ill. July 10, 2007)...16 Bonura v. Sea Land Serv., Inc., 505 F.2d 665 (5th Cir. 1974)...17 Bose Corp. v. JBL, Inc., 274 F.3d 1354 (Fed. Cir. 2001)...4 Dynacore Holdings Corp. v. U.S. Philips Corp., 363 F.3d 1263 (Fed. Cir. 2004)...20 Energy Transp. Grp., Inc. v. William Demant Holding A/S, 697 F.3d 1342 (Fed. Cir. 2012)...14 Gen-Probe Inc. v. Becton Dickinson & Co., No.3:09-cv BEN-NLS (S.D. Cal Nov. 6, 2012)...16 Glenayre Elecs., Inc. v. Jackson, 443 F.3d 851 (Fed. Cir. 2006)...20 Grand River Enters. Six Nations v. King, No. 02-CV-5068, 2011 U.S. Dist. LEXIS (S.D.N.Y. Mar. 17, 2011)...16 i4i Ltd. P ship v. Microsoft Corp., 598 F.3d 831 (Fed. Cir. 2010)...14, 19 Inline Connection Corp. v. AOL Time Warner, Inc., 470 F. Supp. 2d 424 (D. Del. 2007)...1 Interactive Pictures Corp. v. Infinite Pictures, Inc., 274 F.3d 1371 (Fed. Cir. 2001)...4 LG Elecs. U.S.A., Inc. v. Whirlpool Corp., No. 1:08-cv-242, 2010 U.S. Dist. LEXIS (N.D. Ill. Aug. 24, 2010)...6 Martinez v. Altec Indus., No. 3:02-cv-1100-J-32TEM, 2005 U.S. Dist. LEXIS (M.D. Fla. Aug. 3, 2005)...13 Mformation Techs., Inc. v. Research in Motion Ltd., No. C , 2012 U.S. Dist. LEXIS (N.D. Cal. Mar. 29, 2012)...16 McKool ii

4 Case 3:11-cv RBD-TEM Document 305 Filed 07/25/13 Page 4 of 27 PageID Micro Chem., Inc. v. Lextron, Inc., 317 F.3d 1387 (Fed. Cir. 2003)...14, 19, 20 Minebea Co. v. Papst, No. 97-CV-590, 2005 U.S. Dist. LEXIS (D.D.C. June 21, 2005)...16 Minks v. Polaris Indus., 546 F.3d 1364 (Fed. Cir. 2008)...19 Oracle Am., Inc. v. Google Inc., 798 F. Supp. 2d 1111 (N.D. Cal. 2011)...17 PACT XPP Techs., AG v. Xilinx, Inc., No. 2:07-cv-563, slip. op. (E.D. Tex. May 11, 2012)...6 Powell v. Home Depot U.S.A., Inc., 663 F.3d 1221 (Fed. Cir. 2011)...4, 6 Power Integrations v. Fairchild, 711 F.3d 1348 (Fed. Cir. 2013)...7 Quiet Tech. DC-8, Inc. v. Hurel-Dubois UK Ltd., 326 F.3d 1333 (11th Cir. 2003)...10, 12 ResQNet.com, Inc. v. Lansa, Inc., 594 F.3d 860 (Fed. Cir. 2010)...8 Rite-Hite Corp. v. Kelley Co., 56 F.3d 1538 (Fed. Cir. 1995) (en banc)...19, 20 Rosenfeld v. Oceania Cruises, Inc., 654 F.3d 1190 (11th Cir. 2011)...3 Sanofi-Aventis Deutschland GmbH v. Glenmark Pharms., Inc., USA, No. 07-CV-5855, 2011 U.S. Dist. LEXIS (D.N.J. Feb. 3, 2011)...16, 17 Snellman v. Ricoh Co., 862 F.2d 283 (Fed. Cir. 1988)...4 Suffolk Techs. LLC v. AOL, Inc., 2013 U.S. Dist. LEXIS (E.D. Va. April 12, 2013)...18 Trans-World Mfg. Corp. v. Al Nyman & Sons, Inc., 750 F.2d 1552 (Fed. Cir. 1984)...19 TWM Mfg. Co. v. Dura Corp., 789 F.2d 895 (Fed. Cir. 1986)...4, 8 McKool iii

5 Case 3:11-cv RBD-TEM Document 305 Filed 07/25/13 Page 5 of 27 PageID VirnetX Inc. v. Apple Inc., No. 6:10-CV-417, 2013 U.S. Dist. LEXIS (E.D. Tex. Feb. 26, 2013)...16 VirnetX Inc. v. Cisco Sys., No. 6:10-CV-417, 2013 U.S. Dist. LEXIS (E.D.Tex. Mar. 1, 2013)...16 OTHER AUTHORITIES Donald S. Chisum, Chisum on Patents Federal Rule of Evidence , 17 McKool iv

6 Case 3:11-cv RBD-TEM Document 305 Filed 07/25/13 Page 6 of 27 PageID I. INTRODUCTION ParkerVision s damages expert, Paul Benoit, has conducted a careful analysis under the Georgia-Pacific factors to model the parties hypothetical negotiation, quantify the total benefit that Qualcomm has derived from its infringement of ParkerVision s patents-in-suit, and determine how Qualcomm and ParkerVision would have split those benefits. Benoit bases his analyses solidly in the facts and evidence, including the prior negotiations that took place between ParkerVision and Qualcomm regarding the very same receiver technology at issue in this case. Each of Qualcomm s arguments only challenges the factual underpinnings of Benoit s analysis, not his methodology. Thus, each of Qualcomm s arguments go to the weight of the evidence, not the admissibility of Benoit s testimony, and must be rejected at the Daubert stage. II. BACKGROUND A step-by-step examination of the methodology applied by Benoit to calculate the damages to which ParkerVision is entitled demonstrates its reliability and accuracy. To calculate the reasonable royalty, Benoit analyzed each of the Georgia-Pacific factors. Benoit s analysis of Georgia-Pacific factors 1 through 5 considers the various license agreement transactions entered into between ParkerVision and its licensees for use of the patents-in-suit. Ex. 1 at However, unlike Qualcomm s expert, Dr. Gregory Leonard, Benoit did not rely exclusively on his review of market transactions for use of the patents-in-suit. Georgia-Pacific factors 6 through 13 embody many of the traditional valuation concepts reflected in the Income Approach to valuation. The Income Approach estimates an asset s price based on the value of the benefits derived from the use of that asset. Inline Connection Corp. v. AOL Time Warner, Inc., 470 F. Supp. 2d 424, 432 (D. Del. 2007). Applying the Income Approach concepts, Benoit quantifies the benefit that Qualcomm has derived from infringement of the patents-in-suit by segregating the profits attributable to the patents-in-suit from Qualcomm s intellectual property and intangible assets. Ex. 1 at 95, Ex. C, D-4, D-5. Benoit s analysis is grounded in Qualcomm s own, internally-created analysis, a thorough analysis of the McKool

7 Case 3:11-cv RBD-TEM Document 305 Filed 07/25/13 Page 7 of 27 PageID intervening facts and circumstances, as well as Benoit s independent analyses of Qualcomm s revenue and profit from use of the patents-in-suit. See Ex. 1 at 81-82, , Ex. B-P. As discussed in Section III.A, Benoit then, within Georgia Pacific factors 14 and 15, identifies the ratio at which Qualcomm and ParkerVision would have agreed to split the benefits of the patents-in-suit through an application of the Nash Bargaining Solution (NBS). Id. at Thus, by conducting a detailed apportionment analysis, Benoit calculates a total benefit of $.80 per receiver and $1.30 per transceiver. Then, through a careful analysis of the NBS, Benoit derives a split of Qualcomm s incremental profit as result of infringement of 50% to ParkerVision and 50% to Qualcomm. Id. at 105. The table below summarizes Benoit s detailed reasonable royalty analysis for the infringing Qualcomm transceivers: BENOIT S QUANTIFICATION OF TRANSCEIVER ROYALTY RATE Average Sales Price (ASP) of Transceiver Less: Transmitter Functionality Unrelated to the Patents-in-Suit Transceiver Functionality Less: Other Transceiver Functionality Contributed by Qualcomm Total GP Factor Transceiver Revenue Apportioned to Patents-in-Suit Qualcomm s Incremental Profit Margin 8, 9&10 Profit Apportioned to Patents-in-Suit and Other Intangible Assets Percent of Apportioned Profit Allocated to Patents-in-Suit 12&13 Profit Allocated to Patents-in-Suit Convoyed Sales Benefit per Transceiver 6 Benefit from Use of Patents-in-Suit per Transceiver Infringing Transceivers Total Benefit from Use of Patents-in-Suit Implementation Costs 7 Net Benefit from Use of Patents-in-Suit Percent of Net Benefit Retained by Licensee (NBS) Reasonable Royalty for Use of Patents-in-Suit Infringing Transceivers Royalty Per Transceiver McKool

8 Case 3:11-cv RBD-TEM Document 305 Filed 07/25/13 Page 8 of 27 PageID III. ARGUMENT Qualcomm challenges Benoit s opinion regarding damages on five grounds: Asserting that Benoit took a shortcut by utilizing Qualcomm s own, internal documents to measure of the amount of Qualcomm s incremental operating profit that should be apportioned to the patents-in-suit; Contending that Benoit may not consider the parties prior negotiations regarding the very same receiver technology at issue in this case because, according to Qualcomm, the facts surrounding the negotiations do not comply with the requirements of comparability for reliance on a royalty rate from a prior patent license agreement; Arguing that Benoit may not glean information from the parties prior negotiations because the passage of time and intervening events between 1999 and the hypothetical negotiation in 2006 render them fundamentally different; Attempting to undermine Benoit s fact specific analysis of the Nobel-prize winning NBS by comparing it to a rule of thumb an argument rejected by an overwhelming majority of courts to address the issue; and Disregarding Qualcomm s very own documents reflecting that Qualcomm believed that ParkerVision s technology would drive sales of basebands, and claiming that Benoit lacks evidence to support his convoyed sales opinion. Each of Qualcomm s challenges is baseless. Qualcomm does not argue that Benoit has misapplied any of the Georgia-Pacific factors, nor does Qualcomm challenge the Income Approach concepts employed by Benoit or argue that the NBS is fundamentally unreliable. Indeed, Qualcomm cannot make such arguments because it has not offered its own expert opinion as to the quantification of the benefit obtained by Qualcomm from use of the patents-insuit. See Ex. 3 at 90:5-91:2. Nor has it offered its own expert opinion on the appropriate split of benefit to which the parties would have agreed. Id. at 92:20-93:14, 95:4-96:10. Rather, Qualcomm repeatedly resorts to challenging the factual underpinnings of Benoit s analysis not his methodology. But [v]igorous cross-examination, presentation of contrary evidence, and careful instruction on the burden of proof are the traditional and appropriate means of attacking admissible but disputed evidence. See Rosenfeld v. Oceania Cruises, Inc., 654 F.3d 1190, 1193 (11th Cir. 2011) (quoting Daubert v. Merrell Dow Pharms., Inc., 509 U.S. 579, 596 (1993)). A. Benoit s Use of Qualcomm s Own Estimates to Derive the Benefit or Portion of Operating Profit Attributable to ParkerVision s Invention is Proper McKool

9 Case 3:11-cv RBD-TEM Document 305 Filed 07/25/13 Page 9 of 27 PageID Benoit s Methodology is Sound, As Confirmed by Powell v. Home Depot Benoit has conducted a meticulously thorough analysis of the amount of Qualcomm s incremental profit from receivers that is attributable to Qualcomm s infringement of the patentsin-suit. 1 See Ex. 1 at , Exs. C-P. His analysis properly incorporates estimates that Qualcomm itself generated regarding the anticipated incremental profits from its use of the patents-in-suit. Ex. 1 at 94-95, Ex. D-4. Indeed, there is likely no better evidence of the benefit of the patents-in-suit to Qualcomm than the calculations made by Qualcomm s very own personnel unrelated to this litigation. Qualcomm s argument that Benoit may not rely on Qualcomm s own pre-infringement estimates of the benefits that it would derive from use of the patents-in-suit is directly refuted by numerous Federal Circuit cases, 2 most recently Powell v. Home Depot U.S.A., Inc., 663 F.3d 1221 (Fed. Cir. 2011). 3 Qualcomm s internal analysis regarding profit sharing reflects how much of the value of the receiver Qualcomm attributed to the patents-in-suit, i.e., the implied benefit from use of the patents-in-suit. During the prior negotiations, Qualcomm personnel prepared a series of internal analyses in order to value ParkerVision s technology and guide Qualcomm in formulating the offers that it proposed to ParkerVision. See, e.g., Exs In February 1999, Qualcomm prepared an analysis quantifying the amount that it contemplated paying to ParkerVision under a 1 While Qualcomm cites Lucent Techs., Inc. v. Gateway, Inc. and Cornell Univ. v. Hewlett-Packard Co. in support of its claim that Benoit s analysis should be excluded for failing to provide a demand curve or other economic link between consumer demand and ParkerVision s invention, the cases are inapposite because the entire market value rule does not apply here. Dkt. 288 at Regardless, there is no requirement of demand curves. See Bose Corp. v. JBL, Inc., 274 F.3d 1354, 1361 (Fed. Cir. 2001) (affirming damages where plaintiff did not use demand curves). In any event, Benoit established the economic demand for ParkerVision s invention by considering the incremental benefit of infringement to Qualcomm. See Ex. 2 at 2 See, e.g., Interactive Pictures Corp. v. Infinite Pictures, Inc., 274 F.3d 1371, 1384 (Fed. Cir. 2001) (holding damage expert s testimony was not speculative when he relied on infringer s past business plan); Snellman v. Ricoh Co., 862 F.2d 283, 289 (Fed. Cir. 1988) ( The jury justifiably could have calculated a lump sum royalty based on [infringer s] expected sales. ); TWM Mfg. Co. v. Dura Corp., 789 F.2d 895, 900 (Fed. Cir. 1986) (affirming use of a pre-infringement memorandum of projected profits to arrive at a reasonable royalty in part because the infringer used the figures in the memorandum in deciding whether to manufacture and market the infringing device ). 3 Powell reinforced that the infringer s profit expectation may be considered in the overall reasonable royalty analysis. Powell, 663 F.3d at There, just as here, the parties negotiated (in 2004) for Home Depot to access Powell s technology prior to the issuance of the patents-in-suit (in 2006). Id. The jury considered the royalty rates contemplated in 2004 when calculating the reasonable royalty (arriving at a higher rate). Id. at McKool

10 Case 3:11-cv RBD-TEM Document 305 Filed 07/25/13 Page 10 of 27 PageID hypothetical profit sharing agreement. Ex. 4 at This analysis reflects that Qualcomm considered paying ParkerVision 29.3% of the operating profit from the infringing products in a profit sharing arrangement. 4 See id.; Ex. 1 at 26, 69. This 29.3% represents the amount of the benefit that Qualcomm itself calculated would have been derived from the use of ParkerVision s technology, divorced from its negotiation strategies designed to lower the cost of a license to ParkerVision s technology. See Ex. 2 at 145: Benoit reasoned that Qualcomm would never quantify a profit share that would exceed the benefit contributed by the technology. Ex. 1 at 70. The 29.3% of operating profits represents the total benefit for three types of products: receivers, transceivers, and integrated transceivers/basebands. Each of these products benefits from the patents-in-suit at differing levels. Ex. 2 at 120:8-22. To break out the amount of benefit specific to each product type, Benoit looked to the ratios between the different royalty rates contemplated by the parties in the 1999 negotiations for the receiver, transceiver, and integrated transceiver/baseband to determine the relative importance of ParkerVision s technology in each product. See Ex. 1 at Ex. N, N.1; Ex. 2 at 120:25-121:5. Utilizing these ratios and Qualcomm s calculation of a 29.3% share of operating profit, Benoit calculated the percentage of revenue attributable to ParkerVision s technology for each product type. 5 Ex. 1 at Ex. D-4; Ex. 2 at 120:5-121:5. The resulting percentages, 46.6% for receivers, 36.6% for transceivers, and 12.4% for integrated transceivers/basebands, have a weighted average of 17.9%. Id. at Ex. D-4. This 17.9% represents the percentage of revenue attributable to the ParkerVision technology it is not a royalty rate. Ex. 2 at 122: This 17.9% of revenue equates to a total benefit attributable to ParkerVision s technology of as much as 81% of receiver operating profit, 6 62% of transceiver 4 The spreadsheet stated that Qualcomm would pay ParkerVision $352 million between 1999 and Ex. 4 at The total operating profit that Qualcomm estimated for in this analysis was $1,200,000,000. See id.; Ex. 1 at Ex. D-4. Accordingly, the anticipated payments to ParkerVision constitute 29.3% of Qualcomm s estimated operating profit ($352 / $1,200 =.2933). See id. 5 This calculation results in what Benoit calls a royalty rate adjustment factor of Ex. 1 at Ex. D-4. Qualcomm mischaracterizes this adjustment as an attempt by Benoit to inflate the prior negotiations royalty rates by Dkt. 288 at However, as described by Benoit in his deposition, the adjustment factor is not used to derive a royalty rate, but is instead the end result of Benoit s calculation to determine the percentage of revenue that reflects Qualcomm s benefit from the use of ParkerVision s technology. See Ex. 2 at 121:23:25; 129:19-22; 272: The average royalty for the receiver / the average receiver operating profit, or $21/$26 =.811. See Ex. 1 at D-4. McKool

11 Case 3:11-cv RBD-TEM Document 305 Filed 07/25/13 Page 11 of 27 PageID operating profit, and 20% of integrated baseband operating profit. 7 See Ex. 1 at 228, Ex. D-4. Qualcomm and its damages expert, Dr. Leonard, do not contest these calculations. The fact that Qualcomm offered ParkerVision royalty rates reflecting a smaller share of operating profit than the amounts calculated under the internal profit sharing analysis prepared by Qualcomm personnel does not (as Qualcomm claims) somehow undermine the profit sharing analysis. See Dkt. 288 at Indeed, it is unsurprising. It should be expected that in business negotiations Qualcomm s internal calculation of the total potential benefit from ParkerVision s technology would be larger than the amount that Qualcomm sought to share with ParkerVision in royalties. As in the Powell case, it is a fact issue for the jury to determine whether and how the substance of the parties prior discussions should be applied. See Powell, 663 F.3d at Benoit Relies on Reliable Evidence He Relies on the Best Evidence of How Qualcomm Would Value ParkerVision s Technology Qualcomm argues that the data that Benoit utilizes in his analysis, specifically the spreadsheets created by Qualcomm in 1999, renders Benoit s entire calculus unreliable. But for at least the following four reasons, Qualcomm s own spreadsheets and the circumstances surrounding their creation demonstrate their accuracy and reliability. 8 First, as previously noted, the documents relied upon by Benoit were created by Qualcomm employees in an attempt to quantify the benefit of ParkerVision s technology to Qualcomm and to calculate various options for acquiring access to ParkerVision s technology. 9 Second, Qualcomm created at least six 7 Qualcomm asserts that [a]n economist would have performed the required economic analysis to show what actually drove Qualcomm s profits for accused products. Benoit simply did not perform the required analysis to support this foundational aspect of his opinion. Dkt. 288 at 3. But as a check on his 81.1% analysis, Benoit calculated the percentage of operating profit derived from ParkerVision s technology at the time of the hypothetical negotiation without utilizing data from the prior negotiation. Ex. 1 at Using that analysis, Benoit calculated that approximately 90.9% of Qualcomm s operating profit attributable to receiver functionality embodied in Qualcomm s infringing transceivers is enabled by the use of the patents-in-suit. Id. at , Ex. L. The 81.1% number was used to be conservative. Ex. 1 at Qualcomm cannot now distance itself from information it relied on during the ordinary course of business. See PACT XPP Techs., AG v. Xilinx, Inc., No. 2:07-cv-563, slip. op. at *2-3 (E.D. Tex. May 11, 2012) (holding that defendants complaints about plaintiff s expert s interpretation of their internal documents went to weight, not admissibility); LG Elecs. U.S.A., Inc. v. Whirlpool Corp., No. 1:08-cv-242, 2010 U.S. Dist. LEXIS 87721, at *10-18 (N.D. Ill. Aug. 24, 2010) (rejecting complaints about plaintiff s expert s use of defendants internal documents, and noting that [s]ignificantly, [plaintiff s expert] relies on documents from [defendant s], not from [plaintiff s] files ). 9 Qualcomm s attempt to analogize Benoit s analysis to Power Integration falls flat. See Dkt. 288 at In Power McKool

12 Case 3:11-cv RBD-TEM Document 305 Filed 07/25/13 Page 12 of 27 PageID iterations of the spreadsheet that Benoit relies upon demonstrating that Qualcomm employees subjected the documents to scrutiny and editing but despite revisions to other portions of the document, the profit sharing calculations remained constant. See Exs Third, Qualcomm ignores the fact that the authors and recipients of these documents were decision-makers charged with evaluating ParkerVision s technology, vetting Qualcomm s business case regarding a deal with ParkerVision, and formulating Qualcomm s offers to ParkerVision. See Ex. 26 at 76:8-25 (Mr. Kantak explaining that he had the buy-in of senior Qualcomm executives). The authors and recipients of these documents, Prashant Kantak and Andy Oberst, have risen through the ranks at Qualcomm to senior positions, reflecting that they conscientiously represent Qualcomm s best interests. 10 And fourth, beyond pointing to the royalty rates that Qualcomm actually proposed to ParkerVision, which would have been influenced by Qualcomm s motivation to negotiate the cheapest deal possible, Qualcomm has offered no evidence that the numbers reflected in these spreadsheets are unreliable or incorrect. More importantly, the accuracy of Qualcomm s own spreadsheets on which Benoit relies presents a pure question of fact for the jury not a Daubert issue. To the extent that Qualcomm believes that its own documents should not have been relied upon because they are unreliable, incorrect, or unclear, it is free to call Mr. Kantak and Mr. Oberst to the stand and elicit testimony as to why the documents do not accurately represent Qualcomm s 1999 estimates of the benefits that would flow from Qualcomm s implementation of the patents-in-suit. 3. Benoit s Evaluation of the Prior Negotiations Between Qualcomm and ParkerVision is Not a Comparison of Past Patent Licenses Analysis Qualcomm argues that the Federal Circuit requires that damages opinions rely on comparable patent licenses for comparable rights and subject matter. Dkt. 288 at 3. This is certainly true and is precisely why the Court should exclude Leonard s opinions regarding the Integration, the expert relied on documents found on the Internet, with no identifiable source. Power Integrations v. Fairchild, 711 F.3d 1348, 1373 (Fed. Cir. 2013). Here, Benoit relied on documents created by Qualcomm, found in Qualcomm s production, and which Qualcomm has essentially admitted are authentic. See Ex. 7 at Andy Oberst is a Senior Vice President for Qualcomm s semiconductor business and Prashant Kantak served as a Senior Director of Business Development until See Ex. 26 at 30:3-22. McKool

13 Case 3:11-cv RBD-TEM Document 305 Filed 07/25/13 Page 13 of 27 PageID WRF License and the VIA Agreement as ParkerVision has argued. See Dkt. 291 at But the Federal Circuit s requirement of comparability is specific to market transactions, such as patent licenses, reflecting royalty rates that an expert considers to be instructive as to the patents-in-suit under Georgia-Pacific factors 1 or Benoit has not opined that the royalty rates discussed during the negotiations are instructive as to the royalty rates that the parties would have agreed to for use of the patents-in-suit in the hypothetical negotiation. Thus, Benoit does not use the prior negotiations as a comparable license under Georgia Pacific factors 1 and 2 to determine applicable royalty rates. Rather, as explained above, Benoit utilizes the parties prior negotiations in his Income Approach analysis under Georgia-Pacific factors 6 through 13 to quantify the economic benefit that Qualcomm would have obtained through the use of ParkerVision s technology. Ex. 1 at ; see Ex. 2 at 52:3-54:12. He then analyzes the factors that would have impacted that benefit between 1999 and Ex. 1 at This is a completely different analysis than one taken under Georgia Pacific 1 and 2 and therefore is not subject to the comparability requirements of ResQNet. Not only is Benoit s use of the prior negotiation documents not controlled by requirements regarding the use of comparable patent licenses under Georgia-Pacific factors 1 and 2, but Benoit s analysis addresses each of the points raised by Qualcomm. First, Qualcomm states that the 1999 analysis that Benoit utilized in his calculations reflects a proposed transfer of technology to Qualcomm. Dkt. 288 at But Benoit s analysis includes an apportionment for the implementation costs that Qualcomm would have incurred. Ex. 1 at 80, 97, Ex. C, C.1. Second, considerations of exclusivity or non-exclusivity affect the royalty rate to which the parties might agree, but do not affect the benefit that Qualcomm would have derived from the use of ParkerVision s technology. Ex. 2 at 144:7-145:25, 169: Differences in the Market Between 1999 and 2006, Considered Within 11 Compare ResQNet.com, Inc. v. Lansa, Inc., 594 F.3d 860, 870 (Fed. Cir. 2010) (requiring comparability when expert relied on royalties in seven prior licensing agreements to support a reasonable royalty rate), with TWM Mfg., 789 F.2d at 900 (not imposing comparability requirements when expert calculated damages based on infringer s prior profit projections). McKool

14 Case 3:11-cv RBD-TEM Document 305 Filed 07/25/13 Page 14 of 27 PageID Benoit s Analysis, Present Fact Issues for the Jury Similarly, Qualcomm s argument that courts routinely bar reliance on outdated patent licenses plainly applies only to patent licenses. Dkt. 288 at 15. Again, Benoit does not use the prior negotiations as a comparable royalty rate. He instead uses the negotiations and analysis from 1999 to determine the economic benefit that Qualcomm perceived it would receive from the technology. Ex. 2 at 121: In any event, Benoit carefully accounted for the technical and economic changes in the intervening years between the 1999 negotiations and the 2006 hypothetical negotiation and validated that the value of the benefits attributed to ParkerVision s technology not only remained relevant, but increased over time. See Ex. 1 at , For example, Benoit examined the profitability of Qualcomm in and found that it is similar to Qualcomm s profitability between 2006 and today. See Ex. 2 at 259:20-260:7. Where necessary, Benoit adjusted Qualcomm s analysis from 1999 in order to account for the differences between 1999 and Moreover, Qualcomm s arguments yet again go to the weight to be accorded to Benoit s opinions and not their admissibility. Qualcomm argues that [w]hen the parties talked in 1998, Qualcomm did not have a proven direct downconversion receiver. In 2006, it did. Benoit ignores that fundamental difference in bargaining positions driven by the passage of time and intervening events. Dkt. 288 at 4. But Benoit addresses these issues and concludes that the royalty rate to which ParkerVision and Qualcomm would agree in a hypothetical negotiation in 2006 would be higher today than the royalty rates contemplated by the parties in 1998 and 1998 based on a number of factors. Ex. 1 at 101. He specifically considers that by 2006: the patents-in-suit have issued; ParkerVision s technology has been vetted and proven in actual wireless products; the market demand for the benefits of ParkerVision s patents-in-suit has climbed since 1999; and no commercially acceptable non-infringing alternatives exist. Id. at At the Daubert stage 12 For example, Benoit used actual average sales prices and gross profits of the infringing products rather than the 1999 estimates; he accounts for implementation costs in 2006 because know-how would not have been transferred under a bare patent license; he measured Qualcomm s non-patented contributions in 2006, but conservatively utilized Qualcomm s 1999 calculations; and he considered that in 2006 Qualcomm had no commercially acceptable non-infringing alternative See Ex. 1 at 80, 98-99, Ex. D, Ex. G; Ex. 2 at 44:3-14. McKool

15 Case 3:11-cv RBD-TEM Document 305 Filed 07/25/13 Page 15 of 27 PageID it is not the role of the district court to make ultimate conclusions as to the persuasiveness of the proffered evidence and Qualcomm s evidence-based challenge should be rejected. Quiet Tech. DC-8, Inc. v. Hurel-Dubois UK Ltd., 326 F.3d 1333, 1341 (11th Cir. 2003). B. Benoit s Reasoned, Fact-Specific Application of the Nash Bargaining Solution Cannot be Characterized as an Improper Rule of Thumb The NBS is a model described by the Nobel Prize winner John Nash in two classic scholarly papers. See Ex. 8; Ex. 9. The NBS describes an axiomatic solution to a two-player bargaining problem, given known disagreement benefits. See Ex. 2 at 17:8-18:5; Ex. 10 at ; Ex. 11 at 176. The simplicity and flexibility of NBS to consider a range of variables has contributed to its wide-spread application by academics and experts across numerous different disciplines. 13 Academics and experts studying and applying NBS have attempted to refine its axioms, tailor it to specific areas of study, and have incorporated a litany of variables and modifications. See, e.g., Ex. 10; Ex. 11. One of the specific areas of study where economists have applied the NBS is in the patent damages context. See Exs. 10, Thus, application of the NBS to determine a reasonable royalty in a patent damages context is a reliable methodology. As the court in Summit 6 LLC v. Research in Motion Corp. stated when addressing Daubert and rejecting challenges to Benoit s application of NBS: The Court finds Benoit s use of NBS was not an improper 50% rule of thumb, but rather based on his belief that because neither party had a stronger negotiating position, they would have split the profits evenly. Thus, Benoit s analysis is not inherently flawed and unreliable based on his use of NBS. Also, the Court reiterates the Federal Circuit s explanation that a district court should not use Daubert to evaluate the correctness of facts underlying an expert s testimony. Questions about what facts are most relevant or reliable to calculating a reasonable royalty are for the jury. The jury was entitled to hear the expert testimony and decide for itself what to accept or reject. i4i, 598 F.3d at 856 Based on the foregoing, the Court finds that Benoit s testimony was not flawed and should not have been excluded under Daubert. Summit 6, 3:11-cv-367-O, 2013 U.S. Dist. LEXIS 95164, at *36-37 (N.D. Tex. June 26, 2013). 13 Ex. 14, William Darity, Jr., INTERNATIONAL ENCYCLOPEDIA OF THE SOCIAL SCIENCES 540 (2d ed. 2007) ( Nash equilibrium is a fundamental concept in the theory of games and the most widely used method of predicting the outcome of a strategic interaction in the social sciences. ). McKool

16 Case 3:11-cv RBD-TEM Document 305 Filed 07/25/13 Page 16 of 27 PageID Benoit Reliably Applies the Nash Bargaining Solution by Conducting a Careful Analysis Closely Tied to the Facts of this Case Qualcomm s challenge of Benoit s application of NBS must be rejected because (1) Benoit properly adheres to the NBS model as described by Nash; (2) Qualcomm s criticism of Benoit s application of NBS relies on superfluous constructs not found in Nash s papers; (3) Qualcomm relies on mischaracterizations of Benoit s deposition testimony; and (4) Qualcomm resorts to baselessly disparaging Benoit s credentials. First, Benoit reliably applies NBS to evaluate how the parties would have arrived at an agreement in the hypothetical negotiation. Ex. 1 at A sound application of the NBS, as described by Nash in his two foundational papers, requires knowledge of the total profits available to both parties, the amount of disagreement benefits that would be achieved by each party in the event an agreement could not be reached, and the relative bargaining power of the parties. See Ex. 9 at ; Ex. 10 at 248; Ex. 12 at 5-6. Inputting these variables into the NBS equations derives the split of benefits that the negotiators would have agreed to. See id. Benoit s application of the NBS properly adheres to the NBS model as described by Nash and subsequently described in the patent reasonable royalty construct by Weinstein, Jarosz, and Putnam. Benoit defined the terms for each of the NBS variables and performed the NBS equation using those variables. Ex. 1 at , Ex. C, Ex. L; Ex. 2 at 25:13-27:3. Qualcomm assails Benoit s use of NBS, but Qualcomm s damages expert does not offer his own NBS analysis nor offer an opinion as to the split of the benefits that the parties would have agreed to. See Ex. 15at Indeed, the work of Qualcomm s own damages expert Leonard shows that Benoit s methodology is sound: Economic models of bargaining suggest that the split of the gains to trade will be influenced by the companies relative levels of patience When the two companies are sufficiently patient and are roughly equally patient, the gains to trade will be approximately equally divided.. The resulting Mexican standoff yields an equal split of the profits. Ex. 16 at Leonard therefore agrees that depending upon their relative bargaining position, the parties may end up in a standoff where they will agree to an equal split of the McKool

17 Case 3:11-cv RBD-TEM Document 305 Filed 07/25/13 Page 17 of 27 PageID profits. Id. In addition Leonard acknowledges that: At the midpoint of the bargaining range, each party can hurt the other party by equally walking away from the negotiating table. The midpoint of the bargaining range allows both parties to share equally in the benefits of reaching an agreement and is therefore a useful starting place for considering other factors that affect the relative bargaining strengths of the two parties to the negotiation. Ex. 17 at At the end of the day, Qualcomm and Leonard merely take issue with Benoit s conclusions as to how certain facts would or would not impact the symmetry between the parties. 14 In other words, Qualcomm disputes the values Benoit assigned to the variables that go into the NBS equation, but Qualcomm presents no reason why the NBS analysis itself should be excluded. This is an archetypal issue of differing expert opinions regarding the applications of the facts of the case it is not a methodology issue subject to exclusion under Daubert. 15 Second, Qualcomm attempts to impose on the NBS analysis additional constructs included in one scholarly paper that discusses NBS. Qualcomm and Leonard have relied on just one scholarly article discussing the NBS to interpret its requirements, The Nash Bargaining Solution in Economic Modeling by Ken Binmore, et al. See Ex. 15 at Qualcomm faults Benoit for failing to adhere to the Binmore paper. See id. at For example, Qualcomm criticizes Benoit for his lack of familiarity with the terminology status quo utility levels. Dkt. 288 at 17. But the term status quo utility levels appears nowhere in the foundational NBS papers. See generally Exs Qualcomm has pulled the term from Leonard s expert report and the Binmore paper. 16 See Ex. 15 at Merely because Benoit does not use NBS terminology found in one paper preferred by Leonard and not found in Nash s papers or the other scholarly 14 As Binmore recognizes, the choices of the inputs to the NBS equation are matters of modeling judgment. Ex. 11 at 177. An expert exercises modeling judgment regarding the additional information (e.g., time preferences or attitudes toward risk) that is to be embedded in the utility functions. Id. 15 See Quiet Tech., 326 F.3d at 1346 (holding that an expert s failure to include all available data in his model will affect the analysis probativeness, not its admissibility ) (quoting Bazemore v. Friday, 478 U.S. 385, 400 (1986)); Summit 6 LLC, 2013 U.S. Dist. LEXIS 95164, at * See Ex. 11 at 176 (observing that the pair of utility functions u1 and u2 may be referred to variously as the status quo, the disagreement point, or the threat point ). McKool

18 Case 3:11-cv RBD-TEM Document 305 Filed 07/25/13 Page 18 of 27 PageID papers cited by Benoit does not render his understanding of NBS unreliable. In addition to the foundational NBS papers, Benoit has also elected to rely on other scholarly articles to guide his understanding of the NBS, such as Application of Game Theory to Intellectual Property Negotiations by John Jarosz, et al. See Ex. 1 at 103; Ex. 2 at 23:15-21; Ex. 10. Reliance on different, but equally respectable, scholarly articles than the one relied upon by Leonard in criticizing Benoit is not grounds for exclusion under Daubert. Third, Qualcomm s arguments rely on mischaracterizing Benoit s deposition testimony. During his deposition, Benoit began working through his NBS calculation at the request of Qualcomm s counsel, who then upon realizing that Benoit had no problem performing the calculation instructed Benoit not to complete his work and chose not to explore the issue further despite having ample opportunity to do so. Ex. 2 at 42:25-43:11. In his deposition, Benoit explained his correct understanding of the principles and application of NBS. See Ex. 2 at 17:1-18:5, 21:12-24:25, 25:16-29:7, 31:5-34:10, 34:15-37:23, 38:22-40:12-41:8-42:1, 43:23-46:10, 270:2-19, 271:16-275:16. For example, contrary to Qualcomm s arguments, Benoit correctly defined a utility function as the preference that somebody has for something. Id. at 39:20-41:7. He further explained that he modeled the utility function of ParkerVision and Qualcomm in the hypothetical negotiation by defining Qualcomm s benefit from the use of ParkerVision s technology versus other options. Id. at 41:8-16. Fourth, Qualcomm resorts to empty disparagement of Benoit s expert credentials. But Benoit is well-qualified in intellectual property valuation, is knowledgeable in the field of economics, has been qualified to testify on damages in many patent cases, and has applied the NBS in previous patent damages testimony in Federal Court, including one recent case in the Northern District of Texas. See Summit 6, 2013 U.S. Dist. LEXIS 95164, at * Under Federal Rule of Evidence 702, an expert may be qualified by knowledge, skill, experience, training, or education. The threshold for qualification as an expert is low and Benoit more than surpasses it. Martinez v. Altec Indus., No. 3:02-cv-1100-J-32TEM, 2005 U.S. Dist. McKool

19 Case 3:11-cv RBD-TEM Document 305 Filed 07/25/13 Page 19 of 27 PageID LEXIS 46451, 9 (M.D. Fla. Aug. 3, 2005) ( As long as some reasonable indication of qualifications is adduced, the court may admit the evidence without abdicating its gate-keeping function. ). Benoit certainly has the relevant knowledge, skill, experience, training, and education to testify regarding patent damages. See Ex. 1 at Ex. A. Nonetheless, Qualcomm repeatedly takes aim at Benoit for no other reason than because his degree is in accounting and not in economics: Arguing that Benoit cannot conduct the type of patent damages analysis required by the Federal Circuit because he is not an economist. Dkt. 288 at 10. By Qualcomm s logic, Daubert precludes anyone without a degree in economics from opining on patent damages. Numerous cases refute this interpretation of Daubert. 17 Contending that Benoit lacks the expertise to apply the NBS. But Judge Reed O Connor in the Northern District of Texas recently found Benoit qualified to render a patent damages opinion based on an NBS analysis, over the Daubert challenge of the defendants. See Summit 6, 2013 U.S. Dist. LEXIS 95164, at * Asserting that Benoit should not be allowed to offer an opinion based on the NBS because he has no degree or professional experience in economics. 18 Dkt. 288 at As explained below in Section A, Benoit properly analyzed the facts of the case within the NBS construct. In Benoit s words: you can t just decide who has gotten it right based on their background. 19 Ex. 2 at 110:8-9. Claiming that there has been no showing that Benoit is qualified by knowledge, skill, experience, training, or education to perform a patent damages analysis. Id. at 22. Qualcomm wholly disregards Benoit s accomplished resume and damages opinions that have been accepted in at least six other cases. See Ex. 1 at Ex. A. Benoit is well-qualified to testify regarding economic analysis and mathematical modeling for purposes of conducting a patent damages analysis and applying the NBS. 20 See Ex. 1 at Ex. A. 17 See Energy Transp. Grp., Inc. v. William Demant Holding A/S, 697 F.3d 1342, 1356 (Fed. Cir. 2012) (allowing testimony from damages expert with a Masters in Public Accounting); i4i Ltd. P ship v. Microsoft Corp., 598 F.3d 831, (Fed. Cir. 2010) (allowing testimony from damages expert with a JD and MBA); Micro Chem., Inc. v. Lextron, Inc., 317 F.3d 1387, 1390 (Fed. Cir. 2003) (allowing testimony from damages expert with an MBA). 18 Not only is an economics pedigree not required to calculate patent damages, Qualcomm misrepresents Benoit s experience and education in economics. As Qualcomm is well aware, Benoit s accounting and MBA curriculum included economics (including econometrics and managerial economics), mathematics, statistics, and finance; he served as a teaching assistant who ran the econometrics lab at the University of Mississippi; and his work experience includes more than 15 years focusing on intellectual property valuation. Ex. 2 at 6:21-24, 10:6-16, 11:17-12:2. 19 Moreover, in essence Qualcomm asks this Court to create an insurmountable barrier to entry for an inventor who lacks the means to hire an economist. For example, Qualcomm s expert, Dr. Leonard, charges $650/hour. Ex 15 at Qualcomm argues that Benoit lacks the expertise to apply the NBS because when asked in his deposition about the difference between Nash Equilibrium and the Nash Bargaining Solution, dkt. 288 at 17-18, he responded that he would have to read the article to refresh my recollection. Ex. 2 at 29: By Qualcomm s logic, an expert s McKool

20 Case 3:11-cv RBD-TEM Document 305 Filed 07/25/13 Page 20 of 27 PageID Qualcomm s baseless attempt to disparage Benoit s expertise must be rejected. 2. A 50/50 Split Of Incremental Profit Under the NBS Does Not Constitute An Impermissible Rule of Thumb Benoit s application of NBS cannot be construed as an impermissible 50/50 rule of thumb. See Ex. 16 at As Benoit has explained, based on his calculation of the benefit (or incremental profit) that Qualcomm derived from the use of ParkerVision s technology, ParkerVision and Qualcomm would have agreed to evenly split those benefits a fact Leonard recognizes as appropriate in the right circumstances. Benoit explains the elements that he considered with respect to the parties disagreement benefits for purposes of the NBS equation. 21 See Ex. 1 at ; Ex. 2 at 21:12-22:21. Qualcomm and Leonard are free to argue to the jury that the split of incremental profit between the parties should be higher or lower based on the evidence. However, Benoit s testimony should not be excluded merely because he concludes that an even split of incremental profit between the parties is proper based upon his application of the facts of this case to the NBS analysis. The NBS is fundamentally different from the 25% rule. See Ex. 13 at 14; Ex. 16 at ; Ex. 10 at The 25% rule assumes the infringer would agree to pay the patentee a 25% royalty on its profits on the product incorporating the infringing technology, without any analysis or justification for the 25% royalty rate. Uniloc, 632 F.3d at As Uniloc noted, the 25% rule takes no account of the importance of the patent to the profits of the product sold. Id. at The rule represents an essentially arbitrary rule of thumb, rather than a rigorous methodology. Id. at In contrast to the 25% rule, the NBS is specifically tied to the facts of the case. The NBS assigns to each party in a bilateral negotiation the profit it would obtain from not request to refresh his memory regarding a concept would constitute grounds for exclusion under Daubert. According to this reasoning, Qualcomm s damages expert would be precluded from testifying regarding the Georgia-Pacific factors because he could not recall them by memory during his deposition. See Ex. 3 at 90:24-91:11, 92: Additionally, Jeff Parker, ParkerVision s President, CEO, and lead negotiator, has stated that he would have agreed to split the benefits of the patents-in-suit evenly with Qualcomm. See Ex. 18 at 26: 10-27:1, 175:7-19, Although in the 1999 negotiations, Mr. Parker would have been willing to accept something less than a 50/50 split, he recognized that circumstances, such as the issuance of the patents-in-suit, would have improved ParkerVision s bargaining position. See id. at 204:9-206:7. Mr. Parker s testimony provides additional and independent support for the 50/50 split of incremental profit that Benoit calculated. McKool

21 Case 3:11-cv RBD-TEM Document 305 Filed 07/25/13 Page 21 of 27 PageID reaching an agreement ( disagreement benefits ), and splits the remaining profits ( agreement surplus ) according to the parties relative bargaining power. See Ex. 9 at ; Ex. 10 at 248; Ex. 12 at 5-6. Unlike the 25% rule, the facts and circumstances of a case drive the disagreement benefits that ultimately determine the appropriate split of the incremental benefits of the patented technology. 22 The overwhelming majority of courts that have addressed the use of NBS in reasonable royalty calculations in patent cases have correctly admitted testimony based upon the NBS. These courts understood that NBS is not akin to the 25% rule where, as here, it is adequately applied to the specific facts of the case. 23 For example, in Sanofi-Aventis, the defendants moved to exclude expert testimony utilizing the NBS. Sanofi-Aventis, 2011 U.S. Dist. LEXIS 10512, at *34. There, just like Qualcomm, the defendants argued that the NBS is indistinguishable from the 25% rule because it applies a rule of thumb to determine the reasonable royalty rate. Id. at 22 For example, assume an infringer sells an infringing product and makes a $4 profit, but without the patented technology, the infringer could sell a similar product and make a $3 profit. Under the 25% rule, the infringer would pay the patentee $1 (a 25% royalty on $4 profit) regardless of the non-infringing alternatives. Under the NBS, the infringer pays the patentee (assuming equal bargaining power) $0.50 based on an equal split of the $1 incremental benefit over the next-best alternative. If the infringer could sell a similar product and make a profit of $3.98, the infringer would now pay the patentee only $.01 under the NBS, while the 25% rule would dictate a payment of $1. 23 See Summit 6 LLC, 2013 U.S. Dist. LEXIS 95164, at *36-37 (denying Daubert motion and JMOL motion attacking Benoit s application of NBS); Amakua Dev. LLC v. Warner, No. 05-CV-3082, 2007 U.S. Dist. LEXIS 49952, at *64 (N.D. Ill. July 10, 2007)( the court admitted expert testimony that the parties would have evenly split the surplus identified, based upon a bargaining theory consistent with a Nash Equilibrium outcome among parties with equal negotiating power and explained that the proper method of challenging this testimony is on cross-examination ); Sanofi-Aventis Deutschland GmbH v. Glenmark Pharms., Inc., USA, No. 07-CV-5855, 2011 U.S. Dist. LEXIS 10512, at *34 (D.N.J. Feb. 3, 2011) (rejecting assertion that application of NBS equates to use of a 50 percent rule); Mformation Techs., Inc. v. Research in Motion Ltd., No. C , 2012 U.S. Dist. LEXIS 56784, at *15 16, n.19 (N.D. Cal. Mar. 29, 2012) (holding that the damages expert s use of the NBS as a check on the reasonableness of the rate reached through his Georgia-Pacific analysis was not grounds for exclusion of testimony); Ex. 19, Order On Daubert Motions and Motions In Limine at 4-5, Gen-Probe Inc. v. Becton Dickinson & Co., No.3:09-cv BEN-NLS, at *4-5 (S.D. Cal Nov. 6, 2012) (rejecting rule of thumb challenge because the expert s calculations were influenced, appropriately, by the facts of the case, including the competitive environment and Gen-Probe s policy of exploiting its own patents ); VirnetX Inc. v. Apple Inc., No. 6:10-CV-417, 2013 U.S. Dist. LEXIS 35631, at *51-54 (E.D. Tex. Feb. 26, 2013) (rejecting challenge to damages expert s use of NBS in a methodology similar to that applied by Benoit); VirnetX Inc. v. Cisco Sys., No. 6:10-CV-417, 2013 U.S. Dist. LEXIS 28223, *15 (E.D.Tex. Mar. 1, 2013) (holding that VirnetX s use of the NBS was proper because its expert did proffer some explanation as to why Cisco and VirnetX would have deviated from the traditional 50%- 50% profit split ); Grand River Enters. Six Nations v. King, No. 02-CV-5068, 2011 U.S. Dist. LEXIS 27424, at *22 (S.D.N.Y. Mar. 17, 2011); Minebea Co. v. Papst, No. 97-CV-590, 2005 U.S. Dist. LEXIS 11946, at *15 (D.D.C. June 21, 2005) (noting that the court had no doubt about [the expert s] methodology in applying the well-known Nash equilibrium concept). McKool

22 Case 3:11-cv RBD-TEM Document 305 Filed 07/25/13 Page 22 of 27 PageID *36. The plaintiffs argued that the NBS is the standard model in economics for calculating the outcome of a negotiation and that the damages expert reached his result under the NBS after considering the facts of the case, specifically the relationship between the parties and their relative bargaining power, the relationship between the patent and the accused product, the standard profit margins in the industry, and the presumed validity of the patent. Id. The court ruled that the damages expert did not arbitrarily apply a 50/50 profit split akin to the 25 percent rule rejected in Uniloc but rather based his reasonable royalty analysis on the specific facts of this case. Id. at *37. Because Benoit correctly applies the NBS, like the expert in Sanofi- Aventis, the Court should follow the great weight of the authority and permit his testimony. The Court should decline to follow the cases cited by Qualcomm. Oracle is not binding, its reasoning is suspect, its legal conclusions are incorrect, and it applied a Rule 403 framework (not Rule 702). Oracle Am., Inc. v. Google Inc., 798 F. Supp. 2d 1111, 1120 (N.D. Cal. 2011). The Oracle court fundamentally misunderstood the agreement surplus divided between the parties under the NBS, suggesting that in most cases [this] will amount to half the infringer s profit, which will be many times the amount of real-world royalty rates. Id. at In fact, the divided surplus will amount to 50% of the infringer s profit only when 100% of that profit depends on the license. In most cases (including this case), only a portion of the infringer s total profit will depend on a particular license. As shown in the chart included in Section A.1, supra, in this case the incremental profit split is not half the infringer s profit. Furthermore, Oracle did not hold that the NBS involved bad science or lacked rigor it held that the NBS involved too much science and too much rigor. Id. at 1120 ( The [NBS] involves complex mathematical formulas and equations that would surely be incomprehensible to the average juror ). Oracle s analysis is unsound: the presence of math formulas is not a reason to exclude expert testimony. See FED. R. EVID. 702; Bonura v. Sea Land Serv., Inc., 505 F.2d 665, 669 (5th Cir. 1974). And Oracle ultimately excluded the NBS-related testimony under Rule 403 as confusing, not Rule 702. Id. Here Qualcomm makes no Rule 403 argument. McKool

23 Case 3:11-cv RBD-TEM Document 305 Filed 07/25/13 Page 23 of 27 PageID Suffolk is also inapposite here. In Suffolk, the Court rejected the expert s NBS analysis because it was unclear how the NBS was tied to the facts of this particular case. Suffolk Techs. LLC v. AOL, Inc., 2013 U.S. Dist. LEXIS 64630, at *6 (E.D. Va. April 12, 2013). In contrast, here Benoit ties the NBS analysis to the facts of this particular case by examining qualitative considerations other than those addressed in the Georgia-Pacific factors, such as the financial position of the parties, the competitive intensity of the market, and the risk to other aspects of the licensee s business. Ex. 1 at 105; Ex. 2 at 21:12-27:3. Thus, Benoit s NBS analysis applies the rigor sought by the Suffolk court. C. Unlike the VIA Agreement, the Technology at Issue was the Motivation for the Symbol Agreement. Both the Symbol Agreement and the VIA Agreement involved circumstances and considerations materially different from the circumstances and considerations relevant to the hypothetical negotiation between ParkerVision and Qualcomm. Ex. 1 at 64, 67; Ex. 15 at 27-28, As set forth in ParkerVision s Daubert challenge, Leonard s reliance on the VIA Agreement in support of Qualcomm s damages calculations should be excluded. See Dkt. 285 at 3-7. Therefore, the Symbol Agreement better reflects the required comparability as to the licensee s motivation that is missing in the VIA agreement. 24 D. Benoit Presents An Economically Sound And Legally Sufficient Model for Convoyed Sales Qualcomm s brief argument regarding the exclusion of Benoit s opinion on convoyed sales under Georgia-Pacific factor 6 fails on both the law and the facts. At the outset, Qualcomm 24 If the Court determines that Benoit s testimony regarding the Symbol Agreement should be excluded, Leonard s testimony regarding the VIA Agreement must also be excluded. McKool

24 Case 3:11-cv RBD-TEM Document 305 Filed 07/25/13 Page 24 of 27 PageID misstates the law of convoyed sales. An increased royalty rate based on convoyed sales under Georgia Pacific factor 6 is appropriate when sales of the defendants infringing products would promote sales of the defendant s other products. 25 See Micro Chem., Inc. v. Lextron, Inc., 317 F.3d 1387, 1393 (Fed. Cir. 2003). Whether the infringing products actually promoted derivative sales of the defendants other products is a question of fact for the jury. Id. In addition, convoyed sales damages may be warranted when infringing and non-infringing components constitute a functional unit. See Rite-Hite Corp. v. Kelley Co., 56 F.3d 1538, 1550 (Fed. Cir. 1995) (en banc). Qualcomm disregards MicroChem 26 and Rite-Hite (an en banc decision) instead relying on an unpublished pre-rite-hite decision, T.J. Smith, and an inapposite entire market value rule opinion, LaserDynamics. 27 In any event, contrary to Qualcomm s conclusory argument, ParkerVision has advanced abundant evidence that Qualcomm s sales of the infringing transceivers drove demand for sales of basebands. 28 See Ex. 1 at 75-78; see also Ex. 22 at 1-4. Most importantly, the parties prior negotiations reflect larger royalties per unit for use of the patents-in-suit in integrated components (including the baseband) than stand-alone components. Ex. 1 at 76-78, Ex. E.1; Ex. 24; Ex. 25. In addition, the evidence reflects that Qualcomm sells its transceivers and basebands 25 See i4i Ltd. P ship v. Microsoft Corp., 598 F.3d 831, (Fed. Cir. 2010) ( For factor 6, which asks whether the patented technology promotes the sale of other products, Wagner concluded that the infringing custom XML editor was critical to Microsoft s sales generally, as evidenced by internal Microsoft statements that a custom XML editor was one of the most important ways for encouraging users to purchase new Word products. ); Minks v. Polaris Indus., 546 F.3d 1364, 1373 (Fed. Cir. 2008) (holding that associated sales of non-patented items supported higher royalty rate); Trans-World Mfg. Corp. v. Al Nyman & Sons, Inc., 750 F.2d 1552, 1568 (Fed. Cir. 1984). 26 In MicroChem, the Federal Circuit addressed a Daubert challenge and held that the district court properly permitted the plaintiff s damages expert to consider sales of the defendants other products in arriving at a reasonable royalty because [t]hat ruling is consistent with one of the Georgia-Pacific factors -- the effect of selling the patented specialty in promoting sales of other products of the [infringer] MicroChem, 317 F.3d at Qualcomm does not challenge Benoit s damages opinion based on the EMVR, nor could it. The EMVR applies where a plaintiff seeks to include unpatented products in its damages base. The rule of convoyed sales applies when a plaintiff seeks to increase the royalty rate under Georgia-Pacific factor 6, but the royalty base includes only the smallest saleable unit of the patented product. Ex. 24, John Skenyon., Patent Damages Law & Practice 329 (2012). 28 Qualcomm argues that [t]he only evidence of record, however, shows that... Qualcomm s noninfringing baseband sales drive sales of accused receiver chips. Dkt. 288 at 5, As ParkerVision will argue in its upcoming motion in limine, however, Qualcomm s evidence from interview with its customers is inadmissible because the customers were not timely disclosed as required by Rule 26, nor made available for deposition. McKool

25 Case 3:11-cv RBD-TEM Document 305 Filed 07/25/13 Page 25 of 27 PageID as a functional unit and that Qualcomm s sales of infringing transceivers promoted derivative sales of Qualcomm s basebands. See Ex. 1 at 75-78; Ex. 23 at ; see Micro Chem., 317 F.3d at 1393, Rite-Hite, 56 F.3d at Whether and to what extent the sales of the infringing receivers and transceivers drive Qualcomm s sales of basebands is a question of fact for the jury. E. Benoit s Opinion on Indirect Infringement Damages Applies to Damages for Both Induced and Contributory Infringement Benoit offered an opinion on damages for indirect infringement in his report. Ex. 1 at 3, 11-16, 60, 82, Because contributory infringement is one of the two types of indirect infringement (the other being induced infringement), Benoit offered an opinion on damages for contributory infringement. See Dynacore Holdings Corp. v. U.S. Philips Corp., 363 F.3d 1263, 1272 (Fed. Cir. 2004) (stating [i]ndirect infringement, whether inducement to infringe or contributory infringement, requires finding of direct infringement). Benoit relies upon ParkerVision s technical experts to establish many of the requisite facts to show that Qualcomm indirectly infringed via induced and/or contributory infringement. See Ex. 1 at 11-15; Ex 23 at (addressing induced and contributory infringement). As in most cases, ParkerVision s damages for indirect infringement, whether for inducement or contributory infringement, will be the same. Likewise, damages for indirect infringement are generally are the same or similar to damages assessed for underlying direct infringement. 29 Accordingly, Benoit should not be precluded from offering an opinion on contributory infringement because he clearly offered an opinion on damages resulting from induced infringement in his expert report. 29 See Glenayre Elecs., Inc. v. Jackson, 443 F.3d 851, 859 (Fed. Cir. 2006); see also 7 Donald S. Chisum, Chisum on Patents 20.03[7][b][iv] (2001) ( In many instances the appropriate measure of monetary relief against an indirect infringer will be the same as that against the direct infringer. ). McKool

26 Case 3:11-cv RBD-TEM Document 305 Filed 07/25/13 Page 26 of 27 PageID July 25, 2013 Respectfully submitted, McKOOL SMITH, P.C. /s/ Douglas A. Cawley Douglas A. Cawley, Lead Attorney Texas State Bar No Richard A. Kamprath Texas State Bar No.: Ivan Wang Texas State Bar No.: McKool Smith P.C. 300 Crescent Court, Suite 1500 Dallas, Texas Telephone: (214) Telecopier: (214) SMITH HULSEY & BUSEY /s/ James A. Bolling Stephen D. Busey James A. Bolling Florida Bar Number Florida Bar Number Water Street, Suite 1800 Jacksonville, Florida (904) (904) (facsimile) ATTORNEYS FOR PLAINTIFF PARKERVISION, INC. T. Gordon White Texas State Bar No Kevin L. Burgess Texas State Bar No Josh W. Budwin Texas State Bar No Leah Buratti Texas State Bar No Mario A. Apreotesi Texas State Bar No McKool Smith P.C. 300 West Sixth Street, Suite 1700 Austin, Texas Telephone: (512) Telecopier: (512) McKool

27 Case 3:11-cv RBD-TEM Document 305 Filed 07/25/13 Page 27 of 27 PageID CERTIFICATE OF SERVICE I HEREBY CERTIFY that on this day, July 25, 2013, I served the foregoing on all counsel of record via ECF. /s/ Leah Buratti Leah Buratti McKool

28 Case 3:11-cv RBD-TEM Document Filed 07/25/13 Page 1 of 1 PageID PARKERVISION, INC., v. Plaintiff, UNITED STATES DISTRICT COURT MIDDLE DISTRICT OF FLORIDA JACKSONVILLE DIVISION QUALCOMM INCORPORATED, Defendant. Case No. 3:11-cv-719-J-37TEM QUALCOMM INCORPORATED, v. Counterclaim Plaintiff, PARKERVISION, INC., AND STERNE, KESSLER, GOLDSTEIN, & FOX PLLC, Counterclaim Defendants. EXHIBIT 1 TO PARKERVISION S RESPONSE TO QUALCOMM S DAUBERT MOTION TO PRECLUDE PAUL BENOIT DAMAGES TESTIMONY OUTSIDE ATTORNEYS' EYES ONLY UNDER SEAL - SUBJECT TO PROTECTIVE ORDER - CONTAINS CONFIDENTIAL INFORMATION - TO BE OPENED ONLY AS DIRECTED BY THE UNITED STATES DISTRICT COURT FOR THE MIDDLE DISTRICT OF FLORIDA McKool v1

29 Case 3:11-cv RBD-TEM Document Filed 07/25/13 Page 1 of 1 PageID PARKERVISION, INC., v. Plaintiff, UNITED STATES DISTRICT COURT MIDDLE DISTRICT OF FLORIDA JACKSONVILLE DIVISION QUALCOMM INCORPORATED, Defendant. Case No. 3:11-cv-719-J-37TEM QUALCOMM INCORPORATED, v. Counterclaim Plaintiff, PARKERVISION, INC., AND STERNE, KESSLER, GOLDSTEIN, & FOX PLLC, Counterclaim Defendants. EXHIBIT 2 TO PARKERVISION S RESPONSE TO QUALCOMM S DAUBERT MOTION TO PRECLUDE PAUL BENOIT DAMAGES TESTIMONY OUTSIDE ATTORNEYS' EYES ONLY UNDER SEAL - SUBJECT TO PROTECTIVE ORDER - CONTAINS CONFIDENTIAL INFORMATION - TO BE OPENED ONLY AS DIRECTED BY THE UNITED STATES DISTRICT COURT FOR THE MIDDLE DISTRICT OF FLORIDA McKool v1

30 Case 3:11-cv RBD-TEM Document Filed 07/25/13 Page 1 of 4 PageID EXHIBIT 3

31 Case 3:11-cv RBD-TEM Document Filed 07/25/13 Page 2 of 4 PageID Page 1 2 UNITED STATES DISTRICT COURT MIDDLE DISTRICT OF FLORIDA 3 JACKSONVILLE DIVISION X 4 PARKERVISION, INC., PLAINTIFF, 5 -against- 6 QUALCOMM INCORPORATED, 7 8 DEFENDANT. 9 AND RELATED COUNTERCLAIM X DEPOSITION OF GREGORY LEONARD 13 New York, New York 14 Friday, April 26, Reported by: 22 Rebecca Schaumloffel, RPR, CLR 23 Job No:

32 Case 3:11-cv RBD-TEM Document Filed 07/25/13 Page 3 of 4 PageID Page 90 1 G. LEONARD 2 negotiation, a defendant wouldn't be willing 3 to pay more than the value of the benefits 4 they were receiving. 5 Q. The Georgia Pacific factors 6 require consideration of the economic 7 benefits that the technology would have 8 created, right? 9 A. Again, I mean all this is 10 trying -- yeah, it's all trying to get at 11 that same concept, p, which is how much is the 12 technology worth. It is worth something 13 because it provides benefits to the defendant 14 as the defendant used it. So, I mean that is 15 the purpose. p I am just saying as a 16 conceptual matter, that's the purpose p and 17 then there are different ways to think about 18 how to get at that. 19 Q. So your answer to my question 20 is yes? 21 A. I don't know. The way yy you phrase 22 it though still seems to be suggesting g 23 that -- well, my answer is what it is. 24 Q. Does the Georgia Pacific 25 factors -- do the Georgia Pacific require Page 92 1 G. LEONARD 2 we'll take a break. If that's all 3 right. Do you need a break right now? 4 MR. GREENWALD: No. 5 Q. Can you tell me what Georgia 6 Pacific factor 10 is? 7 A. I can never remember based on the 8 numbers, so let me look and I will see corresponds to what letter? A, B, C, D, E, 10 F, G, H, I, J. J, okay. 11 Q. It is page A. Oh, great, thank you. You 13 said 10? 14 Q. Yes. 15 A. Right. So it says "The nature of 16 the patented invention, the character of the 17 commercial embodiment of it as owned by the 18 licensor and the benefits to those who have 19 used the invention." 20 Q. Once a damages expert understands 21 the benefits of the patents-in-suit, the 22 next -- another step in the process is 23 determining how the licensor and the licensee 24 would split those benefits, right? 25 A. I mean, I think I have described Page 91 1 G. LEONARD 2 consideration of the economic benefits that 3 the technology in suit would have created? 4 A. Again, I think that -- 5 Q. Consideration however you would 6 like to consider it, but do they require that 7 consideration? 8 A. I don't know what require. There 9 is certainly Georgia Pacific factors that 10 talk about the benefits over I think in some 11 sense it is little too restrictive. They may 12 talk about prior art or something like that. 13 But, again, I think that's what -- at least 14 as an economist, I hope that the Georgia 15 Pacific factors are trying to get at, which 16 is what is the value of the technology as the 17 defendant used it versus other things they 18 could have done. And then the question is, 19 you know, that's again the maximum amount a 20 reasonable royalty could be and then you go 21 from there. 22 MR. GREENWALD: Let me know when 23 is a good time for our next break. 24 MS. BURATTI: I will wrap up 25 this line of questioning and then Page 93 1 G. LEONARD 2 it in that way myself. I mean, that's my 3 economist's view of it. I think in some 4 sense that's what the hypothetical 5 negotiation that's identified by the Georgia 6 Pacific factors is getting g at as well. 7 Q. And the split between the 8 licensor and the licensee is inherently some 9 ratio, whether it is 100%, 90%, 10%, right? 10 A. Well, once you have identified 11 the benefits and you have completely factored 12 out, again, all the other alternatives and 13 everything else, then the split obviously of 14 that amount has to be between 0 and 100%. 15 MS. BURATTI: I think this would 16 be a good time for a break. 17 MR. GREENWALD: Sure. 18 THE VIDEOGRAPHER: The time is 19 10:50 a.m. This completes tape number 20 one of the videotaped deposition of 21 Dr. Gregory Leonard. 22 (Whereupon, a recess was held.) 23 THE VIDEOGRAPHER: The time is 24 11:07 a.m., and this begins tape 25 number two of the videotaped 24 (Pages 90 to 93)

33 Case 3:11-cv RBD-TEM Document Filed 07/25/13 Page 4 of 4 PageID Page 94 1 G. LEONARD 2 deposition of Gregory Leonard. 3 BY MS. BURATTI: 4 Q. Dr. Leonard, when we left off, 5 we'd been discussing the decision on how 6 benefits would be split between 7 two negotiating parties and, I think you had 8 said that it necessarily has to be somewhere 9 between 0 and 100; is that right? 10 A. You know, I said conditional on 11 having identified the range appropriately. 12 If that range is being split between 13 two parties, then, you know, by definition 14 the range it can't go outside of that. 15 Q. By range, you mean the range of 16 benefits to be split between the parties; is 17 that right? 18 A. Or the bargaining, more generally 19 the bargaining range. 20 Q. What do you mean by bargaining 21 range? 22 A. What's typically defined as the, 23 you know, maximum amount, in a royalty 24 context, the maximum amount the licensee 25 would be willing to pay and the minimum Page 96 1 G. LEONARD 2 between the parties? 3 A. Well, again, because the approach 4 I am using, I am going g right to what the 5 royalty y should be. You know, I haven't 6 calculated a split. It is not necessary. So 7 I mean, as I am sitting here, I can't tell 8 you what the split was because it wasn't part 9 of my analysis to figure that out 10 specifically. 11 Q. So to be clear, you don't know 12 what the benefit -- you haven't quantified 13 the benefits in this case and you haven't 14 identified a split in this case; is that 15 right? 16 MR. GREENWALD: Objection. 17 A. You know, I don't think that's 18 really right because implicitly looking at a 19 market outcome tells you both of those things 20 in combination. So I wouldn't agree with 21 that. 22 Q. So you've identified a market 23 outcome -- sorry, you have identified a 24 market outcome based on a market transaction 25 in this case, but you haven't quantified the Page 95 1 G. LEONARD 2 amount the licensor would be willing to 3 accept. 4 Q. Do you have any idea of what the 5 split of benefit would have been between 6 ParkerVision and Qualcomm related to the use 7 of the patents-in-suit in the hypothetical 8 negotiation? 9 A. Well, again, I think the market 10 outcomes are often a good way to figure that 11 out. So in this particular case, you know, 12 we have got some market outcomes to look at. 13 So I think that's a good way to do it. 14 Q. Based on your analysis of the 15 market outcomes, do you think it would have 16 been a 70/30 split? 17 A. Again, I don't really need to for my yp purposes, p I don't need to identify the 19 split it itself because I am going g directly 20 to the number at the end. So,,y you know, the 21 split is what it is. And I don't need to 22 figure that out exactly because I don't need 23 that to get to the royalty y number. 24 Q. Based on your royalty y number, 25 what would the split have been of the benefit Page 97 1 G. LEONARD 2 benefits or identified the split between the 3 parties of those benefits; is that right? 4 A. Well, again, the market outcomes 5 tell you something about what those benefits 6 must be. And I have looked at a lot of other 7 information as to what those benefits are. 8 But as to specifically doing a calculation of 9 what the split is, that's really not 10 necessary for what I was doing and so I 11 haven't done it. I could go back and do 12 that, I suppose. But it wasn't something 13 that was really of interest for what I needed 14 to do. 15 Q. You said that the market outcomes 16 tell you something about what the benefits 17 must be. Based on your analysis of the 18 market outcomes, what have you identified as 19 the benefits to be split between the parties? 20 A. I'm sorry, say that again. 21 Q. You said that the market outcomes 22 tell you about what the benefits are that 23 will be split between the parties, and you 24 have analyzed the market outcomes. Based on 25 your analysis of the market outcomes, what 25 (Pages 94 to 97)

34 Case 3:11-cv RBD-TEM Document Filed 07/25/13 Page 1 of 4 PageID EXHIBIT 4

35 Case 3:11-cv RBD-TEM Document Filed 07/25/13 Page 2 of 4 PageID 11948

36 Case 3:11-cv RBD-TEM Document Filed 07/25/13 Page 3 of 4 PageID 11949

37 Case 3:11-cv RBD-TEM Document Filed 07/25/13 Page 4 of 4 PageID 11950

38 Case 3:11-cv RBD-TEM Document Filed 07/25/13 Page 1 of 17 PageID EXHIBIT 5

39 Case 3:11-cv RBD-TEM Document Filed 07/25/13 Page 2 of 17 PageID 11952

40 Case 3:11-cv RBD-TEM Document Filed 07/25/13 Page 3 of 17 PageID 11953

41 Case 3:11-cv RBD-TEM Document Filed 07/25/13 Page 4 of 17 PageID 11954

42 Case 3:11-cv RBD-TEM Document Filed 07/25/13 Page 5 of 17 PageID 11955

43 Case 3:11-cv RBD-TEM Document Filed 07/25/13 Page 6 of 17 PageID 11956

44 Case 3:11-cv RBD-TEM Document Filed 07/25/13 Page 7 of 17 PageID 11957

45 Case 3:11-cv RBD-TEM Document Filed 07/25/13 Page 8 of 17 PageID 11958

46 Case 3:11-cv RBD-TEM Document Filed 07/25/13 Page 9 of 17 PageID 11959

47 Case 3:11-cv RBD-TEM Document Filed 07/25/13 Page 10 of 17 PageID 11960

48 Case 3:11-cv RBD-TEM Document Filed 07/25/13 Page 11 of 17 PageID 11961

49 Case 3:11-cv RBD-TEM Document Filed 07/25/13 Page 12 of 17 PageID 11962

50 Case 3:11-cv RBD-TEM Document Filed 07/25/13 Page 13 of 17 PageID 11963

51 Case 3:11-cv RBD-TEM Document Filed 07/25/13 Page 14 of 17 PageID 11964

52 Case 3:11-cv RBD-TEM Document Filed 07/25/13 Page 15 of 17 PageID 11965

53 Case 3:11-cv RBD-TEM Document Filed 07/25/13 Page 16 of 17 PageID 11966

54 Case 3:11-cv RBD-TEM Document Filed 07/25/13 Page 17 of 17 PageID 11967

55 Case 3:11-cv RBD-TEM Document Filed 07/25/13 Page 1 of 17 PageID EXHIBIT 6

56 Case 3:11-cv RBD-TEM Document Filed 07/25/13 Page 2 of 17 PageID 11969

57 Case 3:11-cv RBD-TEM Document Filed 07/25/13 Page 3 of 17 PageID 11970

58 Case 3:11-cv RBD-TEM Document Filed 07/25/13 Page 4 of 17 PageID 11971

59 Case 3:11-cv RBD-TEM Document Filed 07/25/13 Page 5 of 17 PageID 11972

60 Case 3:11-cv RBD-TEM Document Filed 07/25/13 Page 6 of 17 PageID 11973

61 Case 3:11-cv RBD-TEM Document Filed 07/25/13 Page 7 of 17 PageID 11974

62 Case 3:11-cv RBD-TEM Document Filed 07/25/13 Page 8 of 17 PageID 11975

63 Case 3:11-cv RBD-TEM Document Filed 07/25/13 Page 9 of 17 PageID 11976

64 Case 3:11-cv RBD-TEM Document Filed 07/25/13 Page 10 of 17 PageID 11977

65 Case 3:11-cv RBD-TEM Document Filed 07/25/13 Page 11 of 17 PageID 11978

66 Case 3:11-cv RBD-TEM Document Filed 07/25/13 Page 12 of 17 PageID 11979

67 Case 3:11-cv RBD-TEM Document Filed 07/25/13 Page 13 of 17 PageID 11980

68 Case 3:11-cv RBD-TEM Document Filed 07/25/13 Page 14 of 17 PageID 11981

69 Case 3:11-cv RBD-TEM Document Filed 07/25/13 Page 15 of 17 PageID 11982

70 Case 3:11-cv RBD-TEM Document Filed 07/25/13 Page 16 of 17 PageID 11983

71 Case 3:11-cv RBD-TEM Document Filed 07/25/13 Page 17 of 17 PageID 11984

72 Case 3:11-cv RBD-TEM Document Filed 07/25/13 Page 1 of 4 PageID EXHIBIT 7

73 Case 3:11-cv RBD-TEM Document Filed 07/25/13 Page 2 of 4 PageID PARKERVISION, INC., QUALCOMM INCORPORATED, UNITED STATES DISTRICT COURT THE MIDDLE DISTRICT OF FLORIDA JACKSONVILLE DIVISION Plaintiff, v. Case No. 3:11-cv-719-J-37-TEM Defendant. QUALCOMM S OBJECTIONS AND RESPONSES TO PARKERVISION S FIRST SET OF REQUESTS FOR ADMISSION Pursuant to Rule 36 of the Federal Rules of Civil Procedure and the Local Rules for the Middle District of Florida ( Local Rules ), Qualcomm Incorporated ( Qualcomm ) submits the following objections and responses to ParkerVision, Inc. s ( ParkerVision ) First Set of Requests for Admission. General Objections 1. Qualcomm objects to ParkerVision s First Set of Requests for Admission to the extent it purports to impose obligations beyond those imposed by the Federal Rules of Civil Procedure and the Local Rules. 2. Qualcomm objects to ParkerVision s First Set of Requests for Admission to the extent that it purports to request disclosure of information that comes within the scope of the attorney-client privilege, the work product doctrine, any protective order or confidentiality agreement or is otherwise exempted from discovery. Qualcomm hereby asserts all applicable privileges and protections, and excludes privileged and CONFIDENTIAL

74 Case 3:11-cv RBD-TEM Document Filed 07/25/13 Page 3 of 4 PageID information and belief. Qualcomm reserves the right to amend and/or supplement its responses, which are subject to such additional or different information as discovery or further investigation may disclose. 7. Qualcomm reserves the right to make use of, or to introduce at any hearing and at trial, any information responsive to ParkerVision s First Set of Requests for Admission but discovered subsequent to the date of Qualcomm s responses, including, but not limited to, any information obtained during discovery. 8. All of the General Objections set forth herein are incorporated in each of the specific responses set forth below and have the same force and effect as if fully set forth therein. Qualcomm, through its attorneys, and for its responses to ParkerVision s First Set of Requests for Admission, states as follows: REQUEST FOR ADMISSION NO. 1 Admit that all documents produced by Qualcomm in this case with a QCPV prefix are authentic. Response to Request No. 1 Subject to the General Objections, Qualcomm lacks sufficient information to admit or deny this request. Qualcomm has produced over ten million pages of documents in this action. Qualcomm admits that it is unaware of any document produced by Qualcomm in this case with a QCPV prefix that is not authentic. REQUEST FOR ADMISSION NO. 2 Admit that Qualcomm is unaware of any document produced by Qualcomm in this case with a QCPV prefix that is not authentic. 3 CONFIDENTIAL

75 Case 3:11-cv RBD-TEM Document Filed 07/25/13 Page 4 of 4 PageID Response to Request No. 2 Admitted. 4 CONFIDENTIAL

76 Case 3:11-cv RBD-TEM Document Filed 07/25/13 Page 1 of 10 PageID EXHIBIT 8

77 Case 3:11-cv RBD-TEM Document Filed 07/25/13 Page 2 of 10 PageID 11990

78 3:11-cv RBD-TEM Document Filed 07/25/13 Page 3 of 10 PageID 11

79 11-cv RBD-TEM Document Filed 07/25/13 Page 4 of 10 PageID

80 11-cv RBD-TEM Document Filed 07/25/13 Page 5 of 10 PageID

81 3:11-cv RBD-TEM Document Filed 07/25/13 Page 6 of 10 PageID 11

82 3:11-cv RBD-TEM Document Filed 07/25/13 Page 7 of 10 PageID 11

83 3:11-cv RBD-TEM Document Filed 07/25/13 Page 8 of 10 PageID 11

84 3:11-cv RBD-TEM Document Filed 07/25/13 Page 9 of 10 PageID 11

85 3:11-cv RBD-TEM Document Filed 07/25/13 Page 10 of 10 PageID 11

86 Case 3:11-cv RBD-TEM Document Filed 07/25/13 Page 1 of 14 PageID EXHIBIT 9

87 e 3:11-cv RBD-TEM Document Filed 07/25/13 Page 2 of 14 PageID 12

88 e 3:11-cv RBD-TEM Document Filed 07/25/13 Page 3 of 14 PageID 12

89 e 3:11-cv RBD-TEM Document Filed 07/25/13 Page 4 of 14 PageID 12

90 e 3:11-cv RBD-TEM Document Filed 07/25/13 Page 5 of 14 PageID 12

91 e 3:11-cv RBD-TEM Document Filed 07/25/13 Page 6 of 14 PageID 12

92 e 3:11-cv RBD-TEM Document Filed 07/25/13 Page 7 of 14 PageID 12

93 e 3:11-cv RBD-TEM Document Filed 07/25/13 Page 8 of 14 PageID 12

94 e 3:11-cv RBD-TEM Document Filed 07/25/13 Page 9 of 14 PageID 12

95 3:11-cv RBD-TEM Document Filed 07/25/13 Page 10 of 14 PageID 12

96 3:11-cv RBD-TEM Document Filed 07/25/13 Page 11 of 14 PageID 12

97 3:11-cv RBD-TEM Document Filed 07/25/13 Page 12 of 14 PageID 12

98 3:11-cv RBD-TEM Document Filed 07/25/13 Page 13 of 14 PageID 12

99 3:11-cv RBD-TEM Document Filed 07/25/13 Page 14 of 14 PageID 12

100 Case 3:11-cv RBD-TEM Document Filed 07/25/13 Page 1 of 28 PageID EXHIBIT 10

101 Case 3:11-cv RBD-TEM Document Filed 07/25/13 Page 2 of 28 PageID LICE \SING BEST PRACTICES STRATEGIC, TERRITORIAL, AND TECHNOLOGY ISSUES ROBERT GOLDSCHEIDER AND ALAÑ H. GORDON EDITORS EXHIBIT JOHN WILEY & SONS, INC.

102 Case 3:11-cv RBD-TEM Document Filed 07/25/13 Page 3 of 28 PageID This book is printed on acid-free paper. Copyright 2006 by John Wiley & Sons, Inc. Ali rights reserved. Published by John Wiley & Sens, Inc., Hoboken, New Jersey. Published simultaneously in Canada. No part of this publicatiqn may be reproduced, stored in a retrieval System, or transmitted in any form or by any means, electronic, mechanical, photocopying, recording scanning, or otluinvise, except as permitted under Section 107 or 108 of the 1976 United States Copyright Act, without either the prior written permission of the Publisher, or authorizatiate through paymcnt of the appropriate per-copy fee to the Copyright Clearance Center. Inc., 222 Rosewood Drive, Danvers, MA 01923, , fax , nr on the web at coin. Requests to the Publisher for permission should be addressed to the Permissions Department, John Wiley & Sons, lite., III River Street, Hobokeri, NJ 07030, , fax 2( , or online at h:rp,'/iwtvw, wiley.coin/go/perìnissions. Limit oflisbilityfdisclainìer of WarranLy: While the publisher and author have used their best elthroc in preparing this book, they make no representations or warranties with respect m the accuracy or completeness Of the contents of this book -md specifically disclaim any implied w Irranties of merchant tbtlity or fitnl',', tor a particular purpose, No warranty may be created or extended by sales reprcsentativc.s or Written sales materials. The ad ce aid strategies contained herein may not bi. suitable for your situatton You tmuld i.on',ult wtlh a professional where appropriate. Neither the publisher nor author shall be liable fur any loss of profit or any other commercial damages, including but not limited to special, incidental, consequential, or other damages. For general information on our oilier products and services, or technical support, please contact our Cusaorner Cure Department within the United States at , outside the United S rates at or fax Wiley nico publishes its books in a variety at' electronic formats. Some content that appears in print may not he available in electronic books. For more information about Wiley products ysit our Web site at hup:/avww,wiley.com. Library of Congress Cataloging-in-Publication Data; Licensing best practices : strategic, territorial, and technology issues / Robert Goldschcider and Alan H. Gordon, editors. p. cm. Includes bibliographical references arid index, ISBN-13: (cloth: acid-free paper) ISBN-10: (cloth : acid-free paper) I. License agreements. 2, Foreign licensing agrcetncnts. I. Goldschcider, Robert. IL Gordots, Alun H., K152S.L S dc Prinred in the United States of America

103 Case 3:11-cv RBD-TEM Document Filed 07/25/13 Page 4 of 28 PageID CHAPTER 1,,r APPLICATION OF GAME THEORY TO INTELLECTUAL PROPERTY ROYALTY NEGOTIATIONS BY JOHN C. JAROSZ AND MICHAEL J, CHAPMAN1 INTRODUCTION One of the mo.st important elements of intellectual property licensing negotiations is the determination of the compensation to be paid to an owner for access to his ôr her intellectual property. Unfortunately, in many ca'ses, relatively little time or effort is devoted to determining the appropriate level of compensation. As noted by Gregory Battersby and Charles Grimes, "Frequently, the inquiry starts and ends with the last license granted by that property owner for the same or similar property."2 Alternatively, the parties simply rely on "customary" royalty rates for particular industries. Both of these approaches can provide sorne useful guidance for intellectual property licensing negotiations, but historical and/or customary royalty levels sometimes are not enough. Compensation needs to be driven by the particular facts of each situation. The goal of this chapter is to provide intellectual property licensing professionals with another systematic approach to thinking about what the "right" payment might be and to offer them Insights into critical parameters to consider in that approach. "Game theory" constitutes the basis for the approach and the source of the insights. USEFULNESS OF GAME THEORY Game theory is the branch of social science that studies strategic decision making. In this field, a "game" can be described as "any interaction between agents that is governed by a set of rules specifying the possible moves for each participant and a set of outcomes for each possible combination of moves."4 It studies how rational actors behave when their separate choices interact to produce payoffs to each player, i.e., when the benefit that a player will receive depends not only on his or her own choice but on choices made by others.5 241

104 Case 3:11-cv RBD-TEM Document Filed 07/25/13 Page 5 of 28 PageID Ch. 17 Application of Game Theory to Jntellectual Property Royalty Negotiations The licensing of intellectual property is undoubtedly a situation in which the benefits available to each side, i.e., inventor and manufacturer, depend critically on the choices made by each party. A game theoretic approach is extremely helpful in and of itself in thinking about the 'right" price for the intellectual property given the strategies atid outcomes available to both negotiators. Also, it effectively supplements other common intellectual property pricing methods by encouraging the evaluation of negotiation-specific and intellectual property- ;. specific factors that are critical to a fair outcome. One common pricing tool is to adopt a rate that ha been used in a previous setting, either from a party's own experience or from the experience of others in industry. The usefulness ofsuch an approach depends on the availability of comparable transactions. If such technologies can be found, then the royalties reflected in those transactions are likely to provide valuable guidance in setting the price for the intellectual property under consideration. However. there are at least two potential limitations. Firste t can be difficult to find sufficiently similar technologies and transactions. The greater the differences between the benchmark transaction/intellectual property and the transaction/intellectual property under consideration, the less useful the benchmark. Second, even if the benchmark used is close, the adoption of a royalty based on this benchmark may simply replicate the results of the earlier negotiation, regardless of whether the price reflected in the benchmark is sensible or not. It' the initial intellectual property was underpriced because of the weakness of the bargaining position of the seller. then reliance on this royalty rate in a subsequent agreement may perpetuate this resulteven if the seller in the subsequent negotiation was in a stronger bargaining position. Another common tool focuses on the expected profits (or cash flows) that are likely to be generated by use of the intellectual property in question. The basic idea behind "income" methods for establishing a royalty is that parties to u licensing negotiation seek to find a mutually agreeable way to "share" the benefits from making, using, or selling products embodying the technology at issue. The owner of the technology is entitled to compensation for the portion of the benefits of the product or process owing to the intellectual property. The manufacturer should retain certain benefits derived from other attributes of the product or process that incorporate the technology at issue. A commonly used profitbased method is the Profit Split Rule, which provides that the manufacturer should pay one-fourth to one-third of its expected operating profits for products incorporating the patented technology.6 The foundation of income approaches is specific Io the intellectual property. i.e., the profits that are expected to be made from the intellectual property at issue. However, there are at least two potential limitations. First, it can be difficult to isolate the true incremental benefits associated with use of the intellectual property at issue. Second, the income approach provides limited guidance on how those benefits should be allocated between the inventor and manufacturer. For example, the Profit Split Rule is a rule of thumb and may not reflect the actual contributions and/or negotiating positions of the parties involved.

105 Case 3:11-cv RBD-TEM Document Filed 07/25/13 Page 6 of 28 PageID Bargaining Basics 243 A third approach to establi$hing a royalty focuses on the costs required to construct or purchase an alternative technology that performs the same function as the patented technology but that dne not infringe on the intellectual property.7 According to cost approaches. a user ofccrtain patented technology would pay no more for access to that technology than the additional costs that it would incur if it were not able to use the technology. A common problem with this tool is that most cost approaches focus on avoided out-of-pocket capital expenditures. Such examinations tend to underestimate the true value of assets, including intellectual property assets. According to Reilly and Schweihs; Costs describe what the intangible asset owner spent in the original production process, or what the owner would have tu spend us of a certain date to recreate that production process. Cut, by itself, does not tell us how much a buyer would pay ta acquire the intangible asset, or how much a seller would seek to motivate the sale of the intangible asset.. Among the "economic" costs that need to be considered are the costs of unsuccessful design attempts. the period of the design-around, and the going forward impacts of the alternative. In short, out-of-pocket costs may not measure either fair value or fair price for the intellectual property at issue or for the parties involved in the negotiation." As noted above, a game-theoretic approach permits consideration of both intellectual property. and negotiation-specific considerations in setting royalty rates. As a general matter, game theory assumes that all the players involved in a given situation are "rational"in the sense that they will seek to maximize their own rewards (e.g., profits, income, or other benefit) in the circumstances they face. In particular, each of the players involved in the game is assumed to adopt the strategy that will yield the maximum benefits to him or her, given the rules governing interactions within the gaine and given the strategy that the other player has chosen or is expected tu choose. A game-theoretic approach to licensing negotiations requires both sides to think through the "rules of the game" and the payoffs that each party can expect under the scenarios that could play out during the course of negotiations. Armed with this knowledge, the negotiators will be in a much better position to determine the "right" price for the intellectual property because the focus is placed on likely outcomes und provides guidance as to how to bargain to those outcomes. BARGAINING BASICS The prototypical bargaining game involves two players who must negotiate over the allocation of a *&pie?tw in most bargaining games, access to "pie" for both parties is contingent on an agreement (i.e., no agreement, no pie for anyone). Three aspects of bargaining games are particularly worthy of note. First, the players in a bargaining game are faced with an interesting challenge: They want to make the most favorable agreement that they can, while avoiding the risk of

106 Case 3:11-cv RBD-TEM Document Filed 07/25/13 Page 7 of 28 PageID Ch. 17 Application of Game Theory to Intellectual Property Royalty Negotiations no agreement at all." Second, delays in reaching an agreement are costly, which encourages both parties to reach an agreement as quickly as possible. Third, in order to reach an agreement, both parties should approach the bargaining by "looking ahead and reasoning back."2 Our Example I assumes that two families- A SIMPLE BARGAINING GAME. the Hatfields and the McCoys.are spending the afternoon at the beach. A galion of (still frozen) ice cream washes up on the beach, and both families try to claim it. As the families start to argue, a lifeguard walks by und agrees to help resolve the dispute. The lifeguard notices that the ice cream is melting at a rate of one quart every 15 minutes, so the ice cream will be gone at the end of an hour. The lifeguard decides to let the l{atfields and the McCoys negotiate for the ice cream under the following raies. There will be four rounds of offers made. Thé first offer is to be made by the Hatfields immediately (when there is one gallon of ice cream to be divided). The McCoys are free to accept or reject this offer. If it is accepted, then the ice cream is divided and the negotiation stops. If it is rejected, the McCoys will make an offer in 15 minutes (when there are three quarts of ice cream to be divided). The Hathelds are free to accept or reject this offer. Once again, if it is accepted, then the ice cream is divided and the negotiation stops. If it is rejected, the Hatfields will make an offer in 30 minutes (when there are two quarts of ice cream to be divided). The McCoys are free to accept or reject this offer. 1f it is accepted, then the ice cream is divided and the negotiation stops. 1f it is rejected, the McCoys will make an offer in 45 minutes (when there is only one quart of ice cream to be divided). The Hatfields are free to accept or reject this offer. Once again, if it is accepted, then the ice cream is divided and the negotiation stops. If it is rejected, neither family gets any ice cream. What portion of the ice cream should the Hatfields offer to share with the McCoys at the start of these negotiations? To answer this question, it is necessary to look forward to the end of the negotiation and reason buck to the beginning of the negotiations. During the last offer period (45 minutes from now), there will be only one quart of ice cream remaining, and the McCoys have the right to decide how much to offer the Hatfields. If the Hatfields do not agree to the McCoys' offer, then neither party would get any ice cream. Therefore, during the last period, the 1-latfields would be better off accepting any amount of ice cream (e.g., one spoonful) offered by the McCoys.' Thus, the division of ice cream in the last period would be one quart (less one spoonful) of ice cream for the McCuys and one spoonful of ice cream for the Hatfields. Given this division at the 45-minute point, the best deal that the Hattields can expect to get when it makes an offer at the 30-minute point (when there will be two quarts of ice cream left) is to offer to give the McCoys one quart of ice cream and to keep one quart for itself. If the Hatfields were to offer any less than one quart to the McCoys, then the McCoys would simply reject the offer and obtain one quart at the end of the negotiations.

107 Case 3:11-cv RBD-TEM Document Filed 07/25/13 Page 8 of 28 PageID Bargaining Basics 245 Given this divi.sion at the 30-minute point, the best deal that the McCoy family can expect to get when it makes an offer at the 15-minute point (when there will be three quarts of ice cream left) is to offer to give the }Jatfields one quart of ice cream and tu keep two quarts for itself. If the McCoys were to offer any less than one quart tu the Hatfields. then the Hattïelds would simply reject the offer and obtain une quart when they imike an offer at the 30-minute point. Finally, given this division at thd IS-minute points the best deal that the Hatfield family can expect to get when it makes an offer immediately (when there is one gallon uf ice cream left) is to offer the Mcoys two quarts of ice cream and to keep two quarts fur itself. If the Flatfields were to offer any less than two quaris to the McCoys, then the McCuys would simply reject the offer and obtain two quarts when they make an otter at the 15-minute point, Based un this analysis. ii can be concluded that an agreement will be reached immediately, as the Hafflelds offer two quarts of ice cream to the McCoys and keep two quarts for themselves. By looking forward and reasoning back, the parties arc abk' tu reach an agreement that minimizes losses caused by delays and ensures that neither party could do better by delaying an agreement. AN INTELLECTUAL PROPERTY LICENSING GAME. In a typical intellectual property licensing situation. one of the players (the inventor) owns a piece of intellectual property (e.g.. a patenl copyright, or trademark) that the other player (the manuflicturer) seeks to use. The "pie" at issue is the profìtscash flows that are expected (O be generated us a result of the use of the technology by the manulìnturcr. Tu understand tue insights providcd by game theory for intellectual property licensing, we have developed Example 2. Inventor is an individual. He or she has developed a new product, the Widget. The intellectual property to be licensed includes a patent covering the Widget and all know-how required to manufacture and bring the Widget to market. The potential Manufacturer is a company that possesses all the resources, i.e., manufacturing facilities, work force, saks force, and so on. necessary to manufacture and commercialize this product. Given its resources and capabilities, Manufacturer is believed to be the must promising developer of Inventor's intellectual property. The remaining life uf the patent is four years. Both Inventor and Manufacturer believe that the price uf Widgets will be $200 each and that the costs associated with manufacturing and selling Widgets is $100 euch. In addition, 1,000 Widgets are expected to he sold euch year. Thus, total revenues (ignoring discounting, for nuwj associated with making and selling Widgets are expected to be $200,001) per year. und profils are expected to he $100,000 per year. In this example, liweinor and Manufacturer take turns making offerswith only one offer being made each year. In Year J (when there are four years remaining in the life of the puient. Inventor extends an offer to give Manufacturer rights to nnnufacure and sell Widgets fur the remaining life of the patent. Manufacturer is free to accept or reject this offer. 1f the offer s rejected, Manufacturer will have an opportunity to make an offer to Inventor in Year 2 (when

108 Case 3:11-cv RBD-TEM Document Filed 07/25/13 Page 9 of 28 PageID Ch. 17 Applkatlon of Game Theory to Intellectual Property Royalty Negottations only three years of patent life remain). Inventor is free to accept or reject this offer. If the offer is rejected, Inventor will have an opportunity to make an offer to Manufacturer in Year 3 (at which time only two years of patent protèction remain). This process continues until Year 4the final year of patent protection. In Year 4, Manufacturer has the right to make an offer. If this offer is rejected by Inventor, then negotiations end, and both parties get nothing. Therefore, it is in Inventor's best interest to accept any amount of money offered by Manufacturer in Year 4, because the alternative would be to get nothing. Following the logic developed in examining Example I (the Hatfields and McCoys), the offers that should be extended by Inventor and Manufacturer can be determined by looking forward and reasoning back. Manufacturer's offer in Year 4 would give virtually all the available profits to Manufacturer. Inventor would be offered (and would be expected to accept) the smallest possible amount (e.g., one penny) in Year 4 because the only alternative would be to reject the offer and get nothing for its Widget invention. Given these results in Year 4, Inventor's licensing offer in Year 3 would have to allocate at least $100,000 to Manufacturerbecause $100,000 is the amount that Manufacturer will get in Year 4 if no agreement is reached in Year 3. An agreement in Year 3 would generate $200,000 in profits ($100,000 in both Years 3 and 4), so Inventor's offer in Year 3 would give Manufacturer $100,000 and would give Inventor $100,000. In Year 2 (when $300,000 in total profit is available if an agreement is reached), Manufacturer's offer would allocate $100,000 to Inventor (the amount that Inventor could get in Year 3 if Inventor rejected Manufacturer's Year 2 offer) and keep $200,000 for itself. Details regarding all the offers that would be extended in these four-year-long negotiations are provided in Exhibit In this situation, an agreement will be reached in Year I. based on an offer made by Inventor in which $200,000 is allocated to Manufacturer and $200,000 is allocated to Inventor. In other words, the license to which Inventor and Manufacturer are expected to agree provides for an even split of the profits generated by the commercialization and sale of Widgets. Moreover, no losses caused by an inability to reach an agreement are expected. Years of Inventor Share ManufacturerShare Proteo- Annual Total Year lion Profit Profit Offeror Total Annual Total Annual 1 4 $100,000 $400,000 Inventor $200,000 $50,000 $200,000 $50, , ,000 Manufacturer 100,000 33, ,000 6bjth , ,000 Inventor 100,000 50, ,000 50, , ,000 Manufacturer , ,000 Exinarr 17.1 NEGOTIATiON OrFERS

109 Case 3:11-cv RBD-TEM Document Filed 07/25/13 Page 10 of 28 PageID The Nash argafnfng Solution (NBS) 247 A more formal exploration (and explanation) of a game-theoretic approach reflecting the previous discussion follows, based on sound economic principles as well as common sense: it is applicable to many licensing situations. Its bases and conclusions can he extremely useful in helping negotiators arrive at a more informed and sensible price tr the intellectual property to be shared. THE NASH BARGAINING SOLUTION (NBS) In the early 1950s. Nobel Prize winner John Nash formally DESCRIPTION. examined issues raised in bargaining games.'4 Instead of providing descriptive or normative observations on the bargaining process, he, in essence, asked: "What would a good bargaining solution look like?" He reasoned: bine can attack the (hariinijigl problem axiomatically by stating general properths fimt 'any reaonahie si,lutitrn' should possess By specifying enough such properties one excludes all hut one solution.'5 The general properties of a bargain that were identified as critical by Nash were the following. I. Any solution to a (wo-person bargaining problem should be feasible and Pareto efficient. That is. the outcome should be attainable and there should he no other feasible solution that is better than the solution for one player and not worse than the solution for the other player. Neither party should get less in the bargain than it could get without a bargain. In other words, the benefit of the bargain for each player (ll) should he greater than or equal to that player's "disagreement" (or no bargain) profits (d,). The bargaining solution should be independent of the numeric specification of the utility function. That is. if payoffs are measured in a different way, the solution should stay the same. What should matter is the relative utilities of the bargainers, not the specific manner in which they are measured. Eliminating alternatives other than the disagreement profits should have no effect on the outcome. 1t in the first bargain, option A is chosen over B and C, in a second bargain, if C is eliminated. A should be chosen over B again. S. If players I and 2 have un equal bargaining position, then the solution should (real them symmetrically (or equally). Nash proved that in each bargaining situation, these properties lead to a unique (or single) bargaining outcome. That outcome has come to be known as the Nash Bargaining Solution (NUS), with the NBS a function of three variables. The first is the total amount of benefits available to be divided, i.e., the size of the pie. This is also referred to us transferable wealth. The second is each party's disagreement profits, This represents the benefits that a party can obtain in the

110 Case 3:11-cv RBD-TEM Document Filed 07/25/13 Page 11 of 28 PageID ch. 17 Application of Game Theory to Intellectual Properly Royalty Negotiations event that no agreement is reached and corresponds to the benefits associated with the party's next best alternative. The third variable is each party's relative bargaining strength. Stronger parties receive larger slices of pie. The NI3S'6 provides that the optimal solution is one in which benefits/profits are split such that: where: I1: Player i Nash-bargained profits 112. Player 2 Nash-bargained profits Player I disagreement profits Player 2 disagreement profits a: Player I relative bargaining strength O a I fl: total profits The intuition behind these results is relatively straightforward. In the allocation of the total pie available, each party is assigned an amount equal to that which it would be able to receive if no agreement were reached, i.e., Player J obtains at least d and Player 2 obtains at least d. 1f this were not the case, it would not be in the interest of either party to enter into the agreement. After that, each party is entitled to a portion of total remaining profits ( d,), which we will call the "agreement surplus." That "portion" is determined by the relittive bargaining strength of each party. The parameter a reflects Player l's negotiating power relative to Player 2 and ranges between O (when Player 2 enjoys overwhelming bargaining power) and I (when Player I enjoys overwhelming bargaining power). When the negotiating parties are on an equal foocing (which is Nash's original assumption and embodies the fifth property identified above), the value of a is and the NJ3S provides that: where: ri1=d1+a(nd--d2) (17.1) fl,=d2+(l c(fldd1) (17.2) ri= rr1+ ri2 111d1(fld1--d,) (17.4) rj2d2(fid1--d,) (17.5) I1: Player 1 Nash-bargained profits 112: Player 2 Nash-bargaIned profits Player I disagreement profits Player 2 disagreement profits Il: total profits ('7.3) rizfi1l12 (17.6)

111 Case 3:11-cv RBD-TEM Document Filed 07/25/13 Page 12 of 28 PageID The Nash Bargaining Solution (NBS) 249 In Example 2 (Inventor/Manufacturer) discussed earlier, we assumed that both parties izero clisagreeiìient profits (d, = d2 = O) and had equal bargaining power (a = ). That resulted in an even split of the benefits associated with an agreement. There is an intuitive and mathematical appeal to such a result. }owever, in the real world, assumptions of no disagreement profits and equal bargaining power often do not hold. In reality, Inventor may have been able to earn something on his or her intellectual property even if there were no agreement with Manufacturer. Assuming Inventor had the ability to earn $ per year by manufacturing and selling Widgets un his or her OWO during any period in which there was no agreement with Manufacturer, the negotiated outcome would be different. We begin, once again, at the end of the negotiating process. When Manufacturer is preparing to make its offer in Year 4, Manufacturer must give Inventor at least $20,000 (plus $0.01) in order to make it worthwhile for Inventor to agree to a license. Otherwise. Inventor would be better off rejecting Manufacturer's offer and selling Widgets on his or her own. Accordingly, the new Year 4 offer extended by Manufacturer would give $20,000 to Inventor and $80,000 to Manufacturer. In light of these Year 4 results. Inventor's offer to Manufacturer in Year 3 would only need to give Manufacturer $ Accordingly, Inventor's offer in Year 3 would allocate $120,000 to inventor and $80,000 to Manufacturer. In Year 2, Manufacturer would need to offer Inventor the $120,000 that Inventur could obtain in a Year 3 dealand an additional $20,000 to account for the profits that Inventor could obtain on his or her own in Year 2 if no agreement is signed. A summary of the complete bargaining process s provided in Exhibit As shown in this exhibit. Inventor's ability to make money outside of an agreement fundamentally shifts the results of the negotiation. There is, once again, an agreement n Year I, but now Inventor is able to claim most of the benefits generated by the license. Specifically, of the $100,000 per year generated by commercialization of the Widget, Inventor is able to lay claim to $60,000, while Manufacturer is limited to only $40,000. The result can be easily derived using the NBS, with the following parameters: (1) a = : (2) 1, (Inventur) 4 $20,000 = $80,000; (3) tim (Manufacturer) =0; and (4)11 = 4 $1 00,00(1= $400,000. Veau of Protec. Annual Total Year Lion Profit Profit Offeror ExHiBir 17.2 Su'.vIAKyur BARGJ%NIN(. Paot rs inventor Share Manufacturer Share Total Annual Total Annual 1 4 $11X1.000 $4(X),000 Inventor $240,001) $60,000 $160,000 $ ,000 3(0,00O Manufacturer 140,000 46, ,000 53, ,001) Inventor 120,000 60,000 80,000 40,000 4 I 100, ,000 Manufacturer 20,000 20,000 80,000 80,000

112 Case 3:11-cv RBD-TEM Document Filed 07/25/13 Page 13 of 28 PageID Ch. 17 Application o( Game Theory to lntellectuj Property Royalty Negotiations fl,=d,+f(11d,dm) (17.7) = $80,000 f ($400,000$80,000$0) (178) = $Sb,000 f ($320,000) (17.9) =$240,000 (17.10) "M=dM+i (fid,.dm) 7.11) f ($400,000$80,000 0) (17.12) =$of ($320,000) (17.13) =$160,000 (17.14) A different result emerges if Manufacturer, rather than Inventor, has an alternative to licensing. Assume that Manufacturer could earn $50,000 per year in tile absence of a license, while Inventor would carri nothing. Negotiations under these conditions are illustrated in lexhibit 17,3. As one might expect, Manufacturer's ability to make money outside of an agreement with Inventor shifts th&. rcu1t of the negoti ilion in Manufacturer ' favor. Under these conditions, an agreement is reached in Year I, but now Manufacturer is able to claim $75,000 per year generated by commercialization of the Widget, whereas Inventor is able Lo obtain only $25,000. This rlsult could also be derived using the NBS with the following parame ters: (1) a = ; (2) d, $0; (3) da, = 4. $50,000 = $200,000; and (4) fl 4 $100,000 = $400,000. IT,=d,f (fld,d,) (17.15) =0+ f ($400,000 $0 200,O00) (17.16) 0+ f ($200,000) (17.17) = $100,000 (17.18) Manufacturer Years of Protec- Annual Total Inventor Share Share Year tien Profit Profit Olferor Total Annual Total Annual 1 4 $100,000 $400,000 Inventor $100,000 25,000 $300,000 $75, , ,000 Manufacturer 50,000 Th,f?67 250,000 83, , ,000 Invuntor 50,000 25,001) 150,000 75,001) , ,000 Manufacturer , ,000 ExuietT 17.3 NEGOTIATIONS UNDER M,V'urAcruEk ALTERNATIvES

113 Case 3:11-cv RBD-TEM Document Filed 07/25/13 Page 14 of 28 PageID The Nash Bargaining Solution (NBS) 231 fl=d11+4 (fld,dm) (17.19) = $200,000 + I ($ $0 - $200,000) (17.20) = $ ($200,000) (17.21) = $300,000 (1712) Finally, we consider a situation in which both parties have viable options. For this analysis, we will assume that Inventor can earn $20,000 per year and Manufacturer can earn $50,00() per year in any year in which no license is in place. The negotiations under these circumstances are summarized in Exhibit Once again, this reult could also he derived using the NBS, with the following parameters: (I) a = : (2) d, 4. $20,000 = $80,000; (3) dh 4. $50,0000 = $200,000; and (4)11 = 4. $100,000 $ (fl-1i,--dm) (17.23) = $80,001) + 4 ($ $80,000 $200,000) (17.24) = $80,001) + i ($120,000) (17.25) = $140,000 (17.26) (fl-d,-d,,) (17.27) = $ ($400,000 - $80,001) - $200,000) (17.28) = $200,000 4 ($ J 2(1,000) (17.29) = $260.00() (17.30) In this example, Manufacturer's alternative to licensing is more valuable than Inventor's alternative. Accordingly. Manufacturer is able to lay claim to a larger share of the annual proffts uf $ associated with che commercialization of Widgetsobtaining $65,000 per year compared with $35,000 per year for Inventor. As modeled thove, differences in the alternative options available to a party involved in a negotiation are critical to the share of profits that might be claimed by each party. The party with the better alternative obtains a higher share of the prolits. Year Years of Prolec. lion Annual Profit Inventor Share Total Profit Offeror Total Annual Manufacturer Share Total Annual 1 4 S100,fifl() $409,0011 Inventor $ 140,0(10 $35,000 $260,000 $65, (1,001) 100,01)0 MdnufaclurLr 90,000 30, ,000 70,000 ri 2 I (111,01(0 20(1,000 Inventur 70,000 35, ,000 65, ( ,000 M.rnulacturcr 20,(lOO 20,000 60,000 80,000 EXHIBIT 17.4 ÑLc,nIArIcJN Willi BI)TII PARTY OI'r,0N5

114 Case 3:11-cv RBD-TEM Document Filed 07/25/13 Page 15 of 28 PageID Ch. 17 Application of Game Theory to Intellectual Property Royalty Negotiations The allocation of agreement surplus (i.e., 11 - d1 - c12), as noted earlier, is dependent on each party's bargaining strength, and much of that is driven by the relative costs of haggling to each party, i.e., the cost associated with delays in reaching an agreement. In Example E (Hatfields and McCoys) discussed earlier, how might the division of ice cream change if the Hatfields' group consisted of one adult and six (impatient and hungry) children, whereas the McCoys' group included seven adults? Recall that, in the original analysis, the Hatfields offèred two quarts of ice cream to the McCoys at the start of the negotiations, keeping two quarts for themselves. For their part, the lvfccoys were indifferent between accepting this offer right away and getting two quarts during the negotiation 15 minutes later. In the present case, the Hatfields are decidedly not indifferent between the immediate deal (two quarts for each party) and the deäl that they could obtain in 15 minutes (in which they receive only one quart of ice cream). To avoid the possibility of the unpleasant outcome of receiving only one quart of ice cream in 15 minutes, the Hatfields are likely to sweeten the deal for the McCoysgiving them slightly more than two quarts, while keeping almost two quarts for the chi!- dren. In this instance, the additional cost of haggling faced by the Hatfields (namely, unhappy children) causes them to weaken their negotiating demands. Using the NBS structure, a (the bargaining power of the Hatfields) is likely to be much lower than that of the McCoys. In other words, a does not equal. Ta the extent it is lower, the Harfields are entitled to less than of the agreement surplus. Applying unequal bargaining power to Exhibit 7.4 of Example 2 (InventorI Manufacturer), and assuming a, results in the following. fl,=d,ct(fl-dl-.dm) (17.3!) = $80,000 + ($400,000-$80,000- $200,000) (17.32) =$80,000 ($120,000) (17.33) = $120,000 (17.34) -dl-dm) (17.35) = $200,000 + ($400,000- $80,000 - $200,000) (i 7.36) = $200,000 + ($120,000) (17.37) = $280,000 (17.38) With lower bargaining power, Inventor is awarded a smaller share of the annual profits associated with Widget sales. That share now becomes $30,000 per year, versus the case, in which he or she obtained $35,00() per year. EXTENSION TO LICENSING, In a paper published in July 2000, William Chai and Roy Weinstein posited that the basic or symmetrical NBS (i.e., the NES shown in Equations (17.4), (17.5), and (17.6)) can be usefully applied to reasonable royalty determinations in intellectual property litigation. They wrote that:

115 Case 3:11-cv RBD-TEM Document Filed 07/25/13 Page 16 of 28 PageID where: fi,: inventor Plush-bargained profits 11,,: Manufacturer Ntuh-bargained profits d,: Inventar disagreement profits d,,: Manufacturer disagreement profits rl: total profits The Nash Bargaining Solution (NOS) 253 The NIlS is well supported by economic theory and is regarded as one of the simplest yet must fruitful paradigms in game theory. The analytical clarity of the NBS atsti is an iniprwtant justification for its use as another useful tool in calculating a reasonable royalty.'7 In their article, Chol and Weinstein converted the Nash Bargaining Solution into a per unit royalty. A summary uf this conversion, and their conclusion, is provided below. Per Unit Royalty. Total profits are those generated by Manufacturer in the event that an agreement is reached. They can be expressed as: where: f1,= P,,. Q,- C.,(Q,,, (17.39) Pi,: Manufacturer price per unit Q',: Manufacturer quantity C,,IQ,,): Manufacturer total cost function li,,: Manufacturer profits The amount uf profit received by Inventur under an agreement will be equal to a royalty rate multiplied by the total volume of sales of the relevant products. Therefore: where: 11,: Inventur profits rr0: royalty rate per unit Q,: Manufacturer quantity (17.40) incorporating Equations (17.39) and (17.40j with (17.4), (17.5), and (17.6) (and assuming equal bargaining power) permits the calculation of a royalty per dollar based un the PIBS. Í1,=d, (fld,d,) (17.41) rç. Q, (J7.42) (Udg--dM) (17.43) n - ri, (17.44)

116 Case 3:11-cv RBD-TEM Document Filed 07/25/13 Page 17 of 28 PageID Ch. 17 Applicaton of Game Theory to FnicIJectut rroperty Royafly Negotiations Combining Equations (17,43) and (17.44), Manufacturer's profits cart be written: d4f-t- ct+ (I1d,--d.,)= flil, (fl ti/dm) = Il rç Q (17.45) (17.46) That results in; rr, Q1 = - (fl - t1 d41) rr Q,= Il (d1c!1) (17.47) (17.48) Incorporating Equation (17.39) into Equation (17.48): rr Q, = (P4, Q, - C (QM)) (d, - Dividing both sides by Q,: rr = [PM - AC,] + Q,, [d, - tl1j ' where: rr: royalty rateper unit PM: Manufacturer p1-ice per unit ACM: Manufacturer average costs per unit Q;,, : Manufacturer quantity d, : Inventor disagreementprofits dm : Manufacturer disagreement profits ) (17.49) (17.50) Çhoi and Weinsteia noted that the first part f Equation ( ) provides that, as n starting point, a royalty rate should be one-half of the licensee's (or Manufacturer' s) profits. The greater the Manufacturer returns, the greater the royalty. The second part of Equation (17.50) prbvides that the royalty rate should be adjusted up or down depending on thé disagreement profits of the two parties. Il Inventor is not in the business and has no alternative, the rate way he adjusted down. 1f, onifie other hand, inventor is in the business and has relatively littleto gain from a bargain, the royalty rate may be adjusted upward. li the two parties h'ive equal disagreement profits the licensing probes should bc split cquaily is the basic (or symmetric) Nash model suggests. For Example 3, let us assume that the price of tile product is$3.50 per unit, the average cost is $1.00 per unit, and the expected quantity demanded is 40 unies. This implies that expected Manufacturer profits (11= 111f) are 40. ($3.50$1.00) = $100. Let us also assume that a non-bargain nets Manufacturer (c1,) $40 und a non-bargain nets Inventor (d1) $0. Assuming equal bargaining power, the basic NBS royalty tate per unit is:

117 Case 3:11-cv RBD-TEM Document Filed 07/25/13 Page 18 of 28 PageID Tho Nash flargaining Solution (NBS) 255 rr= I$3.50$l.001 2(40) L$0$40] (17.51) =$1.25 ['?] (17.52) = $ $0.51) (17.53) (17.54) Applying that royalty rate to the Manufacturer's projected units result in a royalty payment by Manufacturer tu In Ventor of ri1 = $ (17.55) =$30 (17.56) A royalty payment (or transfer ut wealth) from the manufacturer of $30 results in a unitlue bargaining solution. Inventor is granted one-half ofthe agreement surplus, which is half of the difference between total bargaining profits ($100) and total disagreement profits ($40). Manufacturer retains $70its disagreement profits ($40) plus one-half of the agreement surplus ($30). A variant of Example 3 is one in which Inventor and Manufacturer are both in the business. Although they generally compete, it must be the case that Manufacturer will expand the market place beyond that which would have existed (either through price advantages or product enhancements). Otherwise, licensing wòuld be irrational. Again, for Illustrative purposes. let us make the same assumptions as earlier, except that lnventors non-bargain outcome Id,) is $30. The resulting NBS royalty rate per unit is: 0= t$3.50$l.00j+ =$1.25 {_$10] =$1.25$0.125 = $1.125 That results in a royalty payment of: fl,= $ ) i$30$40) (17.57) (17.58) (17.59) (17.60) (17.61) = $45 (17.62) Here, the royalty payment is higher than in the earlier example. That is because Inventor disagreement profits are positive. In other words, tnventor can generate profits from the patent without granting a license to the Manufacturer. Therefore, the royalty fee he or she commands is higher. Adjusting for that, each

118 Case 3:11-cv RBD-TEM Document Filed 07/25/13 Page 19 of 28 PageID Ch. 17 Applicalion of Came Theory to hitellectual Property Royatty Negotiations party does receive an amount that is equivalent to its disagreement profit plus one-half of the agreement surplus. Inventor profits are $45 ($30 $15). Manufacturer profits are $55 ($40 + $15). As noted, Choi and Weinstein examined the basic (or symmetriçal NBS). In real-world licensing negotiations, however, bargaining power often s not equal.19 In order to account for such inequality, one can derive a per unit royalty rate based on the Mash Bargaining Solution shown in Equations (17.1). (17.2), and (17.3). Using that construct, a royalty rate per unit is: where: rr = a [PM AC(QM)) + (1 - Ct) d!iqm - d.i/qm (17.63) rr: royalty rate per unit PM: Manufacturer price per unit ACM: Manufacturer average costs per unit Q: Manufacturer quantity d1: Inventor disagreement profits dm: Manufacturer disagreement profits a: Inventor relative bargaining strength O a I The per unit royalty rate when there is unequal bargaining strength consists of three elements. First, Inventor receives a portion of the per unit profit, with the portion equal to his or her relative bargaining strength. Second, Inventor receives a share of his or her own per unit disagreement profits, with the share equal to I minus his or her relative bargaining strength. Third, the royalty rate is reduced by a share of the Manufacturer's per unit disagreement profits, with that share equal to Inventor's relative bargaining strength. If Inventor is in a commanding negotiating position, i.e., a = 1, Inventor will receive all the per unit profit less Manufacturer's per unit disagreement profits. Remember, an optimal outcome is one in which each party receives at least what it could receive with no bargain. If Inventor has no bargaining power i.e., a= 0, then the licensor's return will be limited to his or her own per unit disagreement profit. As Inventor's bargaining power falls, he or she obtains less and less of the agreement surplus. Per Dollar Royalty. Equation (17.63) expresses the royalty in per unit terms. In fact, most royalties are expressed as a percent of revenues. As a result, it is useful to consider a per dollar royalty rate. Such a calculation can be done using the procedure outlined in Equations (17.39) to (17.50). In that analysis, however, Equation (17.40) should be replaced with the following: where: fl1 = rr,1. (M Q) (17.64)

119 Case 3:11-cv RBD-TEM Document Filed 07/25/13 Page 20 of 28 PageID fl,: Inventor prallts rrd. royalty rate per dollar Ps,: Manufacturer price per unit Q.,,: Manufacturer quantity The Nash Bargainhig Souuon (NBS) 257 In words, the amount of profit received by Inventor under an agreement will be equal to a royalty rate (expressed as a percentage of total revenue) multiplied by total revenues generated by the sales of the relevant products. Using Equation (17.64) in thc analysis outlined above, we get: where: rr = rr/p (17.65) = a f AI - AC(Q)1/P1, + fl - a) d,/(pq,.,,j d,i(pfqm) (17.66) rrd. royalty rate per dollar rr: n)yalty rate per unit Manufacturer price per unit ACe: Manufacturer average costs per Unit Manufacturer quantity d,: Inventor disagreement profits c: Manufieturer disagreement profits a: Inventor relative bargaining power O a i Equation (17.66) shows that there are three components to the royalty rate per dollar indicated by the asymmetric NBS. The first component is a share (a) of the percentage mark-up charged by Manufacturer. The mark-up is profit per dollar of revenue. The share is equal to Inventor's relative bargaining strength. The second component is a share (1 - U) of the percentage of total bargaining revenues that Inventor could receive in the absence of an agreement. The share is equal to Manufacturer's relative bargaining strength. The royalty rate increases with both of these components. The per dollar royalty rate decreases with the third component, which is a share (a) of the percentage of total revenues that are available to Manufacturer in the absence of an agreement. The share is equal to Inventor's relative bargaining strength. For Example 3, we assumed that the price of the product is $3.50 per unit, the average cost is $1.00 per unit, and the expected quantity demanded Ls 40 units. This implies that expected Manufacturer profits (n = rim) are 40. ($3.50$l.00) = $100. We also assumed that a non-bargain nets Manufacturer d) $40, and a nonbargain nets lnvcntor(d,) $0. In Equation (17.54), we derived a royalty rate per unit of $.75. Given that the schlag price was $3.50, that equated to a royalty rate per dollar uf 21.4%. Assuming unequal bargaining power, say, a is obtained..6, a different result

120 Case 3:11-cv RBD-TEM Document Filed 07/25/13 Page 21 of 28 PageID Ch. 17 Applkation of Game Theory to Intellectual Properly Royalty Negotiations rrd =.6 [$3.50 $i.00}/$ SO.6 (17.67) =.6 [.71] -.6 [.29] (17,68) = (I7.6) =.257 (17.70) =25.7% (17.71) Increased Inventor bargaining power íncrease.s the royalty rate per dollar. In this example, the rate increased (from 21.4 to 25.7%), roughly proportionate with the increase in bargaining power. ESTIMAIJON OF NBS The NBS is highly stylized yet versatile. f well understood and carefully applied, it can provide extremely useful intellectual property pricing bounds. Nonetheless, it is not without limitations. The basic insight of the NI3S can be understood geoinetricafly, as shown in Exhibit Total profits represent the benefits that are to be allocated between Inventor and Manufacturer. It is also said to represenl the level of transferable wealth The starting point for thc 'illocation of profits is that cadi party must receive at least as muôh from a dea] as lie or she would receive if no deal were reached. Accordingly, Inventor must receive at least d, in an agreement, and Manufacturer must receive at least dm. The remaining profits (Total Profits d, -. d) are allo. ited b'iscd on the rel'ttive bargaining power of thi. parues ThL stronger party is allocated more of the surplus. Inventor Olsagroement Protits (d,) '0' 5, :' 4 Agreement Súrplúi hit Increasing Manufacturer Bargaining Power increasing Inventor Bargaining Power Inventor Share EXHIBIT 17.5 BASIc lnsigtt OF Ti-IC NBS

121 Case 3:11-cv RBD-TEM Document Filed 07/25/13 Page 22 of 28 PageID Estimaflon of NBS 59 Although reasonably simple n concept, there are practical issues related to the understanding and application of each of the three relevant parameters. DISAGREEMENT PROFITS (d,). Each party needs to assess the benefits that can be expected/br 1)0th paríhs if no deal is consummated. In other words, the value of each trading party's next best alternative should be determined. In principle. that is simply the net present value of all future direct and indirect inflows associated with the course of action that would be chosen by each party in the event that no agreement is reached. That is often done, ordo-able, by licemcing professiona's and economists. There are challenges, however. One is the idenulication of what the next best afternative might beparticularly for the other party. Ocnerally speaking. negotiating parties will have a fair sense of their OWfl Options bui a limited sense of the other partys. Furthermore, parties are often reluctant to share (us- ct)flfirhl) what their alternatives might be. This informatiun asymmetry can make it difficult in develop reasonable estimates of d, and c!1. Even if parties are aware o1 one another' options. the valuation of such options can he difficult. Judgment is required. und consensus can be difficult to obtain. Third, as u general matter. parties will tend to overvalue their own alternatives and undervalue those of the other party. Such biases make it difficult to reach a meeting of the minds im disagreement profits. A fourth challenge to the valuation of each party's next best alternative is the fact that the uhcrnativcs may be constantly changing and evolving. Faced with the prospect nf a potentially substantial license fee, licensees often mobilize their best subject-matter scientists to design-around the intellectual property at issue. Also, faceti with a recalcitrant panner, licensors often "shop around" their technology to others. Thus. d, and d can he moving targets, particularly in licensing negotiations that take time. A fifth challenge is that the disagreement profit calculation for each party should often include indirect considerations, which are difficult to estimate. For example, consider u manufacturer that frequently licenses-in technology. For such a company, the outcome of a giveh negotiation may have an impact on future negotiationsand one effect ni an "tmsuccessful" negotiation over one piece of intellectual property might contribute tu a reputation for tough dealing that will save money in later negotiations. For an inventor, the failure of a given negotiation may contribute to improved results in subsequent negotiations owing to a similar reputation of toughness. Such considerations may be hard to quantifyadding further complexity io the determination old, and dm. Despite these challenges, it is important to put effort into estimating d, and d as accurately as possible. Because the agreement surplus is calculated as total prolits (fl) minus J, and d., failure to estimate both the tatter two variables leads to a nondeterministie outcome.

122 Case 3:11-cv RBD-TEM Document Filed 07/25/13 Page 23 of 28 PageID Ch. 17 Application of Garne Theory (o Intellectual Property Royalty Negotiations TOTAL PROFITS (n). The next critical piece of the NBS analysis is estimating the profits to be generated by the agreement under consideration. As a general matter, the tools used to prepare such estimates are a normal part of the licensing professional's (or economist's) toolbox, and one would expect that such estimates would be routinely prepared by both parties involved man intellectual properly licensing negotiation. For the purposes of applying the NBS,however, three points should be made about estimates of the profits to be generated by che agreement. First, althoúgh one would expect both parties to have prepared an estimate of the profits to be generated by the agreement, the parties may not have reached the same conclusión, partictilarly ifthe intellectual property at issue is a new technology fo which substantial uncertainty exists with regard to its vulue. As before mnfoimttion 'isymmltrme (and beliets) lead to differences mn ascing the pie to bé split, which lead ro differences in proposed slices of pie. Second, total profits (11), and theit estimates, increase or decrease over time as the future becomes close- amid less uncertain. As moore is learned about a potential market opportunilysuch as the size of the penetrable market, the likely degree of peneir ition, the competltivl lanciscall ( e possmbk. compuitor), the market clearing prices, and supplier cost strucureexpectations of bargaining profits become more refined. Even if beliefs are symmetric, fl changes over time. Third, total profits may not be independent of the outcome of the licensing negätiations at hand. After all, the licensing terms may alter the incentives of the parties (particularly the licensee) in ways that influence total profits. In Example 3, we assumed thnt the price of the product is $3.50 per unit, the average cost of produlion is 1 00 per unit and the expected quantity deminded is 40 units Those assumptions implied that expected Manutac,turer profits (IT = M) are $40 ($3.50 $LO0) = $100. In that example, with equal bargaining power, we detctmincd that the royalty rate would be 21.4%. The royalty fee can be accounted for by the manufacturer ma number of ways.first, it. could treat the total royalty to be paid$21.40asa fixed cost of doing business. In such an event, the royalty would not affect the pricingdecisions of the manufacturer. Alternatively, the manut'acturer could pass Lhe entire cost increase (br some portion) through to its customers. Because demand curves are downward sloping, a price increase will be greeted by a quantity reduction. Depending on the point on the demand curve at which transactions are occurring (i.e., depending on the price elasticity of demand), that puce increase may be more than offset by a quantity ueduction, such that total revenues fall. In addition, it may be greeted by a less than proportional quantity reduction. In either case, Il is unlikely to remain constant. As tile SÍZC of the total pie changes, the expected royalty rate may also change, which may, in turn, alter the size of the pie. As before, it is important to estimate Il as accurately and consistently as possible. The size of the agreement surplus is driven by that estimation. BARGA1NING POWER (a). The third element of the N]3S analysis, a, is the relative bargaining power of the negotiating parties.2 lt reflects the forces at

123 Case 3:11-cv RBD-TEM Document Filed 07/25/13 Page 24 of 28 PageID Estimation of NB5 261 work in a given negotiation (such us the rules of the games the tactics employed by the bargainers. the information structure. and the player? discount rates) that can shift the outcome toward the interests of one party over another. Unlike ri and d, the estirnaticrn of cx does not derive from the tools normally used or developed by licensing proessianal. rather, estimation of involves more subjective usessments ofthe bargaining strength ofthe parties involved. Absent additional inrormation.,ne might reasonably expect relative bargain- Ing strength to he equal. i.e.. c =. The question, therefore, becomes: What might make a different from? One of the main ftrces that niove ct away from is time. More specifically, it is the different ways in which the parties involved in the negotiation react to the passage of time. As noted eatlier, and a is well undcrmood by those in the profession, realworld license negotiations typicatly do not involve such one-time, simultaneous offers as suggested by the NBS. Instead. real-world negotiations take time. Also, while such negotiations are taking place, neither of the parties to the (potential) agreement are enjoying the benefits of the agreement. Moreover, as time passes, the likelihood grows that the value of the intellectual property under consideration could change because of other developments (e.g., the emergence of an alternative technology or a shift in consumer tastes). The relative costs and dangers faced by each party owing to delays in reaching an agreement can be expected to have a direct impact on each party's patience for delaying an agreement and, consequently. its relative bargaining strength. In the early 1980s. Ariel Rubinstein published a paper describing a (sequential or time-consuming) model of bargaining in which the players take turns making offers to each other over the division of a "pie" until agreement is reached.' This paper explored the drivers of the division of the pie between the negotiating pai-lies. Rubinstein showed that the division of the pie depends on two parameters: (lithe time between offers and (2) the negotiating parties' relative discc,unt rates. Rubinstein showed that co the extent that there is a delay between the time arie offer is rejected and the next offer can be made, the offering party has an advantage and will be able tu extract a greater share of the pie. This advantage flows from the fact that value is being lost during the period between offers, and the offering party should be able to capture this value. Of course, as the time between offers diminishes, the extent of such losses declines and a first mover advantage is neutralized. With regard to the negotiating parties' relative discount rates, the less patient party is "rewarded" with a smaller share of the pie. In fact, when offers and countercffers are essentially instantaneous, the allocation of the pie between the parties depends only on the ratio of their discount rates. Ifa party is more patient, his or her bargaining power is greater. Mathematically, it is: cx- (17.72) r, + r3

124 Case 3:11-cv RBD-TEM Document Filed 07/25/13 Page 25 of 28 PageID Ch. 17 Application of Game Theory to Intallectuai Property oyaity Negotiations where: a: Player i bargaining power r1 : Player I discount rate r2 : Player 2 discount raie If r1 = io percent and t2 _ 20 percent, then a =.67. Player I hai twice as much power because he or she is twice as patient. It should be noted that the "discount ratet' reflects the cost of bargaining to each party.2 Such costs include each party's time value of money, a well as the expected costs associated with potential adverse market developments, costs associated with ist 'patent time,' and any other cost that makes an immediate agreenient preferable to an agreement at some point in the future. Some new technologies have very broad windows of opportunity for Manufacturers and are not easily disphced Thcii discount rates are quite low Other technoiogic h tve very limited opportunities. Their discount rates are high.moreover, a prospective Inventor with limited licensing-out opportunities mayview the pie a shrinking at a much faster rate than an Inventor with a myriad of licensing-out opportunities in a very crowded marketplace, Furthermorc,the importance ofgenerating cash flows in general, and these project flows iii particular, will be a function ofeach party's current and projected business andfinancial position. An individual Inventor with limited financial resources and limited licensingprospects will, holding all else constant, be more inclined to reach a deal (and faster) than vili be a large multinational corporation with a myriad of prospective projects. Relatively large and profitable projects, holding all else constant, increase the cost of delaying access Lo those projects. Relative bargaining power can also depend on the nature of the relationship between the parties. For example, if Manufacturer is a major 'customer" of Inventor,this circumstance is likely to provide Manufacturer with some degree of negouating leverage that would affect relative bargaining powet Altem i- tively, if Manufacturer has already made a substantial investment in a project and requires a license to complete the project, this can diminish Manufacturer's bargaining strength. Relative bargaining strength can also be affected by the nature of the negotiations For example an Inventor who is seeking to lìcene a ptece of inlel]l.tutl property is likely to have much more bargaining power if he or she is able to have two potential licensees competing for the license than he or she would have if the licensing ôptions were limited. Estimating a (or the relative costs of haggling) is extremely difficult for several reasons. First, it is not at all clear what should be included in the dccrmiradon or how to weigh each factor. "Relative bargaining strength" is not a concept that is often quantified by licensing professionals or economists. Second, not only is there often a lack of full (and symmetric information), but the relative costs change over time. Moreover, parties are often motivated to present biased information Lo their potential partner. -t

125 Case 3:11-cv RBD-TEM Document Filed 07/25/13 Page 26 of 28 PageID Conctustons 263 Third. the distinction between c/ and ais unclear. Disagreementprofits are drivers of bargaining power. A high d, is often associated with a high a. The influences need to he disentangled. Estimating a requires consideration of the entire circumstances involved in the negotiation and a realistic assessment of the strengths of each party. Fortunately, such assessments are commonly made by licensing professitrnals they are just not usually made as explicitly (or perhaps consciously) as they should be in applying the NES framework. Thinking more systematically about such issues will improve ones ability to apply the NBS and, thereby, think about the "righi" price fur the intellectual property that one is seeking to license. CONCLUSIONS The NBS construct does not attempt to explain the results of licensing negotia- Lions. The underlying work WliS theoretical in nature. Moreover, the predictive value is virtually impossible to test because of the largely private and/or subjective nature of the expected total profits (li), disagreement profits (de), and relative bargaining power (a). On the other hand, the NBS is a reasonable outcome in the face of sensible properties. It also might he thought of as what a fully rational and informed arbi trator might conclude, being armed with the relevant facts (arid expectations) surrounding a real-world license negotiation. At the very least, it provides useful insights as to a reasonable outcome and the steps needed tu get there. Both parties should seek to obtain, assess, and share (to the e.lcnt feasible) certain important information. Such an approach may lead to an Individually and collectively optimal solution. Information should be sought and agreement reached (or as close as possible) with regard to three important parameters. The first is the projected benefits (mostly to Manufacturer) flowing from a consummated license. The necond is the disagreement profits (or next best alternatives) available to each party if a license is not consummated. The third is the objective bargaining strength of each party, with that driven primarily by each party's discount rate. In virtually all licensing Situations, there will be violations as to optimal (and consistent) ratioiudity, full (and consistent) information, equivalent (and canaistent) expectations, accurate (and consistent) views of next best alternatives, and reasonable (and consistent) views of bargaining power. Nonetheless, reducing those violations will result in movement toward an optimal license fee. That fee is one that gives each party to the license a fee equivalent to its benefits of not entering the license pius un equitable share of the net benefits of the bargain.

126 Case 3:11-cv RBD-TEM Document Filed 07/25/13 Page 27 of 28 PageID Cli, 1 Applkation of Game Theory to Intellectual Property Royalty Negotiations Endnotes \Vo are grateful to Ahhinay Muthoo, Christopher J3orek, and Joshua Fougere for their valuable insights nd assistance. Richard Peeks and Mark Pictrzykoski did a masterful job n preparing the document, The views and errors here are tise authors'. Gregory J. Battersby and Chattes W. Crimes, Licensing (loyally Rates 5 (Gaitherburg and Now York: Aspen Law & Business, 2002), 5. The focus of this panr wut be on the determination of a running royalty forro of compensation. That, howeser, cars be fairly easily converted into a lump-sum pay. ment, mitestone payments, or any number of hybrids. Shams P. liargreaves Heap and Y;snis Varcu rakis. Ganse Thea,3': A Critical 1,,iros/uclion J (London and New York: Routiedge, 2000), t. 5, I-Toward Raiffa, John Richardson, and David Metealt'e, Nepotititio,i AnalysLr: Th Science andflrt of Collaboralive flwrlun Makus 53CC smbndesnd London BLIkII.11) Press, 2002), 53. See, e.g., Robert GoIdcheidr, "Litigation Backgrounder for Licensing," Lt's Noutelks 29 (March 1994); 20; Robert Goidscheidcr unti Michael A. ITpstein. "Determinjog the Value of Technology and Setting Royalty Rates," Tise marisa! r'fl'roprh'tar-s' Rig/its 4 (Octoher 1992); Robert Goldsclicidcr, John Juruaz, and Carla S. Mulhem. "tise of the 25 Percent Rule in Valuing IP,"LerNusIs'L'IIL'v37 (December 2002;; 123; asid Richard Rsszgaitis, Valuation and Piking of Tcc/s:uthsgy-&ssed Iistelkcusrd Property (Hoboken, NJ: John Wiley & Sons tac 2003), Robert P. Reilly and Robert P. Schweih, Valuing Iniwigilik' Avons (New York: McGraw-Hill, 1999), 97. Ibid., 97, 121. Ibid., 120: "Ii is worth seiterasing that cost, price, und value arc three separate und distinct valuation concepts'. lt is naive and usually incorrect to :tsume tlusi 'cost equals price equals value.' In fact, the reverse i', often mure likely tu be true; that is. 'Cost does not equal price doc not equal value'." The 'pic" could be st specific rinsount of money ur any mistier benefit that is to he divided between tise negotiating parties. II. Morton D, Davis, Ganse 7'/iemny: A Nontechnical hstroduciii.'n (New York: Duvr Publications, Inc., 1983), 119. Ayinasls K. Dixit asid Barry J. Nalcbuff, T/sinLiisg Strategically: The Crmspdllris's' Edge in Business, Politics, ant! Everyday L{f'e (New York und London: V.V/. Nortisn & Co., 1991), 85. This, again, assumes rationality, i.e., spite is hot a factor. John F. Nash, "The Barg,sining Problem," Ecununzeirier, 18 (l95()j ; Julio F, Nash, "Two-Person Cooperative Games," Ecr'nansesrica 21 (1953): Ibid., 1953, 109, To beaccurutò, Equations (17.3) 10 (17,3) represent the bargaining outcome for a generalized or asymmetric Nash Bargaining Suiution, cisme that relaxes the fifth property above. See Abhinay Muthoo, JJar,çoining Thea, y sv'ith Applications (Cambridge: Cambridge Univeruity Press, 1999), 35-36, William Choi and Roy \Veinstein, "An Anulytìcal Solution Co Reasonable Rmsy;sky Rate Calculations," Idea: 1'he Journal ql Lj,m mid Technology 41 (July 2(11)1>): 49, 52. Choi and Weinstein's formulation implicitly mms,sunscs that price and quantity remain constant, Such an assumption, for reasons noted Isster, may he unrealistic. Muthoo (1999), 35.

127 Case 3:11-cv RBD-TEM Document Filed 07/25/13 Page 28 of 28 PageID About the Authors Iii the NI3S analysis.,pius çz,,is set equal ici J. Accordingly, we define a, a, and ArkI Rubintein. "Perfect quilibiiurn in u J3argaining Model," Econwnetrica 50 I 10-15(1. TIte discount rate used here is not equivtiienl tu u project's financial discount rate, i.e.. the rate ut which the holder uf un asset is indifferent between a lump sum paid today und the prospect uf receiving future cash flaws out into the future. rn fact, that nile should already have been accounted for in calculating total profits and each partys disagreement prurits. ABOUT THE AUTHORS JOHN C. JAROSZ, a Managing Principal at Analysis Group, Inc., specializes in applied microeconomics and industrial organization. He has given econcimic testimony, performed research, and provided strategy consultation in matters involving intellectual property, commercial damages, licensing, and antitrust. Mr. Jarnsz has significant expertise evaluating and testifying on damages in patent. copyright, trademark, trade secret, and unfair competition cases. l-k' is a member of the American Economic Association, the American Law and Economics Association, and the Licensing Executives Society. He has heen a columnist and advisory hoard member of The IP Litigator and is a frequent writer and lecturer on intellectual property and damages issues. DR. MICHAEL J. CHAPMAN. a Vice-President of Analysis Group, Inc., specializes in the application cl microeconomics and industrial organization to a wide variety tif litigation matters and complex business problems. He has performed research, prepared economic testimony, and provided assistance to attorneys in intellectual property. antitrust, antidumping, securities, and commercial dirnages matters. He is a member of the American Economic Association, the American Rar Association, and the Licensing Executives Society. Prior to joining Analysis Group. Inc., Dr. Chapman practiced as an International Trade Attorney at Shearman & Sterling and as a Strategy Specialist with McKin- & Company. He has also taught microeconomics courses at the Massachusells lnstilule of Technology Sloan School of Management, the University of Mkhigan I3usines School, and the University of Wisconsin-Madison Economks Depar!nu.'rìl.

128 Case 3:11-cv RBD-TEM Document Filed 07/25/13 Page 1 of 15 PageID EXHIBIT 11

129 Case 3:11-cv RBD-TEM Document Filed 07/25/13 Page 2 of 15 PageID 12042

130 Case 3:11-cv RBD-TEM Document Filed 07/25/13 Page 3 of 15 PageID 12043

131 Case 3:11-cv RBD-TEM Document Filed 07/25/13 Page 4 of 15 PageID 12044

132 Case 3:11-cv RBD-TEM Document Filed 07/25/13 Page 5 of 15 PageID 12045

133 Case 3:11-cv RBD-TEM Document Filed 07/25/13 Page 6 of 15 PageID 12046

134 Case 3:11-cv RBD-TEM Document Filed 07/25/13 Page 7 of 15 PageID 12047

135 Case 3:11-cv RBD-TEM Document Filed 07/25/13 Page 8 of 15 PageID 12048

136 Case 3:11-cv RBD-TEM Document Filed 07/25/13 Page 9 of 15 PageID 12049

137 Case 3:11-cv RBD-TEM Document Filed 07/25/13 Page 10 of 15 PageID 12050

138 Case 3:11-cv RBD-TEM Document Filed 07/25/13 Page 11 of 15 PageID 12051

139 Case 3:11-cv RBD-TEM Document Filed 07/25/13 Page 12 of 15 PageID 12052

140 Case 3:11-cv RBD-TEM Document Filed 07/25/13 Page 13 of 15 PageID 12053

141 Case 3:11-cv RBD-TEM Document Filed 07/25/13 Page 14 of 15 PageID 12054

142 Case 3:11-cv RBD-TEM Document Filed 07/25/13 Page 15 of 15 PageID 12055

143 Case 3:11-cv RBD-TEM Document Filed 07/25/13 Page 1 of 14 PageID EXHIBIT 12

144 Case 3:11-cv RBD-TEM Document Filed 07/25/13 Page 2 of 14 PageID Copyright (c) 2001 PTC Research Foundation of Franklin Pierce Law Center IDEA: The Journal of Law and Technology J.L. & TECH. 49 AN ANALYTICAL SOLUTION TO REASONABLE ROYALTY RATE CALCULATIONS WILLIAM CHOI AND ROY WEINSTEIN * * Mr. William Choi is a Director at Micronomics, Inc., an InteCap company, located in Los Angeles. He received a B.S. in 1993 from University of California-Riverside and a Ph.D. in 1999 from Duke University. Mr. Roy Weinstein is a Managing Director at Micronomics, Inc., an InteCap company, located in Los Angeles. He received a B.B.A. in 1964 from City College New York and an M.A. in 1967 from University of Chicago. This paper reflects the opinions of the authors and not those of InteCap, Inc. The concepts and theories covered by this presentation are not intended to be all-inclusive on the topic of reasonable royalties. The concepts are for illustrative purposes and may not represent approaches that the authors or InteCap would recommend in a particular matter. The reader should keep in mind that each case should be evaluated in light of its own facts and circumstances. SUMMARY:... This model can be used to supplement the often-used Georgia-Pacific factors to calculate a reasonable royalty rate in infringement cases.... The determination of a reasonable royalty rate for a licensing agreement that doesn't exist - and never existed - is a formidable task for licensing experts and triers of fact.... A Formal Analysis of a Reasonable Royalty Rate... Under these conditions, bargaining for a reasonable royalty rate will not exist unless at least one of two broad conditions exists: 1) the infringer/licensee is able to serve markets that the patent holder/licensor is unable to access; and/or 2) the infringer/licensee produces at lower marginal cost per unit than the patent holder/licensor.... In conclusion, the NBS results in an intuitively appealing reasonable royalty rate that reflects the economic conditions affecting the licensing agreement.... After obtaining the present value of the total profits and the total opportunity cost from the DCF method, the patent holder and infringer can then use the NBS to calculate a reasonable royalty rate.... By supplementing the Georgia-Pacific factors with the NBS as the template for a reasonable royalty rate calculation, reasonable royalty rate experts may now have an additional tool to use in constructing opinions regarding the profitability of the patented technology and the back-up alternatives of the parties in dispute....

145 Case 3:11-cv RBD-TEM Document Filed 07/25/13 Page 3 of 14 PageID TEXT: I. Abstract Today's courts are increasingly encouraging the use of more rigorous scientific approaches to royalty rate calculations. The technique proposed in this study applies a classic, peerreviewed game theory model to a hypothetical negotiation to yield an efficient and fair result. This model can be used to supplement the often-used Georgia-Pacific factors to calculate a reasonable royalty rate in infringement cases. n1 In the context of patent infringement litigation, this model will build on GeorgiaPacific by interpreting evidence and data in ways that reflect actual economic conditions affecting the outcome of the hypothetical negotiation. II. Introduction The determination of a reasonable royalty rate for a licensing agreement that doesn't exist - and never existed - is a formidable task for licensing experts and triers of fact. Since 1970, Georgia- Pacific has served as the conventional template for calculating reasonable royalty rates in such situations. n2 GeorgiaPacific established fifteen factors that can be considered as part of a "hypothetical negotiation" between a willing licensee and a willing licensor. These factors are used to consider what the parties would have contemplated if licensing had been pursued instead of infringement. n3 The timing of the "hypothetical negotiation" is the date when infringement began. n4 Misapplication of the Georgia-Pacific template can produce a royalty rate unsupported by economic theory. From our experience, it appears that licensing experts run down the list and identify some factors that support "high" royalty rates, while others identify those factors that support "low" royalty rates, whichever seems to benefit them most. When this happens, an unsound calculation, shrouded by "reliance" on the GeorgiaPacific factors, can occur. In Gasser Chair Co. v. Infanti Chair Manufacturing Corp., n5 the court explained that "it would be an affectation of research to cite the countless cases which simply reiterate the 'Georgia-Pacific' factors to be considered in determining a reasonable royalty.... To set out those fifteen factors would also needlessly burden this decision." n6 The testimony of licensing experts can be strengthened by a consideration of economic theory, rather than solely the identification of which factors in the Georgia-Pacific list support "high" rather than "low" royalty rates. We are not suggesting that the Georgia-Pacific factors be abandoned, as they provide a good reference and starting point. Instead, we suggest that licensing experts should also include two economic concepts that are often central in licensing agreements: 1) the anticipated profitability of the technology; and 2) the relative bargaining power of the participants. If data permits, these two concepts should be closely examined and used to supplement the Georgia- Pacific factors. Our attempt to narrow the focus on profitability and relative bargaining position is not novel to the discussion of reasonable royalty rate calculations. In Honeywell, Inc. v. Minolta Camera Co., n7 the court replaced the twelfth Georgia-Pacific factor with the anticipated profits and losses that the parties reasonably would have anticipated had they consummated a licensing agreement. n8 Additionally, the court cited relative bargaining position as an important factor. n9 Furthermore, the first two Georgia-Pacific factors relating to established royalties and other comparable agreements were omitted from the Honeywell analysis. n10 The Honeywell modification of the Georgia-Pacific

146 Case 3:11-cv RBD-TEM Document Filed 07/25/13 Page 4 of 14 PageID factors represents an important evolution in reasonable royalty rate calculations. n11 The Honeywell factors should better aid licensing experts in determining "commercially realistic" royalty rates. n12 In this paper, we suggest that the two-person bargaining game described by John Nash n13 accommodates an exclusive reliance on the anticipated profitability of a patented invention and the relative bargaining power of the parties in calculating a reasonable royalty rate. The Nash Bargaining Solution ("NBS") has been called the most fundamental model in bargaining theory. n14 The NBS looks for a sharp prediction of the bargaining outcome based on the bargaining strengths of each side. n15 The NBS is well supported by economic theory and is regarded as one of the simplest, yet most fruitful, paradigms in game theory. n16 The analytical clarity of the NBS is an important justification for its use as a tool in calculating a reasonable royalty rate. III. Nash Bargaining Solution Nash obtained his solution by developing a set of conditions that any reasonable solution must satisfy. n17 These conditions include: 1. Pareto efficiency < -> there should be no other feasible allocation that is better than the solution for one negotiator and not worse than the solution for the other negotiator. n18 2. Negotiators must collectively behave in a rational manner such that neither side gets less in the bargaining solution than could be obtained in disagreement. n19 3. The solution is independent of any numeric specification. n20 When constructing a twoperson bargaining problem, if the way profits are measured changes, the solution changes accordingly so it still corresponds to the same outcome. n21 4. Eliminating alternatives, other than the disagreement profits (i.e., the opportunity costs from licensing), that would not have been chosen, should not affect the solution. n22 5. If the disagreement profits of the two parties are equal in the bargaining problem, then the solution should also treat them equally. n23 Using an ingenious mathematical argument, Nash demonstrated that satisfying these conditions yields a unique solution; a solution where the bargaining outcome rests simply on each negotiator's alternative to negotiating and on the potential benefits of cooperation. n24 The NBS requires only knowledge or estimation of 1) the "disagreement" profits of both the licensee and licensor; and 2) the total profits from a licensing agreement. n25 Once these elements are determined, the NBS yields a unique and efficient compromise. To solve for the NBS, we must first identify the disagreement payoff for the patent holder/licensor, the disagreement payoff for the infringer/licensee, and the total potential profit from licensing. We define d1 as the disagreement payoff for the patent holder, where d1 represents the profit the patent holder/licensor expects to receive if the negotiation fails. Likewise, we define d2 as the disagreement payoff for the infringer/licensee, where d2 represents the profit the infringer/licensee expects to receive if negotiations fail. The exact functional form of these disagreement payoffs depends on specific assumptions about the two firms and the economic conditions present at the time of disagreement. The feasible payoff from licensing is represented by <capital pi>, where <capital pi> is the total profit from licensing. We also define

147 Case 3:11-cv RBD-TEM Document Filed 07/25/13 Page 5 of 14 PageID <pi>1 as the profit for the patent holder/licensor from licensing, and <pi>2 as the profit for the infringer/licensee from licensing. Nash demonstrated that the only point that satisfies the conditions outlined above is the one obtained by solving the following constrained maximization problem: [SEE EQUATION IN ORIGINAL] (1) subject to the following conditions: [SEE EQUATION IN ORIGINAL] (2) [SEE EQUATION IN ORIGINAL] (3) [SEE EQUATION IN ORIGINAL] (4) When transfer payments are permitted between the licensor and the licensee, the bargaining problem can be fully characterized by three factors: 1) the disagreement payoff for the patent holder/licensor; 2) the disagreement payoff for the infringer/licensee; and 3) the total transferable wealth available to the two firms from licensing. n26 Thus, the conditions for the equilibrium payoffs are: [SEE EQUATION IN ORIGINAL] (5) [SEE EQUATION IN ORIGINAL] (6) where and are the equilibrium payoffs for the licensor and licensee, respectively. Solving equations (5) and (6) yields the NBS: [SEE EQUATION IN ORIGINAL] (7) [SEE EQUATION IN ORIGINAL] (8) [SEE EQUATION IN ORIGINAL] (9) Equations (7) and (8) have the following interpretation: when the entities bargain over the partition of total profits (<capital pi>), they first agree to give each other the payment that they would respectively obtain from not reaching agreement and they next split the remaining profits equally. n27 For each firm, the agreement payoff is greater when its own disagreement point is higher than its opponent's disagreement point. Therefore, the relative bargaining power will depend on the outside opportunities available to each side. The fundamental insight of the NBS is that the alternatives to agreement that are available to each side will limit how good a bargain the opposing party can obtain in the negotiation. n28 These alternatives also set a lower limit on the profit each side will accept. n29 Under the NBS, the two parties will divide the bargaining surplus - bounded by each bargainer's threat point or minimum acceptable profit - down the middle, so that each has an equal share. n30 An alternative way of thinking about the NBS is in the framework of "an (implicit) arbitrator who tries to distribute the gains from trade or, more generally, from cooperation in a manner that reflects 'fairly' the bargaining strength" of the two negotiators. n31 Once each side's disagreement payoffs are determined, an arbitrator would apply the NBS to obtain an efficient and fair solution. n32 In the following section, we apply the NBS to the calculation of a reasonable royalty rate. IV. A Formal Analysis of a Reasonable Royalty Rate

148 Case 3:11-cv RBD-TEM Document Filed 07/25/13 Page 6 of 14 PageID A reasonable royalty rate is one "that a licensee would be willing to pay the inventor while still making a reasonable profit from use of the patented invention." n33 Many possibilities exist that can affect the relative bargaining positions between a patent holder/licensor and an infringer/licensee. Other things being equal, if the licensor has suitable alternative licensees, it can threaten to leave the bargaining table and deal with an alternative licensee to obtain the best deal. Likewise, if there are few available substitute technologies, the licensee has fewer outside opportunities and will correspondingly fare worse in the negotiation. We start with a simple case of a non-producing firm that owns a patent, with no substitutes, and only one licensee capable of producing the technology. We will later expand the model by introducing different assumptions about the firms to illustrate how they affect the solution. A. Case 1: One-Supplier World The simplest case is that of a research and development firm (licensor) that is incapable of manufacturing any product embodying its invention. Such a firm can earn profits only through licensing. Furthermore, assume only one company (infringer/licensee) has the production capabilities to exploit the licensor's invention. The NBS can determine how much the licensee can expect to pay in royalties. Since the licensor earns nothing without the licensee, the licensor's bargaining position ultimately depends on the licensee's outside alternatives. If negotiations break down, and the licensee has viable alternatives to the licensor, the licensee has an opportunity to earn profits equal to its opportunity cost (i.e., profits lost to disagreement with the licensor). If the negotiation is successful, the joint profit from licensing is equal to the monopoly profit gained from commercializing the patented invention. or: The NBS set-up and solution is straightforward. The licensor's disagreement payoff, d1, is zero, [SEE EQUATION IN ORIGINAL] (10) The licensee's disagreement payoff, d2, is equal to the licensee's opportunity cost, where the licensee's opportunity cost is the loss of return from not manufacturing the invention. Finally, the joint profits from licensing are equal to the monopoly profit: [SEE EQUATION IN ORIGINAL] (11) where C2(Qm) represents the licensee's cost function, Pm represents price, Qm represents quantity, and the subscript m refers to a monopoly. Additionally, PmQm represents the profit maximizing price and quantity for the licensing agreement in a monopoly market. Applying equations (7) through (9), the NBS for a licensing agreement for the licensor and licensee, respectively, are: [SEE EQUATION IN ORIGINAL] (12) [SEE EQUATION IN ORIGINAL] (13) [SEE EQUATION IN ORIGINAL] (14) To solve for the per-unit royalty, equations (12) and (13) can be rewritten as: [SEE EQUATION IN ORIGINAL] (15) [SEE EQUATION IN ORIGINAL] (16)

149 Case 3:11-cv RBD-TEM Document Filed 07/25/13 Page 7 of 14 PageID where r represents the per-unit royalty. Solving for r yields the following formula for a reasonable royalty rate: [SEE EQUATION IN ORIGINAL] (17) where AC2 is the licensee's average total cost. The first part of equation (17) stipulates that the royalty rate should be established at one-half of the difference between the price and the licensee's average total cost. Thus, the greater the mark-up of the patented technology, the greater the royalty rate for the patented technology. The second part of equation (17) demonstrates that the royalty rate will decrease with the licensee's opportunity cost. In other words, as the licensee's next best alternative becomes more lucrative, the royalty rate paid to the licensor will decrease. B. Case 2: Two-Supplier World An alternative patent infringement scenario concerns two firms - the patent holder and the infringer - where the patent holder possesses production capabilities, but has not initiated production at the time of infringement. Under these conditions, bargaining for a reasonable royalty rate will not exist unless at least one of two broad conditions exists: 1) the infringer/licensee is able to serve markets that the patent holder/licensor is unable to access; and/or 2) the infringer/licensee produces at lower marginal cost per unit than the patent holder/licensor. Without at least one of these conditions, the patent holder has no incentive to license its technology. One scenario where these conditions exist is when the inventor does not possess a comparative advantage in production or sales (i.e., when licensees have access to better distribution facilities, sales staff, or marketing resources). Assume in this case, therefore, that: 1) the licensee can produce at lower costs; and 2) it is in the patent holder's best interest to license the entire market and withdraw from production itself. Under these conditions, the disagreement payoff for the patent holder is the profit it can earn as the high-cost, sole producer of its patented product. The patent holder's disagreement payoff, d1, is represented by: [SEE EQUATION IN ORIGINAL] (18) where C1(Q1) is the patent holder's cost function, P1 is the profitmaximizing price for the patent holder absent the infringer, and Q1 is the profit-maximizing quantity for the patent holder absent the infringer. The disagreement payoff for the licensee remains d2, as explained previously. The joint profit from licensing is identical to the onesupplier case: [SEE EQUATION IN ORIGINAL] (19) where C2(Qm) represents the licensee's cost function, Pm represents price, Qm represents quantity, and the subscript m refers to a monopoly. Additionally, PmQm represents the profit maximizing price and quantity for the licensing agreement in a monopoly market. It is assumed that <capital pi> > d1 and that [SEE EQUATION IN ORIGINAL]. The NBS payoff for the licensor, <pi>*2, is: [SEE EQUATION IN ORIGINAL] (20) The NBS payoff for the licensee, <capital pi>, is:

150 Case 3:11-cv RBD-TEM Document Filed 07/25/13 Page 8 of 14 PageID [SEE EQUATION IN ORIGINAL] (21) The total profit from licensing, <capital pi>, is: [SEE EQUATION IN ORIGINAL] (22) The reasonable royalty, r, is equal to: [SEE EQUATION IN ORIGINAL] (23) Equation (23) provides the general framework for calculating the reasonable royalty rate in this situation. The first part of equation (23) is identical to the first part of equation (17) because the royalty rate increases with the mark-up of the patented technology. The second part of equation (23) factors in the relative bargaining positions of the parties. If both sides have equal disagreement payoffs, then additional profits achieved from licensing are split equally. The royalty rate, however, correspondingly changes with differences in a party's relative disagreement point or bargaining position. Therefore, as one party's outside alternatives improve, the terms of the licensing agreement become more favorable to the party having those outside alternatives to licensing. C. Alternative Cases The solution obtained in equation (23) provides a clear and efficient method of determining a reasonable royalty rate. Additionally, it can be adapted to various situations that may affect the "hypothetical negotiation." For instance, if viable and non-infringing substitutes exist for the patented product, then the elasticity of demand for the patented product is larger, and the patent's market power and profitability both decrease. When substitute goods exist in the marketplace, profits are reduced, as is the difference between price and average total cost in equation (23) [Pm-AC2], thereby leading to a lower reasonable royalty rate. When the profitability of a patented technology is lower, the royalties the patent holder can charge in the licensing agreement are also lower. The existence of substitute products will also lower d1, the patent holder's disagreement payoff, which further decreases the royalty rate the patent holder can command. This result is consistent with, and quantifies, Culbertson and Weinstein's conclusion that a reasonable royalty rate "depends fundamentally upon the extent and nature of substitute products for the patented product." n34 In conclusion, the NBS results in an intuitively appealing reasonable royalty rate that reflects the economic conditions affecting the licensing agreement. By analyzing the total potential profit for licensing and the disagreement payoffs of both parties, the methodology of the NBS provides a clear way to quantify the fair value of a patent to the patent holder/licensor and the infringer/licensee. V. NBS and the DCF Method Thus far, the analysis has focused on a static situation. Although this provides an intuition for calculating a reasonable royalty rate, the analysis does not address the fact that the underlying value of a patent is based on the present value of future economic benefits. Some factors that can limit these benefits include the invention's market potential, the sensitivity of future profits to future production costs, and the period of time benefits will be enjoyed.

151 Case 3:11-cv RBD-TEM Document Filed 07/25/13 Page 9 of 14 PageID A popular choice for calculating future economic benefits is the Discounted Cash Flow ("DCF") method. The objective of the DCF method is to discount, into present value, the expected cash flow from a licensing agreement; then discount, into present value, the expected cash flows for the patent holder and the infringer in the absence of a licensing agreement. It is advantageous to use the DCF method because direct comparisons can be made between total profits and total opportunity costs because the present values of each are measured in today's dollars. After obtaining the present value of the total profits and the total opportunity cost from the DCF method, the patent holder and infringer can then use the NBS to calculate a reasonable royalty rate. The application of the DCF method first requires an estimate of the net cash flows resulting from licensing a patented technology. To estimate these future net cash flows, two things must be known: (1) the specific time period over which the cash flows will be evaluated; and (2) the investment risk to be incurred over that time period. The DCF method covers the time period from the point in time when infringement began until the point in time when the patent expires. The determination of a reasonable royalty also incorporates a fair return on the amount of investment risk accepted in developing the patented technology. This determination of investment risk should consider advancing technology, competing technology, and government regulations that pertain to the patented technology. The weighted average cost of capital ("WACC") can be used to discount cash flow in accordance with these considerations. n35 In general, cost of capital is the cost, measured as a percentage rate, of the various sources of capital required to finance investment, such as intellectual property. n36 A company's cost of capital will be the WACC of each type of capital. n37 Accordingly, WACC is the amount of return that that company must earn on its overall investment as comprised of the monetary assets, tangible assets, intangible assets, and contracts. n38 This expected return is considered a hurdle rate for capital investments. n39 The Capital Asset Pricing Model ("CAPM") can also be used to derive an appropriate return on investments. n40 The CAPM provides a framework for measuring market risk and the premium that investors demand in return for assuming such risk. n41 CAPM can be used to estimate the required rate of return for specific intellectual property by analyzing the required rates of return demanded by investors for stocks that operate within the same industry. n42 Applying the DCF to the NBS to obtain a reasonable royalty rate is straightforward. Equation (23) must be slightly modified to reflect future time periods and appropriate risks specific to the firms and the patented technology: [SEE EQUATION IN ORIGINAL] (24) where the variables <delta>1 and <delta>2 represent each firm's WACC, and <delta>m reflects the risks associated with the patented technology itself. Equation (24) indicates that each firm's disagreement payoffs over time, d1t, and d2t, must be discounted by each firm's WACC. The DCF method requires sufficient information about the estimated cash flows during the relevant period. In practice, the application of the DCF method requires accurate manufacturing, research, and marketing estimates, as close to the time of infringement as possible. Estimates of market size and realistic penetration are also required. Additional estimates concerning investment requirements for additional types and amounts of manufacturing facilities, as well as costs associated with designs and marketing, would be helpful.

152 Case 3:11-cv RBD-TEM Document Filed 07/25/13 Page 10 of 14 PageID The principal problem associated with implementing the DCF method is the reliability of the data estimate inputs. From our experience, many internal financial projections, particularly those used to obtain financing, are reasonable sources of data. The analysis can be assisted through discovery concerning the parties' records and opinions at the time of infringement. Discovery can reveal financial projections, memoranda, marked research, and competitive analysis. A cautionary note: not every forecast is reliable. Carefully discerning when the projections were made and what methods were undertaken should ensure the reliability of such reports. Independent market analyst reports produced by investment banks may also provide market projections that can supplement these internal market forecasts. VI. Conclusion The necessity to provide objective and sound determinations of reasonable royalty rates in patent infringement litigation provides reason to supplement the Georgia-Pacific factors with the approach outlined in this article. Data permitting, John Nash's two-person bargaining game represents a peerreviewed methodology that can be used to calculate a reasonable royalty rate from a hypothetical negotiation. The theoretical support for the NBS is overwhelming and, in the context of patent litigation, the reasonable royalty rate solution derived from the NBS is fair, efficient, and sensible. The method of assigning weights to the Georgia-Pacific factors may produce a result that can be significantly improved and refined by the use of the NBS. Given the requirement that the parties conduct a "hypothetical negotiation" and agree to a hypothetical royalty rate, such a result is not surprising. By supplementing the Georgia-Pacific factors with the NBS as the template for a reasonable royalty rate calculation, reasonable royalty rate experts may now have an additional tool to use in constructing opinions regarding the profitability of the patented technology and the backup alternatives of the parties in dispute. The NBS technique may thus contribute to improving patent infringement litigation fact-finding and damages calculations in the future. FOOTNOTES: n1 Georgia-Pacific Corp. v. United States Plywood Corp., 318 F. Supp. 1116, 166 U.S.P.Q. (BNA) 235 (S.D.N.Y. 1970). The Georgia-Pacific factors are: 1. The royalties received by the patentee for the licensing of the patent in suit, proving or tending to prove an established royalty. 2. The rates paid by the licensee for the use of other patents comparable to the patent in suit. 3. The nature and scope of the license, as exclusive or non-exclusive; or as restricted or nonrestricted in terms of territory or with respect to whom the manufactured product may be sold. 4. The licensor's established policy and marketing program to maintain his patent monopoly by not licensing others to use the invention or by granting licenses under special conditions designed to preserve that monopoly. 5. The commercial relationship between the licensor and licensee, such as, whether they are competitors in the same territory in the same line of business; or whether they are inventor and promoter.

153 Case 3:11-cv RBD-TEM Document Filed 07/25/13 Page 11 of 14 PageID The effect of selling the patented specialty in promoting sales of other products of the licensee; that existing value of the invention to the licensor as a generator of sales of his nonpatented items; and the extent of such derivative or convoyed sales. 7. The duration of the patent and the term of the license. 8. The established profitability of the product made under the patent; its commercial success; and its current popularity. 9. The utility and advantages of the patent property over the old modes or devices, if any, that had been used for working out similar results. 10. The nature of the patented invention; the character of the commercial embodiment of it as owned and produced by the licensor; and the benefits to those who have used the invention. 11. The extent to which the infringer has made use of the invention; and any evidence probative of the value of that use. 12. The portion of the profit or of the selling price that may be customary in the particular business or in comparable businesses to allow for the use of the invention or analogous inventions. 13. The portion of the realizable profit that should be credited to the invention as distinguished from nonpatented elements, the manufacturing process, business risks, or significant features or improvements added by the infringer. 14. The opinion testimony of qualified experts. 15. The amount that a licensor (such as the patentee) and a licensee (such as the infringer) would have agree upon (at the time the infringement began) if both had been reasonably and voluntarily trying to reach an agreement; that is, the amount which a prudent licensee -- who desired, as a business proposition, to obtain a license to manufacture and sell a particular article embodying the patented invention -- would have been willing to pay as a royalty and yet be able to make a profit and which amount would have been acceptable by a prudent patentee who was willing to grant a license. Id. at 1120, 166 U.S.P.Q. at 238. n2 See, e.g., Joy Technologies, Inc. v. Flakt, Inc., 954 F. Supp. 796, , 42 U.S.P.Q.2d (BNA) 1042, (D. Del. 1996), Trio Process Corp. v. L. Goldstein's Sons, Inc., 612 F.2d 1353, 1357, 204 U.S.P.Q. (BNA) 881, 883 (3d Cir. 1980). See also Robert Goldscheider, The Employment of Licensing Expertise in the Arena of Intellectual Property Litigation, 36 IDEA 159, 170 (1996). n3 See Georgia-Pacific, 318 F. Supp. at 1120, 166 U.S.P.Q. at 238 (inferring from the fifteenth Georgia-Pacific factor). n4 See id. at 1125, 166 U.S.P.Q. at 242. n5 943 F. Supp. 201, 40 U.S.P.Q.2d (BNA) 1554 (E.D.N.Y. 1996). n6 Id. at 216, 40 U.S.P.Q.2d at 1566 (implying that it was unnecessary to list the well-known Georgia-Pacific factors in the opinion even though they were being applied by the court) (citations omitted).

154 Case 3:11-cv RBD-TEM Document Filed 07/25/13 Page 12 of 14 PageID n7 Nos. CIV.A , , jury instruction at 69 (D.N.J. Jan. 28, 1992). n8 Id. n9 See id. n10 See id. See also Goldscheider, supra note 3, at 170. n11 See Goldscheider, supra note 3, at 171. n12 Id. at 171. n13 See generally John Nash, The Bargaining Problem, 18 Econometrica 155 (1950) [hereinafter referred to as Nash 1]; John Nash, Two-Person Cooperative Games, 21 Econometrica 128 (1953) [hereinafter referred to as Nash 2]. n14 See Alvin Roth, Axiomatic Models of Bargaining 1, 1 (M. Beckmann & H. P. Kunzi eds., 1979). n15 See generally Nash 1, supra note 13; Nash 2, supra note 13. n16 See Roth, supra note 14. n17 See Nash 2, supra note 17, at n18 See id. See also Roger B. Myerson, Game Theory: Analysis of Conflict (1991) (explaining axiom 8.1). n19 See id. (explaining axiom 8.2). n20 More formally, the solution is independent of any numeric risk-neutral utility specification. n21 See id. (explaining axiom 8.3). n22 See id. (explaining axiom 8.4). n23 See id. (explaining axiom 8.5). n24 See Myerson, supra note 22, at 375. n25 See id. at n26 See Myerson, supra note 22, at 375. Transfer payment is an important assumption that can guarantee the given scale factors in a game will also be the natural scale factors for the NBS. Risk neutrality is also an important assumption when we use transfer payments; however, in the context of firms negotiating over an agreement, the assumption is plausible. n27 See Abhinay Muthoo, Bargaining Theory with Applications (1999). n28 See id. at 105. n29 See id. n30 See id. at n31 Andrew Mas-Colell et al., Microeconomic Theory 838 (Oxford University Press 1995). n32 See id. n33 See Black's Law Dictionary 1330 (7th ed. 1999).

155 Case 3:11-cv RBD-TEM Document Filed 07/25/13 Page 13 of 14 PageID n34 John D. Culbertson and Roy Weinstein, Product Substitutes and the Calculation of Patent Damages, 70 J. Pat. & Trademark Off. Soc'y 749, 750 (1988). n35 See Gordon V. Smith & Russell L. Parr, Valuation of Intellectual Property and Intangible Assets (John Wiley & Sons 3d ed. 2000) (providing a detailed application of WACC). n36 See id. n37 See id. n38 See id. n39 See id. n40 See id. at n41 See id. n42 See id. at 553.

156 Case 3:11-cv RBD-TEM Document Filed 07/25/13 Page 14 of 14 PageID 12069

157 Case 3:11-cv RBD-TEM Document Filed 07/25/13 Page 1 of 9 PageID EXHIBIT 13

158 Bargaining and the Construction of Economically Consistent Hypothetical License Negotiations Jonathan D. Putnam and Andrew B. Tepperman Jonathan D. Putnam is Assistant Professor of the Law and Economics of Intellectual Property at the Centre for Innovation Law and Policy, Faculty of Law, University of Toronto, and a Vice-President at Charles River Associates, an economic consulting firm based in Boston, MA. Professor Putnam holds a PhD in economics from Yale University. (Contact: Faculty of Law, University of Toronto, 84 Queen s Park, Toronto, ON, M5S 2C5, tel: (416) , Jon.Putnam@utoronto.ca.) Andrew B. Tepperman is a Senior Associate in the Intellectual Property practice of Charles River Associates, based in Toronto. Dr. Tepperman holds a PhD in economics from the University of Toronto. (Contact: Charles River Associates, 80 Bloor Street West, Suite 1501, Toronto, ON, M5S 2V1, tel: (416) , fax: (416) , atepperman@crai.com.) Determining the appropriate royalty in a technology licensing negotiation requires technical inputs to an economic problem. The technical aspect arises because the correct royalty depends on the ways in which the technology trades off physical possibilities and constraints, the means by which it interacts with the remainder of the production process or product features, and the dimensions in which its performance is superior or inferior to alternative technologies. The fundamental problem is an economic one because, in a market setting, these technical parameters implicitly define unobserved prices (economists call them shadow prices ) corresponding to the costs and benefits of using the technology. Conducted by informed parties, an appropriate arm s length negotiation brings these component prices out of the shadows by placing an agreed-on, aggregate value on the technology, as well as creating a value-maximizing structure for the contract. In most cases the market (i.e., knowledgeable business people seeking to maximize profits) sets the value of a patented technology through arm s-length bargaining. But in some cases, parties cannot agree on the value of the technology, whether it is necessary for a particular purpose, or even what it is. These cases result in patent infringement litigation. We are thus interested in the special case that often arises in such litigation: Setting a reasonable royalty for infringing sales after the fact, via a hypothetical negotiation between the parties. The purpose of this article is to explore the economics of bargaining and its relation to royalty determination for damages purposes. We argue that economic models of the bargaining process can help shed some light on the difficult problem of valuing a technology when a market-determined value does not already exist. We acknowledge that the economics of bargaining is not as well developed as some other areas of microeconomics, and we are cautious not to draw too sharp conclusions from the economic literature as it currently stands. Nonetheless, we suggest that one can usefully apply certain lessons from economic theory in real situations, and in many cases a little structure from economics goes a long way to help solve the analytical problem. We make this argument by first reviewing some of the results from the economic literature on the subject, and then examining certain key patent decisions to guide our application of the theory. Next, we discuss some of the methods by which courts and practitioners address the negotiation of a hypothetical license, which are suggestive of a variety of approaches, not all of which are economically sound. We thus make two general points: (1) economics can help in setting out a consistent approach for use in practical problems, and (2) it can be used to illuminate some of the arguments used in prior cases. The Economic Treatment of Bargaining Bargaining Solutions Economic models of bargaining situations typically strip away most of the idiosyncratic features that are present in any individual bargaining context, and focus only on the general factors that determine the parties payoffs. In broad terms, these factors can be placed into two categories: those that affect the joint payoff from cooperation, and those that affect the individual payoffs if the parties fail to cooperate. The joint payoff is referred to as the bargaining surplus, while the 8 The Licensing Journal AUGUST 2004

159 latter are variously termed threat points, disagreement payoffs, or outside options. Economically speaking, a bargain is nothing more than an agreed split of the bargaining surplus that is consistent with (and superior to) each party s individual disagreement payoff. Economists have recognized that in many cases bargaining problems may be hard or even impossible to solve, especially when each side has information that is not available to the other. 1 However, in the special case in which both sides to the transaction have complete information, 2 economists have derived some useful ways to attack the question. To take a simple example of the type of problem we have in mind, imagine a licensing negotiation in which the licensor owns a patent and has no production facilities (nor any realistic prospects for obtaining them), and a potential licensee has the necessary capital with which to build production facilities. If the parties agree on a licensing contract the patent in question can produce a product that will generate $10 million in profits. In the baseline case, suppose that the licensee has no alternative production plans, which we take to imply zero profits. Both sides are aware of the market value from cooperating, as well as the value to the licensee of not producing. Then, all other things equal, it is reasonable to expect the parties to split the $10 million surplus 50/50, with the licensee paying the licensor a $5 million royalty. The standard economic models predict this outcome as well. Suppose instead that the potential licensee has access to an inferior (and non-infringing) inhouse technology that will generate just $4 million in profits. Again, both sides are aware of the value of the two options. What agreement will the parties reach? Intuitively, one expects that the availability of the alternative technology enhances the licensee s bargaining power, reduces his willingness to pay, and should result in a lower license payment. 3 Similarly, if the prospective licensee s alternative generated $8 million in profits rather than $4 million, one would expect his willingness to pay to fall further. The basic models used by economists to solve this type of bargaining game reproduce this intuition: An increase in the value of the disagreement outcome for one party leads to a higher payoff for that party. The first and most commonly used model is called the Nash bargaining solution. 4 Under the Nash solution each party receives his disagreement payoff plus one-half of the gains from trade, which is simply the bargaining surplus net of the disagreement payoff. 5 In our initial example, when the licensee possesses a low-value alternative technology, the payoffs to the licensor and licensee are easy to characterize under Nash bargaining: the licensor should receive a payoff (in the form of a royalty payment) of 0 + 1/2[10 4] = $3 million, while the licensee should receive a payoff (in the form of net profits) of 4 + 1/2[10 4] = $7 million. The Nash solution embodies the intuitive property that a party s payoff from reaching an agreement should increase with the value of his outside alternative to the agreement. If the licensee s alternative technology instead produced $8 million in profits, the payoffs would be a royalty payment of 0 + 1/2[10 8] = $1 million to the licensor, with 8 + 1/2[10 8] = $9 million remaining for the licensee. 6 The second principal solution concept for bargaining situations is called the outside option model. Loosely speaking, while the Nash model interprets disagreement payoffs as preserving the status quo, the outside option model treats the parties disagreement payoffs as constraints on moving away from the status quo, in a sense that we make more precise below. For now, we note that the choice of bargaining solution has a real impact on the payoffs that the bargaining parties can be expected to receive. In our benchmark model of licensing, in which the in-house technology yields a profit of $4 million, the outside option model predicts that the parties will simply split the total gross surplus 50/50, as the licensee s value from doing so exceeds the value of the outside option. Thus each party receives a payoff of $5 million, which clearly is also the royalty paid by the licensee. In such a case it is sometimes said that the licensee s outside option is not binding, because, being less valuable than a simple split of the gross surplus, it does not constrain the outcome. On the other hand, when the in-house technology is more valuable ($8 million), this constrains the payoffs that the parties will accept. Here, the outside option model predicts that the licensor and licensee will agree on a payoff that leaves the licensee as well off as if he had selected his outside option. This outcome implies a payoff to the licensee of $8 million, and a $2 million royalty payment to the licensor. To summarize, we observe an important difference between the two models in the way changes in disagreement payoffs are built into the bargaining solutions. A low disagreement payoff (i.e., less than the payoff from a 50/50 split) affects the Nash payoffs, but not the outside option payoffs. A high disagreement payoff constrains a party s payoff in the outside option framework and completely determines that payoff. In this latter case we observe that the royalty payment in our simple licensing example is higher when using the outside option concept than from the Nash solution; in fact, it is twice as high. The reason is that the Nash payoff is always equal to the disagreement payoff plus one-half of the gains from trade; under the outside option model, the licensor captures all of the remaining gains from trade. In our simple AUGUST 2004 The Licensing Journal 9

160 example, the licensee is always better off, or no worse off, under the Nash model than under the outside option model. Exhibit 1 summarizes the payoff to the licensee and the royalty paid to the licensor under the two models, in each of three cases: when the licensee has no alternative, when the alternative has relatively low value, and when the alternative has a high value. Exhibit 1 Division of Surplus under Nash and Outside Option Models (Payoff to Licensee, Royalty to Licensor) Gross surplus = 10 Bargaining Model Value of licensee s alternative technology Nash Outside option 0 (5, 5) (5, 5) 4 (7, 3) (5, 5) 8 (9, 1) (8, 2) Choosing the Appropriate Solution Having reviewed the consequences of using one bargaining solution or the other, we are left to ask: Which is most appropriate to use in practical situations? It turns out that the choice of model depends on the particulars of the situation to be analyzed. Recent research in game theory has shown that both bargaining solutions are special cases of a more general bargaining model in which parties alternate in making offers to each other, and then accept or reject these offers. 7 Interestingly, the Nash solution is the natural outcome of the bargaining model in which the parties can earn their disagreement payoffs before they reach an agreement, while bargaining continues in the hope of permanent agreement. 8 For example, during a labor dispute the owner of a sports league may threaten to use (and may use) replacement players as a temporary measure to enhance its bargaining power. In contrast, the outside option solution arises when taking up the alternative payoff ends bargaining altogether. 9 An example would be a one-time negotiation over the price for one batch of goods, when the buyer has alternative suppliers. The implication is that the proper solution depends on the application at hand. These contrasting examples suggest some general observations that relate to licensing negotiations. First, in a simple situation in which a licensor and licensee are bargaining over a royalty payment for the use of a patent in a one-off production run, it may be more realistic to find the appropriate payment using the outside option solution. This is because by construction, taking up the alternative would render the patented technology irrelevant. Second in the (likely more common) situation when production and sale occur over an extended period of time, one can imagine circumstances under which either model would be appropriate. The outside option model might suit cases in which taking up the alternative technology would lead to a long-term dispute over whether it is infringing or not, and thus to a cessation of meaningful licensing negotiations for a long period, or to a situation in which the licensor is choosing from among exclusive licensees. On the other hand, if the alternative can be used as a stopgap measure simply to tide the licensee over until an agreement is reached, or a better technological alternative can be developed, then the Nash solution may be more accurate. Third, although we might intuitively expect that the licensee s ability to terminate bargaining permanently by taking up an alternative technology would tend to increase its bargaining power, the opposite is true; in general, the licensee s royalty payment is higher under the outside option model than under the Nash model. Choosing the model that properly suits the situation thus has a meaningful impact. These remarks concern actual licensing negotiations. Conversely, we ask how to characterize the hypothetical licensing negotiation contemplated in patent damages cases. The Hypothetical License Royalty Determination in Patent Damages Cases In a successful patent infringement case, the defendant must pay damages adequate to compensate for the infringement, but in no event less than a reasonable royalty for the use made of the invention by the infringer. 10 When the patentee proves that infringement caused lost sales, forced a lower price, or increased costs, he may recover lost profits. If the patentee cannot meet the burden of proof to recover lost profits, then damages take the form of a reasonable royalty. 11 In determining this hypothetical royalty, it is important to contrast the intermediate division of the gains from the use of the invention, which evidently was what Congress intended, with the more onesided divisions that Congress has imposed in other intellectual property contexts, and that generally have been rejected by the courts in the patent context. For example, the copyright statute awards actual damages plus any profits of the infringer that are attributable to the infringement and are not taken into account in computing the actual damages. 12 This procedure awards the surplus to the copyright holder. On the other hand, the courts have also held that: The setting of a reasonable royalty after infringement cannot be treated... as the equivalent of ordinary royalty negotiations among truly willing patent owners and licensees. That view would constitute a pre- 10 The Licensing Journal AUGUST 2004

161 tense that the infringement never happened. It would also make an election to infringe a handy means for competitors to impose a compulsory license policy upon every patent owner. 13 Permitting the infringer to impose a compulsory license and to pay only what he might have had to pay in the course of ordinary royalty negotiations awards 100 percent of the gains over and above the compulsory license fee to the infringer. The patent statute thus contemplates an intermediate sharing of the surplus created by the infringer. The key question is: In the absence of an established royalty, how are damages to be calculated? The landmark case in reasonable royalty determination contemplates a hypothetical negotiation between patentee and infringer as a method of arriving at a figure: The amount that a licensor (such as the patentee) and a licensee (such as the infringer) would have agreed upon (at the time the infringement began) if both had been reasonably and voluntarily trying to reach an agreement; that is, the amount which a prudent licensee who desired, as a business proposition, to obtain a license to manufacture and sell a particular article embodying the patented invention would have been willing to pay as a royalty and yet be able to make a reasonable profit and which amount would have been acceptable by a prudent patentee who was willing to grant a license. 14 This framing of the problem suggests that, to the extent possible, the court should approach the hypothetical negotiation as an attempt to replicate a market situation where none existed. The opinion by the Georgia-Pacific court reinforces this view: Where a willing licensor and a willing licensee are negotiating for a royalty, the hypothetical negotiations would not occur in a vacuum of pure logic. They would involve a market place confrontation of the parties, the outcome of which would depend upon such factors as their relative bargaining strength; the anticipated amount of profits that the prospective licensor reasonably thinks he would lose as a result of licensing the patent as compared to the anticipated royalty income; the anticipated amount of net profits that the prospective licensee reasonably thinks he will make; the commercial past performance of the invention in terms of public acceptance and profits; the market to be tapped; and any other economic factor that normally prudent businessmen would, under similar circumstances, take into consideration in negotiating the hypothetical license. 15 Certainly, this passage indicates that economic determinants of the overall bargaining surplus (such as the anticipated amount of net profits and the market to be tapped ) are germane. The reference to bargaining strength hints at the relevance of other economic alternatives, which we have previously discussed. The Georgia-Pacific court pointed to the anticipated amount of profits that the prospective licensor reasonably thinks he would lose as a result of licensing the patent, which characterizes the patentee s disagreement point in analyzing the hypothetical bargain. We would argue that the court also sought to take the infringer s alternatives into account by allowing for such factors as the net profits that the prospective licensee reasonably thinks he will make, and the utility and advantages of the patent property over the old modes or devices. 16 Building on our economic discussion, we thus find three aspects of the Georgia-Pacific hypothetical negotiation of interest. First, the parties must reach agreement on the eve of infringement, though they did not in fact. Thus, opting out of the hypothetical negotiation is not possible. Second, the patentee s alternatives, and particularly his anticipated lost profits, are to be explicitly considered in arriving at a reasonable royalty. (We return to the appropriate means of dividing compensation between lost profits and a reasonable royalty subsequently.) Third, the infringer s alternatives, and thus his disagreement payoff, are definitely relevant from an economic perspective under any plausible negotiating scenario, and are also considered among the Georgia-Pacific factors. 17 The next sections discuss these three aspects in turn. Mandatory Agreement and the Appropriate Bargaining Solution With respect to mandatory agreement, we note that there is an interesting tension in the legal standard. This standard approximates the market value of the technology on the eve of infringement. The licensor and licensee are assumed to be willing to enter a prudent (i.e., market-consistent) agreement, and moreover are not under any special compulsion to come to terms. 18 But by finding infringement of a valid patent, the court has imputed to the infringer an irrevocable commitment to the patented technology, and to the patentee an irrevocable commitment to license. Constraining the parties to reach agreement means neither can threaten to take up a permanent alternative. In sum, though it is appropriate to account for the parties alternatives (because these reflect real factors that help determine the payment the parties would have arrived at), the treatment of alternatives must be consistent with the constraint to reach an agreement. The Nash bargaining solution is thus the more appropriate of our two solutions; it AUGUST 2004 The Licensing Journal 11

162 allows for alternatives to affect payoffs but does not allow for a permanent breakdown in negotiations. Applying the Nash solution is also consistent with the assumed information structure at the time of the hypothetical negotiation. Economic theory recognizes that bargaining outcomes can be indeterminate when the parties possess private information in certain dimensions. 19 In the context of the hypothetical negotiation we can conveniently sidestep this problem, because for practical purposes there can be no private information at the damages stage of litigation. The validity and infringement of the patent are no longer in question, and the parties involved are assumed to have access to additional information that they might not actually have known on the eve of infringement. 20, 21 Thus we can apply methods appropriate to the case of complete information (such as the Nash solution). The economic significance of using the Nash solution in the hypothetical license situation is that even relatively small outside alternatives have an impact on the division of surplus. Recall from our discussion of bargaining economics that under the outside option solution, the only time a bargaining party s alternative payoff affects the split of surplus is when it acts as a constraint, i.e., when it exceeds the simple 50/50 payoff. If the value of the party s outside option is less, it has no impact on the payoff. In contrast, the Nash solution accounts for even low-value outside alternatives in dividing the bargaining surplus. This can make a great deal of difference in practice. To see this, recall that when the infringer has an alternative technology available that yields profits of $4 million, the outside option solution implies a royalty of $5 million (i.e., half of the total surplus of $10 million), while Nash bargaining yields a royalty of $3 million. Thus, in this simple example the royalty predicted by the Nash solution is fully 40 percent lower than the royalty from the simple split called for by the outside option solution. The reason is that, even though the outside alternative is not valuable relative to the patented technology, it still has some value, which should be accounted for in determining the appropriate royalty. The Patentee s Alternatives We next incorporate the patentee s foregone profits in the disagreement payoff. These profits are clearly not an issue in the situation when the patentee does not compete with the infringer in the product market; in this case the patentee s disagreement payoff would simply be zero (as in the examples above). 22 When the patentee does produce, it may be appropriate to consider the profits he would have received in the absence of licensing, if the patentee can prove that he lost profits. 23 In Panduit, the Sixth Circuit held that foregone profits may be taken into account in setting a royalty rate, even when the patentee has failed to meet this burden of proof. Thus, if a producing patentee s profits would be affected in some way by a licensed infringer, even if the amount is not precisely quantifiable, this effect should be taken into account in setting a royalty. We are willing to believe that there may be occasions where it is appropriate to consider a patentee s possible lost profits in determining a reasonable royalty, even when lost profits have not been proved to the required legal standard. 24 However, for the practical purpose of finding the disagreement point in the bargaining negotiation, a patentee that has failed to prove lost profits cannot hope to prove what he would have earned in the absence of licensing, as these are the same thing. Thus the lost profits standard does not appear to be helpful in resolving the strictly economic problem of royalty determination. In light of this, we argue that there are at least two common cases where the patentee s but for profits should not be used to determine its bargaining power. The first concerns the situation in which lost profits are determined based on market shares. In Mor-Flo the Federal Circuit approved a division of infringing sales, based on a market share approach, between those that the patentee could prove it would have made (for which it recovered lost profits), and those that would have been made by other non-infringing competitors (which earned a reasonable royalty). 25 In itself, this approach is economically justifiable, assuming that the royalty is determined properly. Specifically, patentee profits should not be taken into account in setting this royalty, because by definition there is insufficient reason to believe that the patentee could have made these sales. It is also worth remarking that a correctly determined royalty on these sales does not overcompensate the patentee, as some have argued. 26 The patent statute calls for damages based on the use made of the invention by the infringer, and thus some royalty should be assessed for sales that would have been made by other market participants in the absence of infringement. 27 Patentee profits should also not be taken into account if certain sales have been explicitly excluded from the lost profits category for other reasons. For example, in Rite-Hite the patentee was unable to recover lost profits on 502 infringing sales because it could not prove that its sales staff was likely to have made every infringing sale. 28 Either the patentee s sales staff did not first contact the infringing purchaser in each of these cases, or the customer expressed some interest in products made by other manufacturers. 29 Further, the district court corroborated the division of sales between lost profits and reasonable royalties using a Mor-Flo type market share approach. 30 Both the district court and the Federal Circuit approved a reasonable royalty of 50 percent of patentee per-unit profits on the basis that the patentee would not have wanted to grant a license to a competitor when it earned significant profits itself. 31 Given that these reasonable royalty sales were explicitly carved out of the total number of infringing sales because of difficulties in proving lost profits, it seems to us that it is inappro- 12 The Licensing Journal AUGUST 2004

163 priate to use the patentee s prospective profits as the disagreement point of this negotiation. 32 The Infringer s Alternatives Just as it may be appropriate to consider the patentee s alternatives in the form of the profit it could have earned without licensing, it is relevant to account for the alternatives of the infringer. It is surprising that courts have only recently begun to recognize this logical symmetry in a systematic way. The dissent by Judge Nies in Rite-Hite sets out the argument that the district court erred by not considering other commercial alternatives available to the infringer, and instead limited its assessment to Rite-Hite s side of the hypothetical negotiating table rather than to balance the interests of both parties. 33 That the infringer s alternatives matter is underscored under Nash bargaining by the bargaining power, which even a relatively unattractive alternative creates in a license negotiation. More recently, the Federal Circuit s Grain Processing decision entrenched the consideration of the infringer s alternatives more firmly. 34 In this case the court upheld the district court s denial of lost profits on the basis of a non-infringing substitute that, while not actually on the market at the time infringement began, was for all practical purposes available. The court noted that a fair and accurate reconstruction of the but for market also must take into account, when relevant, alternative actions the infringer foreseeably would have undertaken had he not infringed. 35 These alternative actions must necessarily figure into the hypothetical royalty negotiation. In this case, the availability of a slightly more costly production process acted as a cap on the amount the infringer was willing to pay for a hypothetical license. 36 This recognition supports the economically consistent view that the infringer s alternatives should be taken into account in finding the disagreement point for the purposes of bargaining over the license. Applications Cases The economic framework we have outlined is helpful in interpreting a number of high-profile cases. First, the outcome of Grain Processing can be understood as the outcome of a bargaining process. Having rejected the patentee s lost profits argument in light of the available alternative technology, the Federal Circuit approved a royalty of 3 percent of infringing sales. 37 This was based on the district court s best estimate of the outcome of the hypothetical negotiation, in light of the fact that the alternative would have cost only 2.3 percent more to implement, and was subject to fluctuations. 38 This result is very similar to what would be predicted under the outside option solution. In that solution the infringer s disagreement point acts as a constraint, so that if, for example, the infringer could earn $100 per unit with the patent and $97 without, the appropriate royalty is $3 per unit. As we have remarked, we consider Nash bargaining to be more appropriate than the outside option solution in this context. Nash bargaining would tend to predict a lower royalty because the gains from trade would be shared between the patentee and infringer, over and above the value of the alternative. This may appear to make little difference under the facts of Grain Processing, when the value of the alternative technology was very high; but if for instance the infringer could earn only $40 per unit, the outside option solution would predict a significantly higher royalty than Nash bargaining (i.e., $50 per unit as opposed to $30). Note also that a misapplication of the Grain Processing logic would suggest a royalty of $60 per unit (i.e., $100 - $40) under this alternative set of facts, which clearly is to be avoided as it is not consistent with either of our models of bargaining. Nonetheless, the court s decision on damages is at least on its face economically consistent, if not strictly in line with our preferred method. In contrast, the CAFC s earlier decision in Mahurkar v. Bard appears to endorse an economically inconsistent method of royalty determination. 39 Here, the court upheld a royalty rate of percent, calculated as the infringer s net margin of percent, plus savings of research and development expenditures that would otherwise have been made in the amount of 6.72 percent, less 10 percent to which the infringer was deemed entitled (as a reasonable profit ). 40 If we assume that the 10 percent margin represents the amount that Bard (the infringer) would have demanded as a reasonable profit, the implication is that Bard would have been willing to pay the entire amount over its disagreement point (i.e., 29.16% % - 10% = 25.88%) to license the patent. As we have just mentioned, this reasoning is not consistent with either the outside option or the Nash approach to a reasonable patent royalty. In any realistic negotiation, the infringer s threat must be worth something, and thus it should be able to capture at least part of the gains from trade in the form of a lower rate. The Nash solution predicts that if indeed the available gross surplus is percent as claimed, an appropriate royalty might be in the range of 13 percent. 41 Taking into account the substantial difference between the rate found and the rate that is most consistent with economic logic, we suggest that the method used in this case is in error. Rite-Hite illustrates another pitfall of the hypothetical negotiation framework. Recall that the Rite-Hite court approved a rate of approximately 50 percent of patentee profits, because the patentee would not have been willing to grant a license to a competitor. 42 While this outcome does not seem compatible with any of the models we have reviewed, it is important to point out that court s findings of fact violate one of the key assumptions of bargaining models, namely that the AUGUST 2004 The Licensing Journal 13

164 gains from trade are positive. 43 Gains from trade are positive if the infringer makes sales or earns profits that the patentee would not have made. Gains from trade are negative if, for example, an infringer s entry eroded the price of the patented product so that the combined profits of the patentee and the infringer were less than those of the patentee alone. If the gains from trade are negative, then the patentee rationally would be unwilling to license at any price that is consistent with entry by the infringer. If the patentee is truly unwilling to license, there is by definition no agreement it could reach with a licensee that would leave both parties better off. 44 The bargaining models do not apply in this case. The district court s 50 percent figure is therefore not justified by any economic model of bargaining. In fact, under this reasoning, it is not clear why a licensor in this situation would settle for a royalty any less than 100 percent of its profits (assuming that both firms had the same cost of production). Once one departs from an economic model of bargaining, one is confronted with a range of mostly unpalatable alternatives. One can do as the district court did (and as the Federal Circuit approved): choose a rate that is completely arbitrary; one can award lost profits, despite the plaintiff s failure of proof; or one can award a royalty that does not reflect the patentee s profits. As we have argued, we believe the third approach is preferable in many cases like this. The apparently arbitrary nature of the award in Rite- Hite brings up an additional, non-statutory motive for damages, which is the appropriate level of deterrence. The Federal Circuit has affirmed the Panduit argument that damages must be high enough to ensure that the infringer does not receive a compulsory license when the patentee would not otherwise wish to grant a license to a competitor. 45 Lost profits are the preferable remedy from the perspective of adequately compensating the patentee (and deterring the infringer). If lost profits are not available, alternative measures to increase the reasonable royalty have sometimes been proposed. 46 However, courts seem to recognize that such measures are by their nature discretionary, and not firmly grounded in sound economics. 47 This is a thorny policy problem, so we are reluctant to make a strong case that any single approach can satisfy all of the goals of compensation, deterrence, and economic consistency. Ad Hoc Approaches Considering many of the difficult problems that we have reviewed to this point, practitioners and courts have sometimes foregone a more in-depth economic analysis and settled for various ad hoc approaches to royalty determination. Under the analytical approach, a normal rate of profit is subtracted from the infringer s operating profit, with the balance going to the patentee in the form of a royalty. Given our discussion of the method used in Mahurkar v. Bard, which is exactly the same, this is not a method supported by an economic model of bargaining. It may well be supported by the merit of relative simplicity, but this should not make it attractive. 48 Moreover, it assumes that the entire excess profit is due to the patents in question in the litigation, which is very rarely the case. Likewise, the so-called 25 Percent Rule, under which 25 percent of the infringer s operating profit is attributed to the infringed patents, is too simplistic to capture most realistic situations. 49 Certainly, it does not account for differences in disagreement points across different situations. In the Nash bargaining context, this rule implies that the infringer s disagreement point is invariably higher than the patentee s disagreement point by the exact amount of 50 percent of the gross surplus available. This assumption does not allow either for the possibility that the infringer might have few or no alternatives, or that the patentee might have significant alternatives. One can argue that this is merely a starting point and that a number of the Georgia-Pacific factors could be introduced to add realism. 50 However, it is not clear why a practitioner would not start with a more realistic assessment of the outcome of the hypothetical negotiation, thus necessitating fewer ad hoc adjustments. Conclusion We have suggested in this paper that economic models of bargaining can be useful in understanding the hypothetical negotiation that is contemplated in patent damages cases. We do not suggest that this is the right approach to follow in each case, and we admit that there are numerous relevant factors that are not considered in our simple economic models. Nonetheless, there is merit to starting with an economically consistent split of the bargaining surplus, and then taking into account some of the additional contextual richness by applying several of the Georgia-Pacific factors as necessary. To us, this seems a more reasonable approach than starting with an admittedly arbitrary benchmark such as those espoused by some practitioners or courts, and then making a plethora of adjustments. 1. See Paul Milgrom and John Roberts, Economics, Organization and Management, (Englewood Cliffs, NJ, Prentice Hall: 1992) at The requirement of complete information does not imply that each side must know the future in order to find a solution. Complete information means that each side has access to all available information on which to base a forecast, which given the assumed rationality of the parties implies that expected payoffs will be known by both. If the parties place different valuations on future risk, then these valuations may enter into the structure and payoffs of the contract. For example, other things equal, a more risk-averse party would be willing to accept a reduction in its expected 14 The Licensing Journal AUGUST 2004

165 payoff in exchange for reduced variability in the payoff. This reduction might in practice take the form of a fixed annual payment plus a relatively low running royalty, for example. 3. While we focus here on the implications of the licensee s threat point, it should be emphasized that analogous conclusions would follow from a consideration of the licensor s threat point. 4. The original solution is due to John Nash, Jr., The Bargaining Problem, Econometrica, Vol. 18, (1950). The Nash solution has subsequently been used in a wide range of economic applications, particularly in the field of law and economics, from determining the incentive effects of different allocations of ownership rights over physical assets (see generally Oliver Hart, Firms, Contracts, and Financial Structure (Clarendon Press, Oxford: 1995)), to modeling the efficiency properties of liquidated damages clauses in commercial contracts (see Kathryn E. Spier and Michael D. Whinston, On the Efficiency of Privately Stipulated Damages for Breach of Contract: Entry Barriers, Reliance, and Renegotiation, RAND Journal of Economics, Vol. 26, No. 2 (1995). 5. We note that both here and in our discussion of the alternative bargaining model we abstract away from issues such as impatience that may generally affect the distribution of the gains from trade. The economic literature has shown that the extent to which parties discount future payoffs can affect the bargaining outcome (see generally references cited infra n.8). Because in our application of these models there is usually no reason to believe one party has a greater general degree of patience-related bargaining power than the other, the assumption that the parties are symmetric is likely to be appropriate. However because of the inevitable oversimplifications in these models, in real situations there is scope for applying other qualitative factors that may affect payoffs. See text accompanying ns.15 and One can show more generally that under Nash bargaining a party s bargaining payoff increases at one-half the rate of his disagreement point. 7. See Y. Stephen Chiu and B. Rachel Yang, The Outside Option, Threat Point, and Nash Bargaining Solution, Economics Letters, Vol. 62, (1999). 8. See Y. Stephen Chiu, Noncooperative Bargaining, Hostages, and Optimal Asset Ownership, American Economic Review, Vol. 88, No. 4, (1998); Chiu and Yang, supra n.7 at Id U.S.C See infra n U.S.C. 504(a), (b). 13. Panduit Corp. v. Stahlin Bros. Fibre Works, Inc., 575 F.2d 1152, 1158 (6th Cir. 1978). 14. Georgia-Pacific Corp. v. United States Plywood Corp., 318 F. Supp. 1116, Id., at 1121 (emphasis supplied). 16. Id., at 1120 (emphasis supplied). 17. Under current jurisprudence, the court does not consider what price the infringer would have charged, and what profits he would have received, had the proposed royalty been implemented. In other words, the infringer s profits are treated as fixed. Under many competitive conditions, the impact of a running royalty on the infringer s marginal cost would have induced the infringer to raise its prices, partially offsetting the cost of the royalty. This procedure understates the infringer s capacity to absorb a reasonable royalty and yet make a reasonable profit, and so penalizes the patentee for the infringer s failure to negotiate a royalty. 18. Susan M. Maxwell v. J. Baker, Inc., 86 F.3d 1098, (Fed. Cir. 1996). 19. See supra n The methodology encompasses fantasy and flexibility; fantasy because it requires a court to imagine what warring parties would have agreed to as willing negotiators; flexibility because it speaks of negotiations as of the time infringement began, yet permits and often requires a court to look to events and facts that occurred thereafter and that could not have been known to or predicted by the hypothesized negotiators. Fromson v. Western Lithoplate and Supply Co. and Bemis Co., Inc., 853 F.2d 1568, 1575 (Fed. Cir. 1988). 21. A hypothetical negotiation should take into account the actual facts as they occurred in the matter both before and after the hypothetical negotiations would occur. Sinclair Ref. Co. v. Jenkins Petroleum Process Co., 289 U.S. 689, (1933) (referring to later experiences as a book of wisdom which can correct uncertainties present at the time of hypothetical negotiation). University of Colorado Found. v. American Cyanamid Co., 216 F. Supp. 2d 1188, 1197 (D. Col. 2002). We observe that this ex post construction of the hypothetical negotiation does not conform to a pure economic forecasting model, in which parties make ex ante forecasts of the value of the prospective bargain based only on information available at the time of the bargain. We also observe that courts have generally adopted a hybrid of the ex ante and ex post approaches in constructing hypothetical negotiations. For example, in assessing reasonable royalty damages, the court usually mandates that the parties bargain over the royalty rate based on the expected (not actual) profitability of the use of the invention to the infringer; this royalty rate is then multiplied by the infringer s actual (not expected) sales. These sales reflect the infringer s actual infringing acts, for which it must pay compensation. Similarly, the court generally does not take into account subsequent technological developments that might have enabled the licensee to invent around the patent. And, as we have noted, the court does not attempt to determine how the imposition of the royalty would have altered the licensee s price, output and profit. 22. If the patentee could have licensed exclusively, and infringement destroyed that possibility, then his disagreement payoff would be greater than zero. 23. The patentee must show each of four things: (1) demand for the patented product; (2) absence of acceptable non-infringing substitutes; (3) sufficient manufacturing and marketing capability; and (4) the amount of profit lost. Panduit at 1156, For instance, lost profits may have been rejected because of difficulties in quantification (i.e., Panduit factor (4)). 25. State Industries, Inc. v. Mor-Flo Indus., Inc. and Am. Appliance Mfg. Corp., 883 F.2d 1573 (Fed. Cir. 1989). 26. While the CAFC in State Industries greatly improved its economic analysis by adopting a market share approach, it made a new and serious error by simultaneously awarding lost profits on the patent owner s lost sales and a reasonable royalty on the remaining sales. This could never have happened in the real world and is another example of the patentee being far better off in court. Vincent E. O Brien, Economics and Key Patent Damages Cases, University of Baltimore Intellectual Property Law Journal, Vol. 9, No. 1 (2000), at In general, the patentee s compensation should take into account the effect of the infringing sales on the market price. If the but-for price would have been higher, the but-for quantity (including sales that command a reasonable royalty) would have been lower. This is an application of the law of demand which both courts and economic theory say must be taken into account. Crystal Semiconductor Corporation v. Tritech Microelectronics Int l et al., Nos , 1559, , Fed. Cir. (2001). See also Gregory J. Werden, et al., Quantity Accretion: Mirror Image of Price Erosion from Patent Infringement, Journal of the Patent and Trademark Office Society, Vol. 81 (1999). 28. Rite-Hite Corp., et al. v. Kelley Company, Inc., 774 F. Supp. 1514, 1534 (E.D. Wis. 1991), 56 F.3d 1538, 1554 (Fed. Cir. 1995). 29. Rite-Hite, 774 F. Supp. 1514, Id., at Id., 774 F. Supp. 1514, 1535, 56 F.3d 1538, This conclusion is, however, sensitive to the facts. In addition to losing a sale, the patentee may also lose the option to try to make the sale. For example, a faster, more experienced or larger infringer might contact customers before the patentee was able to do so, though eventually the patentee might also have contacted these customers. In general, the loss of a sales option, as opposed to an actual sale, is insufficient to prove lost profits. By choosing a high royalty rate for such sales, the court may have valued this option more highly than would be justified if the patentee truly had no prospect of making the sale. 33. Id., 56 F.3d 1538, Grain Processing Corp. v. American Maize Prods. Co., 185 F.3d 1341 (Fed. Cir. 1999). 35. Id., at Id., at 1347, Id. at Id., at Dr. Sakharam D. Mahurkar v. C.R. Bard, Inc., 79 F.3d 1572 (Fed. Cir. 1996). 40. Id., at To see this, note that if the patentee s disagreement point is zero by assumption, then the gains from trade are 29.16% % - 10% = 25.88%, and the patentee s payoff (and thus the royalty) is 0 + 1/2 x 25.88% = approximately 13%. 42. See supra, text accompanying n Strictly speaking, we would argue that it is correct to view the gains from trade as positive in this case, because it is not appropriate to consider the patentee s foregone profits for these sales. Thus the gains from trade are only negative here to the extent that one accepts the court s premise, which is that these foregone profits are relevant. More generally, gains from trade will be negative in many competitive situations when the licensee or infringer is destroying industry profit through competing with the patentee. 44. Rite-Hite, 56 F.3d 1538, King Instruments Corp. v. Luciano Perego and Tapematic, 65 F.3d 941, 951 (Fed. Cir. 1995). 46. See, e.g., Mahurkar at King Instruments at Robert Goldscheider, The Employment of Licensing Expertise in the Arena of Intellectual Property Litigation, IDEA: The Journal of Law and Technology (1996), at See id., at 183; Robert Goldscheider, John Jarosz and Carla Mulhern, Use of the 25 Per Cent Rule in Valuing IP, les Nouvelles (2002). 50. See Goldscheider et al at 129. AUGUST 2004 The Licensing Journal 15

166 Case 3:11-cv RBD-TEM Document Filed 07/25/13 Page 1 of 4 PageID EXHIBIT 14

167 iess_b3_h-o 4/12/07 4:13 PM Page 540 Case 3:11-cv RBD-TEM Document Filed 07/25/13 Page 2 of 4 PageID Nash Equilibrium tors and despots (e.g., Adolf Hitler, Joseph Stalin), have been described as narcissistic. In addition narcissism has been examined as a potential factor in political terrorism. In criminology narcissistic personality traits are thought to predict criminal behavior, including murder, rape, assault, spousal abuse, and white-collar crime. Narcissism is also a key feature of a psychopathic personality, which is perhaps the most important personality profile for predicting serious criminal behavior. Several significant issues remain unresolved in the scientific study of narcissism. First, there remains debate over the definition of narcissism. While there is strong agreement on key features of narcissism like grandiosity and low empathy, there is disagreement about the link between narcissism and feelings of depression or unhappiness. Some theorists argue that narcissism contains a component of depression or low self-esteem; others argue that narcissism is related to positive emotions. Still others argue that narcissism is linked to negative emotions and self-perceptions but that these feelings are experienced only at an unconscious level. Second, while there are several theories about the development of narcissism in individuals, there is no firm conclusion about its etiology. Some researchers argue that narcissism results from permissive parenting, while others argue that narcissism is a reaction to cold, controlling parents. Finally, the role of culture in maintaining narcissism is not well understood. Some researchers and theorists have identified a rising tide of narcissism, but the cause of this remains unclear. SEE ALSO Freud, Sigmund; Individualism; Leadership; Neuroticism; Obsession; Personality; Political Science; Psychology NASH EQUILIBRIUM Nash equilibrium is a fundamental concept in the theory of games and the most widely used method of predicting the outcome of a strategic interaction in the social sciences. A game (in strategic or normal form) consists of the following three elements: a set of players, a set of actions (or pure-strategies) available to each player, and a payoff (or utility) function for each player. The payoff functions represent each player s preferences over action profiles, where an action profile is simply a list of actions, one for each player. A pure-strategy Nash equilibrium is an action profile with the property that no single player can obtain a higher payoff by deviating unilaterally from this profile. This concept can best be understood by looking at some examples. Consider first a game involving two players, each of whom has two available actions, which we call A and B. If the players choose different actions, they each get a payoff of 0. If they both choose A, they each get 2, and if they both choose B, they each get 1. This coordination game may be represented as follows, where player 1 chooses a row, player 2 chooses a column, and the resulting payoffs are listed in parentheses, with the first component corresponding to player 1 s payoff: The action profile (B,B) is an equilibrium, since a unilateral deviation to A by any one player would result in a lower payoff for the deviating player. Similarly, the action profile (A,A) is also an equilibrium. As another example, consider the game matching pennies, which again involves two players, each with two actions. Each player can choose either heads (H) or tails BIBLIOGRAPHY American Psychiatric Association Diagnostic and Statistical Manual of Mental Disorders. 4th rev. ed. Washington, DC: Author. Freud, Sigmund. [1914] On Narcissism: An Introduction. In The Standard Edition of the Complete Psychological Works of Sigmund Freud, ed. and trans. James Strachey, vol. 14, London: Hogarth. Lasch, Christopher The Culture of Narcissism: American Life in an Age of Diminishing Expectations. New York: Norton. Morf, Carolyn C., and Frederick Rhodewalt Unraveling the Paradoxes of Narcissism: A Dynamic Self-regulatory Processing Model. Psychological Inquiry 12 (4): Figure 1 A (2,2) (0,0) B H T A (0,0) H (1, 1) ( 1,1) B (1,1) T ( 1,1) (1, 1) W. Keith Campbell Joshua D. Miller Figure INTERNATIONAL ENCYCLOPEDIA OF THE SOCIAL SCIENCES, 2ND EDITION

168 iess_b3_h-o 4/12/07 4:13 PM Page 541 Case 3:11-cv RBD-TEM Document Filed 07/25/13 Page 3 of 4 PageID Nash Equilibrium Figure 3 D C C (2,2) (0,3) (3,0) D (1,1) (T); player 1 wins a dollar from player 2 if their choices are the same, and loses a dollar to player 2 if they are not. This game has no pure-strategy Nash equilibria. In some cases, instead of simply choosing an action, players may be able to choose probability distributions over the set of actions available to them. Such randomizations over the set of actions are referred to as mixed strategies. Any profile of mixed strategies induces a probability distribution over action profiles in the game. Under certain assumptions, a player s preferences over all such lotteries can be represented by a function (called a von Neumann-Morgenstern utility function) that assigns a real number to each action profile. One lottery is preferred to another if and only if it results in a higher expected value of this utility function, or expected utility. A mixed strategy Nash-equilibrium is then a mixed strategy profile with the property that no single player can obtain a higher value of expected utility by deviating unilaterally from this profile. The American mathematician John Nash (1950) showed that every game in which the set of actions available to each player is finite has at least one mixed-strategy equilibrium. In the matching pennies game, there is a mixed-strategy equilibrium in which each player chooses heads with probability 1/2. Similarly, in the coordination game of the above example, there is a third equilibrium in which each player chooses action A with probability 1/3 and B with probability 2/3. Such multiplicity of equilibria arises in many economically important games, and has prompted a large literature on equilibrium refinements with the purpose of identifying criteria on the basis of which a single equilibrium might be selected. Nash equilibria can sometimes correspond to outcomes that are inefficient, in the sense that there exist alternative outcomes that are both feasible and preferred by all players. This is the case, for instance, with the equilibrium (B,B) in the coordination game above. An even more striking example arises in the prisoner s dilemma game, in which each player can either cooperate or defect, and payoffs are as follows: The unique Nash equilibrium is mutual defection, an outcome that is worse for both players than mutual cooperation. Now consider the game that involves a repetition of the prisoner s dilemma for n periods, where n is commonly known to the two players. A pure strategy in this repeated game is a plan that prescribes which action is to be taken at each stage, contingent on every possible history of the game to that point. Clearly the set of pure strategies is very large. Nevertheless, all Nash equilibria of this finitely repeated game involve defection at every stage. When the number of stages n is large, equilibrium payoffs lie far below the payoffs that could have been attained under mutual cooperation. It has sometimes been argued that the Nash prediction in the finitely repeated prisoner s dilemma (and in many other environments) is counterintuitive and at odds with experimental evidence. However, experimental tests of the equilibrium hypothesis are typically conducted with monetary payoffs, which need not reflect the preferences of subjects over action profiles. In other words, individual preferences over the distribution of monetary payoffs may not be exclusively self-interested. Furthermore, the equilibrium prediction relies on the hypothesis that these preferences are commonly known to all subjects, which is also unlikely to hold in practice. To address this latter concern, the concept of Nash equilibrium has been generalized to allow for situations in which players are faced with incomplete information. If each player is drawn from some set of types, such that the probability distribution governing the likelihood of each type is itself commonly known to all players, then we have a Bayesian game. A pure strategy in this game is a function that associates with each type a particular action. A Bayes- Nash equilibrium is then a strategy profile such that no player can obtain greater expected utility by deviating to a different strategy, given his or her beliefs about the distribution of types from which other players are drawn. Allowing for incomplete information can have dramatic effects on the predictions of the Nash equilibrium concept. Consider, for example, the finitely repeated prisoner s dilemma, and suppose that each player believes that there is some possibility, perhaps very small, that his or her opponent will cooperate in all periods provided that no defection has yet been observed, and defect otherwise. If the number of stages n is sufficiently large, it can be shown that mutual defection in all stages is inconsistent with equilibrium behavior, and that, in a well-defined sense, the players will cooperate in most periods. Hence, in applying the concept of Nash equilibrium to practical situations, it is important to pay close attention to the information that individuals have about the preferences, beliefs, and rationality of those with whom they are strategically interacting. INTERNATIONAL ENCYCLOPEDIA OF THE SOCIAL SCIENCES, 2ND EDITION 541

169 iess_b3_h-o 4/12/07 4:13 PM Page 542 Case 3:11-cv RBD-TEM Document Filed 07/25/13 Page 4 of 4 PageID Nasser, Gamal Abdel SEE ALSO Game Theory; Multiple Equilibria; Noncooperative Games; Prisoner s Dilemma (Economics) BIBLIOGRAPHY Cournot, A. A Recherches sur les principes mathématiques de la théorie des richesses. Paris: L. Hachette. Fudenberg, Drew, and Jean Tirole Game Theory. Cambridge, MA: MIT Press. Harsanyi, John C Games with Incomplete Information Played by Bayesian Players. Management Science 14 (3): , , Harsanyi, John C., and Reinhard Selten A General Theory of Equilibrium Selection in Games. Cambridge, MA: MIT Press. Kreps, David, Paul Milgrom, John Roberts, and Robert Wilson Rational Cooperation in the Finitely Repeated Prisoner s Dilemma. Journal of Economic Theory 27: Nash, John F Equilibrium Points in N-Person Games. Proceedings of the National Academy of Sciences 36 (1): Osborne, Martin J., and Ariel Rubinstein A Course in Game Theory. Cambridge, MA: MIT Press. von Neumann, John, and Oskar Morgenstern Theory of Games and Economic Behavior. Princeton, NJ: Princeton University Press. NASSER, GAMAL ABDEL Rajiv Sethi Gamal Abdel Nasser, who served as president of Egypt from 1956 to 1970, was born on January 15, 1918, in the small village of Bani Mor in the Egyptian province of Assiut, where he lived for eight years. He came from a humble and poor background to become one of the most prominent and influential leaders in the Middle East and the third world. His father worked as a mail carrier in the Egyptian Ministry of Communication, a position that required him to move with his family from Bani Mor to Alexandria and finally Cairo, where Nasser lived for ten years. In his memoirs, Nasser spoke proudly of his humble origin. His poor background might have been behind his socialist tendencies and his commitment to improve the living conditions of Egyptian peasants and workers. During his high school years, Nasser participated in student demonstrations against the British occupying forces. After receiving his high school diploma in 1937, Nasser entered the Egyptian Royal Military Academy, which started admitting sons of lower-income families in A year later, he joined the Egyptian army, where he met several of his future colleagues, including Anwar el- Sadat ( ) and Zakaria Mohyi El Deen, both of whom served as his vice presidents, and Abdul Hakeem Amer, who became a minister of defense. In 1942 Nasser was transferred to Sudan, where he and other officers founded the Free Officers, a secret revolutionary organization. The Free Officers was a secular nationalist movement that was opposed to the British occupation of Egypt, the corrupt royal family, and the domination of Egypt s economy and parliament by a small landowning class. In 1948 Nasser was a member of the Egyptian army that along with other Arab armies was sent to Palestine to thwart the establishment of Israel. The humiliating defeat of the Arab armies in the 1948 war raised his awareness of the Palestinian problem and the inefficacy of the existing Arab governments. On July 23, 1952, Nasser and his Free Officers seized power and deposed the king. A year later, the Revolutionary Command Council of the Free Officers promulgated a new constitution, abolished the monarchy, and declared Egypt a republic. Though General Mohammad Naguib ( ) served as the head of the government from 1952 to 1954, Nasser held the real power through his control of the Revolutionary Command Council. In November 1954 Nasser placed Naguib under house arrest, accusing him of knowing about an attempt by a member of the Muslim Brotherhood to assassinate Nasser. In 1956 Nasser was elected president of Egypt, a position he held until his death in As president, Nasser created an authoritarian police state, banning political parties and suppressing political opposition, including the local communists and members of the Muslim Brotherhood. He ruled the country through the Arab Socialist Union, a government-controlled party. Between 1956 and 1966, Nasser introduced several socialist measures, including the nationalization of various industries, private companies, and banks, and he expanded the public sector significantly. He also introduced agrarian reform, including the confiscation of 2,000 square miles of cultivable land from wealthy landowners, which he distributed to Egypt s poor peasants. The aim of these socialist measures was to improve the living conditions of the country s peasants and workers. Nasser contended in his book The Philosophy of the Revolution (1955) that Arab socialism was a prerequisite for Arab unity and freedom and for surmounting the social and economic legacy of colonialism. In addition to his domestic socialist reforms, Nasser adopted an anti-western and anticolonial foreign policy. Initially however, he tried to secure arms from Britain and the United States, and it was only after the two countries declined his request that he acquired such weapons from the Soviet Union and Eastern Europe. Along with Prime 542 INTERNATIONAL ENCYCLOPEDIA OF THE SOCIAL SCIENCES, 2ND EDITION

170 Case 3:11-cv RBD-TEM Document Filed 07/25/13 Page 1 of 1 PageID PARKERVISION, INC., v. Plaintiff, UNITED STATES DISTRICT COURT MIDDLE DISTRICT OF FLORIDA JACKSONVILLE DIVISION QUALCOMM INCORPORATED, Defendant. Case No. 3:11-cv-719-J-37TEM QUALCOMM INCORPORATED, v. Counterclaim Plaintiff, PARKERVISION, INC., AND STERNE, KESSLER, GOLDSTEIN, & FOX PLLC, Counterclaim Defendants. EXHIBIT 15 TO PARKERVISION S RESPONSE TO QUALCOMM S DAUBERT MOTION TO PRECLUDE PAUL BENOIT DAMAGES TESTIMONY OUTSIDE ATTORNEYS' EYES ONLY UNDER SEAL - SUBJECT TO PROTECTIVE ORDER - CONTAINS CONFIDENTIAL INFORMATION - TO BE OPENED ONLY AS DIRECTED BY THE UNITED STATES DISTRICT COURT FOR THE MIDDLE DISTRICT OF FLORIDA McKool v1

171 Case 3:11-cv RBD-TEM Document Filed 07/25/13 Page 1 of 18 PageID EXHIBIT 16

172 Case 3:11-cv RBD-TEM Document Filed 07/25/13 Page 2 of 18 PageID The Columbia SCIENCE AND TECHNOLOGY LAW REVIEW MAKING SENSE OF APPORTIONMENT IN PATENT DAMAGES 1 Elizabeth M. Bailey Gregory K. Leonard Mario A. Lopez Unreasonably large damages awards in patent litigation have been an important force in motivating the movement for patent reform. Apportionment has found support as a solution to problem damages awards. Under apportionment, the portion of the overall value of the product that is attributable to the patented technology is identified. Then, reasonable royalty damages are calculated with reference to this apportioned value of the patented technology rather than the overall value of the product. While the problems that have motivated the apportionment movement are real and serious, apportionment makes sense as a solution only under the assumption that an economically invalid approach to calculating damages is being taken in the first place. A more sensible solution is to require litigants to take an economically valid approach to damages. In addition, when there are complementarities between assets, such that the combined use of two or more assets is worth more than their individual use, no unique way exists to apportion the overall value of the product among the assets (including the patented technology at issue), rendering apportionment infeasible in many cases. We consider these and other issues that surround apportionment. 1 Bailey: NERA Economic Consulting, Elizabeth.Bailey@nera.com; Leonard: NERA Economic Consulting, Gregory.Leonard@nera.com; Lopez: NERA Economic Consulting, Mario.Lopez@nera.com. The views expressed herein are those of the authors and not necessarily those of other economists at NERA. Cite as This work is made available under the Creative Commons Attribution-Noncommercial-No Derivative Works 3.0 License

173 Case 3:11-cv RBD-TEM Document Filed 07/25/13 Page 3 of 18 PageID Vol. XII Columbia Science and Technology Law Review 2011 I. INTRODUCTION Controversial patent damages awards have put patent damages at the center of the debate over patent reform. The award in Lucent Technologies, Inc. v. Gateway, Inc. ( Lucent ) is one such example. In a jury trial, Microsoft was found to have infringed a patent referred to as the Day patent that describes a method to enter information on a computer screen without using a keyboard (e.g., by using a stylus), which Microsoft was found to have used in its date-picker calendar tool in Microsoft Outlook. The jury awarded $358 million in damages to Lucent. On appeal, however, the Court of Appeals for the Federal Circuit ( CAFC ) found that the infringing use of the date-picker tool in Outlook is but a very small component of a much larger software program 2 and concluded that the damages calculation lacked sufficient evidentiary support. 3 The matter was remanded to the lower court for a new trial on damages. In recent years, Congress has considered a variety of legislative proposals 4 designed to address the issue of unreasonably large damages awards. Of particular concern are situations where the patented technology is but one of many technologies and assets that are incorporated into a product. 5 Apportionment has been proposed as a solution to these problems. 6 Under apportionment, the portion of the overall value of the product that is attributable to the patented technology is identified. Then, reasonable royalty damages are calculated with reference to this apportioned value of the patented technology rather than the overall value of the product. In this article, we explore apportionment. While the problems that have motivated the apportionment movement are real and serious, apportionment makes sense as a solution only under the assumption that an economically invalid approach to calculating damages is being taken in the first place. Apportionment relieves a symptom, but not the cause, of problem damages awards. Adoption of apportionment as a damages calculation methodology may arbitrarily reduce reasonable royalty awards, even for 2 Lucent Technologies, Inc., et al. v. Gateway, Inc., et al., 580 F.3d 1301, 1337 (Fed. Cir. 2009), cert. denied, 30 S. Ct (2010). 3 Id. at See, e.g., The Patent Reform Act of 2009, S. 515, 111th Cong. (2009); see also, e.g., S. 610, 111th Cong. (2009); H.R. 1260, 111th Cong. (2009); The Patent Reform Act of 2010, S. Managers Amend., 111th Cong. (2010) (Managers amendment to Senate Bill 515 as reported by S. Comm. on the Judiciary, March 4, 2010). 5 E.g., Patent Reform in the 111th Congress: Legislation and Recent Court Decisions Before the S. Comm. On the Judiciary, 111th Cong. 14 (2009) (statement of Mark A. Lemley, Professor of Law, Stanford Law School), available at (testifying that the Georgia- Pacific factors are open to manipulation and that they do not reflect the contribution of the patented technology to the technology at issue). 6 See, e.g., id. at See also, Brian J. Love, Patentee Overcompensation and the Entire Market Value Rule, 60 Stan. L. Rev. 263, 272 (2007). 256

174 Case 3:11-cv RBD-TEM Document Filed 07/25/13 Page 4 of 18 PageID Vol. XII Columbia Science and Technology Law Review 2011 valuable new inventions counter to the purpose of patent system. A more sensible solution is to require litigants to take an economically valid approach to damages. In that case, apportionment would not be necessary. In addition, an attempt to implement apportionment would face serious practical problems. When there are complementarities between assets, such that the combined use of two or more assets is worth more than their individual use, no unique way exists to apportion the overall value of the product among the assets (including the patented technology at issue). Unless a particular apportionment scheme was specified in patent reform legislation, substantial legal ambiguity would be created and courts, juries, and parties would bear a heavy litigation burden. Again, the alternative of requiring litigants to take an economically valid approach to damages is a much more attractive alternative from the perspective of good public policy. II. APPORTIONMENT RELIEVES A SYMPTOM, NOT THE CAUSE, OF PROBLEM DAMAGES AWARDS The apportionment movement has been motivated by scenarios such as the following. A damages expert provides testimony in which a reasonable royalty damages award is calculated by first determining that the average royalty rate in an industry is, say, 1%, and then applying this royalty rate to a royalty base consisting of the revenue of a product that incorporates the patented technology. 7 However, the patented technology covers only a small component of the product, and many other technologies and assets are required to produce the product. Therefore, the resulting damages award (in dollars) seems out of line with the contribution of the patented technology to the product. This scenario played out recently in Cornell University v. Hewlett-Packard Co. ( Cornell ), presided over by Judge Rader of the CAFC, who was sitting by designation. 7 The approach of using industry averages and supposedly comparable licenses to determine the royalty rate in a given situation often fails to be economically sound in its own right. The supposedly comparable licenses are often not, in fact, comparable. License agreements can vary substantially both in terms of the patented technology being licensed and the economic conditions of the parties to the agreement. Unless the important characteristics are similar across two licenses, they will generally not be comparable. Along the same lines, the economic circumstances surrounding the typical or industry average licensing negotiation that led to the typical or industry average royalty rate are unlikely to correspond to the specific economic circumstances of the patented technology and parties at issue in the litigation. Before an existing license can be used as a benchmark, one must carefully analyze whether it is truly comparable in terms of factors such as the technology covered, the product sold by the licensee, the degree of competition between the licensor and licensee, cross licensing arrangements, and other considerations. The CAFC has pointed out this issue in recent opinions. In Lucent, the CAFC determined that some of the licenses that the plaintiff claimed were comparable were in fact radically different from the hypothetical agreement under consideration for the Day patent. As to the remainder of the licenses, the CAFC wrote that they could not understand how the jury could have adequately evaluated the probative value of those agreements, characterizing the evidence presented as superficial and doubtful that the technology of those license agreements is in any way similar to the technology being litigated here. Lucent, 580 F.3d 1301,

175 Case 3:11-cv RBD-TEM Document Filed 07/25/13 Page 5 of 18 PageID Vol. XII Columbia Science and Technology Law Review 2011 A jury had found that Hewlett-Packard infringed a patent that describes a method to read a component of a processor s instruction reorder buffer ( IRB ). The patented technology was claimed to enhance the throughput of a processor. Cornell s damages expert initially testified that the jury should compute damages using a royalty base encompassing Hewlett-Packard s earnings from its sales revenue from its entire servers and workstations. 8 Yet, the patented technology was a small part of the IRB, which is a part of a processor, which is part of a CPU module, which is part of a brick, which is itself only part of the larger server. 9 In other words, the patented technology related to only a very small component of the overall product that was being used to form the royalty base. Judge Rader ruled in a pre-trial motion that servers were not an appropriate 10 royalty base for calculating the reasonable royalty. What followed was essentially application of an apportionment approach. 11 Cornell s expert testified that the royalty base should be reduced from servers to CPU bricks, which yielded damages of $184 million. 12 The jury awarded this amount. However, Judge Rader was troubled by the size of the damages award. 13 In ruling on a Hewlett-Packard post-trial motion, Judge Rader further reduced the royalty base from CPU bricks to processors, and applied the jury s uncontroverted royalty rate of 0.8 percent 14 to this reduced royalty base. 15 This 8 Cornell Univ. v. Hewlett-Packard Co., 609 F.Supp.2d 279, 284 (N.D.N.Y. 2009). 9 Id. at 283. The CPU brick is Hewlett-Packard s term for the combination of the processor, a temperature controlling thermal solution, external cache memory, and a power converter. 10 Cornell Univ. v. Hewlett-Packard Co., No. 01-CV-1974, 2008 WL , at *4 (N.D.N.Y. May 27, 2008). 11 This type of apportionment attempts to limit the use of the entire market value rule (where the entire value of the product is used as the royalty base) to instances where the patented feature is the basis for demand and to some smaller base otherwise. See Love, supra note 6, at 272 ( To prevent overcompensation and its attendant harms, the entire market value rule must be scaled back to its original role as a special case of the apportionment requirement, such that it may not be applied unless as its name suggests the patent at issue indeed accounts for the entire value of the infringing article. ) See also, Eric E. Bensen & Danielle M. White, Using Apportionment to Rein in the Georgia-Pacific Factors, 9 Colum. Sci. & Tech. L. Rev. 1, (2008) ( Where the patent was for an improvement or component, patentees could satisfy their apportionment burden by showing that the entire market value of the infringing product was attributable to the patented invention... ). 12 Cornell, 609 F.Supp.2d 279, In particular, the court stated that [t]he important point is not the way that Cornell derived this royalty base, but that it exceeded again this court s direction and proceeded to attempt to show economic entitlement to damages based on technology beyond the scope of the claimed invention. Cornell, 609 F.Supp.2d 279, Id. at

176 Case 3:11-cv RBD-TEM Document Filed 07/25/13 Page 6 of 18 PageID Vol. XII Columbia Science and Technology Law Review 2011 reduced damages to $53 million. In short, Judge Rader identified the portion of the revenue of the overall product (the server) that was closely related to the patented technology, namely the processor, and then used that revenue as the royalty base. Reducing the royalty base to which a royalty rate is applied is certainly an effective way to reduce a damages award that is overstated. However, it is also a crude approach. Typically, there are only a finite number of possible royalty bases that could be used, with discrete jumps in size between them. For example, in Cornell, the processor royalty base was less than one-third the size of the CPU brick royalty base. It is unlikely that the dollar royalty amount that results from multiplying a specified royalty rate by each of the small number of possible royalty bases is exactly equal to the value of the patented technology. More fundamentally, the apportionment approach of reducing the royalty base treats the symptom (an overly large damages award), without addressing the underlying cause. The underlying cause of problem damages awards is the approach sometimes taken by damages experts of choosing the royalty rate and royalty base independently of each other, and without reference to the economic value of the patented technology. Rather than directing that apportionment be used to eliminate problem damages awards, it would be preferable for patent reform legislation or the CAFC to require damages experts to take an approach that is consistent with sound economic principles. 16 Then, apportionment becomes unnecessary. Under a sound economic approach, the reasonable royalty award (in dollars) should reflect the incremental value (in dollars) of the patented technology to the defendant as compared to the next best alternative. It should not matter what royalty rate or royalty base are used, as long as the product of the two yields a result (in dollars) that is in line with the patented technology s incremental value. Put another way, the royalty rate and royalty base must be chosen together in order for the reasonable royalty (the multiplication of the two) to make economic sense. It is when the rate and base are chosen independently of each other that problem awards (i.e., awards out of line with the economic value of the patented technology) arise. In the context of the Cornell case, suppose that the patented technology resulted in an increase in a server s processing speed relative to what was achievable with the next best technology. Enhanced speed may result in greater sales of, and higher prices for, servers, which in turn leads to incremental profits due to the patented technology. 17 These incremental profits represent the largest dollar amount that a rational licensee 15 Determining the royalty base for processors was not straightforward because processors were generally not sold separately from CPU bricks. Cornell disputed the method by which Hewlett-Packard s expert had calculated the processor royalty base. 16 Indeed, without performing a proper economic analysis, it is difficult to know if a damages award is too large or the degree to which it is too large. As a result, an adjustment after-the-fact is likely to be ad hoc and out of line with the true economic value of the patent. 17 This analysis would have to take into account any incremental costs associated with using the patented technology as well. 259

177 Case 3:11-cv RBD-TEM Document Filed 07/25/13 Page 7 of 18 PageID Vol. XII Columbia Science and Technology Law Review 2011 would pay for the right to use the patented technology. 18 The choice of the royalty base should be largely irrelevant as long as the royalty rate is set, conditional on the choice of royalty base, so as to reflect the economic value (in dollars) of the patented technology. 19 The CAFC appears to be headed in the direction that we advocate. In its opinion in Lucent, the CAFC stated that [t]here is nothing inherently wrong with using the market value of the entire product [as the royalty base], especially when there is no established market value for the infringing component or feature, so long as the multiplier [i.e., the royalty rate] accounts for the proportion of the base represented by the infringing component or feature. 20 III. IN THE PRESENCE OF COMPLEMENTARITIES, NO UNIQUE APPORTIONMENT OF VALUE EXISTS Taken in the most favorable light, advocates of apportionment argue that when calculating reasonable royalties, a royalty rate should be applied to the incremental value added by the patented technology at issue rather than to the overall value of the product. 21 However, the specific methodology that has been put forward to perform apportionment fails to recognize that patented technologies typically create value that is greater than the sum of the parts, i.e., synergies, and incorrectly attributes all of the synergies to the infringer. 18 The procedure we would typically propose is to first define the royalty base as the sales of the smallest product that (1) incorporates the patented feature and (2) can be separately priced (either through actual arm s length transactions or a reliable approximation). Then, given that base, the royalty rate is chosen to reflect properly the economic value of the patented technology. This choice of royalty base is consistent with real world licensing practices. Licensors prefer to have a royalty base that is easily verifiable and not subject to manipulation. Licensees prefer to limit the royalty base to the smallest possible product to limit the distortionary effects of the royalty tax on their incentives. 19 A potential practical limitation to our proposed approach is that juries may be hesitant to award a very small royalty rate, assuming a large royalty base has been chosen, even if that small royalty rate properly reflects the economic value of the patented technology. For example, a jury might be hesitant to award a royalty rate of percent. For this reason, as a litigation strategy, the defendant may prefer to make the case for a lower royalty base (with a correspondingly higher royalty rate). This practical problem would seem to resuscitate the need for apportionment, at least as a means of reducing the royalty base. However, it should be recognized that a symmetric problem exists for plaintiffs. If a very small royalty base was to be the rule, plaintiffs might have difficulty convincing juries to award a large royalty rate, even if that rate were economically sound given the choice of royalty base. For example, a 50 percent royalty rate may be economically appropriate given a narrow royalty base, yet the defendant might be able to persuade a jury that such a rate was out of line with industry practice ). 21 Lucent Technologies, Inc., et al. v. Gateway, Inc., et al., 580 F.3d 1301, 1339 (Fed. Cir. See, e.g., Love, supra note

178 Case 3:11-cv RBD-TEM Document Filed 07/25/13 Page 8 of 18 PageID Vol. XII Columbia Science and Technology Law Review 2011 The synergies, or additional value obtained by combining assets, are the result of complementarities among the assets. To make the concepts concrete, suppose two companies, A and B, each have an asset. Company A can use its asset to make product A and receive value V A. 22 Company B can use its asset to make product B and receive value V B. Alternatively, the two companies can join forces and make product AB, which will generate total value V AB. Product AB may be an improved version of product A or B, or it may be an entirely different type of product altogether. If V AB > VA + VB, it is said that there are (strict) gains to trade, i.e., the two companies would be better off joining forces than pursuing their respective alternatives. The difference S = VAB VA VB represents the amount of the gains to trade, or the synergies generated by combining their assets. Neither company can access S without the other. 23 The value of product AB can be rewritten as VAB = VA + VB + S. It should be clear that there is no unique way of dividing the value of product AB between the two assets that are used to create it. This is because the value of the synergies S is joint and common to the two assets. Both asset owners can lay claim to S. Any mechanical rule to apportion the synergies between the assets needed to create the synergies is analogous to an accounting rule to allocate joint and common costs among the products that those costs support. Economists have long recognized that any such cost allocation is completely arbitrary. 24 Similarly, any rule to apportion the synergies between the assets is necessarily arbitrary. To see the difficulties, suppose Company B argued that its asset should be valued by comparing the value of product AB to the value of the product that could be produced without Company B s asset (i.e., product A). Under that argument, the value of the asset of Company B would appear to be VAB VA = VB + S. However, Company A could make a similar argument and claim that the value of its asset was VAB VB = VA + S. Yet, both companies cannot be awarded these values, or the sum would exceed the value of product AB. In proposing a framework for apportionment, Love states that when the patent at issue covers only a component of or improvement to the infringing item, the value of the sales or uses of that [infringing] item must be apportioned between the patented invention and the remaining unpatented components. 25 According to this construct, Company B could claim that the value of the unpatented components (i.e., Company B s asset) is V and the apportioned value of Company A s patent therefore should be limited to AB V A 22 Value here would be the net present discounted value of the expected profits from selling the product. 23 To reiterate, V A represents the best Company A could do by itself without access to Company B s asset, and similarly for V B. 24 William J. Baumol, On the Proper Cost Tests for Natural Monopoly in a Multiproduct Industry, 67 Am. Econ. Rev. 809 (1977). 25 Love, supra note 6, at

179 Case 3:11-cv RBD-TEM Document Filed 07/25/13 Page 9 of 18 PageID Vol. XII Columbia Science and Technology Law Review 2011 V A. From above, however, this incorrectly attributes all of the synergies, S, to Company B s asset. The approach proposed by Love would only make economic sense in the unlikely situation in which the combination of Product A with Product B created no synergies (i.e., S = 0 such that V AB = VA + VB ). 26 As discussed above, however, combining patented technologies typically creates value that is greater than the sum of the parts. For example, patent pools often bring together various technologies that are necessary to create the product in question. The stand-alone value of any one patent in the pool may be low or close to zero unless combined with the other patents in the pool. To take another example, consider a patent related to a microprocessor incorporated into mobile phones. A chip that provided some improvement (in speed, efficiency, etc.) may enable other functionality on the phone, such as an improved touchscreen interface, software applications with greater capability, greater video functionality, or improvement of other features of the phone. While apportionment would recommend that the royalty base be limited to the chip portion of the phone, this delineation may miss synergies between the patent at issue and the other features of the mobile phone. It would be incorrect to attribute all such synergies to the infringing company (or, for that matter, the patented feature). Instead, we can ask the more direct question of the additional profits that the manufacturer could be expected to make by incorporating the patented feature into its product. If the patented feature provides only a small improvement over existing technology (or, equivalently, a good non-infringing alternative to the patent exists), then the royalty should be limited. The manufacturer would not be willing to pay much for access to that technology. Using this same methodology, a major innovation would result in a higher royalty. In this way, an economic approach will produce damages awards that are consistent with the purpose of patent laws such that the incentive to innovate is commensurate with the value to society of the innovation. IV. A MARKET-BASED APPROACH FOR DETERMINING HOW TO DIVIDE VALUE BETWEEN TWO ASSET OWNERS Because any mechanical apportionment rule is inherently arbitrary (and thus would simply lead to irresolvable disputes in litigation), a better approach to determining how value should be divided between asset owners is to analyze how the asset owners would negotiate a split of the synergies. This approach is market-based and therefore is not arbitrary. Economic principles suggest that the negotiated payouts to the two companies must satisfy several conditions. First, the payout to company A, π A, must satisfy π A V A or company A would prefer to pursue its alternative (product A). The value 26 Love s model assumes that the various technologies of a product are additive, such that each additional patent adds value independent of other technologies but with no synergies between technologies. This leads him to conclude that the value of the entire product can only be attributed to the patented technology when the value of all the other components is zero. Love, supra note 6, at 276. When synergies exist among the technologies, this is not true. 262

180 Case 3:11-cv RBD-TEM Document Filed 07/25/13 Page 10 of 18 PageID Vol. XII Columbia Science and Technology Law Review 2011 from product A is the opportunity cost that Company A faces (i.e., the foregone profit) in pursuing product AB. Company A must earn a payout from Product AB that exceeds its opportunity cost. Similarly, the payout to company B must satisfy π B VB. The sum of the payoffs must of course satisfy π A + π B = VAB. Under these conditions, each company s payout can be thought of as the value of its alternative (the opportunity cost) plus a fraction of the gains to trade: π π A B = V = V A B + λs + ( 1 λ) S where λ is the negotiated split of the gains to trade. Economic models of bargaining suggest that the split of the gains to trade will be influenced by the companies relative levels of patience (as reflected by the rate at which they discount future expected cash flows). 27 The more patient company is willing to let the negotiations play out longer and therefore receives a larger split of the gains to trade than the less patient company. When the two companies are sufficiently patient and are roughly equally patient, the gains to trade will be approximately equally divided. 28 In this case, λ = Before applying this framework to patent licensing negotiations, we make an observation about patents as assets in this context. Unlike a physical asset, such as a manufacturing plant, the use of an intellectual property asset in one application does not necessarily preclude its simultaneous use in another application. 30 This can reduce the opportunity cost associated with using intellectual property in a given application. For example, suppose Company A s asset is a patented technology. By licensing Company B under its patent, Company A may not need to give up any opportunities to license the same technology to other companies that operate in different markets than Company B. We now consider several licensing examples. Example 1: Dividing Value Between Asset Owners When Both Companies Have Blocking Assets Suppose Company A holds a patent on a technology. Company A has no ability to produce any product itself, and there are no suitable licensees other than Company B. 27 See Ariel Rubinstein, Perfect Equilibrium in a Bargaining Model, 50 Econometrica, Jan. 1982, at As a technical matter, in the Rubinstein model of bargaining, the party that has the opportunity to make the first offer receives a slightly higher split of the gains to trade. This advantage goes to zero as the two parties become infinitely patient, or equivalently, as the time between offers goes to zero. 29 Whether it is appropriate to apply the framework described here to a particular real world situation will depend on the specific facts of the case. 30 Unless, of course, prevented by contractual obligations such as an exclusive license. 263

181 Case 3:11-cv RBD-TEM Document Filed 07/25/13 Page 11 of 18 PageID Vol. XII Columbia Science and Technology Law Review 2011 Company B has important complementary know-how to the patented technology, but cannot use this know-how to produce any product without a license to Company A s patent. If licensed, it could produce and sell product AB and generate profit V AB, which accounts for all expected future economic costs associated with developing and offering product AB. Under these assumptions, V A = 0, V B = 0, and S = VAB. If the two companies are equally patient, they will equally split the profits from the patented product. 31 Essentially, each company holds a blocking asset that is required to produce product AB, and they have no alternative use for their assets (we call this a case of completely blocking assets). The resulting Mexican standoff yields an equal split of the profits. In some situations it may seem counterintuitive that both companies have blocking assets. For example, suppose Company A uses its patented technology to make a product, product A, that is sold in a different market from the market in which product AB would be sold. Suppose further that product A is sold at $500 per unit and generates profits of $250 per unit. Company B, on the other hand, has no means by which to earn a profit on its know-how without getting a license from Company A. Product AB, which combines Company A s patented technology with Company B s know-how, would be sold at $5,000 per unit and would generate profits of $2,500 per unit. One might think that Company B has a blocking asset and thus Company B s know-how must be valued at $2,250 (the $2,500 profit per unit from Product AB less the $250 profit per unit for the product that utilizes Company A s patented technology). 32 However, this is incorrect as a matter of economics. Company A s patent is just as blocking as Company B s knowhow in making Product AB. This puts the two companies on equal footing, each able to block the other. As a result, they would negotiate an even split of the product AB profits (assuming both companies are sufficiently and equally patient). 33 Example 2: Dividing Value Between Asset Owners When Company A Has a Partially Blocking Asset Suppose Company A has a patented technology with no ability to produce a product itself, and there are no suitable licensees other than Company B. Company B has know-how, which allows it to produce a product B that generates profit V B. Licensing Company A s patented technology would allow Company B to offer an improved version of its product, product AB, that adds the product feature enabled by Company Again, assuming that both parties are also sufficiently patient. That is, V V. AB A 33 A related principle is that a company with more than one blocking asset does not get a larger piece of the pie than a company with only one blocking asset. However, in the case where the validity and infringement of patents are uncertain, so that blocking is uncertain, more than one potentially blocking patent (subject to validity and infringement) can strengthen a company s bargaining position because the probability that at least one of the patents turns out to be blocking increases with the number of patents. 264

182 Case 3:11-cv RBD-TEM Document Filed 07/25/13 Page 12 of 18 PageID Vol. XII Columbia Science and Technology Law Review 2011 A s patent. Product AB will generate profit V AB. If product AB is offered, Company B would no longer offer product B. However, if product AB is not offered, Company B has the ability to continue selling non-infringing product B. In this situation, Company A has a partially blocking patent on a product feature. In this case, V A = 0 and S = V AB VB. It is unclear how to implement apportionment in this case, since the value of Company A s patent can only be realized in combination with Company B s know-how. The economic approach recognizes that the two companies are negotiating only over the incremental profit S that adding Company A s patented feature would bring: if the feature is minor, S is a small fraction of V AB and the royalty rate would be relatively small. 34 V. ADDRESSING ROYALTY STACKING A. Apportionment Is a Crude Solution to Royalty Stacking Problems Many complex technology products incorporate multiple different features, each of which may be covered by a patent. For example, in Lucent, Microsoft Outlook was described as an enormously complex software program comprising hundreds, if not thousands or even more, features. 35 Royalty stacking refers to the potential problem that can arise from situations in which a single product may require a license from multiple patent holders. The total royalties paid by the manufacturer is the sum (or stack) of royalties paid to each individual patent holder. Some companies, particularly those manufacturers that produce complex products that incorporate multiple patents, have argued that the sum of the royalties paid to each individual patent holder may leave too little profit for the manufacturing company, reducing their own incentives to innovate. Some have even argued that royalties could exceed the total profit of a product. 36 In such a case, the 34 Note that in this bargaining situation, Company A does not have the ability to expropriate more than S because Company B can turn to its non-infringing alternative, valued at V B, if the royalty were too high ). Lucent Technologies, Inc., et al. v. Gateway, Inc., et al., 580 F.3d 1301, 1332 (Fed. Cir. 36 For example, in the Senate Judiciary Committee hearing on the 2009 Patent Reform Act, Steve Appleton, Chairman and CEO of Micron Technology, Inc., testified: The difficulty is that the current patent litigation system too easily allows damages to be assessed based on the value of the whole product, rather than the contribution of the patent. If we assume thousands of patents relate to this device, the resulting damages under current law would result in an amount that would exceed the total amount of revenue derived from the product. Patent Reform in the 111th Congress: Legislation and Recent Court Decisions Before the S. Comm. on the Judiciary, 111th Cong. 5 (2009) (statement of Steven Appleton, Chairman and 265

183 Case 3:11-cv RBD-TEM Document Filed 07/25/13 Page 13 of 18 PageID Vol. XII Columbia Science and Technology Law Review 2011 magnitude of the combined royalties would deter the introduction of the product, an economically inefficient outcome. 37 The royalty stacking problem has been cited as one reason that apportionment should be applied to calculate damages in patent infringement litigation. 38 Specifically, it is argued that apportionment is necessary because, when a product is covered by hundreds of patents, simple arithmetic shows that each patent can receive only a small royalty or the profit of the product would be exhausted entirely. However, again apportionment treats the symptom, not the disease. While apportionment can be used to reduce the size of the base to which a royalty rate is applied, apportionment does not address the cause of the royalty stacking problem. In contrast, the economic approach to calculating reasonable royalty damages has the potential to address royalty stacking issues directly because it explicitly can take into account multiple patent owners making claims on different synergies being generated by different combinations of technologies within the same product. B. Bargaining Between a Manufacturer and More Than One Patent Owner Having a Fundamental Technology If a manufacturer is negotiating with multiple patent holders, the economic analysis of licensing negotiations is considerably more complex than the situation in which a single manufacturer is negotiating with a single patent holder. Apportionment is ill-equipped to deal with such complexities. While apportionment attempts to divide up the various sources of profits between the contributing assets under the assumption that the whole is equal to the sum of the parts, determining the economic value of a patented technology requires recognizing and accounting for the synergies between technologies. We first consider a manufacturer negotiating with multiple patent owners where each of the patent owners holds a fundamental technology. 39 With multiple patent owners, the manufacturer may negotiate simultaneously with each patent holder or sequentially with each patent holder in turn. The equilibrium of a simultaneous CEO, Micron Technology, Inc.), available at See also, Mark A. Lemley & Carl Shapiro, Patent Hold-up and Royalty Stacking, 85 Tex. L. Rev. 1991, (2007) (explaining that because of the threat of an injunction, negotiated royalty rates can exceed the true economic contribution of the patented technology, especially when the value of the patented technology is small relative to the product s total value). 37 This outcome can result from the negative externalities that exist between owners of complementary patents, whereby each patent owner does not take into account the effects on the others of increasing its royalty. Patent pools are one remedy to this situation in that they internalize the externalities by putting royalty setting under the control of a single entity. 38 Love, supra note 6, at Whether it is appropriate to apply the framework described below to a particular real world situation will depend on the specific facts of the case. 266

184 Case 3:11-cv RBD-TEM Document Filed 07/25/13 Page 14 of 18 PageID Vol. XII Columbia Science and Technology Law Review 2011 bargaining situation in which one manufacturer and N patent owners are dividing up the profits of a product where they all hold a blocking position is, generally, an equal split of the profits among the parties. 40 For example, in the case of three players one manufacturer and two patent holders the equilibrium is that each player receives onethird of the profits. At first glance, it may seem that the sequential bargaining game, where the manufacturer negotiates with one patent owner and then another in turn, should lead to a different outcome because the portion of the profits that any one patented technology receives would seem to depend on the order in which the manufacturer values them. To illustrate, suppose a manufacturer wants to produce product AB, which incorporates patented technology A and patented technology B. Suppose both patented technology A and patented technology B are blocking in the sense that product AB cannot exist without both technologies. Consider the situation in which the manufacturer first values patent A and then, second, values patent B. The value of patent A, conditional on not having patent B, is zero because by itself patent A does not allow for production of product AB. The value of having patent B, conditional on already having patent A, is the entire value of product AB. Now consider the situation in which the manufacturer reverses the order: first valuing patent B and then, second, valuing patent A. By reversing the roles, now patent B is valued at zero while patent A is valued at the entire value of Product AB. This apparent paradox is once again the result of synergies being created when the two patented technologies are combined. An analysis of a sequential bargaining model again suggests how markets would resolve the apparent paradox. Suppose the manufacturer is negotiating with two patent holders, patent holder A and patent holder B, over licenses in order to produce Product AB. The outcome of a sequential move negotiation will generally be the same as the simultaneous negotiation an equal split of profits among the three parties as long as the players are equally and sufficiently patient. 41 This outcome can be explained using the following intuition. Suppose that the manufacturer has previously agreed to a royalty payment of amount R to patent holder A and is now engaged in negotiations with patent holder B. The total available profits to split between the manufacturer and patent holder B is V AB R. Because this is a bilateral negotiation, the equilibrium of the negotiation game is that the two parties will agree on an even split of the available profits (assuming equal and sufficient patience). In other 40 While the outcome of the game depends on the bargaining procedure considered in the model, for a range of possible bargaining models, the unique solution is an equal split of profits. See, e.g., Suchan Chae & Jeong-Ae Yang, The Unique Perfect Equilibrium of an N-Person Bargaining Game, 28 Econ. Letters 221, (1988); Suchan Chae & Jeong-Ae Yang, An N- Person Pure Bargaining Game, 62 J. Econ. Theory 86, (1994); Vijay Krishna & Roberto Serrano, Multilateral Bargaining, 63 Rev. Econ. Stud. 61, (1996); and Sang-Chul Suh & Quan Wen, Multi-Agent Bilateral Bargaining and the Nash Bargaining Solution, 42 J. Mathematical Econ. 61, (2006). 41 See, e.g., Chae & Yang The Unique, supra note 40, at ; Chae & Yang An N- Person, supra note 40, at 88 96; Krishna & Serrano, supra note 40, at 68 76; Suh & Wen, supra note 40, at

185 Case 3:11-cv RBD-TEM Document Filed 07/25/13 Page 15 of 18 PageID Vol. XII Columbia Science and Technology Law Review 2011 words, patent holder B will get a royalty of ( V AB R) / 2 and the manufacturer will get profits of ( V AB R) / 2. Now move back in time and consider the earlier negotiation between the manufacturer and patent owner A. The manufacturer will know what will happen in the later negotiation with patent owner B. Specifically, for whatever royalty R it pays patent owner A, the manufacturer knows that it will end up with ( V AB R) / 2. This means that the pie to be divided between the manufacturer and patent owner A is R + ( VAB R) / 2. Assuming equal and sufficient patience, the negotiated royalty R that splits this pie evenly is V AB / 3. Thus, the three parties each receive an even one-third split of the value of the product, the same as in the simultaneous bargain between the three parties. C. Bargaining Between a Manufacturer, One Patent Owner Having a Fundamental Technology, and One Patent Owner Having an Ancillary Technology We now compare the situation where a manufacturer is negotiating with two patent holders, one of which has a fundamental technology that is necessary to produce a product and the other of which is an ancillary feature that can be incorporated into the product. 42 The manufacturer has two choices. It can produce product A that only incorporates the fundamental technology, or it can produce product AB that incorporates both the fundamental and ancillary technology. Neither product A nor product AB would exist without the fundamental technology. Moreover, assume that neither product would exist without the participation of the manufacturer. While it may be tempting to use an apportionment rule to argue that the value of the fundamental technology is V A, the value of product A, while the value of the ancillary technology is VAB VA, the difference between the value of product AB and the value of product A, this is incorrect as a matter of economics for two reasons. First, it fails to consider that the manufacturer is blocking for product A. In this example, neither the value of product A nor the value of product AB can be realized without the manufacturer. Second, it fails to consider that the incremental value generated by Product AB over Product A could not be obtained without first having the fundamental technology. Thus, both the manufacturer and the fundamental technology owner can lay claim both to the value of Product A and to the incremental value of Product AB over Product A. The ancillary technology owner, however, can lay claim only to the incremental value. Apportionment rules would fail to reflect the complexities of this situation, potentially understating the value of the fundamental patent as V A and overstating the value of the ancillary technology as VAB VA. The economic approach, in contrast, is based on analyzing the likely outcome of a market-based negotiation among the parties. Assuming equal and sufficient patience, the manufacturer and fundamental technology owner would equally split among themselves 42 See generally, Richard J. Gilbert, Antitrust for Patent Pools: A Century of Policy Evolution, 3 Stan. Tech. L. Rev. 1 (2004), (discussing two-way blocking (both parties own patents with blocking positions) versus one-way blocking (a prior technology blocks the implementation of an improvement patent, but not vice-versa) in the context of patent pools). 268

186 Case 3:11-cv RBD-TEM Document Filed 07/25/13 Page 16 of 18 PageID Vol. XII Columbia Science and Technology Law Review 2011 the value of product A and would equally split the added benefit of the product AB over product A with the ancillary patent holder. In this situation, the manufacturer and the 1 1 fundamental technology owner would each receive a value equal to VA + ( VAB VA), 2 3 while the ancillary technology owner would receive its split of the incremental value of 1 product AB over product A, ( VAB V A ). 3 D. Incomplete Information Regarding Patent Infringement We discussed above the intuition for how a sequential negotiation between a manufacturer and multiple fundamental patent owners could lead to the same outcome as a simultaneous negotiation among the parties. The sequential outcome, however, is predicated on the manufacturer being aware of the existence of the second fundamental patent owner B and accounting for the future royalties it would pay for patent B in its negotiation with patent owner A. 43 We now consider the outcome if the manufacturer has less than complete information about the patents that might be asserted against its product. Consider the case in which the manufacturer negotiates with patent owner A, before it knows of the existence of patent owner B. As with the previous example, assume both patents are fundamental technologies. Companies that produce complex technologies may reasonably expect that there is some probability that they will face claims of infringement in the future from currently unknown patent owners. The possibility of additional royalty payments in the future is a factor that a manufacturer would take into account when negotiating with patent owner A, and would reduce the maximum royalty that the manufacturer would be willing to pay patent owner A. Consider, however, the situation in which the manufacturer is completely surprised by patent owner B s assertion of a patent infringing claim. This scenario corresponds to a situation in which apportionment might suggest that the royalty base be limited because the product incorporates many complex technologies and, therefore, royalty stacking issues must limit the royalty that the manufacturer should pay patent holder B. In other words, back when the manufacturer negotiated with patent owner A, it was under the assumption that patent A was the only patent that could be asserted against the manufacturer s product. In that negotiation, assuming both parties were equally and sufficiently patient, the manufacturer and patent holder A would have split V AB equally. When patent owner B unexpectedly sues for infringement, the manufacturer has only 43 Even if the hypothetical negotiation took place before the negotiation with other patent owners, any patents on which royalties are not currently being paid could, in theory, be taken into account based on the sequential bargaining analysis of the previous section. 269

187 Case 3:11-cv RBD-TEM Document Filed 07/25/13 Page 17 of 18 PageID Vol. XII Columbia Science and Technology Law Review 2011 V AB 2 to split with patent owner B. 44 manufacturer and patent owner B each get Thus, after the negotiation with patent owner B, the V AB 4 while patent owner A gets The first thing to note is that the economic approach does, in fact, account for ongoing royalties already being paid on the product, in that the manufacturer pays less to patent owner B than if it was not already paying royalties to patent owner A. 45 Nevertheless, the manufacturer ends up paying too much in royalties relative to the complete information case. Because it is surprised, the manufacturer ends up with only one-quarter of profits, whereas if it had negotiated simultaneously or under complete information, it would have received one-third of the profits. But it should also be noted that the manufacturer does not overpay the second patent owner, since patent holder B receives also receives less than in the complete information case (again, one-quarter rather than one-third). Thus, the effect of royalty stacking in this example is not that the later patent owner is overpaid, but that the manufacturer paid more to the earlier patent owner than it would have in the absence of surprise. 46 Apportionment aimed at limiting plaintiffs royalties (where the plaintiff is the later patent owner) does not address this issue. Parties to license agreements have come up with ways to deal with the prospect of paying royalties to unexpected patent holders. For example, agreements may incorporate a royalty adjustment mechanism, whereby the royalty rate is reduced if the licensee later has to pay royalties to other licensors. In the above example, the license with patent holder A would include a provision that royalty payments to patent holder A would be reduced from 50 percent to 33 percent if another patent holder with a fundamental technology later asserts its patents against the manufacturer s product. This contingency clause restores the efficient outcome of a simultaneous negotiation since total profits would again be evenly split between the manufacturer and the blocking patent holders. V AB 2 Such a provision fully mitigates the risk to the manufacturer in this example.. 47 Put 44 We assume that the manufacturer has no ability to go back and renegotiate with patent owner A. 45 We assume that the manufacturer is paying ongoing running royalties to patent owner A. Royalty stacking issues can still arise in situations where the manufacturer negotiated lump sum payments with previous patent holders as opposed to running royalties. 46 This raises questions about the incentives created by damages awards. If earlier patent holders receive a higher portion of the total profits, patent owners in general would have an incentive to make their patents known early, helping to mitigate royalty stacking issues. 47 When licenses are for ancillary technologies, the analysis becomes more complex. Nevertheless, such contingency clauses can still be included to mitigate the risk of future unknown patent holders. 270

188 Case 3:11-cv RBD-TEM Document Filed 07/25/13 Page 18 of 18 PageID Vol. XII Columbia Science and Technology Law Review 2011 differently, patent holder A helps pay the royalties for the newly asserted patent by reducing the royalties for patent A. 48 VI. CONCLUSION Proponents of apportionment have argued that excessive royalty damages awards can be curtailed by limiting the royalty base upon which royalty damages are awarded. Such apportionment rules, however, would be arbitrary and may under-compensate valuable innovations, particularly when significant synergies exist among technologies. An approach consistent with economic principles would largely eliminate damages in excess of the true economic value of a patent and align damages awards with incentives to innovate. Even under a proper economic approach, however, other factors may still lead to what some might believe are excessive damages awards. Consider, for example, the issue of the time at which a patent is valued under the U.S. law. Under this legal framework, a hypothetical negotiation is assumed to take place between the patent holder and the alleged infringer at the date of first infringement. In certain cases, the infringer may have previously made large sunk cost investments that are specific to the patent at issue, making a switch to a non-infringing alternative relatively more costly. Whether or not the timing of the negotiation results in excessive royalty awards presents separate economic and legal questions. Rather than dealing with these factors by attempting to limit royalty awards through arbitrary rules such as apportionment, employing an economic analysis based on the specific facts of the case will provide the greatest flexibility in identifying the true economic value of a patent in infringement cases. 48 In this case, before patent holder B shows up, the manufacturer and patent holder A shared 50 percent of future profits. When B shows up, the cost of that license is paid equally by the manufacturer and patent holder A, since the amount of future profits falls from 50 percent to 33 percent for both. 271

189 Case 3:11-cv RBD-TEM Document Filed 07/25/13 Page 1 of 5 PageID EXHIBIT 17

190 Case 3:11-cv RBD-TEM Document Filed 07/25/13 Page 2 of 5 PageID 12103