C R I T E R I O N
FRAND Royalties and Good-Faith Negotiations to License SEPs
The Way Towards the Internet of Things:Open Standards vs Silos
European University InstituteFlorence
November 15, 2019
J. Gregory SidakCriterion Economics
C R I T E R I O N
The Way Towards the Internet of Things. Open Standards vs Silos15 November 2019
C R I T E R I O N
Overview
1. Economics of the FRAND contract and FRAND royalties
2. Calculating FRAND royalties for 5G applications using hedonic price analysis
3. Economic ambiguity in the legal analysis of the duty to negotiate FRAND licenses in good faith
C R I T E R I O N
Economics of the FRAND contract and FRAND royalties
C R I T E R I O N
The FRAND contract
• FRAND is a matter of private contract– Overlays the national patent statute– Adds (relative to U.S. patent law):
• nonexclusion• nondiscrimination• fairness
– Competition law, if applicable, varies by jurisdiction
C R I T E R I O N
Important appeals pending
• Unwired Planet– Currently the most thorough and intellectually
rigorous analysis of the FRAND contract
• TCL v. Ericsson– Lacks Unwired Planet’s clarity and cogency
• FTC v. Qualcomm– Aggressive interpretation of an antitrust duty to
license at the component level
C R I T E R I O N
FRAND jurisprudence
• Contract interpretation, not policy making• Economic analysis must be relevant to the
question of fact that informs the question of law to be decided
• The controlling authority is – Contract language– Controlling law (France, New York)
• When is the contractual duty discharged?
C R I T E R I O N
Role of economic evidence
• Are lawyers posing, and expert economic witnesses answering, the correct question?– Example: Confusion over fundamental difference
between a contract offer and contract formation– Normative theories of patent valuation, which
courts do not in fact use (more later on this point)
C R I T E R I O N
Economic meaning of FRAND
• Contract interpretation seeks to give independent meaning to “fair,” “reasonable,” and “nondiscriminatory”– Depends on the specific language of the SSO’s
FRAND obligation, choice of law, and law on contract ambiguity of the controlling jurisdiction
• How can economic analysis assist judicial interpretation rather than be oblivious to it?
C R I T E R I O N
Reasonable
• A reasonable royalty is one that would result from voluntary exchange
• A bargaining range, not a unique point, bounded by– Licensor’s minimum willingness to accept– Licensee’s maximum willingness to pay
C R I T E R I O N
Principles of bargaining theory
C R I T E R I O N
A range of FRAND royalties
C R I T E R I O N
A FRAND royalty for a given SEPwill vary across licenses
C R I T E R I O N
Bargaining range for a RAND royalty (MWA = 0)
C R I T E R I O N
Bargaining range for a RAND royalty (MWA > 0)
C R I T E R I O N
When does the hypothetical negotiation occur?
• Moment of first infringement?• standard rule under U.S. patent jurisprudence
• Moment of standard adoption?• “ex ante incremental value” theory
• Moment when patent holder decided to monetize its technology by offering it to the SSO?
C R I T E R I O N
Fair
• What makes FRAND “fair” and distinguishable from RAND?
• Courts have ignored the question, contrary to conventional contract interpretation principles, without any explanation
• Can economics offer an answer?
C R I T E R I O N
What makes FRAND fair?
• “a division of surplus that is perceived by both parties to be fair maximizes the probability of contract formation”
J. Gregory Sidak, What Makes FRAND Fair? The Just Price, Contract Formation, and the Division of Surplus from Voluntary Exchange, 4 CRITERION J. ON INNOVATION 701 (2019), https://www.criterioninnovation.com/articles/what-makes-frand-fair/
https://www.criterioninnovation.com/articles/what-makes-frand-fair/
C R I T E R I O N
Nondiscriminatory
• Unwired Planet requirements for showing discrimination
• Similarly situated licensees• Disparately treated• Without objective justification
• Specific language of SSO’s FRAND or RAND commitment will vary
C R I T E R I O N
Different SSO wording“A license will be offered, with compensation, to applicants desiring to utilize the license for the purpose of implementing the JEDEC Standard under reasonable terms and conditions that are demonstrably free of any unfair discrimination.”
—JEDEC Patent Policy
Patent owner “undertaking in writing that it is prepared to grant irrevocable licences on fair, reasonable and non-discriminatory(‘FRAND’) terms and conditions”
—ETSI IPR Policy
C R I T E R I O N
Positive legal principles
• A finding of discrimination under U.S. law typically requires evidence of differential treatment of similar comparators
• The defendant may provide justifications for its differential treatment (even in the case of an unqualified prohibition against discrimination)
• Both cost-related and non-cost-related factors might justify differential treatment of similar comparators
C R I T E R I O N
Normative economic principles
• Economic principles might help to determine whether two licensees are similarly situated
• The analysis of differential treatment should focus on the SEP holder’s offer (rather than on the actual terms of the license)
• Economic justifications for differential treatment include the SEP holder’s financial distress, the implementer’s negotiating conduct, or changing market conditions
C R I T E R I O N
Calculating FRAND royaltiesfor 5G applications using
hedonic price analysis
C R I T E R I O N
Challenges for the IoT
• If the 5G application is new, there are no comparable licenses to use
• Patent pools are helpful but are already being challenged in antitrust litigation– Addresses nondiscrimination as it concerns
completely different uses of the SEPs– Establishes royalty ceiling
C R I T E R I O N
Challenges for the IoT, cont’d
• Ex ante (bottom up) calculations are infeasible– No data exist– Contrived depiction of competition
• Top-down calculations start with an assumption of an aggregate SEP royalty
• Hedonic price analysis uses real data from real transactions
C R I T E R I O N
Challenges for the IoT, cont’d
• Hedonic price analysis– What is the incremental value of connectivity in
an automobile that results from standard-essential patents?
– Scientific methodology that analyzes the actual facts of the case (Federal Rule of Evidence 702)
• Uses real data from real transactions• Econometric analysis is replicable
C R I T E R I O N
The theory of hedonic prices“When goods can be treated as tied packages of characteristics, observed market prices are also comparable on those terms. The economic content of the relationship between observed prices and observed characteristics becomes evident once price differences among goods are recognized as equalizing differences for the alternative packages they embody.”
—Sherwin Rosen Hedonic Prices and Implicit Markets:
Product Differentiation in Pure Competition, 82 Journal of Political Economy 34, 54 (1974)
C R I T E R I O N
DDR4 LRDIMM
C R I T E R I O N
DDR4 LRDIMM
C R I T E R I O N
The model
2 0 17] Hedonic Price s and Patent Royalt i e s 627
memory-module product determine its price. A hedonic regression model can reliably estimate the implicit price that consumers demonstrate that they are willing to pay for each of those distinguished features.
There might also be economy-wide or sector-wide effects that influence the price level of memory modules from year to year. To account for these effects, we control for the year in which each price was charged.
1. Using a Hedonic Regression Model to Identify Customers’ Demonstrated Willingness to Pay for a Memory Module’s Features
We specify a hedonic regression model to analyze the effect of each feature on the price of the memory module and separately determine how much customers demonstrated that they were willing to pay for a given feature or for an average memory module at a given time. Our hedonic regression model to identify the incremental value of the LRDIMM standard is expressed in the following equation:
Price = α + β1 × LRDIMM + β2 × DDR4 + β3 × GB + β4 × Year + ε
where LRDIMM is a dummy variable that takes a value of 1 if the memory module is LRDIMM-compliant, DDR4 is a dummy variable that takes a value of 1 if the memory module is DDR4-compliant, GB is a continuous variable for the memory module’s gigabyte capacity, and Year is a variable for each year for which we have price data. The beta coefficients (βi) in our hedonic regression model measure the average effect of a given feature on the final price of a given memory module. For example, β2 measures the average effect that being compatible with the DDR4 standard has on the price of a memory module. Similarly, β1 identifies the incremental value that being compatible with the LRDIMM standard adds to a memory module. The constant term, α, is the base-case product. It measures the value attributable to standard-ization and the product’s basic functionality. By using the next-best standard (namely, DDR3 RDIMM with 4 GB of capacity in 2013) as the base case, we can measure the value of standardization by carefully selecting appropri-ate data to make a valid comparison between similar standardized products. Table 1 reports the results of the hedonic regression on the price data.
where LRDIMM is a dummy variable that takes a value of 1 if the memory module is LRDIMM-compliant, DDR4 is a dummy variable that takes a value of 1 if the memory module is DDR4-compliant, GB is a continuous variable for the memory module’s gigabyte capacity, and Year is a variable for each year for which we have price data.
C R I T E R I O N
The value of the LRDIMM standard628 The Criter ion Jour nal on Innovation [Vol . 2 :601Table 1. Hedonic Regression Results for Memory-Module Prices
Variable Coefficient Standard Error t-StatisticLRDIMM 100.4716*** 35.0722 2.86DDR4 48.7459*** 15.8708 3.07GB 8 61.4273* 31.9943 1.92 16 108.5513*** 34.4684 3.15 32 223.0828*** 45.3401 4.92 64 577.4247*** 63.9095 9.04 128 1457.388*** 69.2066 21.06Year 2014 –9.3930 23.6438 –0.40 2015 –61.0179*** 23.3763 –2.61 2016 –128.272*** 23.0100 –5.57Constant 46.6667* 25.5737 1.82R2 0.8981F-Statistic 141.01Prob > F 0.0000N 171Source: De Dios & Associates (2016). Notes: * indicates statistical significance at the 90-percent confidence level, ** indicates statistical significance at the 95-percent confidence level, and *** indicates statistical significance at the 99-percent confidence level.
Table 1 reports that the patented technologies that allow the LRDIMM stan-dard to operate add $100.47 in value above the RDIMM standard to a given memory module. The LRDIMM variable identifies the incremental value of the LRDIMM standard (excluding value attributable to the module’s giga-byte capacity) to the DDR4 standard, and to market conditions prevailing in the year in which the memory module was sold. LRDIMM prices exceed RDIMM prices by a statistically significant amount, indicating that the LRDIMM variable has explanatory power and that we should include the LRDIMM variable in our hedonic regression model.
2. Testing the Statistical Significance and Robustness of the Hedonic Regression Model to the Inclusion of the LRDIMM and DDR4 Variables
Here, we discuss the statistical significance of our hedonic regression findings and confirm that we have correctly specified our model to identify the incre-mental value of the LRDIMM standard. The t-statistics that our regression
C R I T E R I O N
The Lorenz curve depicting value distribution
020
040
060
080
0C
itatio
n Valu
e (N
o Se
lf-C
itatio
ns)
0 10 20 30 40 50Patent Citation Rank
Google INPHILongitude MicronNetlist RambusRound RR Samsung
C R I T E R I O N
Distribution of value by company
C R I T E R I O N
RAND Royalty
Netlist’s Hypothetical
MWALicensee’s
MWP
Netlist’s Portion of the
Surplus
Licensee’s Portion of
the Surplus
[A] [B] [C]
([A] – [B]) ÷ ([C] – [B])
= [D]1 – [D]
= [E][1] $12.00 $0.00 $34.27 35.02% 64.98%[2] $12.00 $3.00 $34.27 28.78% 71.22%[3] $12.00 $6.00 $34.27 21.22% 78.78%[4] $12.00 $9.00 $34.27 11.87% 88.13%[5] $12.00 $12.00 $34.27 0.00% 100.00%
Division of surplus from a licensing agreement
C R I T E R I O N
The division of surplus will differ with different MWA
C R I T E R I O N
The same hedonic price analysis has been done for the features
of a smartphone
J. Gregory Sidak & Jeremy O. Skog, Hedonic Prices for Multicomponent Products, 4 CRITERION JOURNAL ONINNOVATION 301 (2019).
https://www.criterioninnovation.com/articles/hedonic-prices-for-multicomponent-products/
C R I T E R I O N
Negotiating FRAND licensesin good faith
C R I T E R I O N
Activity rule and closing rule
• What must a party do to stay active in the negotiation?
• When is the negotiation at an end, given that it has not resulted in contract formation?
C R I T E R I O N
Ambiguity across jurisdictions
• American principles of contract formation provide a clear closing rule
• French law, which governs ETSI, is less clear about a closing rule and silent on an activity rule
C R I T E R I O N
Proposed activity rule
To the extent that the SEP holder has not already discharged its contractual obligation to ETSI (such as by already having made a legitimately FRAND offer at the very outset of the negotiation)—a party must revise its bid or ask price by the minimum agreed-upon increment for that party to be deemed still to be negotiating in good faith.
– But how is the minimum bid increment determined?
C R I T E R I O N
Proposed closing rule
A party will be deemed to have made its final offer or counteroffer if it does not sweeten its price relative to its price in the previous round of offer and counteroffer.
C R I T E R I O N
https://www.criterioneconomics.com/
What Makes FRAND Fair? The Just Price, Contract Formation, and the Division of Surplus from Voluntary Exchange, 4 CRITERION JOURNAL ON INNOVATION 701 (2019)
Misconceptions Concerning the Use of Hedonic Prices to Determine FRAND or RAND Royalties for Standard-Essential Patents, 4 CRITERION JOURNAL ON INNOVATION 501 (2019)
Hedonic Prices and Patent Royalties: Epilogue, 4 CRITERION JOURNAL ON INNOVATION 401 (2019)
Hedonic Prices for Multicomponent Products, 4 CRITERION JOURNAL ON INNOVATION 301 (2019)
Judge Selna’s Errors in TCL v. Ericsson Concerning Apportionment, Nondiscrimination, and Royalties Under the FRAND Contract, 4 CRITERION JOURNAL ON INNOVATION 101 (2019)
Why Unwired Planet Might Revolutionize the Resolution of FRAND Licensing Disputes, 3 CRITERIONJOURNAL ON INNOVATION 601 (2018)
Is Patent Holdup a Hoax?, 3 CRITERION JOURNAL ON INNOVATION 401 (2018)
Recent articles on SEPs
https://www.criterioneconomics.com/https://www.criterioneconomics.com/sidak-what-makes-frand-fair.htmlhttps://www.criterioneconomics.com/sidak-misconceptions-concerning-hedonic-prices.htmlhttps://www.criterioneconomics.com/sidak-skog-hedonic-prices-and-patent-royalties-epilogue.htmlhttps://www.criterioneconomics.com/sidak-skog-hedonic-prices-for-multicomponent-products.htmlhttps://www.criterioneconomics.com/sidak-frand-errors-in-tcl-v-ericsson.htmlhttps://www.criterioneconomics.com/sidak-unwired-planet-and-frand-disputes.htmlhttps://www.criterioneconomics.com/sidak-is-patent-holdup-a-hoax.html
C R I T E R I O N
https://www.criterioneconomics.com/
Will the CJEU’s Decision in MEO Change FRAND Disputes Globally?, 3 CRITERION JOURNAL ON INNOVATION 301 (2018)
Citation Weighting, Patent Ranking, and Apportionment of Value for Standard-Essential Patents, 3 CRITERION JOURNALON INNOVATION 201 (2018)
The FRAND Contract, 3 CRITERION JOURNAL ON INNOVATION 1 (2018)
Hedonic Prices and Patent Royalties, 2 CRITERION JOURNAL ON INNOVATION 601 (2017)
Using Regression Analysis of Observed Licenses to Calculate a Reasonable Royalty for Patent Infringement, 2 CRITERIONJOURNAL ON INNOVATION 501 (2017)
Is a FRAND Royalty a Point or a Range?, 2 CRITERION JOURNAL ON INNOVATION 401 (2017)
Fair and Unfair Discrimination in Royalties for Standard-Essential Patents Encumbered by a FRAND or RAND Commitment, 2 CRITERION JOURNAL ON INNOVATION 301 (2017)
Enhanced Damages for Infringement of Standard-Essential Patents, 1 CRITERION JOURNAL ON INNOVATION 1101 (2016)
Recent articles on SEPs, cont’d
https://www.criterioneconomics.com/https://www.criterioneconomics.com/sidak-petrovcic-meo-and-frand-disputes.htmlhttps://www.criterioneconomics.com/sidak-skog-citation-weighting-and-sep-ranking.htmlhttps://www.criterioneconomics.com/sidak-the-frand-contract.htmlhttps://www.criterioneconomics.com/hedonic-prices-and-patent-royalties_1.htmlhttps://www.criterioneconomics.com/regression-analysis-of-observed-licenses.htmlhttps://www.criterioneconomics.com/is-a-frand-royalty-a-point-or-a-range.htmlhttps://www.criterioneconomics.com/fair-and-unfair-discrimination.htmlhttps://www.criterioneconomics.com/enhanced-damages-infringement-standard-essential-patents.html
C R I T E R I O N
https://www.criterioneconomics.com/
A FRAND Contract’s Intended Third-Party Beneficiary, 1 CRITERION JOURNAL ON INNOVATION 1001 (2016)
How Commissioner Vestager’s Mistaken Views on Standard-Essential Patents Illustrate Why President Trump Needs a Unified Policy on Antitrust and Innovation, 1 CRITERION JOURNAL ON INNOVATION 721 (2016)
What Aggregate Royalty Do Manufacturers of Mobile Phones Pay to License Standard-Essential Patents?, 1 CRITERION JOURNAL ON INNOVATION 701 (2016)
Does the International Trade Commission Facilitate Patent Holdup?, 1 CRITERION JOURNAL ONINNOVATION 601 (2016)
Testing for Bias to Suppress Royalties for Standard-Essential Patents, 1 CRITERION JOURNAL ONINNOVATION 301 (2016)
Tournaments and FRAND Royalties, 1 CRITERION JOURNAL ON INNOVATION 101 (2016)
Recent articles on SEPs, cont’d
https://www.criterioneconomics.com/https://www.criterioneconomics.com/a-frand-contracts-intended-third-party-beneficiary.htmlhttps://www.criterioneconomics.com/vestager-standard-essential-patents-antitrust-innovation-policyhttps://www.criterioneconomics.com/aggregate-royalty-to-license-standard-essential-patents.htmlhttps://www.criterioneconomics.com/does-the-itc-facilitate-patent-holdup.htmlhttps://www.criterioneconomics.com/bias-to-suppress-royalties-for-standard-essential-patents.htmlhttps://www.criterioneconomics.com/tournaments-and-frand-royalties-for-standard-essential-patents.html
