Case 5:21-cv-03267-LB Document 1 Filed 05/03/21 Page 1 of 147
`
`
`
`Steve W. Berman (pro hac vice forthcoming)
`HAGENS BERMAN SOBOL SHAPIRO LLP
`1301 Second Avenue, Suite 2000
`Seattle, WA 98101
`Telephone: (206) 623-7292
`Facsimile: (206) 623-0594
`steve@hbsslaw.com
`
`Benjamin J. Siegel (256260)
`Rio S. Pierce (298297)
`HAGENS BERMAN SOBOL SHAPIRO LLP
`715 Hearst Avenue, Suite 202
`Berkeley, CA 94710
`Telephone: (510) 725-3000
`Facsimile: (510) 725-3001
`bens@hbsslaw.com
`riop@hbsslaw.com
`
`Attorneys for Plaintiffs
`
`
`UNITED STATES DISTRICT COURT
`
`NORTHERN DISTRICT OF CALIFORNIA
`
`
`
`
`Case No. __________________
`
`CLASS ACTION COMPLAINT
`
`
`JURY TRIAL DEMANDED
`
`
`
`
`BRYAN LEE, ALEXANDER DALZIEL,
`BENJAMIN MURRAY, JARED VAN
`VALKENBURG, BRETT CLAVIER, BEN
`FABER, ANTHONY ANNESE, MICHAEL
`REILLY, JOSHUA LODGE, ERIC ROMERO,
`JONATHAN CLAY, JOHN HICKS, MARY
`STOCKTON, ZACHARY PRATER, and
`BENJAMIN ANDERSON, individually and on
`behalf of all others similarly situated,
`
`
`v.
`
`MICRON TECHNOLOGY, INC., MICRON
`SEMICONDUCTOR PRODUCTS, INC.,
`SAMSUNG ELECTRONICS CO., LTD.,
`SAMSUNG SEMICONDUCTOR, INC., SK
`HYNIX, INC., and SK HYNIX AMERICA,
`INC.,
`
`
`
`
`Plaintiffs,
`
`Defendants.
`
`010742-11/1496419 V1
`
`
`
`-1-
`
`1
`2
`3
`4
`5
`6
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`24
`25
`26
`27
`28
`
`

`

`Case 5:21-cv-03267-LB Document 1 Filed 05/03/21 Page 2 of 147
`
`TABLE OF CONTENTS
`
`Page
`INTRODUCTION AND OVERVIEW OF THE CONSPIRACY ...................................... 1 
`JURISDICTION AND VENUE ........................................................................................ 19 
`THE PARTIES .................................................................................................................. 21 
`A. 
`Plaintiffs ................................................................................................................ 21 
`B. 
`Defendants ............................................................................................................. 23 
`C. 
`Agents and Non-Party Co-Conspirators ................................................................ 24 
`DESCRIPTION OF DRAM .............................................................................................. 25 
`A.  What is DRAM? .................................................................................................... 25 
`B. 
`How is DRAM Manufactured? ............................................................................. 26 
`C. 
`How is DRAM Sold to Direct Purchasers? ........................................................... 27 
`D. 
`How is DRAM Sold to Indirect Purchasers? ......................................................... 29 
`E. 
`The Markets for DRAM and DRAM Devices. ..................................................... 30 
`DEFENDANTS CONSPIRED TO RESTRAIN COMPETITION FOR
`SALES OF DRAM ............................................................................................................ 43 
`Prior to the Start of the Class Period, Defendants Made
`A. 
`Independent Supply and Capacity Decisions, Leading to Declining
`DRAM Prices ........................................................................................................ 43 
`Between August 2014 and the end of 2015, competition
`1. 
`between Defendants caused DRAM prices to decline .............................. 43 
`Defendants’ Supply and Pricing behavior changed abruptly
`and nearly simultaneously at the beginning of the Class
`Period ......................................................................................................... 43 
`Samsung tried on its own to limit supply and raise
`a. 
`prices and when that failed, it attempted to secretly
`communicate its intention to raise prices to the other
`Defendants to spur joint action. ..................................................... 45 
`In 2015, the Defendants grew supply at the same
`rate as their forecasted growth in demand, but
`starting in 2016 Defendants intentionally restricted
`supply so that it grew slower than demand. .................................. 45 
`The Class Period Begins No Later than June 1, 2016, as
`Defendants’ Efforts to Coordinate to Ensure Capacity Discipline
`
`2. 
`
`b. 
`
`
`
`I. 
`II. 
`III. 
`
`IV. 
`
`V. 
`
`1
`2
`3
`4
`5
`6
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`24
`25
`26
`27
`28
`
`B. 
`
`010742-11/1496419 V1
`
`
`
`-i-
`
`

`

`Case 5:21-cv-03267-LB Document 1 Filed 05/03/21 Page 3 of 147
`
`
`
`1
`2
`3
`4
`5
`6
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`24
`25
`26
`27
`28
`
`C. 
`
`VI. 
`
`2. 
`3. 
`
`4. 
`
`5. 
`
`and DRAM Price Increases That Started in Early 2016 Began to
`Bear Fruit ............................................................................................................... 47 
`From early 2016, Defendants engaged in concerted
`1. 
`signaling to each other through public statements and
`actions in response that effectuated an agreement between
`Defendants to artificially restrict supply growth of DRAM
`– causing DRAM prices to skyrocket. ....................................................... 47 
`Defendants’ public statements. .................................................................. 48 
`In the second half of 2016, Defendants sequentially
`reduced output, thus sacrificing profits and market share
`against each of their own unilateral self-interests. .................................... 68 
`Throughout 2017, Defendants repeatedly reassured each
`other that, unlike in previous periods, each would not
`respond to rising prices and strong demand with increased
`supply growth. Instead, the Defendants would stick with
`their publicly announced plans to keep supply growth
`below demand growth by not adding wafer capacity and
`not seeking to take market share from each other. .................................... 70 
`From 2016-2017, Samsung changed its prior behavior and
`stopped trying to take market share from competitors,
`which is consistent with the collusive objectives identified
`by Confidential Witness 1. ........................................................................ 73 
`End of Conspiracy: Once the Chinese Investigation was
`Announced, Defendants’ Conduct Changed Abruptly .......................................... 75 
`DEFENDANTS’ INCREASED PRICES FOR DRAM CANNOT BE
`EXPLAINED BY MARKET FACTORS ......................................................................... 77 
`A. 
`DRAM costs remained low or stable during the Class Period .............................. 78 
`B. 
`Price increases during the Class Period cannot be explained by the
`technology life cycle or unexpected increases in demand ..................................... 78 
`VII.  THE DEFENDANTS’ DRAM REVENUES INCREASED
`DRAMATICALLY DURING THE CLASS PERIOD ..................................................... 81 
`VIII.  TRADE ASSOCIATION PARTICIPATION PROVIDED MANY
`OPPORTUNITIES FOR DEFENDANTS TO SHARE INFORMATION
`AND COLLUDE, WHICH CONFIDENTIAL WITNESS TESTIMONY
`AND THE DATA INDICATE OCCURRED ................................................................... 82 
`A. 
`Defendants’ Participation in Trade Associations .................................................. 82 
`1. 
`Joint Electron Device Engineering Council (“JEDEC”) ........................... 82 
`2. 
`Semiconductor Industry Association (“SIA”) ........................................... 83 
`3. 
`Korean Semiconductor Industry Association (“KSIA”) ........................... 84 
`
`010742-11/1496419 V1
`
`
`
`-ii-
`
`

`

`Case 5:21-cv-03267-LB Document 1 Filed 05/03/21 Page 4 of 147
`
`
`
`B. 
`
`C. 
`
`IX. 
`
`4. 
`5. 
`
`2. 
`
`World Semiconductor Council (“WSC”) .................................................. 84 
`World Semiconductor Trade Statistics Organization
`(“WSTS”) .................................................................................................. 85 
`Global Semiconductor Alliance (“GSA”) ................................................. 87 
`6. 
`Trade association meetings in which Defendants participated are
`correlated with price increases during the Class Period ........................................ 87 
`Defendants Also Used Third Party Research Firms to Exchange
`Information in Furtherance of the Conspiracy ...................................................... 90 
`THE STRUCTURE OF THE DRAM INDUSTRY IS CONDUCIVE TO
`CONSPIRACY .................................................................................................................. 90 
`A. 
`DRAM is a Commodity Product ........................................................................... 91 
`B. 
`The DRAM Market is Highly Concentrated ......................................................... 92 
`C. 
`The DRAM Market Has High Barriers to Entry ................................................... 95 
`Defendants own the intellectual property for DRAM
`1. 
`through cross-licenses, patents, and joint ventures ................................... 96 
`DRAM manufacturers have the benefit of economies of
`scale ........................................................................................................... 97 
`Demand for DRAM is Inelastic ............................................................................. 98 
`D. 
`DEFENDANTS’ PRIOR COLLUSION IN THE DRAM MARKET, AND
`RELATED MARKETS, MAKES COLLUSION EVEN MORE
`PLAUSIBLE HERE ........................................................................................................ 100 
`Defendants Were Previously Convicted for Fixing Prices of
`A. 
`DRAM ................................................................................................................. 100 
`1. 
`Prior DRAM - Other Regulators ............................................................. 106 
`2. 
`Prior DRAM Civil Cases (Direct and Indirect) ....................................... 107 
`Other Semiconductor Memory Products ............................................................. 107 
`1. 
`SRAM ...................................................................................................... 107 
`2. 
`Flash ........................................................................................................ 108 
`EFFECTS OF DEFENDANTS’ MISCONDUCT .......................................................... 109 
`The Inflated Prices for DRAM Caused by Defendants’
`A. 
`Anticompetitive Conduct Were Passed on to Consumers who
`Purchased DRAM Devices. ................................................................................. 109 
`XII.  CLASS ACTION ALLEGATIONS ................................................................................ 114 
`
`X. 
`
`XI. 
`
`B. 
`
`010742-11/1496419 V1
`
`
`
`-iii-
`
`1
`2
`3
`4
`5
`6
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`24
`25
`26
`27
`28
`
`

`

`Case 5:21-cv-03267-LB Document 1 Filed 05/03/21 Page 5 of 147
`
`
`
`XIII.  VIOLATIONS ALLEGED ............................................................................................. 118 
`FIRST CLAIM FOR RELIEF VIOLATIONS OF SHERMAN ACT,
`15 U.S.C. § 1 ................................................................................................................... 118 
`XIV.  VIOLATIONS OF STATE ANTITRUST LAWS .......................................................... 119 
`SECOND CLAIM FOR RELIEF VIOLATION OF THE DISTRICT OF
`COLUMBIA ANTITRUST ACT D.C. CODE § 28-4501, ET SEQ. (ON
`BEHALF OF THE DISTRICT OF COLUMBIA CLASS) ............................................ 120 
`THIRD CLAIM FOR RELIEF VIOLATION OF THE IOWA COMPETITION
`LAW IOWA CODE §553.1, ET SEQ. (ON BEHALF OF THE IOWA
`CLASS) ........................................................................................................................... 121 
`FOURTH CLAIM FOR RELIEF VIOLATION OF THE KANSAS
`RESTRAINT OF TRADE ACT KAN. STAT. ANN § 50-101, ET SEQ.
`(ON BEHALF OF THE KANSAS CLASS) ................................................................... 121 
`FIFTH CLAIM FOR RELIEF VIOLATION OF THE MINNESOTA
`ANTITRUST LAW, MINN. STAT. § 325D.49, ET SEQ. (ON BEHALF
`OF THE MINNESOTA CLASS) .................................................................................... 122 
`SIXTH CLAIM FOR RELIEF VIOLATION OF THE MISSISSIPPI
`ANTITRUST STATUTE, MISS. CODE ANN. § 74-21-1, ET SEQ. (ON
`BEHALF OF THE MISSISSIPPI CLASS) ..................................................................... 123 
`SEVENTH CLAIM FOR RELIEF VIOLATION OF THE MISSOURI
`MERCHANDISING PRACTICES ACT, MO. ANN. STAT. § 407.010,
`ET SEQ. (ON BEHALF OF THE MISSOURI CLASS) ................................................. 125 
`EIGHTH CLAIM FOR RELIEF VIOLATION OF THE NEW MEXICO
`ANTITRUST ACT, N.M. STAT. ANN. §§ 57-1-1, ET SEQ. (ON
`BEHALF OF THE NEW MEXICO CLASS) ................................................................. 126 
`NINTH CLAIM FOR RELIEF VIOLATION OF SECTION 340 OF THE NEW
`YORK GENERAL BUSINESS LAW (ON BEHALF OF THE NEW
`YORK CLASS) ............................................................................................................... 126 
`TENTH CLAIM FOR RELIEF VIOLATION OF THE OREGON ANTITRUST
`LAW, OR. REV. STAT. §646.705, ET SEQ. (ON BEHALF OF THE
`OREGON CLASS) .......................................................................................................... 127 
`ELEVENTH CLAIM FOR RELIEF VIOLATION OF THE TENNESSEE
`TRADE PRACTICES ACT, TENN. CODE, § 47-25-101, ET SEQ. (ON
`BEHALF OF THE TENNESSEE CLASS) .................................................................... 128 
`TWELFTH CLAIM FOR RELIEF VIOLATION OF THE UTAH ANTITRUST
`ACT, UTAH CODE ANN. §§ 76-10-911, ET SEQ. (ON BEHALF OF
`THE UTAH CLASS) ...................................................................................................... 130 
`THIRTEENTH CLAIM FOR RELIEF VIOLATION OF THE WISCONSIN
`ANTITRUST ACT, WIS. STAT. ANN. §133.01(1), ET SEQ. (ON
`BEHALF OF THE WISCONSIN CLASS) ..................................................................... 131 
`
`010742-11/1496419 V1
`
`
`
`-iv-
`
`1
`2
`3
`4
`5
`6
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`24
`25
`26
`27
`28
`
`

`

`Case 5:21-cv-03267-LB Document 1 Filed 05/03/21 Page 6 of 147
`
`
`
`XV.  VIOLATIONS OF STATE CONSUMER PROTECTION LAWS ................................ 132 
`FOURTEENTH CLAIM FOR RELIEF VIOLATION OF THE DISTRICT OF
`COLUMBIA CONSUMER PROTECTION PROCEDURES ACT, D.C.
`CODE § 28-3901, ET SEQ. (ON BEHALF OF THE DISTRICT OF
`COLUMBIA CLASS) ..................................................................................................... 132 
`FIFTEENTH CLAIM FOR RELIEF VIOLATION OF THE MINNESOTA
`CONSUMER FRAUD ACT, MINN. STAT. § 325F.68, ET SEQ. (ON
`BEHALF OF THE MINNESOTA CLASS) ................................................................... 133 
`SIXTEENTH CLAIM FOR RELIEF VIOLATION OF THE NEW MEXICO
`UNFAIR PRACTICES ACT, N.M. STAT. ANN. §§ 57-12-3, ET SEQ.
`(ON BEHALF OF THE NEW MEXICO CLASS) ......................................................... 134 
`SEVENTEENTH CLAIM FOR RELIEF VIOLATION OF THE OREGON
`UNLAWFUL TRADE PRACTICES ACT, OR. REV. STAT. § 646.605,
`ET SEQ. (ON BEHALF OF THE OREGON CLASS) ................................................... 136 
`EIGHTEENTH CLAIM FOR RELIEF VIOLATION OF THE UTAH
`CONSUMER SALES PRACTICES ACT, UTAH CODE ANN. §§ 13-11-
`1, ET SEQ. (ON BEHALF OF THE UTAH CLASS) ..................................................... 137 
`NINETEENTH CLAIM FOR RELIEF VIOLATION OF THE UTAH UNFAIR
`PRACTICES ACT, UTAH CODE ALL. §§ 13-5-1, ET SEQ. (ON
`BEHALF OF THE UTAH CLASS) ................................................................................ 138 
`TWENTIETH CLAIM FOR RELIEF VIOLATION OF THE VIRGINIA
`CONSUMER PROTECTION ACT VA. CODE ANN. § 59.1- 196, ET
`SEQ. (ON BEHALF OF THE VIRGINIA CLASS) ....................................................... 139 
`PRAYER FOR RELIEF .............................................................................................................. 140 
`DEMAND FOR JURY TRIAL ................................................................................................... 141 
`
` 
` 
`
`010742-11/1496419 V1
`
`
`
`-v-
`
`1
`2
`3
`4
`5
`6
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`24
`25
`26
`27
`28
`
`

`

`Case 5:21-cv-03267-LB Document 1 Filed 05/03/21 Page 7 of 147
`
`
`
`Plaintiffs bring this action on their own behalf, and as a putative class action on behalf of all
`others similarly situated in the United States. Plaintiffs, by and through their attorneys, based on their
`individual experiences, the independent investigation of counsel and consultants, and information
`and belief, allege as follows:
`I.
`INTRODUCTION AND OVERVIEW OF THE CONSPIRACY
`In this Complaint, Plaintiffs, among other things, make the following key allegations:
`
`1.
`
` Based on the statements of a confidential witness, that Sewon Chun, the Samsung executive in
`charge of DRAM, intentionally leaked Samsung’s confidential plans to raise DRAM prices in
`early 2016 to an industry analyst as a signal to Hynix and Micron to also raise prices, only
`after, according to the same confidential witness, Samsung had taken unilateral action in late
`2015 in an attempt to raise DRAM prices by shelving DRAM wafers that failed to stop DRAM
`prices from falling;
`
` That after DRAM prices doubled from 2016 to 2017, the exact same Samsung executive, Sewon
`Chun, repeatedly publicly attributed DRAM price increases to restrictions on supply among the
`Defendants, including stating on April 27, 2017 that “supply and demand continued to be solid
`and price rose strongly due to restrictions of industry supply,” and stating on July 27, 2017 that
`“due to restriction of industry supply, supply and demand remained solid and price continued
`to rise”;
`
` That after Micron’s CEO stated in March 2016 that it would “be foolish for it to be the first
`ones to take capacity off,” Samsung and Hynix each announced the following month, April
`2016, that they were both reducing capacity, and the next month, Micron’s CEO stated that
`“there has been some relatively encouraging news that has been disseminated via various
`channels and from the competitors about . . . capital spending going into next year, particularly
`for DRAM . . . we all are going to either benefit or be hurt by excess supply in the
`marketplace” and so “I actually remain bullish on . . . the DRAM business and the actions of
`the competitors in the marketplace”;
`
`CLASS ACTION COMPLAINT –
`Case No. _____________________
`010742-11/1203459 V1
`
`-1-
`
`1
`2
`3
`4
`5
`6
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`24
`25
`26
`27
`28
`
`

`

`Case 5:21-cv-03267-LB Document 1 Filed 05/03/21 Page 8 of 147
`
`
`
` That after Defendants’ executives made statements in 2016 encouraging their putative
`competitors to minimize supply growth for the benefit of the industry, Defendants then made
`subsequent statements in 2017 attributing the industry’s recent success to a shared commitment
`to restraining supply and encouraging each other to continue to restrain supply, with Micron’s
`CFO on June 6, 2017 stating that continued success “over the course of a multi-year period”
`would require the industry’s competitors to be disciplined about not adding supply because he
`foresaw a “good balance between supply and demand as long as capital discipline is
`exercised,” followed the next month, July 2017, by Samsung stating that “we will refrain from,
`for example, increasing market share, fighting on volume”;
`
` Based on economic analysis, that Samsung, Micron, and Hynix acted against their own unilateral
`self-interest through DRAM output cuts in 2016 that cost each company tens of millions of
`dollars, which would have been irrational without a collusive agreement, with Samsung first
`unilaterally reducing output and costing itself close to $50 million in profits as well as
`significant market share, then followed the next quarter by corresponding output reductions
`from Micron and Hynix that cost them tens of millions of dollars each;
`
` Based on the statements of a confidential witness, that senior executives from Samsung, Micron,
`and Hynix attended private DRAM-specific conferences where they “could talk to competitors”
`about “future volume” and “overall volume expectations”;
`
` Related to Defendants’ prior price-fixing conspiracy, that Micron’s chief strategy officer
`throughout the current conspiracy period, had previously testified at a criminal trial of a Hynix
`executive for the prior conspiracy that he had directly fixed prices with the Hynix executive; that
`a Hynix executive who had pled guilty in 2006 to the prior conspiracy was promoted by Hynix to
`chief marketing officer in charge of all memory products in 2007 while he was still in prison
`and had then remained at Hynix until at least 2015; and statements from a Samsung confidential
`witness that some of the very same executives who participated in the prior price-fixing
`conspiracy are still employed at Samsung;
`
` With regard to the class of DRAM Devices at issue in this indirect purchaser action, personal
`computers, network servers, cellular phones, tablets, computer graphical processing units, and
`CLASS ACTION COMPLAINT –
`Case No. _____________________
`010742-11/1203459 V1
`
`-2-
`
`1
`2
`3
`4
`5
`6
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`24
`25
`26
`27
`28
`
`

`

`Case 5:21-cv-03267-LB Document 1 Filed 05/03/21 Page 9 of 147
`
`
`
`computer DRAM memory products, that each is a type of product for which DRAM constitutes
`approximately 5-50% of the total cost, and often significantly more, as well as specific evidence
`that rising DRAM prices during the class period caused corresponding increases in the price
`of the DRAM Devices, thus plausibly suggesting that the DRAM and DRAM Device markets are
`“inextricably intertwined” and some or all of the overcharges on DRAM caused by Defendants’
`collusive conduct was passed-on to purchasers of these products;
`2.
`Defendants – Samsung, Micron, and Hynix – are the three companies that control
`
`nearly 100 percent of the Dynamic random access memory (“DRAM”) market, with Samsung
`
`controlling the largest share.1 As of mid-2017, these Defendants collectively controlled at least 96%
`
`of worldwide DRAM market share.
`
`3.
`
`In late 2015, Defendants faced a problem. Vigorous competition between Defendants
`
`in 2014 and 2015 had caused DRAM prices to fall dramatically. Confidential Witness 1 (“CW 1”) is
`
`a former U.S.-based Samsung executive.2 He explains that, in late 2015, Samsung made a unilateral
`
`decision to act on its own and stockpile, rather than sell, DRAM wafers in an attempt to limit supply,
`
`with the goal to stop DRAM price erosion and ideally raise DRAM prices. But this legal, unilateral
`
`action did not work and DRAM prices in the industry continued to fall precipitously through the end
`
`of 2015.
`
`4.
`
`In early 2016, Samsung switched from unilateral self-action to attempts to signal to
`
`the other Defendants to embark on concerted action. According to CW 1, in early 2016, Sewon
`
`                                                            
`1 Dynamic random access memory (“DRAM”) is one of the most common forms of
`semiconductor memory. DRAM is made from silicon wafers, and it is widely used as a component in
`digital electronics. “Defendants” include: Micron Technology, Inc. (“Micron Technology”); Micron
`Semiconductor Products, Inc. (“Micron Semiconductor”) (collectively “Micron”); Samsung
`Electronics Co., Ltd. (“SEC”); Samsung Semiconductor, Inc. (“SSI”) (collectively “Samsung”); SK
`Hynix, Inc. (f/k/a Hynix Semiconductor, Inc.) (“SK Hynix Korea”); SK Hynix America, Inc. (f/k/a
`Hynix Semiconductor America, Inc.) (“SK Hynix America”) (collectively “SK Hynix” or “Hynix”).
`2 Confidential Witness 1 is a former Samsung memory executive who worked at Samsung North
`America’s headquarters in San Jose, California during at least part of the conspiracy (Class) period.
`CLASS ACTION COMPLAINT –
`Case No. _____________________
`010742-11/1203459 V1
`
`-3-
`
`1
`2
`3
`4
`5
`6
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`24
`25
`26
`27
`28
`
`

`

`Case 5:21-cv-03267-LB Document 1 Filed 05/03/21 Page 10 of 147
`
`
`
`Chun, an executive vice president of Memory Marketing for Samsung who had responsibility for
`
`how much DRAM Samsung would produce and what types of pricing should be offered to
`
`customers, told Pablo Temprano, a Samsung executive in the United States, to intentionally leak to
`
`Sean Muir, an industry analyst at Cleveland Research, confidential information that Samsung was
`
`raising prices on DRAM.3 Mr. Muir was chosen because he published a monthly report on DRAM
`
`prices that was closely monitored by the Defendants. CW 1 understood based on his conversation
`
`with Mr. Temprano that Mr. Temprano was reluctant to convey the information to Mr. Muir, but that
`
`he eventually did so in a telephone call. CW 1 believed that Samsung’s purpose in intentionally
`
`divulging this confidential information about future prices was to inform Samsung’s chief
`
`competitors that Samsung was raising DRAM prices and that Micron and Hynix should follow suit.
`
`CW 1 observed that shortly after this incident, DRAM prices began to rise.
`
`5.
`
`During the same time period, Micron invited the other Defendants to cut production,
`
`coupling this invitation with a clear reassurance that Micron would not try to take DRAM market
`
`share if Samsung or Hynix cut production. If the industry collectively cut production, then prices
`
`would rise – which is exactly what the data show happened. On March 30, 2016, Micron’s CEO,
`
`Mark Durcan stated that Micron would “be foolish to be the first ones to take capacity off.”
`
`Micron’s CFO, Ernie Maddock, stated that Micron would not unilaterally cut production, explaining
`
`that “it’s a really ill-advised move to be unilaterally cutting production.” But at the same time, Mr.
`
`Maddock reassured Samsung and Hynix that “our focus isn’t on market share. Our focus is on
`
`making sure that we’ve deployed equivalent advanced technology . . . so that we’re not incentivizing
`
`others to play for market share.” Sean Muir of Cleveland Research issued a report on Micron’s
`
`earnings call the following day that emphasized that Micron had “maintained it would be foolish for
`
`                                                            
`3 CW1’s knowledge is based on conversations that he had with Mr. Temprano.
`CLASS ACTION COMPLAINT –
`Case No. _____________________
`010742-11/1203459 V1
`
`-4-
`
`1
`2
`3
`4
`5
`6
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`24
`25
`26
`27
`28
`
`

`

`Case 5:21-cv-03267-LB Document 1 Filed 05/03/21 Page 11 of 147
`
`
`
`them to take capacity offline before Samsung or Hynix saying that at their fixed cost structure it’s
`
`a really bad move to cut production” – indicating that Micron’s statements were understood by
`
`industry participants to be a clear invitation to the other Defendants to be the first ones to cut DRAM
`
`supply.
`
`6.
`
`Within a month, both Samsung and Hynix responded to Micron’s invitation to cut
`
`production by publicly stating that they would cut their DRAM production. On April 20, 2016,
`
`Samsung stated that its DRAM bit growth had decreased by low single-digits over the prior quarter.
`
`Samsung also stated that “we don’t expect there to be major increases in supply of DRAM in the
`
`near future.” On April 27, 2016, Hynix stated that on its earnings call that “DRAM suppliers’ CapEx
`
`execution is projected to decrease as we move into the second half.” Deutsche Bank issued a report
`
`on this Hynix earnings call that emphasized “management toned down its capex expectations for
`
`2016, especially for DRAM. SK Hynix expects its total capex to be down [year over year] in 2016,
`
`with some capex allocated to R&D purposes and not for capacity improvement.”
`
`7.
`
`The following month, on May 25, 2016, Micron’s CEO Mark Durcan recognized that
`
`its competitors were planning to cut supply and Durcan encouraged further production cuts, stating
`
`that “there has been some relatively encouraging news that has been disseminated via various
`
`channels and from the competitors about . . . capital spending going into next year, particularly
`
`for DRAM.” Durcan encouraged his competitors to reduce wafers, stating that “[i]f wafers actually
`
`come down as we’re starting to hear some equipment suppliers talk about, then it could be mid- to
`
`high-teens [supply growth rate], in which case that would be more beneficial” for DRAM prices.
`
`Durcan emphasized that there are only three suppliers in the market, and “we all are going to either
`benefit or be hurt by excess supply in the marketplace.” He expected Micron and its competitors to
`maintain supply discipline because “there’s a natural tightening tendency absent, somebody wanting
`to do something different than that. And so…I actually remain bullish on…the DRAM business and
`the actions of the competitors in the marketplace.” The very next week, Deutsche Bank issued a
`CLASS ACTION COMPLAINT –
`Case No. _____________________
`010742-11/1203459 V1
`
`-5-
`
`1
`2
`3
`4
`5
`6
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`24
`25
`26
`27
`28
`
`

`

`Case 5:21-cv-03267-LB Document 1 Filed 05/03/21 Page 12 of 147
`
`
`
`report on June 1, 2016 that it was “encouraged that both Samsung and SK Hynix confirmed that their
`DRAM capex will decline y/y . . . while emphasizing the importance of profitability.”
`8.
`On July 18, 2016, a Deutsche Bank analyst report noted an unprecedented change in
`
`DRAM industry behavior with production cuts occurring despite industry profitability, stating that
`
`“our historical data show that DRAM capex cuts only happen when the industry has been
`
`unprofitable. However, DRAM makers have cut capex in 2016 despite the industry remaining
`
`highly profitable, proving the industry has become more capex-disciplined.”
`
`9.
`
`In addition to the statements, Bloomberg market data indicates that Defendants took
`
`actions contrary to self-interest that would have been irrational and perilous in the absence of a
`
`collusive agreement. First, Samsung cut DRAM production in Q3 2016, even as demand for DRAM
`
`continued to rise. Samsung’s production cut was against its unilateral self-interest as it cost an
`
`estimated $47 million in immediate profits for Q3 if it had simply maintained its production at Q2
`
`levels. Samsung’s voluntary action sacrificed approximately 5% of the total market share by this
`
`output reduction, a complete reversal of prior behavior in the 2014-2015 period of fiercely
`
`competing to gain market share. Richard Posner recognizes that “declining market shares of leaders”
`
`is a plus factor potentially indicative of cartel conduct.4 Samsung’s output cut is shown in the
`
`following chart.
`
`                                                            
`4 See Richard A. Posner, Antitrust Law, 79-93 (2d Ed. 2001).
`CLASS ACTION COMPLAINT –
`Case No. _____________________
`010742-11/1203459 V1
`
`-6-
`
`1
`2
`3
`4
`5
`6
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`24
`25
`26
`27
`28
`
`

`

`Case 5:21-cv-03267-LB Document 1 Filed 05/03/21 Page 13 of 147
`
`
`
`
`
`10.
`
`Second, in a competitive market, Samsung’s rivals would have either maintained or
`
`increased output to gain market share in the face of rising demand, which is part of the reason why
`
`Samsung’s conduct would have been irrational without knowledge that its rivals would not also
`
`decrease output. And indeed, the following quarter – Q4 2016 – both Micron and Hynix actually cut
`
`their output, as shown in the following chart. Micron and Hynix’s actions are indicative of a
`
`collusive agreement and also against their own unilateral self-interest, as they would have gained
`
`market share and significant profits if they had either simply maintained their output levels or raised
`
`their output levels. Based on available data, Micron and Hynix sacrificed respectively $30 million
`
`and $27 million in lost profits by failing to maintain their output growth rate.
`
`CLASS ACTION COMPLAINT –
`Case No. _____________________
`010742-11/1203459 V1
`
`-7-
`
`1
`2
`3
`4
`5
`6
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`24
`25
`26
`27
`28
`
`

`

`Case 5:21-cv-03267-LB Document 1 Filed 05/03/21 Page 14 of 147
`
`
`
`
`
`11.
`
`In addition to Samsung’s new output restrictions, CW 1 also observed a change in
`
`behavior during the conspiracy period by Samsung for its DRAM pricing. CW 1 regularly attended
`
`quarterly meetings between Samsung Elec