Nursing Home Pension Fund et al v. Oracle Corporation et al, No. 3:2001cv00988 - Document 1619 (N.D. Cal. 2009)

Court Description: ORDER GRANTING DEFENDANTS' MOTION FOR SUMMARY JUDGMENT and DENYING PLAINTIFF'S MOTION FOR PARTIAL SUMMARY JUDGMENT (SI, COURT STAFF) (Filed on 6/16/2009)

Download PDF

Nursing Home Pension Fund et al v. Oracle Corporation et al Doc. 1619 1 2 3 4 5 IN THE UNITED STATES DISTRICT COURT 6 FOR THE NORTHERN DISTRICT OF CALIFORNIA 7 8 9 IN RE ORACLE CORPORATION SECURITIES LITIGATION No. C 01-00988 SI Related Cases: C 01-01237-SI C 01-01263-SI United States District Court For the Northern District of California 10 11 12 / ORDER GRANTING DEFENDANTS’ MOTION FOR SUMMARY JUDGMENT and DENYING PLAINTIFF’S MOTION FOR PARTIAL SUMMARY JUDGMENT 13 On February 13, 2009, the Court heard oral argument on defendants’ motion for summary 14 15 16 17 judgment and plaintiffs’ motion for partial summary judgment. Having considered the arguments of the parties, their papers, the cases submitted after oral argument, and for good cause shown, defendants’ motion is GRANTED and plaintiffs’ motion is DENIED. 18 BACKGROUND 19 20 21 22 23 24 25 26 1. Procedural Background This action was filed in March of 2001. Plaintiffs, a number of purchasers of Oracle stock, allege that Oracle Corporation and three of its top executive officers1 violated section 10(b) of the Securities Exchange Act of 1934 (“Exchange Act”), 15 U.S.C. § 78j(b), and Rule 10b-5, promulgated thereunder. Plaintiffs further allege control person liability against the individual defendants under section 20(a) of the Exchange Act, 15 U.S.C. § 78t(a), and that Henley and Ellison are liable for contemporaneous trading under section 20A of the Exchange Act, 15 U.S.C. § 78t-1(a). After 27 1 28 Lawrence J. Ellison (Chief Executive Officer), Jeffrey O. Henley (Executive Vice President and Chief Financial Officer), and Edward J. Sanderson (Executive Vice President). Dockets.Justia.com United States District Court For the Northern District of California 1 certification and consolidation of related actions, a series of dismissals and filings of amended 2 complaints ensued until, in March 2003, this Court2 dismissed the revised second amended complaint 3 with prejudice for failure to state a claim under Federal Rule of Civil Procedure 12(b)(6), finding that 4 the allegations did not create a strong inference that allegedly false statements were known to be false 5 when made. [Docket No. 166] Plaintiffs appealed, and in November of 2004, the Ninth Circuit reversed 6 the dismissal, holding that the operative complaint met the heightened pleading requirements of the 7 Private Securities Litigation Reform Act (“PSLRA”). See Nursing Home Pension Fund, Local 144 v. 8 Oracle Corp., 380 F.3d 1226 (9th Cir. 2004). 9 On September 2, 2008, this Court issued an order denying plaintiffs’ motion for partial summary 10 judgment on the question of whether defendants made false or misleading statements regarding Suite 11 11i and the company’s financial results for 2Q01. The Court also granted in part and denied in part 12 plaintiffs’ motion for sanctions. The Court held that plaintiffs are entitled to adverse inference 13 instructions with regard to two categories of evidence: defendant Ellison’s email files, and materials 14 created during preparation for the book Softwar. The Court determined that it is appropriate to infer that 15 the emails and Softwar materials would demonstrate Ellison’s knowledge of, among other things, 16 problems with Suite 11i, the effects of the economy on Oracle’s business, and problems with 17 defendants’ forecasting model.3 The Court held that it would take these adverse inferences into account 18 when deciding the parties’ summary judgment motions and directed the parties to revise their briefs in 19 light on this ruling. 20 Now before the Court are defendants’ revised motion for summary judgment, plaintiffs’ revised 21 motion for partial summary judgment against defendant Ellison, and seven Daubert motions to preclude 22 expert testimony. 23 24 25 26 2 This case was originally assigned to the Honorable Martin J. Jenkins. It was reassigned to this Court after Judge Jenkins’ resignation in April of 2008. 3 27 28 The Court noted that these inferences alone would not assist plaintiffs on all elements of their § 10(b) claims, including particularly the element of loss causation. See Sept. 2, 2008 Order, at 12. [Docket No. 1478] 2 1 2. Factual Background4 2 Oracle is the second largest software company in the world. This case arises from plaintiffs’ 3 claims that Oracle and certain of its officers and executives made false and misleading statements about 4 a new product, issued inflated earnings reports for the second quarter of fiscal year 2001, issued a false 5 and misleading forecast about the company’s financial condition for the third quarter of fiscal year 2001, 6 and falsely stated during the third quarter that the company was not being affected by the slowing 7 economy. The four quarters of Oracle’s 2001 fiscal year were: from June 1 to August 31, 2000 8 (“1Q01”); from September 1 to November 30, 2000 (“2Q01”); from December 1, 2000 to February 28, 9 2001(“3Q01”); and from March 1to May 31, 2001 (“4Q01”). United States District Court For the Northern District of California 10 11 A. 12 In May 2000, Oracle released its Applications Suite 11i (“Suite 11i”). “Enterprise applications” 13 are computer programs used to help companies automate their business processes. Enterprise resource 14 planning (“ERP”) applications perform functions such as accounting, human resources, and 15 manufacturing. Customer relationship management applications (“CRM”) perform functions such as 16 managing call centers. Through the late 1990s, businesses that used enterprise applications software 17 could not obtain ERP and CRM applications from the same vendor. Customers generally followed the 18 “best of breed” strategy, buying different applications from several vendors. They would achieve 19 “systems integration” by hiring software engineers to write custom code that would allow their 20 applications to run together. Suite 11i was marketed as a product that would combine ERP and CRM 21 applications. See generally, Decl. of Lawrence J. Ellison in Supp. of Defs. Mot. for Summ. J. ¶¶ 15, 16 22 (“Ellison Decl.”). [Docket No. 932] Oracle’s statements about the functionality of Suite 11i. 23 Suite 11i was first available for sale in 1Q01. In 1Q01 and 2Q01, Oracle reported a total of $435 24 million in revenue from applications licenses, marking increases of 42% and 66% over the same quarters 25 4 26 27 28 The word “voluminous” does not do justice to the record in this case. Defendants have numbered their exhibits sequentially. For ease of reference, the Court refers to their exhibits as DX__. Plaintiffs have used a variety of different numbering systems. The Court refers to exhibits filed in support of plaintiffs’ motion for summary judgment against Ellison as PSJM __, to the exhibits filed in support of plaintiffs’ reply as PReply__, and to the exhibits filed in support of plaintiffs’ opposition to defendants’ motion for summary judgment as POpp.__. 3 United States District Court For the Northern District of California 1 in the prior year (the “year over year” comparison). DSJM 25 at 973930, 26 at 976820. In 3Q01, 2 Oracle reported $249 million from applications licenses, a year over year increase of 25%. DSJM 27 3 at 977030. Plaintiffs contend that Oracle released Suite 11i prematurely, that the parts were neither 4 designed nor engineered to work together, that it did not work in multiple different languages, and that 5 it generally did not work correctly. According to plaintiffs, Oracle officials who knew of these 6 deficiencies nonetheless made repeated statements to the public that misrepresented the functionality 7 of Suite 11i. 8 motions, plaintiffs cite to the following statements as constituting material misrepresentations about 9 Suite 11i by officers of Oracle. In the complaint and in the various memoranda concerning the summary judgment 10 • 11 The right model for enterprise software is “Here are all the pieces. They’ve all been engineered to work together. No systems integration required. You can install it in a matter of months in the largest and most complex operations. All the pieces are there: marketing, sales, web store, service, internet procurement, auctioning, supply chain automation, manufacturing, human resources, everything. And all the pieces fit together.” ... So in the early stages – the very early stages of this release 11i, we’re saying, “We’re right. The rest of the world is wrong;” where there’s all this controversy where we can’t show lots and lots of companies up and running – big companies up and running – they’re just beginning to come live now, already we’re getting tremendous traction in the market. And it will be far and away the biggest success in the history of our company, much bigger than the database. ... We’re trying to make this very, very simple. You engineer all the pieces to fit together, they come out of the box, and all the pieces fit together. It’s still hard to install. You still have to convert your data and train people and do stuff. It’s still not 15 minutes. It’s still a real project to install this. This is taking us 6 months to get just the first factory up and running at General Electric. It [is] still not like you just walk in – it’s not like installing a new word processor. It’s still pretty complicated stuff. ... [A]nd we’re very close. At some time over the next few months, it’ll click. I’m serious. We’ll win every deal. Every deal. ... [Y]ou can get a small team to build an auctioning system over at Commerce One. You get a small team to build internet procurement over at Ariba. But you can’t get a large team to build all of those things such that all of the pieces fit together. And it is a bigger job – you know, building each one of those separately is a smaller job than building each one of those together that work together. That’s a much bigger job. But we are bigger. . . . Our engineering teams are larger. And we’ve done it. The pieces actually work together. And the barrier to entry, I think, is insuperable. I don’t think anyone else can do it. 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 On November 29, 2000, Ellison made the following comments at a conference: POpp. 405 at 12-14, 37-38, 40, 43-44 (quoted in Plaintiffs’ Opposition at 12; plaintiffs cite only the 4 1 2 underlined portions of these statements). • At the December 14, 2000 conference call with analysts, Ellison said, “[Y]ou can buy our complete E-Business Suite, where all the pieces are designed and engineered to fit together, and no systems integration is required. It’s up and running in months. You get the savings in months. It costs you less, and it takes less time to install.” “Demonstrations, we’re still tweaking those, we got those all ready now. We finished up all the training . . . we’re working on partner training. So, yes, I think where we sit right now we’re in pretty darn good shape.”5 POpp. 26, 234430, 234436 (quoted in Plaintiffs’ Opposition at 13). • Mark J. Barrenechea said at a presentation on February 6, 2001, “I think our applications are written in 23 languages. So not only do we have, you know, for example, an E-business Suite, which I’ll tell you more about. But it’s basically ERP and CRM all integrated together. . . . But we also have taken care of the localization requirements of all these countries around the world as well.” POpp. 156, 3285 (quoted in Plaintiffs’ Opposition at 13). • An Oracle “Technical White Paper” by Mark J. Barrenechea, dated February 6, 2001, reads, “To install the Oracle CRM suite, no systems integration is required. And because the CRM suite consists of true Internet applications, every application works in every country, every major language, and every major currency.” POpp. 155 at 106691 (quoted in RSAC ¶ 63). • On February 13, 2001, Sanderson said at a conference, “I think our applications are written in 23 different languages. So not only do we have, you know, for example, an E-Business suite . . . but it’s basically ERP and CRM all integrated together. But it’s written in 23 different languages, including Spanish Spanish and Latin American Spanish, Portugal Portuguese and Brazilian Portuguese, as being four languages. But we also have taken care of the localization requirements of all these countries around the world as well.” POpp. 156 (quoted in Plaintiffs’ Opposition at 13). • On February 21, 2001, Ellison said at the “AppWorld” Conference: “In fact, we recommend that you start with, you try a component of the suite and then you add it in. Now the nice thing is it’s like Lego blocks. Once you have one piece in, the other pieces just snap together. There’s no systems integration required. . . . You just basically turn it on or snap it together.” “It is absolutely, all the pieces within the suite are literally plug and play.” POpp. 438, 14:54:52 (quoted in Plaintiffs’ Opposition at 13). B. Defendants’ statements about Oracle’s financial results for 2Q01. 3 4 5 6 7 8 9 United States District Court For the Northern District of California 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 In a December 14, 2000 press release, Oracle announced 2Q01 earnings of $0.11 per share and 66% growth in sales of Suite 11i applications. POpp. 105, 019764. According to plaintiffs’ expert D. Paul Regan, Oracle arrived at the $0.11 figure for per share earnings through an improper accounting method. Decl. of Shawn A. Williams in Supp. of Expert Report of D. Paul Regan, (“Regan Report”) 5 The speaker of the latter quote is identified only as “Male.” 5 1 ¶ 39. In Regan’s opinion, the accurate figure for Oracle’s earnings per share was $0.10. Id. ¶ 41. By 2 inflating its earnings by a penny, Oracle was able to “beat Wall Street,” i.e. exceed the $0.10 earnings 3 per share that analysts had projected. Id. Regan also opines that Oracle fabricated the 66% figure for 4 growth in Suite 11i applications. In his opinion, the correct figure was 54%. Id. ¶ 62. United States District Court For the Northern District of California 5 6 C. 7 On December 14, 2000, Oracle also issued its public forecast (or “guidance”) for 3Q01. Oracle 8 predicted total license revenue growth of 25%, database revenue growth of 15%-20%, applications 9 revenue growth of 75%, and earnings per share (“EPS”) of $0.12. PSJM 1 at 3221-22. Oracle’s forecast for 3Q01 10 Oracle based its 3Q01 guidance on its internal forecasting method. Oracle used what it calls a 11 “bottom-up” method for deriving its internal forecasts. It began with information from Oracle’s 12 salespeople, which was incorporated into a summary of all deals Oracle was working on at a given time. 13 DX 59 at 74.6 Using a computer program called Oracle Sales Online (“OSO”), salespeople would enter 14 the account names of all ongoing deals, the potential dollar amount of each potential sale, and their 15 predictions for when each sale would likely close. Id. at 75. The sum of all sales that could close in a 16 quarter – including those that are about to close and those that are still just leads – are referred to as the 17 “pipeline.” DX 60 at 93. Regional managers reported information entered by sales representatives in 18 their regions to their supervisors. DX 59 at 75-76. Supervisors recognized that salespeople had a 19 tendency to “sandbag,” i.e. significantly underestimate the size of their projected sales in order to ensure 20 that they would meet or exceed expectations. DX 63 at 566-67; 64 at 101. The forecasting method 21 therefore allowed supervisors to take into account, in addition to data in OSO, their own judgment of 22 what sales were likely to be. Id. at 565-67. Their judgment was based on direct contact with regional 23 managers, sales representatives, and customers. Id. Information was relayed in this way up to the heads 24 of Oracle’s business units, each of which would make a forecast for its unit. (For software licensing, 25 those units were North American Sales (“NAS”), Oracle Product Industries (“OPI”), and Oracle 26 27 28 6 Plaintiffs object that defendants’ characterization of this and other deposition testimony is misleading. The Court has relied on the underlying deposition testimony, not defendants’ characterization. 6 United States District Court For the Northern District of California 1 Services Industries (“OSI”).) The business units would submit their forecasts to Oracle’s financial 2 department, which would consolidate them. DSJM 65 at 102-3. 3 Jennifer Minton, Senior Vice President of Global Finance and Operations, played a central role 4 in the forecasting process by consolidating all the field forecasts into a single report. Id. 104-05. 5 Minton then adjusted the consolidated field report to create the “Upside” report, which was Oracle’s 6 consolidated or “potential” forecast. Id. She arrived at the Upside report by adjusting the field data 7 based on her own judgment, which was informed by her weekly meetings with business unit 8 representatives and her conversations with field finance representatives. Id. at 123-23. Another factor 9 Minton considered when preparing the Upside report was the “conversion ratio” (the percentage of the 10 pipeline that was actually converted into sales) from the corresponding quarter in the prior year. Id. at 11 122. Although Minton’s adjustment was referred to as “Upside,” she would also adjust the consolidated 12 report downward if necessary (for example, if she found out that there had been error in the forecasts 13 submitted to her, or if the historical conversion ratio suggested that the conversion rate for the current 14 quarter would be lower than projected). Id. at 137. 15 During the time period at issue here, Oracle’s method for generating the potential forecast had 16 been a reliable but conservative predictor for the company’s performance: in the seven quarters before 17 3Q01, Oracle had met or exceeded its forecast and analyst projections for quarterly earnings per share 18 (“EPS”):7 19 20 21 22 23 24 /// 25 26 27 28 7 See evidence summarized at DX 42, 43. Plaintiffs object to these exhibits as improper summaries of voluminous evidence under Federal Rule of Evidence 1006. The Court disagrees. The declarations of Ivgen Guner [Docket No. 934] and Bruce Deal [Docket No. 931] lay sufficient foundation for these exhibits to be admissible as summaries of voluminous evidence. Plaintiffs’ objection is OVERRULED. 7 1 Quarter Date of Estimate Oracle Internal Forecast EPS Analyst Forecast EPS Actual EPS 4Q99 Mar. 1999 0.0693 0.080 0.090 1Q00 June 1999 0.0401 0.040 0.040 2Q00 Sept. 1999 0.0495 0.050 0.065 3Q00 Dec. 1999 0.0553 0.070 0.085 6 4Q00 Mar. 2000 0.1375 0.130 0.155 7 1Q01 July 2000 0.0761 0.070 0.085 8 2Q01 Sept. 2000 0.0966 0.090 0.1108 9 Another dynamic which affected Oracle’s ability to forecast its sales is a phenomenon called the 10 hockey-stick effect.9 Knowing that software vendors report their earnings on a quarterly basis, 11 purchasing customers expect that they can extract the lowest possible price for the product by waiting 12 until late in the quarter to finalize deals. DX at 70. This effect is even more exaggerated at the end of 13 the fiscal year, so the most prominent hockey stick effect occurs in the forth quarter. Id. Consequently, 14 Oracle generates most of its new license revenue in the last days of each quarter. Id. 2 3 4 United States District Court For the Northern District of California 5 15 Returning to the public guidance for 3Q01 issued on December 14, 2000, Minton’s Upside report 16 for December 11, 2000 reflected potential total license revenue growth of 33% and potential EPS of 17 $0.1282. See DX 1 at 213092. Oracle’s policy was to round earnings to the nearest penny, see DX 68 18 at 329:1-4, so the public guidance from the Upside report should have been $0.13. Instead, Oracle 19 issued a more conservative estimate: earnings per share of $0.12 and total license revenue growth of 20 25%. Oracle’s December 11 “pipeline report” indicated that its total software pipeline was 52% larger 21 than the previous year. DX 11. 22 23 D. 24 Between January 22 and 31, 2001 Ellison sold 29 million shares of Oracle stock. This amounted The Trades 25 26 8 27 As noted, the accurate figure for Oracle’s 2Q01 earnings per share is disputed. 9 28 Imagine quarterly sales plotted on a chart with a sharp upswing at the end of the quarter. The shape resembles a hockey stick. 8 1 to 2.09% of his Oracle holdings. On January 4, 2001, Henley sold 1 million shares of Oracle stock, 2 about 7% of his total Oracle shares. Ellison has filed a declaration stating that he sold his stock in order 3 to exercise options that were going to expire within nine months and could only be sold during certain 4 trading windows within that period. See Ellison Decl. ¶¶ 9-13 [Docket No. 932] United States District Court For the Northern District of California 5 6 E. 7 On February 25, 2001, three days before 3Q01 ended, Oracle officials received notification that 8 major deals scheduled for 3Q01 had been lost. For example, Oracle’s General Business West area 9 reported that it had lost 70 transactions, worth about $10 million, in the past few days. NAS and OSI 10 similarly reported that they were failing to close major deals. Defendants contend that the news of failed 11 sales was a surprise and that until the very last days of the quarter, they had expected Oracle to meet its 12 guidance. According to defendants, the hockey stick effect led them to expect that the majority of 3Q01 13 deals would be closed at the end of the quarter; the shortfall was caused when major customers decided 14 at the last minute to postpone their purchases. Plaintiffs vigorously dispute defendants’ version of the 15 facts. According to plaintiffs, Oracle’s internal indicators alerted officials throughout the quarter that 16 it would miss its guidance and that the likelihood of a shortfall was apparent to insiders, but not 17 disclosed to the public. The Crash 18 19 F. 20 In a press release dated March 1, 2001, Oracle made the following disclosures “based upon 21 preliminary financial results” for the third quarter of fiscal year 2001: Oracle’s earnings per share 22 would be $0.10 (as opposed to the $0.12 it had predicted on December 14, 2000) and applications 23 growth would be 50% (the prediction was 75%). POpp. 196. The press release quoted Ellison as saying 24 “License growth was strong in this first two months of Q3, and our internal sales forecast looked good 25 up until the last few days of the quarter. However, a substantial number of our customers decided to 26 delay their IT spending based on the economic slowdown in the United States. Sales growth for Oracle 27 products in Europe and Asia Pacific remained strong. The problem is the US economy.” Id. Also on 28 March 1, 2001, Ellison and Henley held a conference call with investors in which they discussed The March 1, 2001 announcement 9 1 Oracle’s earnings miss. 2 3 G. 4 On March 2, 2001, Oracle’s stock price dropped to a closing price of $16.88 per share from a 5 closing price of $21.38 per share on March 1, 2001. POpp. 168, ¶ 43. Relying on the expert opinion 6 of Bjorn I. Steinholt, plaintiffs characterize the drop as “highly statistically significant.” Id.10 The drop in Oracle’s stock price 7 United States District Court For the Northern District of California 8 LEGAL STANDARD 9 Summary judgment is proper “if the pleadings, depositions, answers to interrogatories, and 10 admissions on file, together with the affidavits, if any, show that there is no genuine issue as to any 11 material fact and that the moving party is entitled to a judgment as a matter of law.” Fed. R. Civ. P. 12 56(c). The moving party bears the initial burden of demonstrating the absence of a genuine issue of 13 material fact. See Celotex Corp. v. Catrett, 477 U.S. 317, 323 (1986). The moving party, however, has 14 no burden to negate or disprove matters on which the non-moving party will have the burden of proof 15 at trial. The moving party need only demonstrate to the Court that there is an absence of evidence to 16 support the non-moving party’s case. See id. at 325. 17 The burden then shifts to the non-moving party to “designate ‘specific facts showing that there 18 is a genuine issue for trial.’” Id. at 324 (quoting Fed. R. Civ. P. 56(e)). To carry this burden, the 19 non-moving party must “do more than simply show that there is some metaphysical doubt as to the 20 material facts.” Matsushita Elec. Indus. Co., Ltd. v. Zenith Radio Corp., 475 U.S. 574, 586 (1986). 21 “The mere existence of a scintilla of evidence . . . will be insufficient; there must be evidence on which 22 the jury could reasonably find for the [non-moving party].” Anderson v. Liberty Lobby, Inc., 477 U.S. 23 242, 252 (1986). 24 In deciding a summary judgment motion, the evidence is viewed in the light most favorable to 25 the non-moving party, and all justifiable inferences are to be drawn in its favor. Id. at 255. “Credibility 26 determinations, the weighing of the evidence, and the drawing of legitimate inferences from the facts 27 10 28 The Court recognizes that defendants move to exclude the expert report and testimony of Steinholt on damages and loss causation. 10 1 are jury functions, not those of a judge [when she] is ruling on a motion for summary judgment.” Id. 2 The evidence presented by the parties must be admissible. Fed. R. Civ. P. 56(e). Conclusory, 3 speculative testimony in affidavits and moving papers is insufficient to raise genuine issues of fact and 4 defeat summary judgment. Thornhill Publ’g Co., Inc. v. GTE Corp., 594 F.2d 730, 738 (9th Cir. 1979). 5 6 United States District Court For the Northern District of California 7 DISCUSSION 1. Defendants’ Motion for Summary Judgment 8 Section 10(b) of the Securities Exchange Act of 1934 makes it unlawful “[t]o use or employ, in 9 connection with the purchase or sale of any security . . . any manipulative or deceptive device or 10 contrivance in contravention of such rules and regulations as the Commission may prescribe.” 15 11 U.S.C. § 78j(b). Pursuant to this section, the Securities and Exchange Commission promulgated Rule 12 10b-5, which makes it unlawful “[t]o make any untrue statement of fact or to omit to state a material fact 13 necessary to make the statements made, in light of all the circumstances in which they were made, not 14 misleading.” 17 C.F.R. § 240.10b-5. “The scope of Rule 10b-5 is coextensive with the coverage of 15 § 10(b).” SEC v. Zandford, 535 U.S. 813, 815 n.1 (2002). “In a typical § 10(b) private action a plaintiff 16 must prove (1) a material misrepresentation or omission by the defendant; (2) scienter; (3) a connection 17 between the misrepresentation or omission and the purchase or sale of a security; (4) reliance upon the 18 misrepresentation or omission; (5) economic loss; and (6) loss causation.” Stoneridge Inv. Partners, 19 LLC v. Scientific-Atlanta, 128 S.Ct. 761, 768 (2008) (citing Dura Pharms., Inc. v. Broudo, 544 U.S. 336, 20 341-42 (2005)). 21 Plaintiffs claim that defendants violated § 10(b) by: (1) falsely reporting Oracle’s financial 22 results for 2Q01; (2) misrepresenting the functionality of Suite 11i; (3) issuing forecasts for 3Q01 that 23 had no reasonable basis; (4) repeating the 3Q01 forecast despite their knowledge of facts seriously 24 undermining that forecast; and (5) denying the effects of the slowing economy on Oracle’s business. 25 26 27 /// 28 11 United States District Court For the Northern District of California Objections to evidence11 1 A. 2 Plaintiffs rely on analyst reports and newspaper articles to prove that defendants made false 3 statements about Suite 11i, repeated their 3Q01 forecast throughout the third quarter, and denied that 4 the economic downturn was affecting Oracle. These documents constitute hearsay as they are out of 5 court statements offered to prove the truth of the matter asserted: that Oracle officials made specific 6 fraudulent statements to analysts. See Fed. R. Evid. 801(c). They are thus inadmissible and cannot be 7 considered in support of plaintiffs’ opposition at summary judgment unless they fall within a hearsay 8 exception. See In re Cirrus Logic Sec. Litig., 946 F. Supp. 1446, 1469 (N.D. Cal. 1996) (“It is plainly 9 unfair to hold defendants liable for the reporting of their statements by third parties without independent 10 corroboration of the accuracy of the reported statements.”); see also In re Cypress Semiconductor Sec. 11 Litig., 891 F. Supp. 1369, 1374 (N.D. Cal. 1995) (excluding newspaper articles and analyst reports 12 offered in securities litigation to prove that defendants made purportedly false statements) (citing Larez 13 v. City of Los Angeles, 946 F.2d 630, 643 (9th Cir.1991) (holding that newspaper article offered to prove 14 that defendant made statement quoted in article was hearsay)). 15 Plaintiffs have failed to respond to defendants’ objections to this category of evidence and have 16 suggested no hearsay exception whereby the statements by Oracle officials contained in newspaper and 17 analyst reports could be admitted. Accordingly, defendants’ objections to the following statements are 18 SUSTAINED: 19 i. Statements about Suite 11i 20 • After Oracle Executive Vice President Sandy Sanderson visited the offices of Salomon Smith Barney, the investment firm reported on January 10, 2001 that Suite 11i “is pre-integrated and fully interoperable out of the box, helping to lower consulting costs and time-to-value.” POpp. 211 (quoted in RSAC ¶ 58) (Plaintiffs also offer this exhibit to prove that Sanderson reiterated the 3Q01 guidance. It is also inadmissible for that purpose.) • On February 9, 2001, Bloomberg News reported, in an article headlined “Oracle Shares Fall on Concern Economic Earnings Outlook May Turn Grim,” “Oracle is still upbeat about its prospects for earnings growth, which will be fueled by a new suite of Internet-friendly business software dubbed Oracle 11i, spokeswoman Jennifer Glass said. ‘We haven’t changed our projections at all,’ 21 22 23 24 25 26 27 28 11 The parties raise numerous objections to evidence submitted in support of the instant motions. Unless otherwise discussed in this order, the Court has either not relied on the disputed evidence or has not used it for the purposes to which either party objects. 12 1 Glass said. ‘This slowdown is going to provide new opportunities for Oracle as companies need to streamline and be more strategic about the technology they buy.’” POpp. 377, 141679 (quoted in RSAC ¶ 65(c)). (Plaintiffs also offer this exhibit, labeled PSJM 3, at ex. D, to prove that Glass stated that the economy was not having a negative effect on Oracle’s business. It is also inadmissible for that purpose.) 2 3 4 5 ii. Statements about the effects of the economy on Oracle’s business 6 • Bloomberg News included the following quote in an article that ran on December 14, 2000, “The economy is slowing,’ Henley said in an interview. ‘It’s just not having a negative impact on our business.’” PMSJ 3 at ex. A. • A December 15, 2000 Bloomberg article reported: “The economic slowdown isn’t hurting Oracle, said Oracle Chief Executive Larry Ellison, because the company has spent the past three years updating its product line to focus on software that helps companies use the Internet to cut costs and boost efficiency.” PMSJ 3 at ex. B (quoted in RSAC ¶ 45(d)). • A January 10, 2001 report by Salomon Smith Barney: “Oracle sees robust demand for both its database and applications business. . . . Oracle says it is also seeing sustained demand for its database product, despite industry-wide concern over contracting IT budgets.” PMSJ 7. • January 10, 2001 RealMoney.com interview quoting Sanderson as responding to question about whether Oracle could repeat its performance on applications sales, “You know, it’s a big hill to climb. Every year we climb that hill. I expect we’ll do it again. Our pipelines are strong, we’re well positioned from a products perspective, and so it’s all about execution.” POpp. 208 (also PMSJ 6 at 141677). • A January 11, 2001 Bloomberg article reported, “Company spokeswoman Stephanie Aas today said Oracle has yet to see any signs that its business is being hurt by the economic slowdown or reported cuts to information-technology budgets.” PSJM 3 at ex. C (also POpp. 207 at ex. C) (quoted in RSAC ¶ 60). • A February 7, 2001 First Union Securities report stating, “Oracle is not seeing the effects of a slowing economy at this point, but next several weeks will be critical.” POpp. 215 (quoted in RSAC ¶ 65(a)). • A February 8, 2001 First Union Securities report that stated, “Oracle is not seeing the effects of a slowing economy at this point, but next several weeks will be critical. CFO Henley commented that Oracle is not seeing a decline in sales at this point as a result of reduced corporate spending, although this issue has plagued several other large technology companies. While the sales pipeline apparently shows no signs of weakness at this point, we note the next several weeks will be critical for the company as many potential customers will likely make decisions to buy or defer purchase during the activity-intensive final weeks of 3Q01.” POpp. 209 at 91531-32 (also PSJM 5). • Deutsche Banc reported on February 8, 2001, after a meeting with Henley, “According to management, it has yet to see macro-related weakness in its business. That said, the full impact of the current macro environment may not be evident until the end of the quarter, as revenue is typically back-end loaded 7 8 9 United States District Court For the Northern District of California 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 13 1 for Oracle.” “Barring a severe economic downturn, management sees continued growth driven by strong demand in key segments such as supply chain, customer relationship management, and collaboration.” POpp. 208 (also PSJM 4 at 91536) (quoted in RSAC ¶ 65(b)). 2 3 • Portions of various reports on comments made by Henley at the AppsWorld conference in New Orleans on February 21, 2001. See PSJM 8 (POpp. 216), 10 (POpp. 213), 11 (POpp. 214), POpp. 212 (PSJM 9).12 6 iii. Intra-quarter repetitions of the 3Q01 forecast 7 • January 10, 2001 RealMoney.com interview (described above: POpp. 208 / PMSJ 6 at 141677). • Salomon Smith Barney report dated January 10, 2001 (described above: POpp. 211). • January 11, 2001 and February 9, 2001 Bloomberg articles (described above: PSJM 3 at ex. C / POpp. 207 at ex. C). • February 7, 2001 First Union Securities report (described above: POpp. 215). • February 8, 2001 Deutsche Banc report (described above: PSJM 4 at 91536 / POpp. 208). • February 8, 2001 First Union Securities report (described above: POpp. 209 at 91531-32 / PSJM 5). • Portions of various reports on comments made by Henley at the AppsWorld conference in New Orleans on February 21, 2001. See PSJM 8 (POpp. 216), 10 (POpp. 213), 11 (POpp. 214), POpp. 212 (PSJM 9). B. Functionality of Suite 11i 4 5 8 9 United States District Court For the Northern District of California 10 11 12 13 14 15 16 17 18 19 20 21 22 i. Falsity It was no secret before and during 3Q01 that Suite 11i was an imperfect product. The following problems were discussed by analysts and reported in financial publications:13 • 23 24 25 The inherent instability of a new, untested product, see DX 127 at 307330 (May 8, 2000 Business Week article: “Even after the suite ships, consultants such as Gartner Group Inc. warn corporate customers that it probably won’t be stable enough to handle the most crucial jobs until the end of the year.”); DX 207 (November 15, 2000 Business Wire press release entitled “Oracle Applications Users Ask Oracle Corp. Executives about Quality, Customer Support, 26 12 27 Portions of this exhibit would be admissible; this is discussed in more detail infra. 13 28 These reports are not hearsay to the extent they are offered not for their truth but to prove the reports were made. 14 1 Functionality and Pricing:” “11i is not yet working optimally.”); DX 130 at 5916 (December 13, 2000 CIBC Markets Corp: “We also expect that customers would prefer not to be among the early adopters, waiting until some of the initial bugs get worked out of the software, which may take a quarter or so.”); DX 210 at 309148 (February 22, 2001 CIBC World Markets report: “[T]he current version of 11i is noted to have many [] bugs (close to 5,000). . . . We think that the delays in the upgrade cycle pose a near-term risk for applications sales”); 2 3 4 5 • Unfavorable comparisons with best of breed products, see DX 245 at 85851 (September 15, 2000 Deutsche Banc report: “some 11i modules still may fall short on functionality compared with best-of-breed rivals”); • Lack of references, see DX 149 at 419950 (November 8, 2000 Robertson Stephens, Inc: “Although the company has announced a number of 11i customer wins, none of the bigger names, including BellSouth, GE, and Lucent, have gone live yet. We believe it will take several more quarters for the company to implement these customers and to use them as references to win additional business.”); DX 208 at 419977 (December 12, 2000 J.P. Morgan Equity Research: “We believe Oracle’s 11i e-business suite continues to have bugs thereby limiting the number of notable customer references for the new product.”); • 15 Lack of integration, see DX 140 at 9467 (December 4, 2000 GartnerGroup report: “Release of Oracle Applications r.11i has prompted inquiries about its robustness . . . [Oracle’s] track record of facilitating integration between multiple products is not strong, nor is Oracle building its ERP and ERP-complimentary applications . . . for easy integration with other products. . . . 11i as a complete ERP suite remains suitable primarily for risk-tolerant, early-adopter-oriented enterprises.”). 16 The fact that problems with Suite 11i were known to the market raises a serious question as to 17 whether plaintiffs can show that any of defendants’ purportedly false statements about the product were 18 materially misleading. In a fraud-on-the-market case such as this, “an omission is materially misleading 19 only if the information has not already entered the market.” In re Convergent Techs. Sec. Litig., 948 20 F.2d 507, 513 (9th Cir. 1991) (citing In re Apple Computer Sec. Litig., 886 F.2d 1109, 1114 (9th Cir. 21 1109)). “If the market has become aware of the allegedly concealed information, ‘the facts allegedly 22 omitted by the defendant would already be reflected in the stock’s price’ and the market ‘will not be 23 misled.’ ” Id. (quoting In re Apple Computer Securities Litig., 886 F.2d at 1114). 6 7 8 9 United States District Court For the Northern District of California 10 11 12 13 14 24 25 The Court need not decide this issue, however, because the Court agrees with defendants that there is not a genuine factual dispute on loss causation. 26 27 28 15 United States District Court For the Northern District of California 1 ii. Loss causation 2 The parties dispute plaintiffs’ burden at summary judgment in demonstrating the existence of 3 a factual dispute on loss causation, the sixth element of a private § 10(b) action. Loss causation is the 4 causal connection between the defendant’s material misrepresentation and the plaintiff’s loss. Metzler 5 Inv. GMBH v. Corinthian Colleges, Inc., 540 F.3d 1049, 1062 (9th Cir. 2008) (citing Dura, 544 U.S. 6 at 342). “A plaintiff bears the burden of proving that a defendant’s alleged unlawful act ‘caused the loss 7 for which the plaintiff seeks to recover damages.’” In re Gilead Sciences Sec. Litig, 536 F.3d 1049, 8 1055 (9th Cir. 2008) (quoting 15 U.S.C. § 78u-4(b)(4)). Put another way, “[t]o establish loss causation, 9 ‘the plaintiff must demonstrate a causal connection between the deceptive acts that form the basis for 10 the claim of securities fraud and the injury suffered by the plaintiff.’” Id. (citing In re Daou Sys., Inc., 11 411 F.3d 1006, 1025 (9th Cir. 2005)); see also Metzler, 540 F.3d at 1063 (the plaintiff must show that 12 “the practices that the plaintiff contends are fraudulent were revealed to the market and caused the 13 resulting losses”) (discussing Daou). “The misrepresentation need not be the sole reason for the decline 14 in value of the securities, but it must be a ‘substantial cause.’” Gilead, 536 F.3d at 1055 (citing Daou, 15 411 F.3d at 1025). 16 Analysis of loss causation calls on courts to perform a “balancing act” between allowing 17 plaintiffs to link “each and every bit of negative information about a company to an initial 18 misrepresentation that overstated that company’s chances for success” and exacting such a high standard 19 as to “eliminate the possibility of 10b-5 claims altogether.” In re Williams Sec. Litig., 558 F.3d 1130, 20 1140 (10th Cir. 2009). At one end of the spectrum, it is clear that the plaintiff need not prove that the 21 defendant admitted a fraud. Metzler, 540 F.3d at 1064. At the other extreme, it is equally clear that the 22 plaintiff must do more than show that the market was “merely reacting to reports of the defendant’s poor 23 financial health generally.” Id. at 1063. 24 Plaintiffs argue that Oracle officials’ misrepresentations that Suite 11i was fully functional led 25 to unrealistically high earnings expectations, which were corrected when the market recognized the true 26 state of Oracle’s flagship product on March 1, 2001. According to plaintiffs, the March 1 analyst call 27 “disclosed the negative effects of 11i” and “communicated to investors that issues concerning the 28 functionality, i.e., bugs and lack of stability of Suite 11i . . ., had not in fact been cured as defendants 16 1 2 This is not an accurate characterization of the content of the phone call. In fact, Ellison 3 repeatedly told analysts that the miss in applications sales was caused by nervousness about the slowing 4 economy, not problems with Oracle’s products: 5 6 7 8 9 10 United States District Court For the Northern District of California had reported at the end of 2Q01.” Pl. Opp. to Def. Mot. for Summ. J. at 46. It really appears to be economic factors, where people actually need the database, where actually they were getting ready to sign deals, people just delayed. Where there is no question they were going to buy, they were going to go ahead and buy but they are trying to push it out as long as possible. . . . [T]his was not a matter of they’re not going to buy; they would just like to wait 30 days or 60 days. They’re just looking at the economy. Everyone is trying to get a read on this economy and everyone is being slow to act in light of the economy. DX 393, 419804-05. Later in the call, Ellison said: 12 I do know that some of these transactions can only be deferred for a short period of time. Because these are projects that are going through an implementation cycle where they actually are using the software, or about to start using the software so they have to buy. On the other hand, there are, you know, other projects that can be deferred for three or six months. 13 Id. at 419807-08. He repeated that applications sales were delayed only temporarily: “They can only 14 delay so long . . . [] because they’re actually using the software and the applications are growing.” Id. at 15 419808. And once more: 11 16 18 All the indications that we have are that people want to do these projects. People want to put in the e-business suite. They’re actually – if you look at budgets, the database budgets, are going up with the exceptions of the dot.coms. The database budgets are going up everywhere. So all of that looks very, very good. It’s just this umbrella of uncertainty that is causing people to defer decisions. 19 Id. at 419812. When asked for his impression for now long it was taking for CEOs to decide whether 20 to buy the “suite” (presumably Suite 11i), Ellison responded: 17 21 22 23 24 25 26 27 28 Well, again, some of these guys are moving incredibly fast. So we have an example of going from the first meeting to deal in 30 days or 60 days. We’ve got several examples of those. We got, you know, my favorite example which I cited, GE Power, from contract to live on manufacturing financial e-business suite, redoing all the business processes in five months. So we are moving very, very quickly with a variety of customers. Id. at 419814. Ellison also emphasized that Oracle had failed to close deals because of the economy, not because the company was losing out to competitors: “[T]hese were not cases of deals that we lost competitively, or deals are going away. They’re just being shut down. At some point they’re you know, the customers are going to have to buy.” Id. at 419815. He repeated this point at the close of the 17 United States District Court For the Northern District of California 1 conversation: “These are not deals that we lost competitively. These are not deals where they decided 2 not to buy. These are literally deferrals because of economic uncertainty. . . . [A]s we wear on in Q4 3 a bunch of these deals should come in.” Id. at 419817. At no point during the phone call did Henley 4 or Ellison say that there was anything wrong with the Suite 11i or suggest that the cancelled sales were 5 due to anything other than customers’ fear of making a big investment at a time when the economy was 6 uncertain. 7 Plaintiffs also claim that analyst reports after the March 1 conference call demonstrate that the 8 market understood the announcement to have revealed problems with Suite 11i. Defendants respond 9 with overwhelming evidence that the market understood the announcement as disclosing that the 10 earnings shortfall was caused by the economic downturn. See DX 406 (Banc of America Securities), 11 407 (Bloomberg), 410 (Salomon Smith Barney), 409 (Lehman Brothers), 465 (FAC/Equities), 466 12 (Prudential), 467 (Wit SoundView).14 Defendants also cite evidence that because analysts interpreted 13 the earnings to be a bad “omen” for other applications vendors, see DX 410 at 91361, they downgraded 14 their ratings across the enterprise software industry. See DX 451-57. 15 For the most part, plaintiffs do not accurately describe the evidence they cite in support of their 16 argument that the market recognized the March 1 announcement as revealing problems with Suite 11i. 17 Plaintiffs refer to a March 2, 2001 Wall Street Journal article as evidence that analysts attributed the 18 miss to problems with Suite 11i. In fact, the article did not discuss Oracle’s applications sales. The 19 relevant portion reads: 20 22 The reasons for the shortfall were at least as troubling to analysts as its magnitude. Oracle’s database software, a mainstay product line used as the foundation for many other business programs, had flat to negative growth over the year-earlier period. While database sales have been slowing at Oracle for many quarters, analysts were surprised by the abruptness of the latest downturn. 23 POpp. 436. That is, the shortfall was “troubling” because it was due to slower sales in Oracle’s 24 “mainstay product line” of database software, not because there were problems with Oracle’s new 25 applications product. 21 26 The same article also reported that Oracle’s application software had grown 50%, rather than 27 28 14 The analyst reports are not hearsay if offered to prove merely that the reports were made. 18 1 the 75%-100% predicted. As Ellison and Henley did the day before, the writers attributed the shortfall 2 to the slowdown in the dot.com sector: 3 Oracle had already been hurt by a falloff in orders from dot-com start-ups, many of which used Oracle databases to build new Web services. The company had been counting on conventional companies taking up the slack, using Oracle software to develop new electronic-business applications and improve their internal efficiency. Still, the CEOs held off signing the purchase orders. “That was true, even where it was acknowledged that this deal would save the company money,” Mr. Ellison said. “We have a lot of nervous senior executives looking at this economy and being very cautious.” Bob Austrian, a Banc of America Securities analyst who had cut his numbers for Oracle earlier this week, said the announcement showed that “the economic downturn has become severe enough that it has become a shock. And shocks always impair purchasing decisions.” 4 5 6 7 8 9 United States District Court For the Northern District of California 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 Id. Plaintiffs’ citation to a March 2, 2001 Los Angeles Times article is similarly unhelpful. Reporting on the earnings miss, the article repeated Ellison’s representations that the shortfall was due to the economy. It concluded, “In December, the company said it wasn’t being hurt by the slowdown because corporations were buying its applications software to cut costs and boost efficiency.” PReply 173. The writers did not comment on the functionality of Suite 11i. Out of the flurry of news and analyst reports on Oracle’s earnings miss, plaintiffs cite only two that discussed problems with Oracle’s applications products. The first is a March 2 report by the financial firm UBS Warburg. The writers thought that Oracle’s announcement the day before had put too much emphasis on the economic slowdown and stated, “we believe there may have been further contributors.” POpp. 426. According to UBS Warburg, the “major shortfall” came from the “flat to slightly negative” growth in Oracle’s database business. Id. After a discussion of various factors other than the economy that might have contributed the database shortfall, the writers suggested that problems with Oracle’s applications products could been a contributing factor: On a somewhat more positive note, the applications business did considerably better than the database business. Granted, the applications business missed our estimates as well, but growth for these products came in around 50% year over year. . . . We also believe that the weakness in Oracle’s applications business is because the company’s applications are not yet ready for prime time. At Oracle’s applications conference last week, we learned that over 200 patches had been developed for the CRM product for the latest version. Furthermore, many customers we talked with indicated that although the 28 19 1 2 CRM product showed promise, the SCM15 products are not even on the radar. Although [f]eedback from these customers suggested that they were impressed with the idea of a fully integrated suite, we were unable to find any that had fully integrated and gone live on the suite. 3 POSJM 426. The second report is a March 16 market research summary by Banc of America 4 Securities. Under the subheading “Are Oracle’s problems entirely the economy? We don’t think so,” 5 the writers state, “[o]n the applications side, especially in light of Oracle’s weaker than expected 3Q 6 applications growth, we believe the economy may only explain 20-30% of the weakness. The rest, in 7 our view, is a result of the product set not yet reaching a competitive level of functionality, relative to 8 9 United States District Court For the Northern District of California 10 best-of-breed vendors.” PSJM 113 at 2710.16 These two reports, neither of which indicates that the writers learned new information about the 11 15 12 Presumably “supply chain management.” 16 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 As evidence that the market linked the shortfall to Suite 11i, plaintiffs also cite an e-mail from Oracle employee Karen Houston stating, “Over the past couple of weeks, several stories have been published on 11i product quality issues. . . . Additionally, we have seen several stories recently in the US that link our lower than expected apps sales to quality issues of 11i.” POpp. 420 at 158596. Defendants object that Houston’s statement is hearsay to the extent that it is offered to prove the truth of what the news reports said. See Docket No. 1586. The Court agrees that Houston’s repetition of the out of court statements in the news reports is inadmissible hearsay. In any event, the e-mail is dated March 21, 2001 and therefore is not evidence of how the market reacted to the March 1 disclosures. In addition, the e-mail does not indicate that the market learned of “product quality issues” with Suite 11i through the March 1 disclosures. Plaintiffs also cite to the following passage from Softwar: An Intimate Portrait of Larry Ellison and Oracle, a profile by Matthew Symonds: “It didn’t take a genius to see that not everything that was going on could be explained by the weakening economy and edgy CEOs waiting for ‘visibility’ to return. For anyone who wanted to see, there was mounting evidence that it wasn’t only the economy that prospective Oracle applications customers wanted to see stabilize.” PSJM 12 at 201. Defendants object that Symonds’ book is hearsay. See Docket No. 1585. The Court agrees. Plaintiffs repeatedly cite to Softwar for the truth of the matters asserted therein but offer no basis for the admissibility of these statements. The Court also notes that Symonds’ statement about what analysts said is hearsay within hearsay. Finally, plaintiffs cite what purports to be an e-mail chain (dated March 22, 2001) of Oracle employees commenting on a draft of an article. The cited portion of the article reads, “Oracle blamed the economic slowdown in the United States for affecting its business, with many of its enterprise customers deferring purchases. However its lowered sales performance across the board also stems from several factors unique to Oracle: database pricing, 11i quality and the suite approach to selling applications.” PReply 177. Defendants object that Oracle e-mails about news reports are hearsay. See Docket No. 1586 at 5. The Court agrees that the e-mail is hearsay to the extent it is offered to prove what an article stated. I n addition, it is also not evident when, if ever, the article was published. In any event, an article from March 22 is not probative of what the market learned from the March 1 disclosures. Defendants’ objections to these documents are SUSTAINED. 28 20 United States District Court For the Northern District of California 1 functionality of Suite11i through the March 1 conference call, are the closest plaintiffs come to citing 2 evidence that the market recognized that quality problems with Suite 11i contributed to the miss in 3 Oracle’s applications forecast. The Court will assume for the sake of argument that a rational factfinder 4 could conclude from these two reports that the market linked Oracle’s miss to the following problems 5 with Suite 11i: that it was “not yet ready for prime time” because it still required patches, that SCM 6 products were not available, that customers had not yet “gone live” with Suite 11i, and that it was not 7 competitive with “best of breed” products. The problem for plaintiffs is that none of these issues had 8 been hidden from the market. As discussed above, defendants have cited abundant evidence that 9 beginning with the release of Suite 11i and continuing through late February of 2001, public reports had 10 discussed these deficiencies with the product. Thus, even those analysts who linked the miss to 11 deficiencies with Suite 11i did not do so on the basis of information that had previously been hidden 12 from the market. 13 In sum, there is no evidence that on March 1, Oracle revealed previously undisclosed facts about 14 Suite 11i to the market or that the market recognized the earnings miss as being caused by previously 15 undisclosed problems with this product. Plaintiffs’ only possible loss causation theory for Suite 11i is 16 therefore that Oracle revealed the truth about Suite 11i on March 1 by announcing that the year over 17 year growth in applications would be 50%, not the 75% the company had projected. Ninth Circuit 18 precedent is clear, however, that an earnings miss alone is not sufficient proof of loss causation. 19 The missing causal link here is similar to that in Metzler, in which the alleged fraud involved 20 the manipulation of student enrollment figures to obtain federal funding. The Ninth Circuit held that 21 the plaintiffs had not pled loss causation because they had not alleged facts in support of their claim that 22 a press release revealed the purportedly improper financial aid practices. See 540 F.3d at 1063 23 (plaintiffs must allege that the market learned of and reacted to the fraud, “as opposed to merely reacting 24 to reports of the defendant’s poor financial health generally.”). It was not sufficient for the plaintiffs 25 to point to a euphemistic reference in a press release to “higher than anticipated attrition” and allege that 26 the market understood this statement to reveal that the defendant company had overstated its 27 enrollments: “So long as there is a drop in a stock’s price, a plaintiff will always be able to contend that 28 the market ‘understood’ a defendant’s statement precipitating a loss as a coded message revealing the 21 United States District Court For the Northern District of California 1 fraud. . . . Loss causation requires more.” Id. at 1064. Here, Henley and Ellison did not make even a 2 euphemistic reference to problems with Suite 11i during the March 1 call. Instead, they repeatedly 3 assured analysts that the only problem was the economy. As a matter of logic, Oracle cannot have 4 revealed the fraud by repeating the purported misrepresentations about the functionality of Suite 11i. 5 The Ninth Circuit’s decisions in Gilead and Daou confirm that loss causation requires more than 6 a company’s announcement of a missed financial projection. The fraud alleged in Gilead involved off- 7 label marketing of a drug, a practice that purportedly accounted for 75% to 95% of the defendant 8 company’s revenue from the drug and inflated the company’s stock price. 536 F.3d at 1058. The 9 plaintiffs pled loss causation through their allegations that the company released an FDA warning letter 10 that disclosed the off-label marketing, causing physicians to write fewer prescriptions for the drug, 11 which in turn led to decreased revenues and ultimately to the company’s announcement of lower than 12 expected revenues, after which the stock price dropped. Id. 13 In Daou, the alleged fraud was that the company was reporting revenues before they were 14 earned. Daou held that loss causation was established through the allegations that the defendants 15 revealed “figures showing the company’s true financial condition,” including (1) that its operating 16 expenses and margins were deteriorating, (2) that it would have to report a loss of $0.17 a share, and 17 (3) the existence of $10 million in unbilled receivables in its work in progress account. 411 F.3d at 18 1026. Notably, the $10 million appeared as “the direct result of prematurely recognizing revenue.” Id. 19 After these revelations, an analyst noted, “You have got to question whether they are manufacturing 20 earnings.” Id. Thus, the plaintiffs properly alleged that the market recognized the disclosures as 21 revealing the defendants’ allegedly improper accounting practices. 22 In conclusion, there is an absence of evidence that on March 1, Oracle revealed previously 23 concealed information about Suite 11i or that analyst reports about the March 1 announcement linked 24 the miss in Oracle’s applications earnings to previously concealed deficiencies with Suite 11i. 25 Plaintiffs’ only possible theory for loss causation is that the earnings miss itself revealed the truth about 26 Suite 11i to the market, but plaintiffs cite no case in which an earnings miss alone was sufficient to 27 prove loss causation. Thus, the Court concludes that plaintiffs have not identified evidence that could 28 lead a juror to conclude that defendants’ alleged misrepresentations about Suite 11i were a “substantial 22 1 cause” of the decline in value of Oracle’s stock. See Gilead, 536 F.3d at 1055. The Court GRANTS 2 defendants’ motion for summary judgment on this issue. United States District Court For the Northern District of California 3 4 C. 5 Plaintiffs contend that Oracle violated Generally Accepted Accounting Principles (“GAAP”) in 6 2Q01 through two accounting frauds. The first alleged accounting fraud involved Oracle’s method of 7 accounting for customer overpayments.17 According to plaintiffs’ expert D. Paul Regan,18 by 2Q01, 8 Oracle had accumulated at least $144 million in “unapplied cash” – cash receipts that the company could 9 not apply to an invoice. A significant amount of the unapplied cash was from overpayments made by 10 Oracle’s customers. Regan opines that Oracle improperly inflated its 2Q01 financial results by 11 transferring the unapplied cash to its bad debt reserve and subsequently applying the overpayments to 12 “debit memo” invoices. The debit memos made the overpayments appear to be refunded or “applied,” 13 creating the impression that Oracle had reduced its bad debt reserve by $20 million. The end result was 14 that, in Regan’s opinion, Oracle made an improper adjustment of $20 million to its revenue and pre-tax 15 earnings, allowing the company to overstate its earnings per share by $0.01. See Regan Report. Oracle’s financial results for 2Q01 16 The second purported 2Q01 accounting fraud involved a deal with computer company Hewlett 17 Packard (“HP”). According to Regan, Oracle improperly recognized $19.9 million in revenue and pre- 18 tax earnings on November 30, 2000 (the last day of 2Q01). Regan opines that HP agreed to buy 19 software that it did not need from Oracle and that it did so pursuant to an agreement that Oracle would 20 buy $30 million in hardware from HP over the following quarter. In Regan’s opinion, the evidence 21 shows that products sold to HP under the agreement lacked a valid business purpose. This arrangement, 22 according to Regan, allowed Oracle to overstate its 2Q01 earnings per share by $0.01. Id. 23 In sum, Regan opines that each of the two accounting frauds – overstatement of earnings related 24 to the transfer of customer overpayments and the deal with HP – independently allowed Oracle to 25 26 27 28 17 Customer overpayments are caused by mistakes such as duplicate payments of an invoice, payments on amounts that were credited, amounts paid where the debt had been cancelled, and payments of unnecessary tax. 18 Defendants have not moved to exclude this expert opinion. 23 1 overstate its 2Q01 earnings per share by $0.01. Eliminating either one of these improper practices 2 would have caused Oracle to report earnings per share of $0.10, rather than the $0.11 that it did report. 3 Id. 4 United States District Court For the Northern District of California 5 i. Pleading 6 As an initial matter, defendants argue that plaintiffs did not plead these theories of accounting 7 fraud in the operative complaint and may not introduce them into the case at summary judgment. In 8 their revised second amended complaint (“RSAC”), plaintiffs allege that in 2Q01, defendants “created 9 phony sales invoices and improperly recognized revenue from past customer credits and overpayments 10 it had held in reserve without informing its customers. . . . Oracle held the money in what it called its 11 ‘unapplied account.’” RSAC ¶ 8. Plaintiffs also allege that Oracle’s practice was to not refund 12 customer overpayments, or to do so only at the request of the customer. Id. ¶ 36. The Court finds that 13 plaintiffs sufficiently pled debit memo accounting fraud though these allegations. Defendants point out 14 that the plaintiffs did not specifically allege that the customer overpayments were moved to Oracle’s 15 bad debt reserve. This distinction is not dispositive, however, as it can be expected that some details 16 of an accounting fraud will materialize during discovery. The crux of plaintiff’s theory was that Oracle 17 created the appearance of revenue through improperly accounting for customer overpayments. This 18 theory was alleged in the RSAC. 19 In contrast, the Court agrees with defendants that plaintiffs did not allege their second accounting 20 fraud theory – the RSAC is devoid of any reference to HP. Plaintiffs do not dispute that they did not 21 allege the fraudulent HP swap. Instead, they argue that defendants have “been fully aware of the HP 22 allegations for years of litigation” and that plaintiffs detailed this theory in their contention interrogatory 23 responses. Pl. Opp. at 10 n.15. 24 The PSLRA amended the Exchange Act to apply a heightened pleading standard to private class 25 actions. See Oracle, 380 F.3d at 1230. “To avoid dismissal under the PSLRA, the Complaint must 26 ‘specify each statement alleged to have been misleading, the reason or reasons why the statement is 27 misleading, and if an allegation regarding the statement or omission is made on information and belief, 28 the complaint shall state with particularity all facts on which the belief is formed.’” Id. (quoting 15 24 United States District Court For the Northern District of California 1 U.S.C. § 78u-4(b)(1)). Permitting plaintiffs to add an unpled fraud theory to the case now, when the 2 case is at summary judgment, would effectively dispense with the “formidable” pleading requirements 3 of the PSLRA.19 See Metzler, 540 F.3d at 1055; see also In re Stratosphere Corp. Sec. Litig., 66 F. 4 Supp. 2d 1182, 1201 (D. Nev. 1999) (“To allow [p]laintiffs to amend their [complaint] at this late stage 5 of the proceedings would render the particularity requirement for pleading securities fraud a nullity.”). 6 Plaintiffs cite In re JDS Uniphase Corp. Sec. Litig., 2007 WL 2429593 (N.D. Cal. Aug. 24, 7 2007) for the proposition that if defendants cannot show undue delay, bad faith, or dilatory motive, this 8 Court may consider plaintiffs’ unpled allegations at summary judgment. JDS Uniphase is inapt because 9 the unpled allegations in that case consisted of nineteen purportedly false statements. There was no 10 suggestion that the plaintiffs had failed to allege a fraudulent scheme at issue in the case. Here, the 11 unpled allegation concerns an entirely new theory of accounting fraud. Accordingly, the Court agrees 12 with defendants that the HP swap accounting fraud is unpled and GRANTS defendants’ motion for 13 summary judgment on this claim. 14 15 ii. Loss causation 16 Plaintiffs argue that Oracle’s debit memo accounting fraud, which purportedly allowed the 17 company to overstate its 2Q01 earnings per share by $0.01, was revealed to the market on March 1 when 18 Oracle reported its earnings miss. According to plaintiffs, the accounting fraud allowed Oracle to 19 conceal the problems with Suite 11i by overstating its applications sales. Importantly, Regan opines that 20 the HP fraud (not the debit memo fraud) allowed Oracle to report its 2Q01 applications business growth 21 rate as 66%, while the accurate figure was 54%. See Decl. of Shawn Williams in Supp. of Expert Report 22 of D. Paul Regan, ex. 2 (Regan Rebuttal Report) at 6-7. Oracle disclosed this fraud, according to 23 plaintiffs, when it revealed the truth about Suite 11i on March 1.20 24 25 26 27 28 19 Plaintiffs do not dispute defendants’ contention that allowing plaintiffs to amend their complaint at this juncture would be futile in light of the five year statute of repose for §10(b) claims. See 28 U.S.C. § 1658. 20 Plaintiffs also argue that the false 2Q01 earnings were the basis for Oracle’s 3Q01 forecast and rendered the forecast unreliable. According to plaintiffs, the fraud was revealed to the market on March 1 when Oracle announced its earnings miss. The Court considers this theory in more detail in 25 1 This theory of loss causation fails for several reasons. First, as discussed above, there is no 2 evidence that the market recognized the March 1 disclosures as revealing previously undisclosed 3 information about Suite 11i. Second, plaintiffs’ expert links the inflated applications growth rate to the 4 deal with HP, not the debit memo fraud. As the debit memo fraud is the only 2Q01 accounting fraud 5 that remains in the case, there is no evidence that Oracle misstated its applications revenues in 2Q01. 6 Third, there is no evidence that the market understood the March 1 earnings miss as revealing that 7 Oracle had misstated its earnings for 2Q01. To the contrary, in their post-March 1 reports, analysts 8 continued to report that Oracle’s 2Q01 earnings per share were $0.11, not the $0.10 that plaintiffs claim 9 is the accurate figure. See DX 465, 466, 473, 475, 478, 480, 482, 483, 484, 485, 486.21 There is no United States District Court For the Northern District of California 10 evidence that Oracle has ever restated its 2Q01 earnings. 11 None of the cases cited by plaintiffs support their argument that Oracle’s 3Q01 earnings miss 12 revealed the 2Q01 accounting fraud. For example, the plaintiffs in In re Impax Labs., Inc. Sec. Litig. 13 pled loss causation by alleging that the company revealed the accounting fraud that led to its erroneous 14 revenue statements for 1Q04 and 2Q04 by announcing its actual financial results for those quarters. See 15 2007 U.S. Dist. LEXIS 52356 (N.D. Cal. July 18, 2007) (Ware, J.). The disclosure “explicitly 16 pertained” to the company’s results for 1Q04 and 2Q04. Id. at *17-18. Here, in contrast, the March 1 17 disclosure made no mention of Oracle’s 2Q01 results. 18 Accordingly, the Court finds that plaintiffs have failed to establish that there is a triable issue 19 as to whether the inflation of Oracle’s 2Q01 earnings per share through the purported debit memo fraud 20 was a substantial cause of plaintiffs’ loss. Defendants’ motion for summary judgment on this issue is 21 GRANTED. 22 23 24 25 26 27 28 conjunction with plaintiffs’ argument that there was no reasonable basis for the 3Q01 forecast. 21 The analyst reports are not hearsay if offered to prove that the reports were made. 26 D. 2 Plaintiffs argue that Oracle’s December 14, 2000 public forecast of its 3Q01 earnings was 3 materially false. In order for a financial projection to give rise to 10b-5 liability, the plaintiff must 4 prove that “(1) the statement is not actually believed, (2) there is no reasonable basis for the belief, or 5 (3) the speaker is aware of undisclosed facts tending seriously to undermine the statement’s accuracy.” 6 Provenz v. Miller, 102 F.3d 1478, 1487 (9th Cir. 1996) (citation omitted); see also In re Adobe Systems, 7 Inc. Sec. Litig., 787 F. Supp. 912, 919 (N.D. Cal. 1992) (“10b-5 liability for a projection requires that 8 there be either no reasonable basis for believing that the projection was accurate or the awareness of 9 undisclosed facts tending seriously to undermine the accuracy of that projection.”) (emphasis in 10 United States District Court For the Northern District of California Oracle’s December 14, 2000 public guidance for 3Q0122 1 original). 11 Defendants contend that Oracle’s December 14, 2000 public guidance was based on its internal 12 forecast, as compiled by Minton. This forecasting method had proven to be consistently reliable at 13 predicting Oracle’s quarterly performance – in the seven quarters before 3Q01, Oracle had met or 14 exceeded the forecast. Defendants argue that because the public guidance was based on Oracle’s proven 15 internal forecasting method, there can be no factual dispute that Oracle had a reasonable basis for its 16 public guidance. Plaintiffs respond with five theories for why Oracle had no reasonable basis for the 17 public guidance. The Court will consider each in turn. 18 19 i. Defendants’ failure to take into account the end of the dot.com boom 20 Plaintiffs argue that by December of 2000, the speculative period known as the “dot.com bubble” 21 had burst and that the U.S. economy was slowing. Plaintiffs claim that the beginnings of the dot.com 22 bust rendered Oracle’s forecast fundamentally unreliable because it was based on the unreasonable 23 assumption that Oracle would convert as much of its pipeline in 3Q01 as it had in the boom economy 24 of 3Q00. 25 26 27 28 22 For the purposes of this discussion, the Court assumes that this projection was not accompanied by meaningful cautionary language and therefore does not fall within PSLRA’s “safe harbor” provision for forward looking statements. See 15 U.S.C. § 78u-5(c); see also Employers Teamsters Local Nos. 175 & 505 Pension Trust Fund v. Clorox Co., 353 F.3d 1125, 1132 (9th Cir. 2004). 27 United States District Court For the Northern District of California 1 The first point of contention is whether, as plaintiffs claim, the 3Q00 conversion ratio was the 2 “foundation” of Oracle’s 3Q01 public guidance. See Pl. Reply at 4. Defendants contend that the 3 historical conversion ratio was only one of several factors that Oracle used to arrive at its forecast. They 4 cite Minton’s testimony that she considered the conversion ratio in conjunction with conversations with 5 heads of Oracle’s business units, field forecasts, and information she received about the status of 6 especially large potential deals. See DX 333 at 122-26. Plaintiffs concede that the field reports were 7 a factor that Oracle used in arriving at its internal forecast. See Pl. Mot. at 4. The evidence plaintiffs 8 rely on for their characterization of the conversion ratio as the “foundation” of the forecast does not 9 support this point.23 10 Next, the parties dispute how Oracle used the 3Q00 conversion ratio in calculating its 3Q01 11 public guidance. Plaintiffs contend that Oracle “mechanically” applied the prior year’s ratio to the 12 current year’s data. Again, plaintiffs’ basis for this contention is not clear. The conversion ratio for 13 3Q00 was 53%, while the conversion ratio Oracle applied on December 11, 2000 was 48%. DSJM 1 14 at 213095. Plaintiffs do not explain why, if Oracle mechanically applied the prior year’s conversion 15 ratio, there was a five percentage point difference between the ratios applied at this point in 3Q00 and 16 3Q01. 17 18 19 20 21 22 Plaintiffs cite an Oracle e-mail chain from January 8, 2002 that forwarded a message from Jennifer Minton with the following statement: forecast co[n]version ratios – forecast as a % of pipeline. We track this for every forecast period within a quarter. This enables us to evaluate conversion rates. The conversion rates have been declining over historical periods due to the economic recession. When we were going through the dot.com bubble the field would generally “sandbag” their forecast. By evaluating historical trends, Jeff and I would be able to determine what the true forecast was by applying historical conversion rates to the pipeline. As a side note, my upside analysis was usually spot on! 23 24 25 26 27 28 23 Plaintiffs cite Minton’s deposition testimony at 122:1-24, 132:11-136:3 and 157:14-159:17 (PSJM A) in support of their contention that the upside adjustment was entirely based on the prior year’s conversion ratio. In fact, in each of the cited portions of her deposition, Minton testified that the historical conversion ratio was one of a variety of factors she considered – “a number of data points,” as she put it. PSJM A at 159:17. 28 United States District Court For the Northern District of California 1 POpp. 226 at 132078-79.24 This description of the forecasting process gives no information about 2 how Minton applied “historical conversion rates” to the pipeline. She does not state, as plaintiffs 3 contend, that she mechanically applied the exact same conversion rate from the prior year without regard 4 to any other factors. There is therefore no factual dispute that the 3Q00 conversion ratio was just one 5 of several factors that Oracle used to determine the 3Q01 public guidance.25 6 The parties also dispute whether the end of the dot.com boom necessarily rendered Oracle’s 7 forecasting system unreliable. Plaintiffs focus on a statement Ellison made in Matthew Symonds’ 8 profile Softwar: “LE writes: As I’ve said before, our forecasting system is not clairvoyant. Our 9 forecasting does statistical extrapolations based on historic trends. If something that’s outside our 10 mathematical model of the business changes, like a war in the Middle East, our forecasting becomes 11 inaccurate.” POpp. 17 at 226, fn. Ellison was questioned about this statement at his deposition, as 12 follows: 13 14 15 16 17 18 19 Q: [Y]ou stand by that? A: Absolutely. Q: Now, it wouldn’t – a war in the Middle East is just one example. There could be who knows how many examples; right? A: Price of oil goes over a hundred dollars a barrel, lots of things. Q: And the point you are making is that while your forecasting system does extrapolations based on historic trends, if something is happening that would suggest that the historical trend is not necessarily reliable, then your forecasting will not be reliable; is that right? A: Right. Major macroeconomic change; sudden – sudden growth in the economy or sudden shrinkage in the economy would cause – you can’t extrapolate anymore. POpp. J at 384:16-385:7. 20 Plaintiffs argue that it was evident in December 2000 that the end of the dot.com boom 21 constituted a “major macroeconomic change.” According to plaintiffs, Ellison has therefore admitted 22 24 23 24 25 26 27 28 Defendants object to that this document is hearsay to the extent it is offered to prove that the forecast was inaccurate. Defendants’ objection is OVERRULED. An internal e-mail chain written by Oracle employees is admissible as an admission by a party opponent. See Fed. R. Evid. 801(d)(2). Plaintiffs also cite exhibit 14, which purports to be notes taken in conjunction with Oracle’s Special Litigation Committee investigation: “Minton explained that prior to Q3 of FY 2001, Oracle thought that it could model out its business, but with the current economic downturn, no analytical models can predict one quarter to the next.” PSJM 14 at 609541. The Committee’s notes are inadmissible hearsay because plaintiffs rely on them to prove the truth of what Minton said. 25 The Court recognizes that plaintiffs’ expert Alan Goedde opines that the historical conversion ratio was the basis for Minton’s forecast. The reliability of Goedde’s opinion will be addressed presently. 29 United States District Court For the Northern District of California 1 that Oracle’s 3Q01 forecasting system was unreliable. 2 Plaintiffs’ argument fails for several reasons. First, Ellison did not define the term “major 3 macroeconomic change.” The examples he gave were war in the Middle East and oil prices rising to 4 over a hundred dollars a barrel. There is no evidence that he also meant that Oracle’s forecasting system 5 became unreliable during an economic slowdown. 6 Second, the evidence shows that when it issued its 3Q01 forecast, Oracle had a reasonable basis 7 for believing it could ride out the economic downturn. Plaintiffs cite several indicators that show the 8 economy was slowing, including the precipitous decline of the NASDAQ between 4Q00 and 3Q01 and 9 the Federal Reserve’s reduction of the federal funds rate twice in one month (an action it had not taken 10 in ten years). See PSJM at exs. 23 & 15 at 15-16 (citing Federal Reserve Board press releases); 16 at 11 ex. 6 (Yahoo! Finance chart showing NASDAQ drop from approximately 5,000 to approximately 12 2,100). Plaintiffs also cite evidence that Oracle’s sales to dot.coms were diminishing. Oracle had 13 already sold its database to the larger dot.coms, so the remaining customers were smaller companies – 14 a “fishing hole [that was] drying up.” PSJM E at 78:7-79:24.26 The difficulty with plaintiffs’ focus 15 these indicators is that between 3Q00 and 2Q01, Oracle’s license sales had increased despite the 16 declines in the NASDAQ and the declines in revenues from dot.com customers, as summarized in the 17 following table: 18 19 Quarter NASDAQ average for quarter Percentage of Oracle revenues from dot.coms Increase in license growth over prior year 1Q00 2633.25 6.82% 8.2% 2Q00 2951.26 6.37% 17% 3Q00 4044.59 10.63% 29% 4Q00 4085.98 9.56% 21% 20 21 22 23 24 25 26 27 28 26 Plaintiffs argue that assistant vice presidents “in NAS ‘began to voice concern’ in December and in January ‘were reluctant to raise their forecast’ as ‘deals began to shrink and get delayed.” Pl. Mot. at 8. They cite exhibit 31, which appears to be a slideshow prepared for a managers’ meeting on April 2, 2001. The slide quoted by plaintiffs reads “Month of December . . . AVPs beginning to voice concern; Month of January – AVPS reluctant to raise their forecast.” PSJM 13 at 179337. This statement is hearsay to the extent plaintiffs rely on it for the truth of what assistant vice presidents were saying in December, 2000 and January, 2001. Defendants’ objection to this evidence is SUSTAINED. 30 1 1Q01 3927.92 6.61% 30% 2 2Q01 3413.08 6.01% 24% 3 3Q01 2591.54 2.89% 5.3% 4 5 6 7 8 9 United States District Court For the Northern District of California 10 11 12 13 14 15 16 17 18 19 20 21 22 23 See PSJM 135. As this evidence shows, the pattern did not hold true in 3Q01, but at issue is what Oracle officials knew at the time they issued the forecast. Plaintiffs do not explain why Minton’s forecast had been accurate in the first two quarters of 2001, when the economy was already weakening. In addition, Oracle intentionally gave a conservative forecast, projecting earnings per share of $0.12, rather than the $0.1282 (which per Oracle’s policy, would have been rounded to $0.13) that Minton had projected. Finally, plaintiffs do not explain why Oracle’s internal forecasting system should have taken the end of the dot.com boom into account when applying the 3Q00 conversion ratio to the pipeline. Minton’s bottom up analysis began with data from sales representatives and direct contact with potential customers. This approach could reasonably have been expected to account for larger economic changes through the direct input of sales representatives, and in the past had proved effective at doing so. In sum, plaintiffs have not cited evidence from which a rational factfinder could conclude that the end of the dot.com boom rendered Oracle’s internal forecasting system so unreliable that there was no reasonable basis for the forecast Oracle provided to the public on December 14, 2000. Contrary to plaintiffs’ assertion, the 3Q00 conversion ratio was only one of several factors that Oracle used to determine its forecast, plaintiffs have not cited evidence that the 3Q00 conversion ratio was “mechanically” applied in 3Q01, Oracle’s record leading up to 3Q01 suggested that it would outperform NASDAQ despite its decrease in dot.com customers, and Oracle’s bottom up method of forecasting provided a means of taking into account effects that larger economic factors would have on sales. 24 25 ii. Defendants’ failure to take into account changes in Oracle’s products 26 Plaintiffs claim that Oracle’s December 14, 2000 public forecast had no reasonable basis because 27 Oracle failed to take account of the fact that it was selling different products than it had in 3Q00. 28 Specifically, plaintiffs point to evidence that in 3Q00, “applications” products made up 18.6% of 31 United States District Court For the Northern District of California 1 Oracle’s pipeline (with “technology” making up the remainder); in 3Q01, Oracle predicted that 2 applications would make up 31.5% of its pipeline – an increase of 13%. PSJM 15 at 85. Plaintiffs also 3 cite evidence that Oracle historically had been less successful at selling applications, which 4 consequently had a lower conversion ratio than technology products. PSJM 15 at ex. 7. According to 5 plaintiffs, the larger percentage of applications in the pipeline meant that the 3Q00 conversion ratio was 6 not a reliable predictor of 3Q01 performance. 7 Plaintiffs’ argument suffers from the same flaw as their contention that the dot.com bust rendered 8 the 3Q00 conversion ratio unreliable: plaintiffs have not put forward evidence that the conversion ratio 9 was the “foundation” of the internal forecasting system or that it was applied “mechanically” by Minton. 10 In addition, defendants note that plaintiffs’ evidence shows that applications also made up a greater 11 percentage of the pipeline in 1Q01 and 2Q01 than it had the previous year. PSJM 15 at ex. 6. Plaintiffs 12 do not explain why, if an increase in applications made the internal forecast unreliable, Oracle exceeded 13 its internal forecast in 1Q01 and 2Q01 despite the increase in applications. Plaintiffs also do not cite 14 evidence that Oracle ever analyzed its pipeline data by applying different conversion rates to its 15 projected applications and technology sales. As discussed above, Oracle’s internal forecasting system 16 had proved consistently reliable, even though it did not separate its projections for applications and 17 technology. Plaintiffs cannot create a factual dispute about whether there was any reasonable basis for 18 the December 14 forecast by arguing, with the benefit of hindsight, that Oracle should have used a 19 different formula for analyzing its sales data. 20 21 iii. Defendants’ failure to take into account deficiencies with Suite 11i 22 Plaintiffs argue that the December 14 public guidance was fundamentally unreliable because 23 Oracle failed to take into account problems with Suite 11i that indicated that this product would be 24 difficult to sell. According to plaintiffs, Oracle therefore should not have used the 3Q00 conversion rate 25 to project applications sales in 3Q01. 26 Defendants do not dispute that before and during 3Q01, there were problems with Suite 11i. 27 See Def. Opp. to Pls. Mot. for Summ. J. at 22 (citing analyst reports of bugs, the need for patches, 28 functional gaps, advice not to buy the product, and the lack of positive customer references). Plaintiffs’ 32 1 contention that these problems made Oracle’s forecast fundamentally unreliable nonetheless fails for 2 at least two reasons. First, according to plaintiffs, Oracle had difficulty selling Suite 11i throughout 3 1Q01 and 2Q01, but, as noted above, the internal forecasting system for those quarters proved accurate. 4 Second, the Court again returns to the point that the historical conversion ratio was not the only factor 5 that Oracle used to arrive at its internal forecast. If the problems with Suite 11i made it less likely that 6 sales representatives would close deals in 3Q01, the bottom up forecasting process provided a 7 mechanism for Oracle to take that information into account. 8 United States District Court For the Northern District of California 9 iv. Ellison’s directive to increase risk 10 Plaintiffs contend that the December 14, 2000 public guidance had no reasonable basis because 11 of a change Ellison made to Oracle’s forecasting system. According to plaintiffs, on October 4-5, 2000, 12 Minton assigned Patricia McManus and James English, both finance personnel, to provide training in 13 a new forecasting technique that radically changed Oracle’s forecasting method. PSJM 154. On 14 October 6, 2000, McManus sent the following e-mail to English: 15 16 17 18 19 20 21 22 23 24 25 Jim, we can talk through the recommendation, let me know your thoughts . . .27 Recently Larry28 has changed the way that he is interpreting our forecast. He would like us to reflect our numbers as follows . . . Worst – this is our bottom threshold – minimum 80% probability for opportunity – our old thinking of “commit.” Forecast – some risk included, potentially 50% of the time you make it and 50% of the time you don’t – minimum 60% probability for an opportunity Best – the top threshold for the quarter – minimum 40% probability for an opportunity This is a change from our current thinking in that our forecast has not usually had a significant amount of judgment. It was the amount that you believed you could deliver at a minimum. That emphasis has shifted to “worst” and now our forecast is a number that includes more risk than in the past. Best case should not be our entire quarter pipeline. As deals cross the 40% win probability threshold, they enter our best case. The win probabilities are guidelines and should be adjusted as you move through the quarter.[] Obviously, a 60% win probability the last week of the quarter is probably not a “forecast” item. Common sense has to prevail. The win probabilities are in OSO and should be reviewed for accuracy with the reps. The win probabilities have to be updated on a timely basis. We will incorporate this in our OSO IIi training that is tentatively scheduled the week of 10/23. Please let me know if you have any questions! 26 27 27 All ellipses in original. 28 28 Ellison stated at deposition this refers to him. PSJM D at 423:24-425:1. 33 United States District Court For the Northern District of California 1 PSJM 17. According to plaintiffs, the McManus e-mail is evidence of a company-wide directive from 2 Ellison for sales people to add “risk” to their forecast – that is, report less conservative sales projections. 3 While divisions had formerly just given their “worst” (or “commit”) number, they would now be 4 required to give their “worst,” their “forecast” (which would be based on their judgment) and their 5 “best” estimate.29 Plaintiffs contend that the directive rendered the 3Q01 public guidance unreliable 6 because Minton continued to add upside to her reports to account for sandbagging, even though sales 7 representatives’ tendency to underestimate their sales had now already been accounted for through the 8 new method articulated in the McManus e-mail. 9 Plaintiffs’ argument fails because they cite no evidence that the new system in fact introduced 10 more risk into the forecasts. Plaintiffs do not cite evidence of how Minton used these numbers (e.g. that 11 she made the forecast less reliable by including only the “best” estimates).30 In addition, in eight of the 12 nine weeks of 3Q01 that Minton projected a conversion rate, her prediction was equal to or lower than 13 the conversion rate from the same point in 3Q00. DSJM 143. If plaintiffs are correct that Ellison’s 14 directive caused sales representatives to report inflated projections, Minton’s 3Q01 weekly reports 15 should have shown higher conversion rates than those from 3Q00. Accordingly, plaintiffs’ evidence 16 that by October of 2000, Ellison had directed sales representatives to begin reporting their “worst,” 17 “forecast,” and “best” figures does not create a triable issue on whether Oracle’s forecast had no 18 reasonable basis.31 19 20 29 21 22 23 Defendants contend that no such directive was issued or followed, pointing out that the e-mail itself is the only document which mentions such a directive. Plaintiffs present no evidence that the trainings referred to took place, or what their content was. The Court will assume for the purposes of this discussion that Ellison did issue a directive for sales representatives to include three figures in their forecast and that the directive was followed. 30 24 25 26 27 Plaintiffs contend that Minton “took no account of the directive,” implying that she proceeded to add upside to the field forecast despite the new reporting, redundantly correcting for sandbagging. Plaintiffs’ characterization of the evidence is not accurate. In support of their claim that she specifically ignored the directive, plaintiffs cite a portion of her deposition testimony in which she stated that the “criteria for the forecast column” in her internal forecasting reports did not change during fiscal years 2000 and 2001. See PSJM B 19:17-19. Plaintiffs do not cite deposition testimony on the issue of whether she was aware of or responded to a directive from Ellison. 31 28 The Court recognizes that plaintiffs’ expert Alan Goedde opines that Ellison’s directive had an effect on the field forecasts. The reliability of Goedde’s opinion will be addressed presently. 34 1 Fraudulent 2Q01 accounting 2 Plaintiffs argue that Oracle’s overstatement of its reported revenue in 2Q01 rendered the 3Q01 3 forecast unreliable. According to plaintiffs, “the falsely reported $0.11 EPS for 2Q01 was undisputedly 4 the false basis for 3Q01 projections of $0.12 EPS[.]” Pls. Opp. to Defs. Mot. for Summ. J., at 47. 5 Plaintiffs point out that in a conference call with analysts discussing the December 14 forecast, Henley 6 linked Oracle’s 3Q01 projection of $0.12 cents per share to the company’s 2Q01 results: 7 8 9 10 United States District Court For the Northern District of California v. 11 In the area of per share, we think it would be 12 cents. Now, again, that’s based on some models, and I think most of the people’s estimates are around that. And also, just historically, the third quarter is slightly better than the second quarter. That’s the way, sequentially, it’s worked here. And if you look back at the last three years, sequentially, the third quarter, split adjusted, has been one cent a share better than the second quarter. So we did 11 cents in the second quarter. So I would assume, 12 cents would be a reasonable number at this point. I don’t think history is going to be a lot different, here. PSJM 1 at 3222. 12 Plaintiffs are correct that Henley told analysts that Oracle’s 3Q01 earnings per share projection 13 of $0.11 was reasonable because, based on historical patterns, the company could be expected to exceed 14 its 2Q01 earnings per share of $0.11 by a penny. This statement does not demonstrate, however, that 15 Oracle had no other basis for its forecast. In fact, the full quote shows that Henley first stated that the 16 12 cents projections was “based on some models.” To establish the existence of a triable issue as to 17 whether the 3Q01 forecast is actionable, plaintiffs would have to cite evidence showing that those 18 models were without a reasonable basis, which they have failed to do. 19 20 vi. Expert opinion of Dr. Alan G. Goedde 21 Defendants move to exclude the expert testimony and reports of Dr. Alan G. Goedde, who 22 opines that Oracle’s 3Q01 forecast lacked a reasonable basis.32 Specifically, they seek to exclude his 23 opinions that (1) “Oracle failed to make adjustments to its forecast process to account for the negative 24 32 25 26 27 28 Defendants also move to exclude Goedde’s opinion that “Oracle was facing a major macroeconomic change heading into Q3 2001 resulting in a decline in the market for its products.” Goedde’s opinion about the economy in 3Q01 does not create a factual dispute on the issue of whether there was a reasonable basis for Oracle’s 3Q01 public guidance. Assuming that Goedde is correct that the economy was changing in 3Q01, he does not explain why these changes would not be accounted for in Minton’s “bottom up” forecasting process. In addition, he does not explain how Oracle’s forecasting process had proven accurate in previous quarters even though the NASDAQ had already begun to decline. 35 United States District Court For the Northern District of California 1 impact of the changes in the economy and the market for Oracle’s products in violation of accepted 2 forecast principles” and (2) in 2Q01, “Larry Ellison directed significant changes to Oracle’s forecast 3 process that resulted in more ‘risk’ being included in future Field Forecasts and rendered Jennifer 4 Minton’s upside adjustments unreliable.” See Goedde Report 1, at 4.33 5 Federal Rule of Evidence 702 provides that expert testimony is admissible if “scientific, 6 technical, or other specialized knowledge will assist the trier of fact to understand the evidence or to 7 determine a fact in issue.” Fed. R. Evid. 702. Expert testimony under Rule 702 must be both relevant 8 and reliable. Daubert v. Merrell Dow Pharms., Inc., 509 U.S. 579, 589 (1993). When considering 9 evidence proffered under Rule 702, the trial court must act as a “gatekeeper” by making a preliminary 10 determination that the expert’s proposed testimony is reliable. Elsayed Mukhtar v. Cal. State Univ., 11 299 F.3d 1053, 1063 (9th Cir. 2002), amended by 319 F.3d 1073 (9th Cir.2003). As a guide for 12 assessing the scientific validity of expert testimony, the Supreme Court provided a nonexhaustive list 13 of factors that courts may consider: (1) whether the theory or technique is generally accepted within a 14 relevant scientific community, (2) whether the theory or technique has been subjected to peer review 15 and publication, (3) the known or potential rate of error, and (4) whether the theory or technique can be 16 tested. Daubert, 509 U.S. at 593-94; see also Kumho Tire Co., Ltd. v. Carmichael, 526 U.S. 137 (1999). 17 18 a. Timeliness 19 Goedde’s expert opinions became more refined as litigation progressed. On August 26, 2007, 20 after filing his expert report (on May 27, 2007) and his rebuttal report (on June 22, 2007), he supplied 21 a supplemental declaration in which he stated that he had used erroneous calculations when he generated 22 exhibit 4 to his rebuttal report.34 See Goedde Report 3. [Docket No. 1030] In conjunction with the 23 August 26 declaration, he submitted a new graph (exhibit 1) and opined that it supported his conclusion 24 that “the prior year conversion rate was the basis for Ms. Minton’s Potential Forecast and therefore 25 33 26 27 All references to Goedde Report 1 are to exhibit 1 to the declaration of Douglas R. Britton in support of the expert report of Alan G. Goedde. The Goedde Rebuttal Report (Goedde Report 2) is attached as exhibit 3 to the same declaration. 34 28 Per the parties’ stipulation, expert and rebuttal reports were exchanged on May 25, 2007 and June 22, 2007, respectively. 36 United States District Court For the Northern District of California 1 Oracle’s guidance in 3Q FY01.” Goedde Report 3 at ¶ 4 & ex. 1. 2 Goedde’s third report went beyond merely correcting a computational error in his rebuttal report, 3 however. He also “performed additional analysis to test the correlation” between Minton’s forecast and 4 the historical conversion ratio. Goedde Report 3 at exs. 2, 3. To that end, he included two new graphs 5 that purported to prove Minton arrived at her guidance by applying the prior year’s conversion ratio to 6 the pipeline. Goedde Report 3 at exs. 2, 3. 7 Goedde also used his third report to supplement his opinion about how Ellison’s purported 8 directive changed Oracle’s forecasting method. He stated, “Based upon my review of defendants’ 9 rebuttal expert reports, I learned for the first time that defendants are denying that Oracle’s sales force 10 actually followed Mr. Ellison’s directive.” Goedde Report 3 ¶ 7. He therefore performed additional 11 analysis that he claimed supported his conclusion that Oracle’s U.S. license division forecasts changed 12 after the directive because more managerial judgment was inserted into the field reports. 13 On November 6, 2007, Goedde filed a fourth declaration, which contained “more detailed 14 statistical analysis” to supplement his August 26 declaration and address arguments raised in the rebuttal 15 report of defendants’ expert. Goedde Report 4 ¶ 5. [Docket No. 1403-2] Specifically, he performed a 16 new computation in his analysis of Minton’s forecast that purported to correct for seasonal patterns. 17 Finally, in conjunction with the renewed summary judgment motions, Goedde filed a fifth 18 declaration, on November 16, 2008, in which he consolidated his opinions from the prior four reports 19 into a single declaration. Goedde Report 5. [Docket No. 1542] 20 The Court agrees with defendants that Goedde’s augmentation of his opinion after he submitted 21 his rebuttal report was not proper. Federal Rule of Civil Procedure 26(e)(1) provides that “[f]or an 22 expert whose report must be disclosed under Rule 26(a)(2)(B), the party’s duty to supplement extends 23 both to information included in the report and to information given during the expert’s deposition.” 24 Thus, after realizing that Goedde’s rebuttal report analysis contained a computational error, plaintiffs 25 had a duty to correct the mistake through a supplemental disclosure. Plaintiffs exceeded the permissible 26 scope of a supplemental disclosure, however, by including additional analysis to shore up Goedde’s 27 original opinion. Rule 26(a)(2)(B) requires that expert witnesses disclose a report containing “a 28 complete statement of all opinions the witness will express and the basis and reasons for them.” See 37 United States District Court For the Northern District of California 1 Fed. R. Civ. Pro. 26(a)(2)(B)(i) (emphasis added). The expert report therefore defines the metes and 2 bounds of an expert’s trial testimony. Here, Goedde’s additional analysis vastly complicated his opinion 3 of how Minton arrived at her upside adjustment, with each supplemental declaration containing more 4 involved discussion of his statistical analysis than the last. Rule 26’s expert disclosure requirements 5 would be obviated if experts could continually refine their opinions in this manner. 6 Goedde’s additional analysis regarding Ellison’s directive is similarly problematic. Originally, 7 Goedde did not use statistical analysis of the field forecasts to support his conclusion that the directive 8 inflated the forecasts; this dimension of his analysis was introduced after the rebuttal reports were 9 exchanged. It is simply not plausible that he did not realize until reading the rebuttal reports of 10 defendants’ experts that defendants contend Ellison did not issue a company-wide directive. Goedde’s 11 initial report anticipates that the existence of this directive will be disputed. See Goedde Report 1 at 22 12 (“There is no evidence that I have reviewed to indicate that the field did not follow Ellison’s directive. 13 To the contrary, there is reason to believe that it did.”35) There is no reason that Goedde could not have 14 performed these computations in his initial analysis. 15 Accordingly, the Court finds that the new opinions offered in Goedde’s third, fourth, and fifth 16 reports are untimely. All of these opinions are STRICKEN, to the extent they include analysis that was 17 extends beyond the analysis Goedde performed in his initial and rebuttal reports. Exhibit 1 to Goedde’s 18 third report, which corrected the erroneous exhibit in his rebuttal report, is not stricken. 19 20 b. 21 22 23 24 25 26 27 28 Goedde’s opinion that Oracle’s 3Q01 forecast was unreliable because Minton did not take economic factors into account Turning to the elements of Goedde’s reports that were timely, the overarching problem is that his opinion is based on a selective reading of the record. A premise of Goedde’s analysis is that “[a]ccording to Ms. Minton’s testimony, the week 2 conversion rate for [3Q00] multiplied by the Pipeline at week 2 of [3Q01] was the underlying basis for the ‘upside’ portion of the [3Q01] revenue forecast that formed the basis of Oracle’s Public Guidance.” Goedde Report 1 at 20. In other words, 35 In support of this contention, Goedde cited the McManus e-mail and Ellison’s deposition testimony that he wanted all three forecast figures (worst, forecast, best) in the system in order to have “more data to work with.” Goedde Report 1 at 22. 38 United States District Court For the Northern District of California 1 Goedde relies on Minton’s testimony for his conclusion that Minton arrived at her upside by applying 2 the conversion ratio from the prior year. As the foregoing discussion demonstrated, plaintiffs cite no 3 testimony by Minton that the conversion ratio was the basis for her upside or that she mechanically 4 applied the conversion ratio from the prior year. Goedde proceeds to opine that “the application of this 5 historical conversion ratio” was flawed because 3Q00 was a boom economy while the economy was in 6 decline in 3Q01, see Goedde Report at 20, but provides no further basis for his conclusion about how 7 Minton used the historical conversion ratio or what role this factor played in her upside. 8 In his rebuttal report (and exhibit 1 of his third report), Goedde for the first time attempted to 9 substantiate his conclusion that the historical conversion ratio was the foundation of Minton’s upside. 10 He provided a graph that purports to compare her actual forecasts throughout 3Q01 with what her 11 forecasts would have been if she had used only the historical conversion ratio. See Goedde Report 2 12 at 76 & Goedde Report 3, ex. 1. Tellingly, the graph shows that throughout 3Q01, Minton’s actual 13 forecast was different from the forecast that Goedde claims she would have arrived at had she 14 mechanically applied the 3Q00 conversion ratio. See id. Goedde provides no analysis in his initial or 15 rebuttal reports in support of his conclusion that this chart shows that Minton’s upside “was based 16 almost exclusively on historical conversion analysis.” Goedde Report 2 at 76. To the contrary, the 17 graph shows that some factors other than the historical conversion ratio must have played a role in 18 Minton’s upside. 19 The Court finds that Goedde’s opinion on how Minton used prior year conversion ratios and the 20 importance of this factor in her upside is not reliable because Goedde ignored deposition testimony by 21 Minton that contradicted his conclusion and because his statistical analysis does not support his 22 conclusion. Accordingly, the Court GRANTS defendants’ motion to exclude Goedde’s expert reports 23 and testimony on this issue. See Brooke Group, v. Brown & Williamson Tobacco Corp., 509 U.S. 209, 24 242 (1993) (“When an expert opinion is not supported by sufficient facts to validate it in the eyes of the 25 law, or when indisputable record facts contradict or otherwise render the opinion unreasonable, it cannot 26 support a jury’s verdict.”). 27 28 39 1 c. Goedde’s opinion that the 3Q01 forecast was unreliable as a result of Ellison’s directive 2 Defendants move to exclude Goedde’s opinion that Ellison’s directive for sales representatives 3 to designate their worst, forecast, and best figures (assuming for the sake of argument that the directive 4 was issued) rendered the 3Q01 forecast unreasonable. The Court agrees with defendants that Goedde’s 5 opinion on this issue is unreliable. In his initial report, he opines that Minton adjusted for Ellison’s 6 change by decreasing and then eliminating her upside adjustment between January 15, 2001 and 7 February 5, 2001. Goedde Report 1 at 23. According to Goedde, Minton’s forecast was unreliable 8 9 United States District Court For the Northern District of California 10 11 12 13 14 15 16 17 before she made these changes. Goedde does not provide any basis for his conclusion that Minton made changes to the upside because of Ellison’s directive, rather than because of her own judgment that the additional upside was not warranted. Second, Goedde assumes that Ellison’s directive resulted in “the virtual removal of ‘sandbagging,’” by sales representatives, thereby rendering Minton’s upside redundant. Id. at 24; see also Goedde Report 2 at 83 (“Mr. Ellison directed the sales force to submit forecasts which exceeded their comfort levels, effectively changing Oracle’s forecasting process from a ‘bottom-up’ process to a ‘top-down’ process.”). Goedde points to no evidence supporting this conclusion that the directive had an effect on the field forecast. He therefore assumes what he sets out to prove: that a directive from Ellison changed the behavior of the field. For these reasons, the Court finds that Goedde’s opinions on the effect of Ellison’s directive is 18 19 20 21 not based on discernible “methods and methodology” and therefore is not reliable. Daubert v. Merrell Dow Pharms., Inc., 43 F.3d 1311, 1316 (9th Cir. 1995). Defendants’ motion to strike Goedde’s expert reports and exclude his testimony on this issue is GRANTED. Without Goedde’s expert report, there is no factual dispute as to whether any of the factors 22 23 24 identified by plaintiffs rendered the 3Q01 forecast without a reasonable basis. Defendants’ motion for summary judgment on this issue is GRANTED. 25 26 27 28 /// 40 Intra-quarter repetitions of the 3Q01 forecast36 1 E. 2 Plaintiffs claim that Oracle repeated the December 14 forecast during 3Q01. See POpp. 217 at 3 3279-80.37 They argue that these repetitions of the forecast were materially false because defendants 4 were “aware of undisclosed facts tending seriously to undermine the [forecast’s] accuracy.” Provenz, 5 102 F.3d at 1487. Plaintiffs identify three categories of facts that were allegedly known to defendants. 6 The Court will consider each in turn. 7 United States District Court For the Northern District of California 8 i. Actual results from flash reports 9 Plaintiffs contend that Oracle officials learned information from “flash reports” – snapshots of 10 key financial indicators – in 3Q01 that alerted them that the company was likely to miss its public 11 guidance. The first flash report, which reported actual data from December 2000, was issued internally 12 on January 17, 2001. POpp. 277. It reported license revenue growth of 35%; the public guidance had 13 forecast license growth of 25%. Id. The January 17 report also showed that in December, Oracle had 14 achieved 19% of the quarterly forecast for license revenue. Id. This figure was in line with prior 15 quarterly results: in 3Q00, Oracle had achieved 16% of its forecast license revenue by December; in 16 3Q99, it had achieved 19% of the forecast license revenue by December. Id. 17 The parties dispute whether the relevant figures in the January 17 flash report should have been 18 adjusted to eliminate the revenue from a deal with internet trading site Covisint. In the first few days 19 of December 2000, Oracle executed a $60 million applications deal with Covisint – the largest 20 transaction in Oracle’s history. Plaintiffs argue that the January 17 figures should be adjusted to 21 eliminate revenue from the Covisint deal. According to plaintiffs, quarterly trends were only apparent 22 after Covisint was eliminated from the December data. They cite, for example, deposition testimony 23 by Minton stating that considering the underlying license revenue without the Covisint deal provided 24 another “data point” for analyzing the December results. POpp. GGG at 312:23-313:4. 25 36 26 27 28 Again, the Court assumes without deciding that these statements were not accompanied by meaningful cautionary language and therefore were not subject to PSLRA’s safe harbor. 37 This transcript of a speech by George Roberts is not dated. Plaintiffs have submitted sufficient other evidence to support a finding that Roberts gave this speech in mid-February 2001. See PSJM V at 341:20-343:6; see also PSJM 218. 41 United States District Court For the Northern District of California 1 Although removing an aberrant transaction may allow for a more accurate comparison with 2 historical data, the inquiry here is whether the January 17 flash report confirmed or undermined the 3 3Q01 public guidance concerning anticipated revenues. There is no question that the Covisint deal, 4 even though it was unusual, would be “booked” as 3Q01 revenue and would therefore help Oracle meet 5 its forecast. 6 Plaintiffs also note that the January 17 flash report indicated that two of the three U.S. divisions 7 showed negative year over year revenue growth: NAS and OSI reported revenue growth of -24% and 8 -81%, respectively. A second flash report, reporting actual results from January 2001, was issued 9 February 8, 2001. POpp. 281. It indicated that NAS and OSI were still reporting negative revenue 10 growth: -33% and -17% respectively. Both flash reports also reported, however, that OPI, the third U.S. 11 division, was reporting significant growth: 243% in December and 235% in January. Plaintiffs do not 12 explain why negative growth in two U.S. divisions should have alerted Oracle that it would miss its 13 3Q01 guidance, particularly when a third U.S. division was reporting significant growth. 14 In sum, if Covisint is not excluded from the flash reports, plaintiffs point to no evidence that the 15 actual results for December and January alerted Oracle that it would miss its 3Q01 forecast. On the 16 contrary, these reports indicate that license revenues were in line with prior years as a percentage of 17 quarterly results and that in December, license revenue growth was exceeding the forecast by 10%. 18 More fundamentally, plaintiffs’ focus on the flash reports does not take into account the hockey- 19 stick effect. Given that Oracle was known to earn the vast majority of its quarterly revenue in the final 20 weeks of a given quarter, it is not apparent that the results from the first or second months of the quarter 21 are good indicators of how the quarter would turn out. 22 For these reasons, the Court finds that there is no factual dispute as to whether the reports of 23 Oracle’s performance in December 2000 and January 2001 contained information that seriously 24 undermined the 3Q01 forecast. 25 26 ii. Pipeline growth 27 Plaintiffs argue that throughout 3Q01, Oracle received weekly data showing that pipeline growth 28 was declining, and that this information warned officials that the company would not meet its forecast. 42 1 Oracle had forecast license revenue growth of 30% (constant dollars). Oracle’s actual pipeline growth 2 (i.e. year over year percentage increase) in the weeks comprising the third quarter is summarized as 3 follows:38 4 5 12/11/00 12/25/00 1/15/01 1/22/01 1/29/01 2/5/01 2/12/01 2/19/01 2/26/01 2/27/01 2/28/01 52% 34% 34% 31% 31% 32% 29% 28% 34% 34% 34% 6 7 8 9 United States District Court For the Northern District of California 10 Plaintiffs pay particular attention to the figures from December 25 and February 5, pointing out that pipeline growth had dropped from a high of 52% at the beginning of the quarter to 34% and 32%, respectively. See Pl. Mot. for Summ. J. at 23. In other words, plaintiffs argue that Oracle officials should have known that the company would miss its guidance when they learned that the pipeline was growing at 4% and 2% above their projection for license growth. This does not make sense. 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 Plaintiffs focus on the “comfort gap,” which is the difference between the forecasted revenue growth and the pipeline growth. The comfort gap was narrow in 3Q01, particularly between December 25 and January 29, when it did not exceed 4%. Plaintiffs point out that this is a much smaller figure than the average comfort gap for the five quarters prior to 3Q01, which was 13.4%. See PSJM 16, ex. 13. The Court agrees with defendants that plaintiffs do not cite evidence from which a rational factfinder could conclude that a narrow comfort gap alerted Oracle officials that they would miss their guidance. Plaintiffs rely on testimony from Ellison’s deposition as evidence that the comfort gap was a figure that Oracle officials relied on to evaluate the reliability of the forecast. Ellison, however, stated only that the relevant information is whether the pipeline exceeds the forecast, not that he used the difference between these figures to assess the accuracy of the forecast. See, e.g., PMSJ C at 420:12-14 (“I would say as long – as long as the pipeline growth is greater than – than the revenue growth, you should be able to meet your numbers.”). Accordingly, the Court finds that there is no triable issue as to whether Oracle’s pipeline growth, which exceeded the forecast for all but two weeks of the quarter, provided defendants with information 26 27 28 38 See PMSJ 16 at ex. 20. Plaintiffs provided the actual dollar amounts of the pipeline. The Court has divided plaintiffs’ figures for 3Q01 by their figures for 3Q00 to determine the percentage growth over the prior year. See also PSJM 69 at 440076, PSJM 70 at 440093. 43 1 that seriously undermined the 3Q01 forecast. 2 United States District Court For the Northern District of California 3 iii. Pipeline information and forecasts from the U.S. license divisions 4 Plaintiffs argue that intra-quarter reports from the U.S. license divisions informed Oracle that 5 it would miss its guidance. Plaintiffs cite considerable evidence that NAS and OSI reported negative 6 trends during 3Q01 and that their 4Q01 forecasts showed year over year declines.39 Plaintiffs do not 7 explain, however, why the declines in NAS and OSI would necessarily undermine the 3Q01 forecast, 8 especially in light of the major Covisint deal reported in OPI. In addition, each of the U.S. divisions 9 continued to stick by their forecasts despite reporting declines. See, e.g., POpp. 329 (1/17/01 email from 10 OPI finance director to Minton that OPI’s forecast “holding at $150M but I’ll be a bit more aggressive 11 and say a likely of $150-$170.”); POpp. 240 (1/11/01 email from NAS’s finance director to Minton, 12 discussing pipe shrinkage, lack of big deals, and drop in technology spending, but stating “[o]ur Q3 13 forecast of $346M has not changed” and revising only NAS’s upside); POpp. 305 (1/18/01 email from 14 OSI finance director saying Nussbaum “still confident in the 225, as long as none of the very large 15 opportunities drop out” and “[M]y sense still is that we’ll come out around 210 as it stands today”); 16 POpp. 327 (2/14/01 email from OSI finance director reaffirming that she still thought OSI would come 17 in between $210 and $220 million). 18 For these reasons, the Court concludes that a reasonable juror could not infer from the data 19 showing losses in OSI and NAS that defendants had information seriously undermining the 3Q01 20 projection. 21 Accordingly, the Court finds that plaintiffs have failed to identify the existence of a triable 22 dispute as to whether the Oracle officials had knowledge of actual facts seriously undermining their 23 3Q01 forecast when they repeated the forecast in 3Q01. Defendants’ motion for summary judgment on 24 25 26 27 28 39 Defendants object that exhibit 236, a slideshow apparently presented at a meeting on April 1, 2001, is hearsay to the extent it is offered for the truth of what assistant vice presidents in NAS said in December of 2000. The Court agrees and SUSTAINS defendants’ objection. Defendants also object that POpp. 315, a summary of “lost big deals” in 3Q01, is an improper summary of voluminous exhibits under Federal Rule of Evidence 1006. The Court agrees. The Court agrees that the chart is an improper summary because it contains attorney argument. Defendants’ objection is SUSTAINED. 44 1 this issue is GRANTED. 2 3 F. 4 During 3Q01, defendants made the following statements, which plaintiffs characterize as 5 6 misrepresentations that the slowing economy was not affecting Oracle: • In the December 14, 2000 conference call with analysts, Henley said, “And then lastly, . . . the economy . . . [a]t this point, we see no impact or slowing in our business. We have seen no slowing of our business. We believe that (the US economy) is slowing down, from what we can tell.” “So, you know, if we were to have a – I guess, a hard landing, a recession, depression, I mean, certainly, that could have some impact. But as long as we’re simply slowing and going into more of a soft landing, we continue to believe that our business should do quite well.” POpp. 26 at 234423-24. • In a December 15, 2000 interview with Radio Wall Street, Henley said: “[T]he economy right now even though it’s slowing doesn’t seem to be affecting us. We see no difference in demand for our upcoming third fiscal quarter. . . . If the economy got really really bad then obviously that would probably have some effect on all of us. But so far we look pretty hard at indicators. We’re seeing no softening in our business.” DX 163 at 141660-61. • In an e-mail dated February 26, 2001, Oracle spokeswoman Stephanie Aas sent an internal e-mail after the AppsWorld conference in New Orleans on February 21, 2001 stating, “Analysts closely monitored comments on the economy which were consistent with recent statements: we do not expect the slowing economy, barring a serious slide to a recession, to significantly impact near term guidance. These comments were met with a degree of relief, as some analysts were anticipating the event to be an opportunity to take down guidance.” See POpp. 212 (PSJM 9).40 7 8 9 United States District Court For the Northern District of California 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 Statements about the effect of the economy on Oracle Defendants contend that their statements about the effects of the economy on Oracle’s business were forward looking and were therefore subject to the PSLRA safe harbor. According to defendants, the statements about the economy were so “intertwined” with statements about Oracle’s ability to make its forecast that they constituted projections. The Court disagrees. Forward-looking statements include statements “containing a projection of revenues, income . . ., earnings . . . per share,” and statements “of future economic performance.” 15 U.S.C. § 78u-5(i)(1)(A) & (C). While the statements were made in the context of discussions of the 3Q01 forecast, the speakers were conveying information about whether 40 Defendants object that Aas’s repetition of analysts’ reports of what Henley said constitutes hearsay within hearsay. The Court agrees that the portion of the e-mail that summarizes analyst reports is inadmissible hearsay if offered to prove the truth of what the reports say Henley said. The portion of the e-mail cited above, however, is admissible as proof of Oracle’s public statements about the company’s financial outlook in February of 2001. 45 1 2 As these are not forward-looking statements, plaintiffs must “demonstrate that a particular 3 statement, when read in light of all the information then available to the market, or a failure to disclose 4 particular information, conveyed a false or misleading impression.” In re Convergent Tech. Sec. Litig., 5 948 F.2d 507, 512 (9th Cir. 1991). They must also show that defendants engaged in “knowing” or 6 “intentional ” conduct or acted with “deliberate recklessness.” See South Ferry LP, No. 2 v. Killinger, 7 542 F.3d 776, 782 (9th Cir. 2008). In the securities context, “an actor is reckless if he had reasonable 8 grounds to believe material facts existed that were misstated or omitted, but nonetheless failed to obtain 9 and disclose such facts although he could have done so without extraordinary effort.” Howard v. Everex 10 United States District Court For the Northern District of California Oracle was being affected by the slowing economy in the present.41 Sys., Inc., 228 F.3d 1057, 1063 (9th Cir. 2000) (citation and alterations omitted). 11 The Court finds that plaintiffs have failed to cite evidence from which a reasonable juror could 12 conclude that Henley was deliberately reckless in making these statements. Although plaintiffs cite 13 ample evidence showing that certain sectors in Oracle were losing revenue as a result of dot.coms going 14 out of business, Henley had a basis for believing that Oracle’s overall business would be immune from 15 the downturn. As discussed above, Oracle had continued to see substantial year over year increases in 16 license revenue even as NASDAQ and dot.com sales steadily declined. Plaintiffs have failed to show 17 that any of the internal data available to defendants through 3Q01 seriously undermined their forecast 18 for that quarter. 19 20 2. Defendants’ Motion for Summary Judgment on Plaintiffs’ Control Person Claim 21 Defendants move for summary judgment on plaintiffs’ § 20(a) claim against Ellison, Henley, 22 and Sanderson. Plaintiffs do not dispute that § 10(b) liability is a prerequisite for controlling person 23 liability under § 20(a). See Paracor Finance, Inc. v. General Elec. Capital Corp., 96 F.3d 1151, 1161 24 (9th Cir. 1996) (“To establish ‘controlling person’ liability, the plaintiff must show that a primary 25 violation was committed and that the defendant ‘directly or indirectly’ controlled the violator.”) (citation 26 omitted). Accordingly, defendants’ motion for summary judgment on this claim is GRANTED. 27 41 28 The Court notes that in the prior round of summary judgment briefs, defendants stated that these were not forward-looking statements. See Def. Nov. 9, 2007 Mot. at 32. [Docket No. 1406] 46 United States District Court For the Northern District of California 1 3. Defendants’ Motion for Summary Judgment as to Ellison and Henley 2 Defendants move for summary judgment on plaintiffs’ third cause of action, violation of section 3 20A of the Exchange Act, which provides a cause of action against “[a]ny person who violates any 4 provision of this chapter or the rules or regulations thereunder by purchasing or selling a security while 5 in possession of material, nonpublic information.” 15 U.S.C. § 78t-1(a). The elements of a section 20A 6 violation are “(1) trading by a corporate insider; (2) a plaintiff who traded contemporaneously with the 7 insider; and (3) that the insider traded while in possession of material nonpublic information, and thus 8 is liable for an independent violation of the Securities Exchange Act of 1934.” Simon v. American 9 Power Conversion Corp., 945 F. Supp. 416, 435 (D.R.I. 1996) (citing In re VeriFone Sec. Litig., 11 F.3d 10 865, 872 (9th Cir. 1993)). “Claims under Section 20A are derivative and therefore require an 11 independent violation of the Exchange Act.” Johnson v. Aljian, 490 F.3d 778, 781 (9th Cir. 2007). 12 Defendants Ellison and Henley argue that if the Court grants summary judgment on plaintiffs’ § 10(b) 13 claim, they are also entitled to summary judgment on plaintiffs’ § 20A claim because plaintiffs cannot 14 establish an independent violation of the Exchange Act. 15 Plaintiffs contend that they can demonstrate the “independent violation” required for their § 20A 16 claim by showing that Henley and Ellison engaged in insider trading in violation of Rule 10b-5. 17 Defendants argue that insider trading cannot serve as a predicate offense for section 20A purposes. In 18 Johnson, however, the Ninth Circuit held that the plaintiffs had met the “independent violation” 19 requirement of § 20A by alleging that the defendants engaged in illegal insider trading in violation of 20 § 10(b) of the Exchange Act. See 490 F.3d at 779; see also In re JDS Uniphase Corp. Sec. Litig., No. 21 C 02-1486, U.S. Dist. LEXIS 76936 *13 (N.D. Cal. Sept. 27, 2007) (holding § 10(b) insider trading 22 claims were sufficient to serve as predicate violation). Defendants therefore are not entitled to summary 23 judgment solely on the basis that plaintiffs’ § 10(b) claim for false statement fails. 24 Henley and Ellison also argue that they are entitled to summary judgment on plaintiffs’ § 20A 25 claim because there is no factual dispute that they did not possess “material, nonpublic information” at 26 the time of their trades. See United States v. O’Hagan, 521 U.S. 642, 651-52 (1997). Plaintiffs argue 27 that Henley and Ellison possessed the following insider information: (1) data from Oracle’s U.S. 28 divisions rendered the 3Q01 forecast unreliable; (2) Oracle’s pipeline had “collapsed;” (3) Oracle’s 47 United States District Court For the Northern District of California 1 3Q01 forecast had no reasonable basis because of economic changes, Ellison’s directive to add risk, the 2 overstatement of 2Q01financial returns due to the debit memo fraud, the higher mix of applications in 3 the pipeline, and Minton’s mechanical application of the historical conversion ratio to the 3Q01 4 pipeline; and (4) problems with Suite 11i were affecting applications sales. 5 The foregoing discussion has established that plaintiffs have failed to establish the existence of 6 a genuine factual dispute on the first three categories of information. Assuming for the sake of argument 7 that there are disputed facts about whether Henley and Ellison had material insider information about 8 Suite 11i at the time of their trades, plaintiffs would have to prove that the revelation of the concealed 9 information about Suite 11i was a substantial cause of plaintiffs’ loss. See Johnson, 490 F.3d at 782 10 (citing Ambassador Hotel Co., Ltd. v. Wei-Chuan Inv., 189 F.3d 1017, 1025 (9th Cir. 1999) (“To prove 11 violation of either Section 10(b) or Rule 10b-5, the private plaintiff must demonstrate that the alleged 12 fraud occurred ‘in connection with the purchase or sale of a security’. Once this foundational 13 requirement has been met, the plaintiff must prove five elements: (1) misrepresentation (or omission, 14 where there exists some duty to disclose); (2) materiality; (3) scienter (intent to defraud or deceive); (4) 15 reliance; and (5) causation. The plaintiff must prove both actual cause (‘transaction causation’) and 16 proximate cause (‘loss causation’).”). As discussed above, defendants have demonstrated that there is 17 no triable issue as to loss causation for the purportedly concealed information about Suite 11i. 18 19 Accordingly, the Court GRANTS defendants’ motion for summary judgment on plaintiffs’ § 20A claim against Henley and Ellison. 20 21 4. Adverse Inference 22 The Court rejects plaintiffs’ argument that the Court should apply the adverse inference broadly. 23 If plaintiffs had been able to establish the existence of a factual dispute on the issue of whether the 3Q01 24 forecast had a reasonable basis, for example, the adverse inference would have helped plaintiffs prove 25 that Oracle officials were aware of those problems. The inference cannot help plaintiffs with the 26 absence of evidentiary support for plaintiffs’ allegations. 27 28 48 1 CONCLUSION 2 For the foregoing reasons and for good cause shown, the Court hereby GRANTS defendants’ 3 motion for summary judgment on plaintiffs’ § 10(b), § 20(a), and § 20A claims and DENIES plaintiffs’ 4 motion for partial summary judgment. The Court also GRANTS defendants’ motion to exclude the 5 expert report and testimony of Alan Goedde. All other evidentiary objections are DENIED as moot 6 unless discussed in this order. 7 8 IT IS SO ORDERED. 9 United States District Court For the Northern District of California 10 Dated: June 16, 2009 11 SUSAN ILLSTON United States District Judge 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 49

Some case metadata and case summaries were written with the help of AI, which can produce inaccuracies. You should read the full case before relying on it for legal research purposes.

This site is protected by reCAPTCHA and the Google Privacy Policy and Terms of Service apply.