IP Law Daily, TRADE SECRETS—9th Cir.: Teradata claims against SAP revived on appeal, (Dec 20, 2024)

By Justin Marcus Smith, J.D.

The Ninth Circuit declined to expand the Epic Games and Microsoft narrow exception to the per se rule.

In an unpublished opinion, disputes of material fact precluded summary judgment for defendant SAP on Teradata Corporation (Teradata) claims that SAP violated antitrust law and misappropriated Teradata trade secrets, held the United States Court of Appeals for the Ninth Circuit. The exclusion of an expert’s testimony about market definition and market power was an abuse of discretion in the district court’s disposition of summary judgment. Teradata also raised material issues of fact on the trade secret misappropriation claim that only a jury could decide. The Ninth Circuit accordingly reversed the district court’s summary judgment and remanded (Teradata Corp. v. SAP SE, No. 23-16065 (9th Cir. Dec. 19, 2024)).

Background. In 2018, Teradata Corporation (Teradata), an enterprise data and warehousing (EDW) analytics company, sued SAP SE, SAP America, Inc., and SAP Labs, LLC (collectively, SAP). Teradata alleged that SAP, an enterprise resource planning (ERP) software vendor: 1) illegally tied sales of its business-management software to sales of its back-end database engine in violation of Section 1 of the Sherman Act, 15 U.S.C. § 1; and, 2) misappropriated Teradata’s trade secrets in violation of the California Uniform Trade Secrets Act, Cal. Civ. Code § 3426. SAP counterclaimed that Teradata infringed various SAP patents.

In 2008, SAP and Teradata began a joint venture known as the “Bridge Project” to integrate SAP front-end applications with the Teradata Database’s back-end computing architecture. The companies executed two agreements to protect their intellectual property: 1) a software development cooperation agreement, which restricted disclosures of each party’s confidential information, and a mutual non-disclosure agreement (NDA), which specified how to maintain respective confidences.

When the Bridge Project encountered technical difficulties, Teradata’s senior engineer proposed incorporating the Teradata “batched merge” method of efficiently aggregating data and sent SAP a relevant design document labeled “Teradata Confidential.” The Bridge Project ultimately produced Teradata Foundation, apparently a sort of shim that enabled SAP’s front-end application and Teradata’s back-end computer architecture to talk to each other. Meanwhile, SAP had been developing its own EDW product, SAP HANA, but two months after SAP launched SAP HANA, it terminated the Bridge Project and stopped supporting Teradata Foundation.

In 2015, SAP released an updated version of its ERP application, S/4HANA, and combined it with HANA in a single sales offering. Customers could either buy a full license or a restricted “runtime” license. The cheaper runtime license would not allow export of HANA data to third-party products. About 88 percent of SAP customers buy HANA with a runtime license.

In the course of this litigation in the district court, Teradata presented an expert report. The Teradata expert opined that 1) the relevant antitrust product market for S/4HANA was “core ERP products for large enterprises” with HANA part of a market defined as EDW solutions with online analytical processing capabilities for large enterprises. Applying those definitions, the expert concluded SAP had ERP market power harming competition in the EDW market.

SAP moved for summary judgment on Teradata’s claims and sought to exclude portions of the expert’s testimony. The district court excluded portions of the expert’s testimony on market definition, market power, and harm to competition. The court deemed his methods unreliable and found that undisputed facts contradicted his opinion. Based on those findings, the court determined that Teradata failed to create a material dispute on its tying claim. It also concluded the trade secret claim failed because Teradata had not designated the batched merge method as confidential, and, in any event, the parties’ agreements gave SAP the right to use the method in its own products. The district court did not fully resolve the patent counterclaims and entered a partial final judgment for SAP under Fed. R. Civ. P. 54(b) (see Antitrust News, Nov. 12, 2021).

Teradata appealed to the United States Court of Appeals for the Federal Circuit. However, that court found it lacked jurisdiction over SAP’s patent-infringement counterclaims because they were not compulsory. They did not arise out of the same “transaction or occurrence” as the Teradata claims. The Federal Circuit transferred the appeal to the Ninth Circuit. The Ninth Circuit said it evaluated Teradata’s tying claims under both the per se rule and the rule of reason. The FTC, meanwhile, filed an amicus brief avoiding a position on the merits but objecting to the district court’s “overly rigid framework” in defining a relevant antitrust market (see Antitrust News, May 31, 2022).

Expert exclusion. The Ninth Circuit found an abuse of discretion in the district court’s decision to exclude the Teradata expert testimony. The Ninth Circuit reasoned that the expert used reasonable quantitative methods, including an “aggregate diversion ratio analysis,” to check his qualitative analysis of SAP’s business documents in defining the relevant markets. There was no reason to exclude the expert’s testimony about market definitions and SAP market power.

The district court’s decision appeared to be based on a substantive rule that “large enterprises” is “too imprecise to describe a properly defined market.” The Ninth Circuit said that was “legally erroneous” because plaintiffs do not need to specify markets by “precise ‘metes and bounds.’” There will usually be some “fuzziness” in attempting to define a market.

A second interpretation of the district court decision was that the expert did not explain how he arrived at his definitions. Here, SAP argued the expert did not explain why he defined “large enterprises” as those with “1,000 to 1,500 employees and over 125 users” when the documents he relied on did not correspond to those measurements. Even if so, the Ninth Circuit said the district court’s Daubert analysis was still flawed. “Large” was a sufficiently intuitive concept such that failure to explain it did not necessarily cast doubt on the expert’s method. Quantitative inconsistencies in the meaning of “large” merely illustrated the difficulty of providing a “metes and bounds” definition, and the expert provided additional generalities to create a gestalt, like “high annual revenues” and “large number of staff.”

Turning to the expert’s quantitative methods, the Ninth Circuit reasoned any flaws in his use of an aggregate diversion ratio analysis were not a basis for exclusion where the analysis was “merely confirmatory.” To the extent the district court based its ruling on a general rejection of aggregate diversion ration analysis, the Ninth Circuit said that was unreasonable. The Ninth Circuit said the tool is commonly used to estimate demand responsiveness and monopolist recapture of sales, if any, after a significant price increase.

The district court was more specific about faulting the expert’s use of customer relationship management (CRM) data capturing competition for a given sales opportunity on the basis that it could not measure cross-elasticity of demand. The Ninth Circuit said the few courts to have considered this have “endorsed the use of customer relationship management and other non-price data to calculate the aggregate diversion ratio.” The Ninth Circuit added this is logical and practical because data recording actual customer responses to price changes is sparse at best.

The expert’s tying market method concluded the relevant market consisted of only Oracle and SAP, while the tied-market aggregate diversion ratio analysis included more than just the minimum number of participants and brought SAP into the market definition. The Ninth Circuit said the difference in method was “grounded in economic logic and well-established market-definition principles.” Broadening the market was appropriate for the purpose of ascertaining the universe of harmed competitors. A trier of fact would not have to accept this, but the expert sufficiently explained himself to satisfy Fed. R. Civ. P. 702.

More fundamentally, the Ninth Circuit said the district court abused its discretion by apparently focusing on the expert’s aggregate diversion ratio method as its sole ground for excluding his tied-market testimony. The primary foundation for the tied-market testimony was the expert’s qualtitative analysis of the deposition testimony and documentary record.

Competitive harm. Turning to harm to competition in the tied market, the Ninth Circuit held the district court was unreasonable in finding the expert’s harm testimony unreliable. Here, the expert opined that the tie distorted customer choices by causing HANA sales that would not have otherwise occurred. The expert based this opinion on SAP business documents and sales data to understand S/4HANA leverage in selling HANA. The district court faulted the expert for failing to analyze how the tie effected Oracle, Microsoft, IBM, and Amazon. The Ninth Circuit held, instead, that an expert can extrapolate market-wide harm to competition based on volume to tied-product sales and coercion of particular customers. The expert provided evidence of both. He also provided qualitative evidence of the impact of the tie on the other major competitors. The Ninth Circuit found nothing to offend R. 702 here.

Second, the district court rejected as “unwarranted” the expert’s assumption that HANA was always sold as an EDW. The district court reasoned that HANA in runtime license form was not an EDW because of the inability to exchange data with other software products. It also faulted the expert for failing to identify HANA full licenses as EDW, but the court said a jury could infer that consumers could use both runtimes and full licenses as EDWs. There was some evidence for this. The expert’s conclusion was reasonable, and it was not the role of the court to try the facts now.

SAP urged that the expert failed to distinguish between tied and non-tied HANA sales, but the Ninth Circuit found the expert also addressed this issue in concluding that the tie caused HANA sales that would not have otherwise occurred.

Summary judgment. Having determined that the district court abused its discretion in excluding the experts’ market definition and competitive harm testimony, the court, applying de novo review, held summary judgment to be inappropriate.

As a preliminary matter, tying arrangements are subject to a “modified” per se analysis only if: 1) the defendant has market power in the tying product market; and, 2) the tying arrangement affects a “not insubstantial volume of commerce” in the tied product market. Moreover, tying arrangements meeting these two requirements are unreasonable as a matter of law.

SAP argued the Teradata tying claim had an acceptable nexus with procompetitive technological innovation and the case fit under the Epic Games and Microsoft narrow exception to the per se rule. See Epic Games, Inc. v. Apple, Inc., 67 F.4th 946, 997 (9th Cir. 2023); United States v. Microsoft Corp., 253 F.3d 34 (D.C. Cir. 2001). However, the court distinguished that “the tying and the tied products here are not technologically or physically integrated.” The court said it appreciated SAP’s concern about juxtaposing the tying per se rule with the general per se rule, but it noted the Supreme Court has made it clear that certain tying arrangements are unreasonable per se. The Ninth Circuit concluded here that it had no basis for expanding the Epic Games narrow exception to apply to software markets generally.

In any event, Teradata’s expert raised a material dispute of fact under either approach. Under per se analysis, the expert said SAP had market power in the core ERP market for large enterprises based on SAP’s market share, high profit margins, and high barriers to entry and switching costs. The court said a trier of fact could find enough ERP market power to coerce large enterprise HANA purchases.

Under the rule of reason, the expert’s testimony also raised a triable issue about whether the tie has substantial anticompetitive effects in the tied market of EDW products. The expert opined the tie would eventually foreclose at least 65 percent of the large-enterprise EDW market. A reasonable juror could find the tie has substantial anticompetitive effects.

The expert also testified that HANA prices were at supracompetitive levels, based on other evidence that this was the result of anticompetitive behavior, without evidence of procompetitive demand expansion. A jury could find market foreclosure. A differences-in-differences regression analysis, a standard econometric tool, also supported market foreclosure estimates beyond mere correlation.

Trade secrets. The Ninth Circuit concluded that disputed issues of material fact also precluded summary judgment on Teradata’s misappropriation of trade secrets claim on both theories of failure to designate the batched merge method as confidential and SAP’s putative right to use it in its own products.

First, Ninth Circuit saw that the Bridge Project NDA specified confidentiality. Each page of the design document sent to SAP was clearly marked “Teradata Confidential.” Nothing in the NDA required, as SAP urged, a detailed description of the trade secret beyond the stated words “batched merge,” yet the NDA actually detailed the essential elements. A jury could decide this dispute.

Second, the district court found the batched merge method was an “input” for SAP use outside the Bridge Product; however, as Teradata pointed out, a different section of the software development cooperation agreement said partner materials remained vested exclusively in Teradata. This, as well as disputes over the meaning of a software “tool,” and application of the covenant of good faith and fair dealing under New York law, all presented questions for a jury to decide.

The Case is No. 23-16065.

Judge: VanDyke, L.

Attorneys: Jack Williford Londen (Morrison & Foerster, LLP) for Teradata Corp. Kristin L. Cleveland (Klarquist Sparkman, LLP) for SAP SE.

Companies: Teradata Corp.; SAP SE

Cases: TechnologyInternet TradeSecrets AlaskaNews ArizonaNews CaliforniaNews HawaiiNews IdahoNews MontanaNews NevadaNews OregonNews WashingtonNews