|This document is available in two formats: this web page (for browsing content), and PDF (comparable to original document formatting). To view the PDF you will need Acrobat Reader, which may be downloaded from the Adobe site. For an official signed copy, please contact the Antitrust Documents Group.|
Claude F. Scott, Esq.
UNITED STATES DISTRICT COURT
NORTHERN DISTRICT OF CALIFORNIA
SAN FRANCISCO DIVISION
FILED UNDER SEAL PURSUANT TO
TABLE OF CONTENTS
TABLE OF AUTHORITIES
AlliedSignal, Inc. v. B.F. Goodrich Co., 183 F.3d 568 (7th Cir. 1999)
Ball Memorial Hospital v. Mutual Hospital Insurance, Inc., 784 F.2d 1325 (7th Cir. 1986)
Brown Shoe Co. v. United States, 370 U.S. 294 (1962)
Brunswick Corp. v. Pueblo Bowl-O-Mat, Inc., 429 U.S. 477 (1977)
California. v. American Stores Co., 697 F. Supp. 1125 (C.D. Cal. 1988)
California v. Sutter Health System, 130 F. Supp. 2d 1109 (C.D. Cal. 2001)
Cargill, Inc. v. Montford of Colo., Inc., 479 U.S. 104 (1986)
FTC v. Cardinal Health, Inc., 12 F. Supp. 2d 34 (D.D.C. 1998)
FTC v. Coca-Cola Co., 641 F. Supp. 1128 (D.D.C. 1986)
FTC v. Elders Grain, 868 F.2d 901 (7th Cir. 1989)
FTC v. Heinz Co., 246 F.3d 708 (D.C. Cir. 2001)
FTC v. Indiana Federation of Dentists, 476 U.S. 447 (1986)
FTC v. Libbey, 211 F. Supp. 2d 34 (D.D.C. 2002)
FTC v. PPG Industrial, Inc., 628 F. Supp. 881 (D.D.C. 1986)
FTC v. Staples, Inc., 970 F. Supp. 1066 (D.D.C. 1997)
FTC v. University Health, 938 F.2d 1206
Forsyth v. Humana, Inc., 114 F.3d 1467 (9th Cir. 1997)
Georgia v. Pennsylvania R.R.Co., 324 U.S. 439 (1945)
Hawaii v. Standard Oil Co., 405 U.S. 251 (1972)
Hospital Corp. of America v. FTC, 807 F.2d 1381 (7th Cir. 1986)
Image Tech. Services, Inc. v. Eastman Kodak Co., 125 F.3d 1195 (9th Cir. 1997)
In re Brand Name Prescription Drugs Antitrust Litigation, 186 F.3d 781 (7th Cir. 1999)
In re High Fructose Corn Syrup Antitrust Litigation, 295 F.3d 651 (7th Cir. 2002)
In the Matter of Weyerhaeuser, 106 F.T.C. 172 (1985)
Metro Mobile CTS, Inc. v. NewVector Communications, Inc., 892 F.2d 62 (9th Cir. 1989)
Olin Corp., 113 F.T.C. 400 (1990), aff'd, 986 F.2d 1295
Olin Corp. v. FTC, 986 F.2d 1295 (9th Cir. 1993)
Philadelphia National Bank, 374 U.S. 321 (1963)
Rebel Oil Co. v. Atlantic Richfield Co., 51 F.3d 1421 (9th Cir.)
Rothery Storage & Van Co., v. Atlas Van Lines, 792 F.2d 210 (D.C. Cir. 1986)
State of California v. American Stores Co., 872 F.2d 837 (9th Cir. 1989)
Tampa Electric Co. v. Nashville Coal Co., 365 U.S. 320 (1961), citing Times-Picayune Public Co. v. United States, 345 U.S. 594 (1953)
Times-Picayune Publishing Co. v. United States, 345 U.S. 594 (1953)
United States v. Archer-Daniels-Midland Co., 866 F.2d 242 (8th Cir. 1988)
United States v. Baker Hughes Inc., 908 F.2d 981 (D.C. Cir. 1990)
United States v. E.I. duPont de Nemours & Co., 351 U.S. 377 (1956)
United States v. Eastman Kodak Co., 63 F.3d 95 (2nd Cir. 1995)
United States v. Grinnell Corp., 384 U.S. 563 (1966)
United States v. Marine Bancorp., 418 U.S. 602 (1974)
United States v. Microsoft, 253 F.3d 34 (D.C. Cir. 2001)
United States v. Rockford Memorial Corp., 717 F. Supp. 1251 (N.D. Ill. 1989), aff'd 898 F.2d 1278 (7th Cir. 1990)
United States v. Rockford Mem'l Corp., 898 F.2d 1278 (7th Cir. 1990)
United States v. Syufy Enterprises, 903 F.2d 659 (9th Cir. 1990)
United States v. United Tote, 768 F. Supp. 1064 (D. Del. 1991)
15 U.S.C. § 18
15 U.S.C. § 25
15 U.S.C. § 26
Clayton Act § 7
Fed. R. Civ. P. 65
Areeda, Antitrust Law ¶ 914a (rev'd ed. 1998)
Areeda & Hovenkamp, Antitrust Law ¶518.1b at 534
Carl Shapiro, Mergers with Differentiated Products, 10 Antitrust 23 (Spring 1996)
JDEC-01E-057715, McKinsey,Memorandum to The Board of Directors, 2, (8/10/2001)
Jonathan B. Baker, Unilateral Competitive Effects Theories in Merger Analysis, 11 Antitrust 21, 22-23 (Spring 1997)
U.S. Dept. of Justice & FTC, 1992 Horizontal Merger Guidelines (rev. 1997):
William J. Kolasky & Andrew R. Dick, The Merger Guidelines and the Integration of Efficiencies Into Antitrust Review of Horizontal Mergers, 71 Antitrust L.J. 207, 228 (2003)
Oracle Corporation proposes to acquire PeopleSoft, Inc. through a hostile cash tender offer valued at $7.7 billion. If Oracle is not stopped, head-to-head competition between these two companies for procurements will vanish, leaving two dominant firms offering the high-function Human Resources Management ("HRM") and Financial Management Services ("FMS") functionality needed by this nation's most complex, global enterprises. Enterprise customers, who currently rely on the rivalry between Oracle and PeopleSoft to provide innovative products at competitive prices, will be left with a dramatically altered competitive landscape. Most customers will be left with two choices from which they can purchase mission critical software.
Objective observers in the marketplace, including even Oracle before it launched its hostile takeover attempt, recognize that there is a category of enterprise software that is produced only by Oracle, PeopleSoft or SAP AG ("SAP").1 Plaintiffs will call major customer witnesses to explain that because of the business realities of the HRM and FMS markets, the transaction would substantially reduce competition. Plaintiffs will also present testimony from the consultants who aid these customers in the procurement process ratifying these customer views. Substantial corroborating evidence of the takeover's likely anticompetitive effects will be presented from Oracle documents and executives' deposition testimony. Finally, Plaintiffs will introduce expert testimony that will provide the Court with an economic and technical framework to evaluate the factual testimony that the Court will hear.
Oracle's rebuttal arguments fail. Neither entry by mid-market software companies nor new firms will overcome the substantial anti-competitive effects this transaction would likely produce. In addition, Defendant's alleged efficiencies are unsubstantiated and are best understood as part and parcel of an anticomeptitive reduction in output. When the evidence is in, the likely anti-competitive effects of the proposed acquisition will be readily apparent.
For these reasons, the transaction violates Section 7 of the Clayton Act, 15 U.S.C. § 18. To prevent this substantial threat to competition, the United States and the Plaintiff States seeks injunctive relief under Fed. R. Civ. P. 65 and, respectively, 15 U.S.C. § 25 and 15 U.S.C. § 26.
Oracle Corporation, headquartered in Redwood Shores, California, is the second largest software company in the world with more than 40,000 employees, software sales in over 87 countries, and almost $9.5 billion in revenues in fiscal year 2003. Oracle is also the world's third-largest vendor of Enterprise Software Applications ("ESAs") including high-function HRM and FMS software, generating $ [REDACTED] in licensing revenue and $[REDACTED] in maintenance revenue in FY2003.
PeopleSoft, headquartered in Pleasanton, California, is the second-largest vendor of ESAs, including high-function HRM and FMS software, generating $[REDACTED] in licensing revenue and $[REDACTED] in maintenance revenue in FY2003. PeopleSoft acquired J.D. Edwards in 2003. Prior to the merger, J.D. Edwards, a provider of mid-market software, had abandoned an attempt to sell high-function ESAs.
On June 6, 2003, Oracle made an unsolicited $16.00 per share cash tender offer for all outstanding shares of PeopleSoft. On or about June 18, 2003, Oracle increased the offer to $19.50 per share, and on or about February 4, 2004, to $26.00 per share. Oracle reduced the offer on May 14, 2004, to $21.00 per share, which values the acquisition at approximately $7.7 billion. The merger would combine the second and third-largest of the three vendors of high-function HRM and FMS software.
For Oracle, this transaction is a no-lose proposition. Oracle seeks to acquire market share and an ongoing revenue stream without competing for it, while concurrently decimating its chief competitor. Oracle effectively plans to shut down sales of PeopleSoft products by no longer marketing to new customers and to end development of major upgrades to PeopleSoft products, reducing consumer choice in the market place. In addition, the transaction will eliminate Oracle's closest competitor for not only installed-base customers, but also for new license sales, allowing Oracle to raise license and maintenance fees. But even if the transaction is stopped, the effect of Oracle's prolonged and unwelcome assault on PeopleSoft will be to slow PeopleSoft's growth, to the primary benefit of Oracle.
Enterprise Software Applications provide the tools necessary to automate key business processes. ESAs enable more efficient management of the most basic functions of any organization. Major corporations and other institutions use ESAs to run and monitor their daily operations.
Organizations use four "pillars" of ESAs. Organizations manage employees through HRM applications, including payroll planning, recruiting, training, and benefits administration. Organizations maintain financial records through FMS applications, including general ledger, accounts payable, accounts receivable, and asset management. Organizations track inventory through supply-chain management ("SCM") applications, including modules to control inventory and manage manufacturing and distribution. Organizations maintain customer relationships through customer relationship management ("CRM") applications, including modules to manage the entire life-cycle of a sale, from customer prospects, through marketing, to customer service and support.
Within each of these pillars a variety of software modules provide various functions. PeopleSoft, for example, has 70 modules within its HRM pillar and 30 to 40 modules each within its FMS, SCM, and CRM pillars. A customer can license all modules available in the pillar or, as is more common, only those it needs.
ESAs may be combined into an enterprise resource processing suite ("ERP"), which usually consists of two or more pillars. This often provides a customer with easier implementation and significant cost savings. Customers need not purchase all four pillars when licensing these ERP suites.
Oracle's proposed acquisition of PeopleSoft combines two of only three firms that develop and distribute "high-function" HRM and FMS software. The HRM and FMS enterprise software products developed by Oracle, PeopleSoft and SAP are in a different class from mid-market software products even though both perform some of the same functions (e.g., general ledger). This distinction is well recognized throughout the enterprise software business, as Plaintiffs will show through the testimony of customer, industry analyst, and expert witnesses and documents.
Plaintiffs' expert Professor Marco Iansiti is expected to testify about the principal characteristics that distinguish high function software from other "mid-market" or "out of the box" products. The principle differences are:
Buying high-function ESAs is a lengthy, costly, and complex process. Although each sales cycle is somewhat different, Oracle has identified nine steps customers generally follow in the procurement process: (1) [REDACTED]; (2) [REDACTED]; (3) [REDACTED]; (4) [REDACTED]; (5) [REDACTED]; (6) [REDACTED]; (7) [REDACTED]; (8) [REDACTED]; (9) [REDACTED].2 Not all customers follow each of these steps, but it is a generally reliable overview for how the process works.
During the procurement process's initial stages, the customer may turn to industry analysts who produce reports, studies, consumer surveys, data on vendor R&D, vendor recommendations, and benchmark studies that estimate the total cost of ownership.3 Due to the complexity of the software and the high-risk nature of an enterprise software decision, customers often hire a "Big Five" consulting firm4 to assist in evaluating, selecting, implementing, and upgrading high-function software.
Once a customer has decided to license an enterprise software solution, the customer will perform a ROI calculation and establish whether it has the budget to invest in an enterprise solution. Vendors like Oracle and PeopleSoft will not engage in the lengthy and costly procurement process without some assurance that the customer will be able to commit the substantial resources required to license, implement, and manage the software. At this point, many customers send vendors RFIs to determine whether their software products can meet the customer's specific needs. Thereafter, the customer sends out requests for proposals to certain vendors. The RFPs are highly specific, can run over one hundred pages, and in some cases call for a vendor's capabilities on over hundreds of specific business requirements.
The customer and its consultants evaluate the responses and invite qualifying vendors for a demonstration. In a typical sales cycle the decision to exclude certain vendors from presenting a demonstration is based on performance, not price. Therefore, vendors not invited to a demo based on their perceived lack of functional performance have little, if any, price constraining influence on the vendors invited to present demonstrations.
Customers and their consultants test and evaluate the software during the vendor demonstrations. Each demonstration may last several days, providing the customer time to score each product's ease of use, performance, graphical interface, and reporting capabilities. Based on this information, vendors can assess which other rivals' products can credibly meet the customer's needs and refine their estimate of their chances of winning a particular contract. The vendors incorporate this knowledge into their bidding behavior. Vendors that lack functionality do not have a price constraining affect because the customer has already informed the vendors of its needs during the RFP and demonstration stages, and vendors know which other vendors' particular products satisfy those needs.
High-function HRM or FMS software generally costs the customer millions of dollars to license, implement, maintain, and keep current. The total cost of ownership ("TCO") associated with implementing ESAs includes: (1) application licensing fees; (2) new hardware costs; (3) costs to install, configure, and customize the software; (4) training costs; (5) maintenance service and update fees; and (6) internal staff costs. Although license and maintenance fees for the applications are a small portion of an ESA system's TCO 20 percent is typical vendors compete on license fees and other terms during the final round of negotiations. The maintenance contracts generally include software updates and worldwide 24-hour support. The fee for the annual maintenance contract is usually a set percentage of the license fee, and Oracle charges [REDACTED] percent of the initial license fee for maintenance for its ESAs.5 Therefore, the license fee affects customers' ongoing maintenance support costs.
Because high-function HRM and FMS applications are mission critical, customers are highly risk averse when it comes to ESA selection and implementation. These enterprises do not want to invest in an unproven software package. Customers making purchase decisions rely heavily on a vendor's history of successful installations in their same industry and with customers of similar size.
Section 7 of the Clayton Act prohibits acquisitions that may substantially lessen competition. 15 U.S.C. § 18. The lawfulness of an acquisition turns on its potential to create or enhance market power the ability of one or more firms to raise prices above competitive levels for some significant period of time. United States v. Archer-Daniels-Midland Co., 866 F.2d 242, 246 (8th Cir. 1988), cert. denied, 493 U.S. 809 (1989).
By proscribing transactions that "may" substantially lessen competition, Congress "indicate[d] that its concern was with probabilities, not certainties." FTC v. Heinz Co., 246 F.3d 708, 713 (D.C. Cir. 2001) (quoting Brown Shoe Co. v. United States, 370 U.S. 294, 323 (1962)). Section 7 does not require proof that higher, anticompetitive prices will occur in the affected market. Hospital Corp. of America v. FTC, 807 F.2d 1381, 1389 (7th Cir. 1986) (Posner, J). "All that is necessary is that the merger create an appreciable danger of such consequences in the future. A predictive judgment, necessarily probabilistic and judgmental rather than demonstrable is called for." Heinz, 246 F.3d at 719 (quoting Hospital Corp. of America, 807 F.2d at 1389). "[D]oubts are to be resolved against the transaction." FTC v. Elders Grain, 868 F.2d 901, 906 (7th Cir. 1989) (Posner, J.) (citing Philadelphia Nat'l Bank, 374 U.S. at 362-63).
Antitrust courts historically have evaluated the likelihood that a merger would create or enhance market power through inference based on market structure. In conventional merger analysis: markets are defined, participants in the market are identified, shares are assigned, concentration is measured, and anticompetitive effects are inferred. See United States v. Marine Bancorp., 418 U.S. 602, 618-23 (1974); Philadelphia Nat'l Bank, 374 U.S. at 363); United States v. Baker Hughes Inc., 908 F.2d 981, 982-83 (D.C. Cir. 1990); Olin Corporation v. FTC, 986 F.2d 1295 (9th Cir. 1993); FTC v. Heinz, 246 F.3d at 715. If the merger significantly increases concentration, the merger is presumed illegal, and the burden of producing evidence shifts to the defendant to show why the increased concentration is not indicative of anticompetitive effects. If the defendant produces sufficient evidence, the burden shifts back to the plaintiff to produce additional evidence demonstrating that the merger will produce anticompetitive effects. The plaintiff retains the ultimate burden of proof at all times. Heinz, 246 F.3d at 715; Baker Hughes, 908 F.2d at 983.
In merger analysis, the ultimate question is always about the creation or enhancement of market power. Market definition and market shares are a means to understanding the market power analysis, not an end in themselves, and as the Supreme Court reminds us, not a necessary predicate to antitrust liability under Section 1 of the Sherman Act:
Since the purpose of the inquires into market definition and market power is to determine whether an arrangement has the potential for genuine adverse effects on competition, proof of actual detrimental effects, such as a reduction in output, can obviate the need for an inquiry into market power, which is but a surrogate for detrimental effects.
FTC v. Indiana Federation of Dentists, 476 U.S. 447, 460-461 (1986); accord Ball Memorial Hosp. v. Mutual Hosp. Ins., Inc., 784 F.2d 1325, 1336 (7th Cir. 1986) ("market share is just a way of estimating market power, which is the ultimate consideration. When there are better ways to estimate market power, the court should use them"). The 9th Circuit follows this view. See, e.g., Rebel Oil Co. v. Atlantic Richfield Co., 51 F.3d 1421, 1434 (9th Cir.), cert. denied, 116 S. Ct. 515 (1995); Forsyth v. Humana, Inc., 114 F.3d 1467, 1475 (9th Cir. 1997).
The rule should be no different under Section 7 of the Clayton Act. Cf. United States v. Rockford Mem'l Corp., 898 F.2d 1278 (7th Cir. 1990) (section 7 and section 1 standards converge in merger cases). Because Section 7 deals with likely future effects of a transaction ("may substantially lessen"), rather than with current effects of challenged conduct, it is often necessary to infer those effects from market structure. Thus, courts define markets and measure shares to understand the potential of a merger to create or enhance market power. When it is available, they may also examine direct proof that a merger will likely result in higher prices to a significant number or group of customers. Where, as here, the market structure analysis and direct evidence of competition both indicate that the merger will increase prices and/or reduce quality, the two approaches each reinforce the conclusion that the transaction violates Section 7.
Oracle's proposed acquisition of PeopleSoft can be enjoined under either the direct or inferential approach to finding market power.6 Plaintiffs will set out ample evidence of a relevant market, and demonstrate that there are only three significant participants in the market. "Three to two" mergers are generally condemned by the courts under the structural approach. H.J. Heinz Co., 246 F.3d at 717 (D.C. Cir. 2001). But Plaintiffs will also offer direct evidence that the proposed acquisition would likely harm consumers.
The seminal Supreme Court cases define product markets by the "reasonable interchangeability of use" or the "cross elasticity of demand" between the product itself and substitutes for it, see Brown Shoe Co. v. United States, 370 U.S. 294, 325 (1962), which involves examining (1) the availability of substitute products, and (2) the degree to which customers are willing to switch to those substitute products. United States v. E.I. duPont de Nemours & Co., 351 U.S. 377, 400 (1956). The Court also teaches that the market "must be drawn narrowly to exclude any other product to which, within a reasonable variation in price, only a limited number of buyers will turn." Times-Picayune Publishing Co. v. United States, 345 U.S. 594, 612 n.31 (1953). The central question is whether an increase in price for one product would cause enough buyers to turn to other products so as to make the price increase unprofitable.
The Horizontal Merger Guidelines have refined the assessment of potential switching in the face of a price increase. See U.S. Dept. of Justice & Federal Trade Commission, 1992 Horizontal Merger Guidelines § 1.1 (rev. 1997) ("Merger Guidelines"). The Merger Guidelines set out the hypothetical monopolist test: If there were only one seller of a product7 (or group of products), would the monopolist exercise market power by significantly increasing price? If so, that product (or group of products) constitutes a product market. The Ninth Circuit has adopted the reasoning and the market power principle articulated in the Merger Guidelines. See, e.g., Image Tech. Servs., Inc. v. Eastman Kodak Co., 125 F.3d 1195, 1203-04 (9th Cir. 1997); Rebel Oil Co., Inc. v. Atl. Richfield Co., 51 F.3d 1421, 1434 (9th Cir. 1995); Olin Corp. v. FTC, 986 F.2d 1295, 1299-300 (9th Cir. 1993); California v. Sutter Health Sys., 130 F. Supp. 2d 1109, 1120, 1128-32 (C.D. Cal. 2001).
Market definition should reflect market realities. Metro Mobile CTS, Inc. v. NewVector Communications, Inc., 892 F.2d 62, 63 (9th Cir. 1989). Courts look to customers' perceptions of the marketplace, the existence of special classes of customers who desire particular products and services, and the defendants' documents reflecting the "business reality" of "how the market is perceived by those who strive to profit in it," FTC v. Coca-Cola Co., 641 F.Supp. 1128, 1132 (D.D.C. 1986), vacated as moot, 829 F.2d 191 (D.C. Cir. 1987), and industry or public perception of separate markets. See United States v. Grinnell Corp., 384 U.S. 563 (1966); Brown Shoe, 370 U.S. at 325, Olin, 986 F.2d at 1299, 1302-03; Rothery Storage & Van Co., v. Atlas Van Lines, 792 F.2d 210, 218 n.4 (D.C. Cir. 1986) ("industry or public recognition of the [market] as a separate economic unit matters because we assume that economic actors usually have accurate perceptions of economic realities"); Merger Guidelines § 1.11. Finally, the market definition need not be perfect, as it is "always possible to take potshots," at a market definition, Rockford Memorial, 898 F.2d at 1285; the definition exercise should assist the Court in understanding competitive effects, and need not be measured by "metes and bounds." See Tampa Elec. Co. v. Nashville Coal Co., 365 U.S. 320, 332 (1961), citing Times-Picayune Pub. Co. v. United States, 345 U.S. 594, 611, 73 (1953).
The markets at issue in this case, as alleged in Plaintiffs' Complaint, are traditional product markets based on product characteristics, not price discrimination markets based on customer characteristics. Compare Merger Guidelines § 1.11 with §1.12. The prevalence of price discrimination is important in understanding how competition functions in this industry, but does not necessarily imply a market definition based on customer characteristics. Cf. In re Brand Name Prescription Drugs Antitrust Litigation, 186 F.3d 781 (7th Cir. 1999) (discussing price discrimination as it relates to market power). To be sure, a price discrimination market definition is supported by Professor Elzinga, is legally permissible, see, e.g., United States v. Eastman Kodak Co., 63 F.3d 95 (2nd Cir. 1995) (rejecting price discrimination market on factual grounds), Merger Guidelines § 1.12, and is amply supported by the factual record in this case. It is not, however, the primary theory of market definition alleged by Plantiffs.8
Put another way, markets defined by product characteristics may contain elements of price discrimination and such markets do not require uniform pricing to avoid the price discrimination label. Indeed, the opposite result would be odd; very few markets are characterized by uniform pricing, and very few antitrust cases describe price discrimination markets. We now explain how the hypothetical monopolist or cartel would employ the price discrimination mechanism to exercise market power over high-function HRM and FMS products.
The market definition should account for how markets actually function on a day to day basis, or it would have little utility for its intended purpose: elucidating the competitive effects analysis. In high-function HRM and FMS software markets, the sales process tells the story. As explained above, and as witnesses will testify at trial, the industry practice is to purchase high-function software through a lengthy and detailed sales process, which by its nature generates information sufficient for price discrimination by the vendors. The uncontroverted fact is that all customers reveal a significant amount of detail about their functional requirements, current software and hardware platforms, and business processes. It is in the customers' interest to do so because the vendors need this information to show how their products are the best functional fit for the business process the software will automate. Vendors know the functional capabilities of their competitors, and can therefore identify their likely competition for the upcoming bid. If a customer has complex functional requirements both the customer and the vendors will know that the customer's best alternatives will be Oracle, PeopleSoft and SAP, and in some instances fewer than all three firms.
The price competition generally starts after the qualified vendors are selected by functional fit with the customer's needs. By the time a vendor is required to offer a final price, it usually has substantial knowledge about its value to the customer and the capabilities of rivals in setting its price. As Oracle's documents make clear, Oracle's prices are driven by the competition for the particular bid, and how intensely the customer prefers each vendor's product. Abritrage is impossible, because the software product as implemented in that customer's systems cannot be resold. In addition, the software license cannot be resold or sublicensed without the vendor's permission.
The price discrimination mechanism helps to explain why a hypothetical monopolist or cartel would significantly raise the price of high-function HRM and FMS software. In some markets, the existence of some customers that could readily switch to other products protects all customers from exercises of market power, because attempts to raise prices to just a subset of the customers would be defeated by arbitrage. In high-function software markets, however, customers that might readily switch to other products for example customers with functional requirements that would make them indifferent between Lawson and Oracle would not be subject to a price increase, but neither would such customers affect the price the hypothetical monopolist would charge the vulnerable customers. Through price discrimination, the hypothetical monopolist would raise prices of the high-function products, but not to all uniformly, and not to all customers.
Oracle's Discount Approval Forms ("DAFs") show unequivocally that Oracle, PeopleSoft and SAP are each other's primary competition for high-function bidding opportunities. Professor Elzinga will testify that he has reviewed and analyzed the DAFs, which are generated by the Oracle sales force when seeking permission to discount the price of the product at issue. The DAF identifies the name of the customer, the discount being requested, and the justification for the request. Smaller discounts may be approved by sales managers, while the largest discounts must be approved at the highest executive levels of the company.
The DAFs clearly show that Oracle views it primary pricing constraint as PeopleSoft, followed by SAP. To reject the government's product market, Oracle must argue that this reality would be no different if the current price constraint from PeopleSoft and SAP were to disappear that Lawson, legacy systems, the "do nothing" option and the various also-rans have the functionality to equally keep prices in check. In Professor Elzinga's analysis, Oracle's DAFs identified over [REDACTED] opportunities in the time frame under study, and competition with PeopleSoft and SAP was mentioned as a justification for the deepest level of discount [REDACTED] times, while the next most frequent justification, [REDACTED], is mentioned [REDACTED] times. The reason Oracle discounts so frequently against PeopleSoft and SAP, and not against other firms, is because Oracle's products are functionally more like SAP and PeopleSoft's products than they are like mid-market products. For this reason, Oracle could successfully form a cartel with PeopleSoft and SAP, and the cartel would not be rendered ineffective by competition from other vendors.9
The decision to purchase and implement ERP software is so costly, and the risk of a failed implementation is so enormous, that many potential customers turn to the "Big 5" consulting firms for advice on matching potential vendors' capabilities and their business automation needs. Plaintiffs will present testimony at trial from Nancy Thomas of IBM, Perry Keating of BearingPoint, and Lynne Anderson of Gap Gemini Ernst & Young to discuss their firms' software selection and implementation services, how the ERP selection process works, and how vendors fit the functional needs and requirements of their clients. In addition, David Dortenzo of Deloitte Consulting testified by deposition, and Christy Bass of Accenture appears on Oracle's witness list. All representatives of the Big 5 firms confirm that Oracle, PeopleSoft, and SAP offer products in FMS and HRM that are scalable, proven, multinationaland multi-currency, setting them apart from all other vendors, including Lawson, AMS and Microsoft.
Business Process Outsourcers, such as ADP and Fidelity Employer Services, offer services that automate HRM business functions. These outsourcers use SAP, PeopleSoft, and Oracle as their software platforms, and can therefore be viewed as value-added resellers of high-function software. Since the outsourcers require SAP, PeopleSoft or Oracle software to service their clients with high-function requirements, they are not in a position to undercut a price increase by the hypothetical monopolist of high-function HRM software.
ADP, thought to be the leading outsourcer, recently announced that it licensed SAP's HR software for use as the platform for its worldwide service.10 Although ADP offers payroll and HR services in several countries around the world, [REDACTED]. In order to compete worldwide, ADP sought [REDACTED].11 [REDACTED].12
The decision to outsource instead of license software is generally considered a fundamental question of how a business operates, rather than as a competitive option that serves as a pricing constraint at the end of a software sales cycle. As Mr. Ellison testified,"[REDACTED]."13
Customers that require high-function HRM and FMS software will not switch to mid-market software to avoid a price increase. Oracle appears to be arguing that mid-market software vendors, including Lawson, SSA/Baan, Microsoft, and AMS can discipline Oracle's prices on its high-function software. These mid-market vendors, however, offer basic HRM and FMS functionality that is not designed to handle large data sets, large numbers of transactions and concurrent users, and multiple languages and currencies. Customers of high-function software need its superior functionality, and for them mid-market software is not an acceptable option.
The geographic market is the United States. This geographic market is supported by the facts and law as applied to this case because it is large enough to capture all sellers of the relevant products (i.e., Oracle, PeopleSoft, and SAP), such that a hypothetical cartel consisting of the products of those firms could profitably raise prices above a competitive level without being constrained by sellers outside the United States. Defendant contends that the relevant geographic market encompasses at least the United States and Europe, because SAP is a German firm. However, that proposed market definition ignores market facts and, moreover, it does not rebut Plaintiffs' showing that the merger is likely to cause higher prices and other anticompetitive effects in the United States.
Under the Merger Guidelines, a "relevant geographic market" is defined using the same "hypothetical monopolist or cartel" test used to define a relevant product market. The smallest geographic area in which the hypothetical monopolist could raise prices constitutes a relevant geographic market. Merger Guidelines, § 1.21. See also Rebel Oil, 51 F.3d at 1434 (A "market" is any grouping of sales whose sellers, if unified by a monopolist or hypothetical cartel, would have market power in dealing with any group of buyers.") (citing Areeda & Hovenkamp, Antitrust Law ¶518.1b at 534).
The only effect of Defendant's proposed market definition is to inflate SAP's share of ERP sales relative to Oracle's and PeopleSoft's shares of such sales. However, this calculation underestimates the nature of competition among Oracle, PeopleSoft and SAP in the United States, as evidenced by customer witnesses and Defendant's own documents. Moreover, it obscures the competitive significance of head-to-head competition between Oracle and Peoplesoft. Oracle's business records show that, in the United States, (1) Oracle competed against PeopleSoft more often than any other competitor and (2) Oracle offered buyers its largest discounts when it competed against PeopleSoft for the sale. Thus, Defendant's proposed market definition ignores market facts and should be disregarded.
The proposed transaction is presumed illegal under Section 7 of the Clayton Act because the combined entity would have a significant market share in a sufficiently concentrated market. See Philadelphia Nat'l Bank, 374 U.S. at 363. "Statistics that indicate excessive post-merger market share and market concentration create a presumption that the merger violates the Clayton Act." State of Cal. v. American Stores Co., 872 F.2d 837, 842 (9th Cir. 1989) rev'd on other grounds, 495 U.S. 271 (1990), relevant part reinstated, 930 F.2d 776, 777 (9th Cir. 1991); see Baker Hughes, 908 F.2d at 982 (D.C. Cir. 1990); California v. Sutter Health System, 130 F. Supp. 2d 1109, 1118 (N. Cal. 2001). In other words, the Plaintiffs establish a prima facie case of a § 7 violation by demonstrating "that the merger would produce 'a firm controlling an undue percentage share of the relevant market, and [would] result [ ] in a significant increase in the concentration of firms in that market.'" H.J. Heinz Co., 246 F.3d at 715 (quoting Philadelphia Nat'l Bank, 374 U.S. at 363) (modifications in original).
Based on the structure of this industry alone, that is, a reduction from three firms to two for most customers,14 the Plaintiffs are presumptively entitled to injunctive relief. Indeed, no court has allowed such a merger, absent a strong argument that one of the companies is failing. According to the D.C. Circuit Court of Appeals, "no court has ever approved a merger to duopoly under similar circumstances." H.J. Heinz Co., 246 F.3d at 717 (D.C. Cir. 2001).
A growing number of courts, including courts of this Circuit, apply the Merger Guidelines' approach for assessing pre- and post-merger concentration through use of the Herfindahl-Hirschman Index ("HHI"). See, California. v. American Stores Co., 697 F. Supp. 1125, 1128-30 (C.D. Cal. 1988), relevant part aff'd 872 F.2d at 841-42; see also, e.g., H.J. Heinz Co., 246 F.3d at 716; AlliedSignal, Inc. v. B.F. Goodrich Co., 183 F.3d 568, 574 (7th Cir. 1999). The HHI for a market is calculated by summing the squares of the individual market shares of all firms participating in the market. See Merger Guidelines § 1.5. When the HHIs are above 1800, and an acquisition would result in an increase of more than 50 points in the HHI, the acquisition is presumed to be "likely to create or enhance market power or facilitate its exercise." Merger Guidelines § 1.51(c). "Sufficiently large HHI figures establish [a] . . . prima facie case that a merger is anti-competitive." H.J. Heinz Co., 246 F.3d at 716.
Oracle's acquisition of PeopleSoft would produce concentration in the high-function FMS and HRM software markets significantly above the Merger Guidelines' threshold for presuming that the transaction would substantially reduce competition. Plaintiffs, through their expert Professor Elzinga, will present evidence of this at trial. The high-function FMS market HHI is [REDACTED] and will rise by [REDACTED] points post-merger to [REDACTED], and the combined Oracle-PeopleSoft share will be [REDACTED]%. The high-function HRM market HHI is [REDACTED] and will rise by [REDACTED] points post-merger to [REDACTED], and the combined Oracle-PeopleSoft share will be [REDACTED]%.15 Both post-merger markets would be highly concentrated and experience large increases in concentration, with the combined firm controlling an undue share of the markets at issue.
Customer testimony, along with Oracle's business documents and other evidence, colorfully illustrates how head-to-head competition with PeopleSoft has benefitted consumers, and shows how the removal of PeopleSoft from the market will harm competition. This evidence shows that (1) bidding competition lowers prices to consumers; and (2) bidding strategy is affected by the fit between the customer's needs and the functional offering of the vendor. There will be numerous examples throughout the record; here are just two.
Mary Glover, Greyhound Lines, Inc.'s Vice President for IT Services, will testify at trial about a recent bidding competition between Oracle and PeopleSoft. Greyhound is a large intercity bus transportation company, with revenues of over $1 billion per year. Greyhound serves more than 3,700 destinations in the U.S. and Canada and employs approximately 15,000 people. [REDACTED].16 [REDACTED].17
Oracle will argue that prices are influenced by a host of factors beyond its control,20 but competition is the final determinant of price, and, in the relevant markets, Oracle's greatest source of competition is PeopleSoft.
As Oracle's "Discount Approval Forms" and other documentation confirm, Oracle and PeopleSoft are often each other's closest competition in the relevant markets, but their products are not perfect substitutes for each other. In economic terms, they are differentiated. Each firm offers high-function software with, for example, different modules, different user interfaces, different ability to interact with database software applications and different industry expertise. When products are differentiated, so that offerings of different market participants are imperfect substitutes, competition may be localized. See Carl Shapiro, Mergers with Differentiated Products, 10 Antitrust 23 (Spring 1996). In this situation sellers will compete more directly with those rivals who sell closer substitutes. See FTC v. Libbey, 211 F. Supp. 2d 34 (D.D.C. 2002) (approving preliminary injunction when the merging firms were each others closest substitutes in the restaurant glassware supply market); FTC v. Staples, Inc. 970 F. Supp. 1066, 1075-80 (D.D.C. 1997) (finding sub-market of office-supply superstores within a broader market that consisted of ordinary office supplies; condemning three to two merger among office-supply superstores). A merger between firms selling products that are relatively close substitutes may diminish competition by enabling the merged firm to profitably raise prices above pre-merger levels.21 In pre-merger competition between Oracle and PeopleSoft, some of the sales lost due to a rise in the price of one of the merging firms' products would be diverted to the other merging firm's product. A combined Oracle/PeopleSoft enables the merged firm to recapture these sales within the merged entity, thus rendering profitable a price increase that would not have been profitable pre-merger.22
This concern is increased here because Oracle, PeopleSoft, and SAP are the only providers that compete in the procurement process for high-function software. As the Discount Approval Forms demonstrate, customers often consider Oracle and PeopleSoft the top two competitors in head-to-head bidding competition. This head-to-head competition at the end of the procurement process, after other competitors have been excluded by the customer, results in substantial price competition as Oracle and PeopleSoft attempt to out-bid one another. A merged Oracle/PeopleSoft would need to outbid only the third closest rival, SAP, for these particular licenses.23 See H.J. Heinz, 246 F.3d at 718-19 (enjoining merger, in part, when it would eliminate competition between only two firms competing for second position grocery store baby foods slots). The customer, instead of paying the price that just outbids the second best-bidder, will now be forced to pay the price that just out-bids the third best bidder. See Jonathan B. Baker, Unilateral Competitive Effects Theories in Merger Analysis, 11 Antitrust 21, 22-23 (Spring 1997). A combined Oracle/PeopleSoft will be able to unilaterally raise prices and substantially lessen competition.
Professor McAfee will explain at trial that the merger would likely result in increased prices for high-function HRM and FMS software because competition from PeopleSoft currently poses a substantial constraint on Oracle's pricing.
Oracle, PeopleSoft and SAP all set customer prices for high-function software according to a "price calculator" or "price list" that specifies a list price based on a prospective customer's size and purchasing volume. "[T]he list price is usually the starting point for the bargaining and the higher it is (within reason) the higher the ultimately bargained price is likely to be." In re High Fructose Corn Syrup Antitrust Litigation, 295 F.3d 651, 657 (7th Cir. 2002). Oracle, PeopleSoft and SAP set and update their list prices interdependently, keeping tabs on one another's pricing structures and worrying about deviations. Oracle, PeopleSoft and SAP also commonly offer discounts off list price as competitive circumstances dictate. When competition requires a discount of more than around percent, sales representatives must formally apply for executive approval of the discount. The most direct evidence for head-to-head competition between Oracle and PeopleSoft is to be found in the many Oracle discount approval forms which describe competition from PeopleSoft as the "Justification" for the requested discount the DAFs referenced earlier.
In addition to exploring a number of case studies that evidence the intensity of such head-to-head competition, Professor McAfee undertook a broad empirical investigation of this competition using regression analysis. He found that, across a wide range of procurement competitions involving PeopleSoft, SAP and other companies, the presence of PeopleSoft as a competitor is associated with an increase in Oracle's discount by a statistically and economically significant amount.
As Professor McAfee will explain at trial, Oracle offers a customer discount to avoid losing sales. The DAFs confirm that the potential to lose sales to PeopleSoft is the most significant driver of Oracle's discounting. The merger would relieve Oracle of its most significant and frequent competitive threat, allowing Oracle to profitably scale back its discounting. Professor McAfee will explain that the merger would broadly result in substantial price increases given the information that Oracle and SAP have about customers and the differences in their high-function software products. Professor McAfee will present a model of the merger of Oracle and PeopleSoft which captures salient features of competition and is calibrated to market share data for HRM and FMS products provided by Professor Elzinga. In this model, the merger of Oracle and PeopleSoft raises price by anywhere from five to twenty-eight percent.
Arbitrage is an economic and legal impossibility in this industry. The software licenses prohibit resale of the product and, moreover, the installed applications are of no use to anyone but the intended customer. Recognizing that actual arbitrage cannot affect the prices paid by consumers, Oracle's experts will likely claim that consumers learn about discounts offered to others and can use this information to negotiate similar deals, which will simulate the effect of economic arbitrage. This strained analogy fails. A consumer's ability to strike a good deal depends on the breadth of alternatives to which the consumer could turn; it is the threat of losing a sale to an attractive alternative that drives Oracle's discounts. Learning that another consumer with broader alternatives has struck a better deal is of little help to a consumer with fewer or less attractive alternatives.
The potential for adverse unilateral effects is not limited to increases in initial license fees, but also includes increases in ongoing maintenance and other related fees. Enterprises typically negotiate long-term maintenance contracts when purchasing the initial license. Enterprises on maintenance contracts receive ongoing software support, including bug patches and updates. The annual maintenance fee is often pegged as a percentage of the enterprise's total license purchases from the vendor. Thus, maintenance fees will be subject to adverse unilateral effects as well as license fees.
Competition to retain one's own base and to convert the installed base customers of rivals is an important aspect of competition in the markets for high-function FMS and HRM software. Oracle's marketing head Keith Block put the point succinctly: "[REDACTED]."24 Competition means Oracle and PeopleSoft each know that they must keep quality high: provide their current customers good service; accurate, efficient and timely updates; and useful enhancements. If not, they would (eventually) risk losing installed-base customers and suffering in reputation that could hurt them in winning new customers. The merger would end such competition between Oracle and PeopleSoft. As a consequence, installed-base, high-function FMS and HRM customers would likely face higher maintenance fees, lower maintenance quality, or both.
To demonstrate that the merger's likely anticompetitive effects will be transitory, Oracle will argue that new vendors will enter the high-function markets. See Areeda & Hovenkamp, Antitrust Law ¶ 941b (Supp. 2004) ("The thrust of the low entry barrier claim is not that competition cannot be 'substantially lessened,' but rather that a substantial lessening of competition will not produce as substantial or as durable an effect as if entry barriers were higher."); cf. Olin Corp. v. FTC, 986 F.2d 1295, 1305 (9th Cir. 1993) (placing entry barrier burden on company "to rebut a prima facie case of illegality."). The defendant must show that entry would be timely, likely, and of sufficient magnitude to prevent a significant exercise of market power. See Rebel Oil Co. v. Atlantic Richfield Co., 51 F.3d 1421, 1440-41 (9th Cir. 1995) (citing Merger Guidelines' "timely, likely, and sufficient" test); Olin Corp., 986 F.2d at 1305 n.10. It is important to note that defendant's burden is not simply to show entry could happen although the technical possibility alone is a huge challenge here but that it likely would happen, i.e., that it is both technically possible and economically sensible. See FTC v. Cardinal Health, Inc., 12 F. Supp. 2d 34, 56 (D.D.C. 1998) ("[E]ntry is to be considered 'likely if it would be profitable at premerger prices, and if such prices could be secured by the entrant.'") (quoting Merger Guidelines § 3.1).
In the high-function HRM and FMS software applications markets, it is unlikely that substantial entry would undercut any post-merger price increase. None of the mid-market HRM and FMS providers, best-of-breed providers, or niche providers with a specific industry focus would be likely to enhance their product's scalability, flexibility, and functionality, and secure the required customer recommendations in a timely manner to prevent the combined Oracle/PeopleSoft from raising prices above pre-merger levels. Any firm attempting to reposition into the high-function HRM or FMS software market faces a formidable series of challenges to creating a high-quality product and then becoming a credible participant at the product review stage of the procurement process.
There are several impediments to timely, likely, and sufficient entry. First, potential entrants must overcome substantial complex technical obstacles impeding them from creating a high-quality differentiated enterprise software suite. United States v. United Tote, 768 F. Supp. 1064, 1075-77 (D. Del. 1991) (finding technology a barrier where likely software entrants adapting mid-market racetrack gambling software to large racetrack gambling software market needed 18 to 24 months to re-tool their software product; customers demanded 100% reliability; and past attempted entrants demonstrated repeated technological failures); Areeda & Hovenkamp, Antitrust Law ¶ 1128c4 (Supp. 2004) ("A firm is not likely to enter a market for which it lacks the capital, technical resources, marketing channels, or skills."). As discussed below, well financed potential entrants with sophisticated software solutions in their own markets have repeatedly failed to gain a toe-hold in the high-function HRM and FMS markets. Access to capital alone simply cannot create a working product in this industry.
Second, any potential sufficient entry is many years away even under the best of circumstances. The time needed to re-tool, expand, and debug a mid-market suite or best-of-breed application for the high-function market would take several years. At trial, Mr. Bergquist will testify about PeopleSoft's entry into the FMS and CRM product space, which took many years to develop, beta-test, fix, gain reference customers, and market to a broader customer base. This time frame exceeds both the Merger Guidelines two-year test, and any time horizon examined by the courts. See Syufy, 903 F.3d at 665 (analyzing post-merger entry at 13 months and 24 months); FTC v. PPG Indus., Inc., 628 F. Supp. 881, 885 (D.D.C. 1986) (finding high entry barriers when acquiring technical expertise delayed entry for two- to six-years), aff'd in part, 798 F.2d 1500 (D.C. Cir. 1986); Olin Corp., 113 F.T.C. 400 (1990) (finding high entry barrier when entry would take five- to ten-years), aff'd, 986 F.2d 1295; cf. In the Matter of Weyerhaeuser, 106 F.T.C. 172, 287 (1985) (finding five-year time frame untimely) .
The Court will hear the testimony of Richard Allen, the former CFO of JD Edwards, a failed entrant into high-function markets who will explain why even a seemingly well-positioned entrant could not break into the "upmarket." In addition, we discuss below prospects for entry by Microsoft, Lawson and AMS. The evidence from these firms refutes Oracle's argument that the threat of entry will cure this anticompetitive transaction.
Before being acquired by PeopleSoft in 2003, J.D. Edwards was the fourth largest provider of ESAs in the world and was thought to offer the highest functioning mid-market product. J.D. Edwards failed in their attempt to enter the "upmarket," and since it was perhaps best positioned to offer a competitive threat to Oracle, PeopleSoft and SAP, its failure is particularly instructive.
After investing hundreds of millions of dollars over several years, J.D. Edwards, [REDACTED], ceased its attempts to enter the high-function market in 2001. J.D. Edwards never secured the all too critical reference from an high-function customer needed to persuade other high-function customers that J.D. Edward's new product satisfied their needs. The product lacked [REDACTED]. Further, the product was not sufficiently [REDACTED]. Finally, the system had [REDACTED].25 The technical barriers were simply too formidable for J.D. Edwards to overcome, despite hundreds of millions of dollars in capital investments, hundreds of dedicated employees, and years of research and development.
Microsoft has no plans to compete against Oracle in the relevant markets. Despite its dominance in the desktop operating system market, it entered the ESA business only in 2001 with its acquisition of Great Plains Software, and continued its expansion by acquiring Navision, a Danish company, in 2002. Both of these companies sold mid-market software solutions aimed at general business users. Microsoft, which tellingly uses SAP software for its own business needs, has remained focused on the small and mid-market segments and has no plans to enter the large enterprise market. Specifically, Microsoft has no intention to [REDACTED].
Microsoft's products do not have high-function characteristics, including lacking some basic functions such as inter-company trading, along with deficiencies in terms of performance, scalability, and other areas. Microsoft believes it would take [REDACTED] to try to improve its products to have high-function characteristics, and even then those products [REDACTED]. Even if it could develop the code, Microsoft would need to test the software in a real world environment, correct any bugs, and secure a credible peer recommendation before selling the product.
Microsoft's current ESA development is called "Project Green." According to Microsoft, Project Green products, will initially have [REDACTED]. Microsoft estimates that Project Green will not be able to [REDACTED]. Microsoft does not intend for Project Green to serve [REDACTED].
Lawson offers both HRM and FMS mid-market software focused on three industries (sometimes referred to as "verticals"): retail, health care, and public sector (state and local government). Oracle views Lawson as a [REDACTED].26 PeopleSoft believes that Lawson has [REDACTED].27
The industries where Lawson competes most effectively do not often require HRM or FMS software with the most highly complex functional features (i.e. multiple lines of business or multinational locations) that are at the core of the Plaintiffs' relevant markets. Mr. Ellison [REDACTED]."28
Lawson's vendor characteristics also make its successful expansion into the relevant markets unlikely. Lawson's annual research and development expenses are far less than SAP's, Oracle's, or PeopleSoft's. Oracle recognizes that [REDACTED].29 SAP's internal business documents recognize [REDACTED].30
AMS is another niche software vendor, with a narrow focus on government-specific financial management functionality. AMS's products are typically purchased only by [REDACTED], and AMS does not [REDACTED]. Oracle's AMS witness, Ms. Morea, will testify at trial that the functionality for public sector organizations is substantially different than private sector functionality, and that [REDACTED]. In addition, Ms. Morea will confirm that AMS has no plans to [REDACTED]. Finally, Ms. Morea will testify that AMS has made [REDACTED].
Oracle urges the Court to focus on anything but the markets and the harm the Plaintiffs have charged in the Complaint. As best we can understand their experts' reports, Oracle claims that the "real" competition is for the technology stack, and, besides, the integration layer of the stack will change in such a way as to make the applications market different, sometime in the future.
To rebut the presumption of anticompetitive effects, Oracle must come forward with evidence, not arguments. See, e.g., Baker Hughes, 908 F.2d at 982 (after government establishes presumption that the transaction will substantially lessen competition, the burden of producing evidence to rebut this presumption then shifts to the defendant). Professor Teece believes [REDACTED], but this is not evidence of how the harm in the markets, as alleged, will be undermined. Professor Teece all but concedes this when he claims that the acquisition will [REDACTED]. While antitrust law no longer condemns bigness for its own sake, neither does it cite it as a reason to permit anticompetitive acquisitions.
Perhaps Professor Teece is best understood as saying that unless one understands how a product competitively interacts with all of its complements, it cannot be a proper antitrust market. That argument fails as a matter of law. In United States v. Microsoft, 253 F.3d 34, 52-54 (D.C. Cir. 2001), the D.C. Circuit upheld the market for operating systems, which are sold to consumers bundled with its compliments. The court held that operating systems had no short-term substitutes that consumers were willing to use, and therefore met the legal test for a product market. Id. The same is true here. Market definition has a set legal meaning, which the courts routinely apply to software and other high-tech markets, despite criticisms that such industries are too dynamic and fast moving to understand.
Mr. Kutnick's argument about the "integration layer" is similarly flawed. Mr. Kutnick argues that changes he perceives in the integration layer will, at some unspecified point in the future, make gluing together "best of breed" applications a more attractive proposition than it is today.31 Against that argument the Court must weigh the evidence from Mr. Ellison. According to Mr. Ellison's testimony, the best of breed approach is fundamentally flawed. The best of breed approach is "[REDACTED]"32 than an integrated suite, "[REDACTED]."33 Under Baker Hughes, Mr. Kutnick must offer evidence, not mere argument and opinion on where he thinks the market is heading. Mr. Ellison has already provided the evidence that undercuts the theory, and, accordingly, the defense should carry no weight.
Oracle will present the testimony of an official from Emerson Electric, and perhaps some other customers who do not demand high-function products. Emerson's business is organized like a [REDACTED] in other words, Emerson does not demand the functionality that is the core of the product markets alleged by the Plaintiffs. Even Oracle's Chairman and CFO recognizes that [REDACTED]"34
The relevance of Emerson Electric's testimony is obscure. Customers who do not demand the product do not inform an understanding of the competitive realities of the marketplace. Perhaps Oracle means to argue that Emerson's size and complexity make it a good sales opportunity for a suite of EAS products, and the fact that Emerson has not embraced Oracle's technology with unlimited zeal should prove that no one really needs high-function products. Emerson's software choice does not constrain the price of any other customer's high-function software procurement because of the price discrimination mechanism. Emerson's particular IT configuration is of no moment, and should be given no weight by this Court.
Only "extraordinary efficiencies" may rebut a showing that a transaction is likely to substantially reduce competition in a highly concentrated market, such as the high-function HRM and FMS markets. Heinz, 246 F.3d at 720; see FTC v. University Health, 938 F.2d 1206, 1223 ("to overcome a presumption that a proposed acquisition would substantially lessen competition [defendant] must demonstrate that the intended acquisition would result in significant economies"). Moreover, courts must test claims of such extraordinary efficiencies by "rigorous analysis" to ensure that they "represent more than mere speculation and promises about post-merger behavior." Heinz, 246 F.3d at 721.
Specifically, Oracle must show that the transaction will achieve significant efficiencies that are: (1) verifiable, i.e., supported by "credible evidence," Staples, 970 F. Supp at 1089; see University Health, 938 F.2d at 1223; (2) not attributable to reduced output or quality, see Rockford Mem'l, 717 F. Supp. at 1290, 4A Areeda, ¶ 974b3, at 64; (3) merger specific, see Cardinal Health, 12 F. Supp.2d at 62-63; and (4) greater than the transaction's substantial anticompetitive effects, see Cardinal Health, 12 F. Supp.2d at 63.
Oracle's speculative efficiency assertions do not satisfy these mandatory requirements and thus cannot overcome the prima facie illegality of the merger.
Oracle makes two types of efficiency claims the transaction purportedly will produce (1) cost savings and (2) innovation benefits. Neither type of claim was formulated using verifiable facts or methods. The underlying computations for Oracle's purported $[REDACTED] savings claim reside in a ten-month old computer spreadsheet (prepared in July 2003). The spreadsheet's projections are at the aggregate company-wide level; nothing is broken down by product, product group, or operating division. The spreadsheet does not explain the factual bases for the projections or cite supporting documents. Nor has Oracle identified any specific verifiable data that it used to formulate its cost savings claims. To the contrary, Oracle's 30(b)(6) efficiencies witness and co-President Safra Catz, testified that she did not [REDACTED].35 The Plaintiffs' rebuttal witness, Professor Mark Zmijewski, will provide testimony that Oracle has not provided sufficient factual bases for its alleged cost savings efficiencies, which factual bases is essential to their verification.
Oracle's innovation claims are equally unverifiable speculation. The only Oracle innovation claim that goes beyond a generalization is that there are purportedly certain specific PeopleSoft product features that the transaction will enable Oracle to replicate for its own software. However, Oracle has not specifically estimated [REDACTED], or determined the resources required to build each feature, with or without acquiring PeopleSoft.36
The transaction's hostile nature does not excuse Oracle from presenting verifiable efficiency claims. The law does not lower the bar for efficiency claims for non-friendly transactions that substantially reduce competition, as this one will. Moreover, the trial record will reflect that Oracle has not seriously attempted to use its formidable in-house business and outside expert resources to document the purported facts and methods used to make its efficiency claims.
Oracle's efficiency claims also are not cognizable because their core is apparently tied to reduced output and consumer choice. Cost savings from anticompetitive reductions in output or services do not count as efficiencies. See Rockford Mem'l, 717 F. Supp. at 1290 (rejecting some headcount reductions as efficiencies because "savings in these areas would occur . . . from a drop in production"); Guidelines at § 4 ("merger-specific efficiencies . . . do not arise from anticompetitive reductions in output or service").37 Simply put, saving money by reducing quantity and quality is not a cognizable efficiency under the antitrust laws.
Oracle's July spreadsheet strongly suggests that [REDACTED]. Roughly [REDACTED] percent of Oracle's purported cost savings are from reductions in sales and marketing and research and development. If achievable at all, these reductions appear largely tied to Oracle's halt to the active marketing, promoting and selling of PeopleSoft's software. For example, the number of projected PeopleSoft and J.D. Edwards customers served by Oracle [REDACTED] by fiscal year 2008, from [REDACTED] to [REDACTED]. Licenses of PeopleSoft and J.D. Edwards software to new customers [REDACTED].
The alleged planned cuts also will reduce consumer choice. One of Oracle's central purported justifications for the deal is to incorporate features of PeopleSoft's software that are not currently in Oracle's products.38 However, when these features will reemerge in Oracle's software is entirely speculative. Nor has Oracle apparently seriously determined the resources needed to support PeopleSoft's existing customers. Oracle's Executive Vice President of Applications Development alluded to these realities when testifying about an April 2004 conversation with Larry Ellison: [REDACTED].39
Oracle's efficiency claims contain at least two other core flaws. First, Oracle does not subtract the substantial integration costs associated with merging the two companies from its $[REDACTED] assertion. "Cognizable efficiencies are assessed net of costs produced by the merger or incurred in achieving those efficiencies." Guidelines § 4. The July spreadsheet predicts $[REDACTED] in merger expenses, more than [REDACTED] is notable because, as Oracle's co-President Catz testified, [REDACTED].40
Second, not all of the cost savings, even as alleged by Oracle, are specific to Oracle's acquisition of PeopleSoft. See Heinz, 246 F.3d at 722 (there must be efficiencies "that cannot be achieved by either company alone because, if they can, the merger's asserted benefits can be achieved without the concomitant loss of a competitor"). Oracle's $[REDACTED] amount includes potential synergies attributable to PeopleSoft's July 2003 acquisition of J.D. Edwards.41
The Plaintiff States brought this action under Section 16 of the Clayton Act, which authorizes them to seek injunctive relief against threatened loss or damage arising from the proposed acquisition. 15 U.S.C. §§ 18, 26; see Georgia v. Pennsylvania R.R.Co., 324 U.S. 439, 447 (1945). The threatened loss or damage from the proposed acquisition gives rise to the Plaintiff States having standing under two different capacities: (1) as parens patriae on behalf their citizens and each states' general welfare and economies and (2) as representatives of government agencies that are past purchasers or likely future purchasers of high-function HRM and FMS software. See Hawaii v. Standard Oil Co., 405 U.S. 251, 259-61 (1972).
With regard to Plaintiff States' standing as parens patriae, proof of a probable substantial lessening of competition will establish the threatened loss or damage to the general welfare and economy of each plaintiff state in which the relevant competition takes place. In its Answer, Oracle admitted that it markets and sells its products throughout the United States and in each of the Plaintiff States. Oracle's Answer - Allegation 4. Oracle also admitted that its sales "represent a regular, continuous and substantial flow of interstate commerce, and have had a substantial effect upon interstate commerce" with and in each of the Plaintiff States. Id. At trial, Plaintiffs will offer evidence that both commercial and public sector customers spend a substantial amount of money every year on high-function HRM and FMS software.
The Plaintiff States also have standing as representatives of governmental entities in their proprietary capacities. The governmental entities represented include state departments, bureaus, agencies, and political subdivisions that have purchased or are likely future purchasers of high-function HRM and FMS software. Georgia v. Pennsylvania R.R. Co., 324 U.S. at 447. To establish antitrust injury as private plaintiffs, the Plaintiff States must prove that the injury is "of the type the antitrust laws were designed to prevent and that flows from that which makes defendants' acts unlawful." Cargill, Inc. v. Montford of Colo., Inc., 479 U.S. 104, 114 (1986) (quoting Brunswick Corp. v. Pueblo Bowl-O-Mat, Inc., 429 U.S. 477, 489 (1977)). Oracle admitted that it either sells, licenses, or markets HRM or FMS software within each of the Plaintiff States. Oracle's Answer - Allegation 4; Defendant's Response to Plaintiff's First Set of Requests for Admissions. In all ten Plaintiff States, there are current or likely future public sector purchasers of high-function HRM and FMS software. At trial, representatives from state and local entities will be called as witnesses at trial as examples of public sector customers in the relevant market. In addition, there will be exhibits offered at trial, in the form of discount approval forms and requests for proposals that prove state and local entities are customers in the relevant market and would be harmed by a reduction in competition among the three high-function HRM and FMS software vendors. Accordingly, Plaintiff States will establish antitrust injury to their proprietary interests and will be entitled to injunctive relief.
For the reasons stated above, the Court should permanently enjoin Oracle's proposed acquisition of PeopleSoft, and enter judgment for the Plantiffs.
1 Plaintiffs have elected to give this software the short-name of high-function HRM and FMS software for purposes of delineating relevant product markets in this case.
2 Ciandrini CID Dep. 81:22-83:16 (discussing CID exhibit 4).
3 The industry analysts include AMR Research, Gartner Group, Giga Information Group, IDC, and Meta Group.
4 The "Big Five" are Accenture, BearingPoint, Cap Gemini, Deloitte Consulting, and IBM.
5 Ellison CID Dep. at 75.
6 Clayton Act § 7 is "a prophylactic measure, intended 'primarily to arrest apprehended consequences of intercorporate relationships before those relationships could work their evil . . . .' "Brunswick Corp. v. Pueblo Bowl-O-Mat, Inc., 429 U.S. 477, 485 (1977). The term "direct proof" is used in Sherman Act cases to refer to harm that has already occurred, and can be observed by the court without inferring market power from market shares. Here, of course, the harm cannot be literally observed because it has not happened, and thus the term "direct" is intended as the category of proof that does not rely on inferring market power through market shares.
7 The Merger Guidelines discuss a hypothetical monopolist, but a hypothetical cartel serves the same analytical function: would exercising market power be profitable if there were no price competition over this product?
8 Price discrimination markets would be permissible in this case, and using that approach would not affect the outcome of the analysis. Typical price discrimination markets consist of customers buying the same physical product as all other customers, but may be charged a higher price because they have less propensity to substitute to other products and can be segregated from the larger group. To sustain a price discrimination market, it is essential that arbitrage would not defeat the differential pricing scheme. Merger Guidelines § 1.12.
9 Professor Preston McAfee, Plaintiffs' primary expert on competitive effects, will also rely on the DAFs, and will also testify that the information that becomes available during the sales process allows Oracle to price discriminate.
10 Fidelity Employee Services Company (sometimes referred to as "FESCo") is in a similar posture. Fidelity offers outsourcing services, and has landed [REDACTED]
11 Zena Brand Dep. 40:16-41:11.
12 [REDACTED] Currently, [REDACTED] for the U.S. that is based on PeopleSoft source code, which ADP has modified to fit its service model. Enterprise HR is a US-only product, and has no global functionality. US clients can purchase ADP services elsewhere in the world, but those services lack high-function attributes.
13 Ellison CID Dep. 252:8-13.
14 For certain industries where a specialized software vendor operates, such as the federal government, some customers face a reduction in competition from four firms to three. Even a four to three merger may substantially lessen competition. See United States v. Rockford Memorial Corp., 717 F. Supp. 1251, 1280 & 1291-92 (N.D. Ill. 1989) (enjoining four to three merger), aff'd 898 F.2d 1278 (7th Cir. 1990).
15 Professor Elzinga analyzed data produced by Oracle, PeopleSoft and other software vendors during the course of the merger investigation before the Antitrust Division. Elzinga measured market shares of license sales of at least $500,000. This size-of-deal proxy was intended to screen out mid-market software sales, but did not screen perfectly, and the market share data contains some sales that are not in the relevant market. For example, Microsoft's license sale to [REDACTED] was included, but as Microsoft will explain at trial, that software lacks the functionality to meet the market definition.
17 ORLITE 0125888-125891 (Discount Approval Form for Greyhound Inc.)
18 Ciandrini CID depo, 212:6-9.
19 Id. at 202:7-15.
20 One such factor is the customer's option to delay purchase, by remaining longer on a legacy system. As Professors Elzinga and McAfee will explain at trial, this option does not currently pose a significant constraint on Oracle's pricing.
21 See Staples, 970 F. Supp. at 1083 ("The merger would eliminate significant head-to-head competition between the two lowest cost and lowest priced firms in the superstore market. Thus, the merger would result in the elimination of a particularly aggressive competitor in a highly concentrated market, a factor which is certainly an important consideration when analyzing possible anti- competitive effects.").
22 Such a price increase will be profitable post-merger if rival sellers are unlikely to reposition their product offerings so as to replace the loss in localized competition. For a further discussion see Areeda, Antitrust Law ¶ 914a (rev'd ed. 1998).
23 In the services and financial services industries, for example, SAP's product is not a particularly good substitute for the PeopleSoft product.
24 E-mail from Keith Block dated October 19, 2003 with the subject line "[REDACTED]," ORCL-EDOC-00685563-4.
25 See JDEC-01E-057715, [REDACTED]
26 Ellison CID Dep. at 196:18-197:7.
27 See [REDACTED], PowerPoint presentation, [REDACTED], PS-3180042-61.
28 Ellison CID Dep. at 219:9-10 (referencing ORCL-EDOC-00144396 - 409 "[REDACTED].").
29 Ellison CID Dep. at 236:11-12; 237:13-20.
30 [REDACTED], SAP01186-01197 at 01191.
31 Best-of-breed is "an industry term, meaning that [customers] go out and find individual vendors who offer the best marketing package, the best sales package, the best service package, the best financial package, the best H.R. package, the best payroll package, you find the best suppliers and you buy from the best suppliers, ignoring the integration issues, you just get the best parts." Ellison CID Dep. at 133-34. These discrete "parts" are not designed to work together, and each application operates on its own database. Therefore, companies that purchase best-of-breed software must connect, or "integrate," the parts to permit them to communicate with each other and form a working system. See Ellison CID Dep. at 131-139.
32 Ellison CID Dep. at 136.
33 Id. at 138.
34 Henley Dep. 240:13-18.
35 Catz Dep., 64:15-66:4.
36 E.g., Wohl dep, 485:2-9. ("[REDACTED]."); Catz Dep., 108:9-14, 109:9 - 111:11.
37 See also 4A Areeda, ¶ 974b3, at 64 (reduction in personnel should not be considered an efficiency when a firm is simply "reducing its output"); William J. Kolasky & Andrew R. Dick, The Merger Guidelines and the Integration of Efficiencies Into Antitrust Review of Horizontal Mergers, 71 Antitrust L.J. 207, 228 (2003) ("reductions of output, for example, will normally be accompanied by reductions in (total) costs, but this cost reduction is not an efficiency").
38 Oracle's Response to the United States' Interrogatory 2.
39 Wohl Dep., 494: 4-15 (emphasis added).
40 Catz Dep., 413:14 - 415:16.
41 In addition, Oracle cannot demonstrate that any efficiencies associated with the transaction will offset its substantial anticompetitive effects.