Market-Derived Patent Data: What Data is Important and How Does It Impact Value? Mike Pellegrino President, Pellegrino & Associates, LLC mike@pellegrinoandassociates.com Some Major Headlines Google Buys 1,023 IBM Patents to Bolster Defense of Android Facebook Is Said to Buy 750 IBM Patents to Boost Defense Facebook Buys AOL Patents From Microsoft for $550 Million AOL Announces Close of $1.056 Billion Patent Transaction with Microsoft Intel to buy InterDigital patents for $375 million Acacia Subsidiary Acquires Patents Related to Cellular HSPA and LTE Technology from a Major Corporation Why is all of this occurring? 2015 Pellegrino and Associates, LLC. All rights reserved. 2 1
Economic justification Why? Authorized monopoly Patents allow for abnormal profits via price skimming Right patent is worth more than most companies! Defensive posturing drives significant recent patent activity Microsoft received 86 patent grants in 2000 Now within the top 5 recipients in any given week for new U.S. patent grants 2015 Pellegrino and Associates, LLC. All rights reserved. 3 Microsoft s Scorecard 2015 Pellegrino and Associates, LLC. All rights reserved. 4 2
Microsoft s Result 2015 Pellegrino and Associates, LLC. All rights reserved. 5 Ignorance is NOT Bliss What happens when a patent-centric consumer-based business stops investing in IP: 2015 Pellegrino and Associates, LLC. All rights reserved. 6 3
Ignorance is NOT Bliss This happens in pharmaceuticals too: 2015 Pellegrino and Associates, LLC. All rights reserved. 7 This is Big Business >94% of all patents assigned at time of award to companies Top 300 patent holders own >half of the issued patents Lots of money at stake At $10,000 per patent, prosecution fees top $4 billion per year Does not include internal staff time (e.g., patent review committees), consultants, etc. >2,500 patent lawsuits enter court system each year >$7.5 billion dedicated to patent litigation, damages, productivity losses, etc. 2015 Pellegrino and Associates, LLC. All rights reserved. 8 4
Main U.S. Patent Class Generalized Hot Development Areas Patent Count Gov t Funded Abandonment Rate 438 Semiconductor Device Manufacturing: Process 65,983 598 11.02% 257 Active Solid State Devices 60,801 665 8.97% 370 Multiplex Communications 54,525 303 6.75% 514 Drug, Bio Affecting And Body Treating Compositions 50,056 2,780 27.25% 455 Telecommunications 43,444 115 7.87% 435 Chemistry: Molecular Biology And Microbiology 41,890 4,424 22.54% 345 Computer Graphics Processing And Selective Visual Display Systems 709 Electrical Computers And Digital Processing Systems: Multicomputer Data Transferring 35,871 116 10.56% 34,399 155 6.46% 424 Drug, Bio Affecting And Body Treating Compositions 32,943 1,914 19.90% 428 Stock Material Or Miscellaneous Articles 32,546 467 20.48% 2015 Pellegrino and Associates, LLC. All rights reserved. 9 New Market Developments Rise of non-practicing entities (NPEs) WiLAN, Conversant, Rambus, InterDigital, Round Rock Research, Intellectual Ventures Interesting business models Own IP Do not practice it directly in the market Assert patent rights against practicing entities Seek historical damages, future royalties, and/or one-time payments 2015 Pellegrino and Associates, LLC. All rights reserved. 10 5
New Developments Brilliant, if controversial monetization strategy As a NPE, there is no practical risk of countersuit NPE does not sell any infringing products E.g., if WiLAN sues Cisco, Cisco can t countersue for infringement Removes cross-licensing lever for alleged infringer Works to NPE advantage Cross licensing creates little economic activity Twist: Large companies are realizing value from NPEs! 2015 Pellegrino and Associates, LLC. All rights reserved. 11 NPE Twist Many large companies hold remarkably strong patent portfolios E.g., Nokia has thousands of patents After several thousand patents, incremental economic opportunities erode E.g., Nokia inks deal with Apple for most of portfolio Nokia loses ability to enforce against Apple for new patents Nokia could then have a long-term challenge monetizing its IP 2015 Pellegrino and Associates, LLC. All rights reserved. 12 6
NPE Twist What if Nokia sells a portion of its IP to a NPE? Nokia receives cash up front Or Nokia receives royalties from NPE This is incremental revenue that Nokia would not have had otherwise NPE can sue Apple or others Apple and others have no lever against Nokia Large companies in effect outsource enforcement efforts Create incremental value by doing so Micron, Nokia, and others have pursued this strategy Impact: NPEs have created more fluid patent marketplace 2015 Pellegrino and Associates, LLC. All rights reserved. 13 Data-Derived Patent Metrics 7
Why Consider Macro Patent Data? Key innovations typically rely on patents Such data is publicly available Patent data is authoritative Published by U.S. government Statistical analysis of trends in data informs market on variety of levels Impact: Insights open doors to new influencers on patent value (both positive and negative) 2015 Pellegrino and Associates, LLC. All rights reserved. 15 Analysis Data Sources Integrate a variety of U.S. patent office data (USPTO) Patent grants (back to 1976) Published patent applications (back to 2000) Recorded patent assignments (back to 1980) Maintenance fee events (back to 1980) European Telecommunications Standards Institute (ETSI) All data related to cellular-telephone related standards-essential patent declarations 2015 Pellegrino and Associates, LLC. All rights reserved. 16 8
About this data Data sources have limitations Errors in data that the USPTO and ETSI receive Some errors are trivial (i.e., misspelled name) Some errors are more serious (wrong dates on patents) Much of the data is impossible to change Patents are legal documents and require costly legal documents to make edits or corrections Each data source tracks data differently Mapping must occur, which is a noisy process Such noise does not alter any conclusions materially 2015 Pellegrino and Associates, LLC. All rights reserved. 17 Grant activity Application activity Application/grant (A/G) ratio Patent priority Citation activity Transaction activity Maintenance rates Metrics of Interest 2015 Pellegrino and Associates, LLC. All rights reserved. 18 9
Grant Activity Rate of change in grants indicates current/historic market significance For example: Few patents in button making Many patents in multiplex communications Grant activity has increased over time Conclusion: Patents for multiplex communications are of greater interest to the market than they are for button making Button making patents are likely worth less and/or there is no buyer 2015 Pellegrino and Associates, LLC. All rights reserved. 19 Grant Activity Location of art indicates market significance Perform search for diabetes Most patents are in chemical compound and biotech art classes Little possible relevance if end device is a diagnostic for measuring glucose levels In such a situation, most diabetes patents bring little relevance to analysis Data informs us that market is more interested in treating disease via drugs versus devices Impact: Lack of data can be helpful in opining on patent value 2015 Pellegrino and Associates, LLC. All rights reserved. 20 10
Application Activity Patent applications are leading indicator to patent grants Rate of change in application indicates current/future market significance For example: Many new patents in soft tissue repair for spinal discs Conclusion: Possible growth market for an orthopedic company Like patent grant activity, location of art indicates market significance 2015 Pellegrino and Associates, LLC. All rights reserved. 21 Application/Grant Ratio A/G ratio provides evidence of market significance over time Total patent applications / total patent grants Interpretation <1 : Market interest is waning as new filings lag grants =1 : Market s interest is stable and maintained as new filings keep up with patent grants >1 : Market interest is growing as new filings exceed grants Impact: A company s A/G ratio can tell you whether their pace of innovation is slowing or not. Same holds true for broader market 2015 Pellegrino and Associates, LLC. All rights reserved. 22 11
Patent Priority Patents have an associated priority Think of patent priority like a place in line Patent priority is based on a date That date is the point where invention novelty evaluation begins Patents with earlier priority dates are closer to line front Patents with earlier priority dates might exhibit a greater degree of novelty and be more foundational in nature Patents with later priority dates might exhibit a lesser degree of novelty and be more incremental in nature Impact: Foundational patents generally worth more 2015 Pellegrino and Associates, LLC. All rights reserved. 23 Patent Priority Profile 2015 Pellegrino and Associates, LLC. All rights reserved. 24 12
Patent Citations Patent citations are references to other granted patents or published patent applications Similar to references in academic publications Have relevance to the patent of interest Much interest exists in using citations to indicate patent quality value Some academic research indicates that companies that have high degrees of citation to their patents have greater market values 2015 Pellegrino and Associates, LLC. All rights reserved. 25 Patent Citations A high degree of citations is indicative of market significance That intelligence can be valuable Helps identify possible licensees and acquirers Provides basis for inspiration to other inventors in the market P&A remains skeptical to a direct connection between patent citations and patent quality or patent value Market players like metric though Impact: Highly cited patents generate greater market interest 2015 Pellegrino and Associates, LLC. All rights reserved. 26 13
Transaction Activity Transactions may comprise: Mergers Assignments Security interests (e.g., corporate financing) And other rights conveyances Searches that yield many transactions may indicate: List of possible licensees or buyers for given technology Active interest and market relevance in related technologies Quality of companies buying patents may provide significant credibility boost to patented concept 2015 Pellegrino and Associates, LLC. All rights reserved. 27 Maintenance Rates Patents are enforceable only when maintained If patent owner does not maintain patent, then no enforcement lever exists Patent maintenance is expensive Gets incrementally more expensive at each maintenance event Large companies can spend tens of millions of dollars per year on maintenance fees Strong relationship between valuable patents and maintained patents Foolish to let a patent lapse that is generating value Patent maintenance rates can inform on economic lives of assets Communications companies maintain >98% of their patents Large pharmaceutical companies abandon >70% of their patents 2015 Pellegrino and Associates, LLC. All rights reserved. 28 14
Apple v. Samsung Putting empirical data to work The Question Apple and Samsung have been slugging it out in court for years Many publications tout the strength of Apple s patent portfolio Some claim it is the most powerful in consumer electronics Is this true? Apple generates remarkable profits from its products Do Apple s profits emerge primarily from patents or other items? 2015 Pellegrino and Associates, LLC. All rights reserved. 30 15
Patent Activity Volume Volume of data even for Apple and Samsung is daunting Apple has >7,000 U.S. patents Samsung has >50,000 U.S. patents across all related entities If one spent one hour reading each patent, it would take nearly 30 years to complete that assignment Naturally, cannot afford the cost of such efforts Perfect opportunity for large data set analysis 2015 Pellegrino and Associates, LLC. All rights reserved. 31 Analysis Methodology Objective comparative analysis of patent portfolios is: Complex Labor intensive USPTO employed nearly 8,000 examiners in 2012 Handled 565K applications in 2012 (about 70 per examiner) Cost $1.6 billion in 2012 Just one art class may require team of 10 people to monitor weekly changes 2015 Pellegrino and Associates, LLC. All rights reserved. 32 16
Analysis Methodology Need a common metric that is easy to use In this case, patent count is the metric Patent counts have warts Does not speak to actual utility if claimed inventions However, we want macro, not micro conclusions Many warts go away when focused on macro conclusions and on maintained patents 2015 Pellegrino and Associates, LLC. All rights reserved. 33 Mobile Phone Patents Start with relative position of each company Top U.S. classes that include mobile phones include 370 Multiplex communications 375 Pulse or digital communications 455 Telecommunications Focus is on patents between January 1, 2000 and December 31, 2012 2015 Pellegrino and Associates, LLC. All rights reserved. 34 17
Patent Class 370 Top companies include many mobile phone manufacturers Where is Apple? Ranked 93 rd 2015 Pellegrino and Associates, LLC. All rights reserved. 35 Top companies include many mobile phone manufacturers Where is Apple? Ranked 51 st Patent Class 375 2015 Pellegrino and Associates, LLC. All rights reserved. 36 18
Patent Class 455 Top companies include many mobile phone manufacturers Where is Apple? Ranked 51 st here too 2015 Pellegrino and Associates, LLC. All rights reserved. 37 All Together Now Apple generated 0.27% of the total patents Certainly not data signature of a thought leader against Samsung 2015 Pellegrino and Associates, LLC. All rights reserved. 38 19
Apple Portfolio Composition 2015 Pellegrino and Associates, LLC. All rights reserved. 39 Apple Portfolio Observations Apple has a remarkable focus on design and touch elements Top classes are D14 and 345 Greatest emphasis is in design patents Emerge from about a dozen key people at Apple 45% of Apple s patents are parent applications Vertically integrated portfolio Vertically integrated portfolios are generally worth less Data suggests: Apple is brilliant at industrial design And branding/marketing, but that is outside of our scope today Basic science behind advanced features and functions of iphone emerges from other company s labs 2015 Pellegrino and Associates, LLC. All rights reserved. 40 20
Samsung Portfolio Composition 2015 Pellegrino and Associates, LLC. All rights reserved. 41 Samsung Portfolio Observations Samsung has one of largest portfolios in the world Remarkably diverse Semiconductors, communications technologies, display technologies, etc. Portfolio composition says more basic research comes from Samsung Empirical evidence suggests this too Apple buys >$8B in parts from Samsung Fortune 350-sized company if relationship were standalone Samsung buys no parts from Apple 2015 Pellegrino and Associates, LLC. All rights reserved. 42 21
Samsung Portfolio Observations 80% of Samsung s patents are parent applications Horizontally integrated portfolio Horizontally integrated portfolios are generally worth more Apple is cash rich, but patent poor compared to Samsung 2015 Pellegrino and Associates, LLC. All rights reserved. 43 Benchmarking the Portfolios Study commonality of prosecution efforts at the class subclass level using patent counts Areas where there are few patent count differences indicates reasonable parity in development efforts Areas where there are large patent count differences indicates significant disparity in development efforts Can move advantage to the company with the larger portfolio Indicates an attack point for the weaker company 2015 Pellegrino and Associates, LLC. All rights reserved. 44 22
Benchmarking the Portfolios 2015 Pellegrino and Associates, LLC. All rights reserved. 45 Interpreting the Data 68.7% of Apple s portfolio overlaps 19.08% of Samsung s Apple has greater concentration of patents where Samsung competes Samsung is much more diversified than Apple Samsung is outgunning Apple more than 2:1 on raw patent production Samsung was outgunning Apple in key classes by margin of 10:1 or greater E.g., Class 370, Samsung has 825 patents, Apple has 67 This data signature was common across the Apple/Samsung portfolio Impact: Apple is a likely patent buyer while Samsung is not 2015 Pellegrino and Associates, LLC. All rights reserved. 46 23
Apple s Buying Spree If you can t invent it, buy it Apple has record of doing this SoundJam itunes FingerWorks Gestures in iphones Placebase maps Lala.com itunes, icloud Siri Siri And many others Apple does not have the deep bench that Qualcomm, Nokia and others on the telecommunications side 2015 Pellegrino and Associates, LLC. All rights reserved. 47 Apple has bought many patents Continues to do so to this day Mostly in key areas where it has technical deficiencies It is a patch Does not fix structural issues Samsung buys few patents at all Apple s Buying Spree 2015 Pellegrino and Associates, LLC. All rights reserved. 48 24
Individual Patent Analysis Patent Value Some claim a patent s value is unknown until tested in court Ample evidence to dispute such a claim Nevertheless, surviving in litigation has direct impact on patent value Patent value can swings millions of dollars based on litigation record Surviving variety of post-grant patent reviews is crucial 2015 Pellegrino and Associates, LLC. All rights reserved. 50 25
Public Use/On Sale Bar Prior public disclosures Inventorship issues Novelty issues Obviousness issues Abandonment issues Foreign patent limitations Interference hearings Typical IPR/PGR Warts 2015 Pellegrino and Associates, LLC. All rights reserved. 51 Good Patent Signatures Early priority date Clean inventorship Seasoned patent examiner Seasoned patent prosecution agent Part of larger patent prosecution family Active market for inventions (e.g., transactions) High degree of market inspiration (e.g., forward citations) Active patent prosecution history without material changes to patent claims 2015 Pellegrino and Associates, LLC. All rights reserved. 52 26
Patent Examiner Analysis Analyze typical grant activity from patent activity Impact: Experience = Greater search fidelity, stronger claims Assigned Patent Applications 450 400 350 300 250 200 150 100 50 0 Patent Examiner Prosecution Experience 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 Year 2,000 1,800 1,600 1,400 1,200 1,000 800 600 400 200 Cumulative Assigned Applications Patent Application Count Cumulative Patents 2015 Pellegrino and Associates, LLC. All rights reserved. 53 Patent Priority Patents with earlier priority dates may be more valuable Impact: Less impeaching art = less invalidity risk Patent Priority 35 30 120.00% 100.00% Patent Count 25 20 15 10 80.00% 60.00% 40.00% % of Cumulative Patents 5 20.00% 0 1990 1993 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2011 Year 0.00% 2015 Pellegrino and Associates, LLC. All rights reserved. 54 27
Patent Transaction Activity Companies actively transact for valuable patents Impact: Art classes with assignments = market interest/value Patent Transaction Activity 60 120% 50 100% Patent Count 40 30 20 80% 60% 40% % of Cumulative Patents Sold 10 20% 0 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 Year 0% 2015 Pellegrino and Associates, LLC. All rights reserved. 55 Consider patent class 705 Enforcement Economics Data Processing: Financial, Business Practice, Management, or Cost/Price Determination 4,157 cases filed in parent patent class 705 164 cases result in an award or injunction 53 cases result in damages 47 cases result in injunction 23 cases include litigation costs 2015 Pellegrino and Associates, LLC. All rights reserved. 56 28
Class 705 Damages Record 36 cases analyzed between 2011-2014 Actual damages/enhanced damages awarded About 33% of awards come from jury Average damages award is substantial >$19M over past 3.5 years 2013 award is lower because of several outliers Median damages award: $6,675,000 Eastern district of Texas is optimal Judge Davis had most cases, average award of $6.5M Impact: At trial, damages are high Year Average Damages Award ($) 2011 30,124,606 2012 11,135,554 2013 3,290,191 2014 37,002,417 Total 19,385,989 2015 Pellegrino and Associates, LLC. All rights reserved. 57 Class 705 Damages Trial Record 36 cases analyzed between 2011-2014 Impact: Trial is not something to fear Decider 2011 (M$) 2012 (M$) 2013 (M$) 2014 (M$) Total (M$) Judge Roy Payne 328.6 328.6 Jury 62.4 49.6 8.1 53.1 Judge David Folsom 26.8 26.8 Judge Robert Payne 17.0 17.0 Judge Virginia Covington 9.7 9.7 Judge Mark Hornak 8.1 8.1 Judge Leonard Davis 6.6 6.6 Judge Rodney Gilstrap 0.6 5.3 2.5 2015 Pellegrino and Associates, LLC. All rights reserved. 58 29
Litigation Analysis: Microsoft Average Microsoft litigation duration Slightly faster than what is typical 80% of cases close within 3 years of filing Litigation Duration Years 120.00% 100.00% 80.00% Mean 1.90 60.00% Median 1.53 40.00% Maximum 13.38 20.00% # Cases 534 0.00% 0 Cumulative % of Cases Microsoft Patent Litigation Duration Frequency (as Defendant) 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0 5.5 6.0 6.5 7.0 7.5 8.0 8.5 Litigation Duration (years) Frequency Cumulative % 120 100 80 60 40 20 Number of Cases 2015 Pellegrino and Associates, LLC. All rights reserved. 59 Litigation Analysis: Google Average Google litigation duration Litigation duration is about average 80% of cases close within 2.5 years of filing Litigation Years Duration Mean 1.53 Median 1.21 Maximum 7.07 # Cases 147 Cumulative % of Cases 120.00% 100.00% 80.00% 60.00% 40.00% 20.00% Google Patent Litigation Duration Frequency (as Defendant) 0.00% 0 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0 5.5 6.0 6.5 7.0 7.5 8.0 8.5 Litigation Duration (years) 40 35 30 25 20 15 10 5 Number of Cases Frequency Cumulative % 2015 Pellegrino and Associates, LLC. All rights reserved. 60 30
Litigation Analysis: Apple Average Apple litigation duration Apple closes cases quickly on average 80% of cases close within 2 years of filing Litigation Years Duration Mean 1.17 Median 0.80 Maximum 7.48 # Cases 280 Cumulative % of Cases 120.00% 100.00% 80.00% 60.00% 40.00% 20.00% Apple Patent Litigation Duration Frequency (as Defendant) 0.00% 0 0 0.5 1 1.5 2 2.5 3 3.5 4 4.5 5 5.5 6 6.5 7 7.5 8 8.5 Litigation Duration (years) 120 100 80 60 40 20 Number of Cases Count Cumulative % 2015 Pellegrino and Associates, LLC. All rights reserved. 61 Practical Implications A patent with a good signature is worth litigating If one gets to trial, the economic benefits of doing so are favorable Large companies do not want to get to trial Many settle early in the litigation process 2015 Pellegrino and Associates, LLC. All rights reserved. 62 31
Tying Data to Patent Value What Do We Do With The Data? Patent valuation does not start or end on statistical analysis No direct connection in absolute terms (e.g., dollars) Statistics inform on a variety of factors important to valuation process New insights and value influencers may include: Litigation value Licensee analysis Economic life of the patent Exit analysis (i.e., possible buyers) Competitors at the company level Competitors at the product or service level Market interest in the technology This may impact forecast growth rates 2015 Pellegrino and Associates, LLC. All rights reserved. 64 32