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.
IN THE UNITED STATES DISTRICT COURT
THE UNITED STATES' REDACTED
The United States submits this memorandum in support of its motion for a temporary restraining order to prevent the acquisition by SunGard Data Systems, Inc. ("SunGard") of substantially all of Comdisco, Inc.'s ("Comdisco") disaster recovery solutions assets.(1) SunGard is scheduled to acquire these assets pursuant to a commercial agreement to be consummated after an order is entered by the Bankruptcy Court, after a hearing scheduled for October 23, 2001. Unless the Court enjoins this transaction, competition in the market for shared hotsite disaster recovery services for large scale enterprise computer processing centers ("shared hotsite services") in North America, a market SunGard describes "as an oligopoly with three major competitors,"(2) is likely to be substantially lessened, resulting in higher prices and reduced services, to the detriment of numerous corporate and government agency customers, in violation of Section 7 of the Clayton Act, 15 U.S.C. § 18.(3)
Today, numerous businesses, government agencies and other entities depend on computer data processing systems to run mission-critical operations.(4) In many instances, these companies run their most vital computer applications on IBM mainframe and other high-end computing platforms because the operations they perform require the high levels of performance and reliability provided by such systems. These functions include processing and storing transaction information, maintaining customer accounts, controlling production resources, inventory and shipping, and maintaining financial and administrative records. Because of the essential role in their operations played by the applications run on mainframe and other high-end computing platforms, companies require an effective disaster recovery plan to reduce the potentially devastating impact on their business of a disaster that destroys or disables their computer capacity.(5) Recent events have underscored the types of risks that are present, and the importance of disaster recovery services.
Disaster recovery vendors serve this critical need for companies by selling a service that enables the restoration of computer applications at another location if a natural disaster, major power outage, or other event causes their customer's primary data centers to become unavailable. Shared hotsite services are the primary service sold by disaster recovery vendors to companies that are dependent on mainframe and other high-end platforms. Because hotsites are shared by multiple clients, they provide cost-effective disaster recovery protection for large companies. (6)
SunGard and Comdisco are two of only three significant vendors in North America that offer shared hotsite disaster recovery services to companies that utilize large scale enterprise and/or mixed platform data processing centers. The proposed transaction will create a duopoly in which SunGard will have a greater than % market share with significant barriers to entry. If completed, the transaction is likely substantially to lessen the competition that has provided customers with competitive prices and services in the shared hotsite services market.
A TRO is warranted because (1) the United States is likely to prevail on the merits of its Section 7 action; (2) without a TRO, the public will suffer irreparable harm from lost competition between the parties and the lack of an effective remedy once SunGard acquires Comdisco's disaster recovery solutions and Comdisco ceases to operate as an independent business; (3) granting the TRO will not substantially harm the defendants; and (4) the TRO will further a strong public interest by maintaining competition in the shared hotsite services market.
Unless a TRO is granted, the defendants will close as soon as the sale is authorized by the Court on Tuesday, October 23, 2001. The United States has conducted an accelerated review of this matter. An investigation was opened on August 18, 2001 and the parties submitted their Hart-Scott-Rodino filings three days later. The United States rapidly served subpoenas on the parties, conducted numerous interviews and reviewed hundreds of boxes of documents. By way of comparison, a typical investigation of this complexity takes four to five months. This past weekend, the Department conducted two depositions of two high-level Comdisco employees that were important to the Department's analysis. Upon reviewing these depositions and the other materials submitted, the Department then filed this action.
I. FACTUAL OVERVIEW
SunGard is a Delaware corporation with its principal place of business in Wayne, Pennsylvania. SunGard is a major supplier of information technology. These solutions include investment support systems, trade processing, risk and asset management, as well as business continuity or disaster recovery services. SunGard is a large provider of shared hotsite disaster recovery services, deriving approximately $ million in revenues from this business.(7) SunGard now is seeking to further consolidate its already strong position in the shared hotsite services market by purchasing Comdisco's assets.
Comdisco is a Delaware corporation, headquartered in Rosemont, Illinois. Comdisco is also a major supplier of shared hotsite services, as well as a variety of other electronics equipment leasing and computer services. Comdisco's shared hotsite business generated approximately $ million in 2000.(8)
On July 15, 2001, Hewlett-Packard Company ("HP") entered into an acquisition agreement to purchase substantially all of the assets of Comdisco's business that provides disaster recovery planning and services ("Availability Solutions") for $610 million. On July 16, 2001, Comdisco filed a voluntary Chapter 11 bankruptcy petition with the U.S. Bankruptcy Court for the Northern District of Illinois, Eastern Division (Case No. 01-24795). By Order of the Bankruptcy Court, dated August 9, 2001, Comdisco's assets were set to be sold by auction on October 11, 2001. The order provided that the HP agreement constituted the auction floor.
SunGard was selected as the successful bidder. SunGard's acquisition proposal will be presented to the Bankruptcy Court for approval at the Sale Hearing scheduled for October 23, 2001.
Disaster recovery services encompass a range of solutions that allow firms to restore their business data processing and communications at another location if their primary data center becomes inaccessible because of a natural disaster, fire, terrorist act, or other cause. The firms continue their operations at this alternate location until their primary data center can be restored or rebuilt. Business data processing and communications applications vary in both (1) the types of computer platforms on which they run and (2) the degree to which they are mission-critical. Some applications are so critical that they require virtually instantaneous recovery, others may be restored within a few days, and still others have a "recovery time objective" ("RTO") of a week or more. To meet these various needs, there are different types of disaster recovery services and technologies, ranging from those that provide nearly immediate restoration of critical applications to services which require several months to implement. Accordingly, businesses often utilize a mix of solutions to meet their varying criticality and RTOs.
Shared hotsite services are one prominent type of disaster recovery service. Shared hotsite services are remote facilities that contain a broad range of the computer systems and communications facilities that would be necessary for a client to recover business applications if its own data center becomes unavailable.(9)
These facilities, effectively, allow a client to replicate its computer data center operations at a hotsite geographically separate from the data center to be recovered, so as to avoid the risk that its hotsite will be disabled by a regional disaster affecting the client. (10)
Shared hotsite services are purchased by organizations that process critical business applications on mainframe computers and/or large scale computer platforms. Typically, customers purchase shared hotsite services for critical business applications with RTOs of 16 to 96 hours.(11)
When a disaster disrupts or destroys a client's data center, a shared hotsite client takes its backup tapes from a secure location (also usually remote from its data center) to the hotsite.(12) Once delivered, the client's personnel, assisted by skilled hotsite technicians, load the software on the facilities' computers, transfer the backup tape data to the hotsite's computer storage systems, and commence operations.(13) Delivering and loading the tapes, and configuring the hotsite's computers to a customer's data processing environment, generally requires between 24 and 96 hours (although hotsite vendors can provide additional services to reduce this minimum time).(14) Typically, hotsite vendors allow clients to use their facilities for up to six weeks. During this time, the client either restores its data center or makes alternative arrangements.(15)
When clients are not using a hotsite to restore a data center, the hotsite is used, almost continuously, to rehearse and test customer recovery plans. Comprehensive testing is critical to effective disaster recovery, and customers contend to schedule their rehearsals or testing.(16)
Shared hotsite services are operated on a subscription basis. Because only a small number of customers are likely to experience disasters simultaneously, hotsite vendors can sell the same physical assets and services to many customers.(17)
If multiple simultaneous disasters occur, hotsite vendors typically make their facilities available on a first come, first served basis, or allocate capacity among customers contending for the same computer systems.(18) If a client's primary hotsite is occupied, the client can be diverted to alternate hotsites owned by the vendor.(19) Because the hotsite computing capacity is shared among multiple clients, each client is charged only a relatively small fraction of the total cost of the hotsite facility.(20)
There are only three major providers of shared hotsite services -- SunGard, Comdisco and IBM. Their respective hotsite businesses' each have multiple facilities that house a broad range of computers and storage devices for multiple platforms, and are interconnected with high-speed, high-bandwidth communications networks. These assets cost each company hundreds of millions of dollars.
SunGard: SunGard owns three disaster recovery megacenters. SunGard's facilities offer the following computing platforms: IBM mainframes, IBM AS/400, RS/600, Digital Equipment Corporation ("DEC"), HP, Intel-NT, Intel-Windows, Stratus, Sun, Tandem and Unisys.(21) Since 1997, SunGard's lease payments for its business continuity equipment have totaled approximately $ .(22)
Comdisco: Comdisco has 28 disaster recovery sites with 1.5 million square feet in North America. Comdisco provides services for users of the following platforms: IBM mainframes, AS/400, IBM RS/6000, DEC, HP, Intel-NT, Sequent, Stratus, Sun, Tandem and Unisys.(23)
IBM is the only other major provider of disaster recovery hotsite services in the United States. IBM offers similar computing platforms to those provided by SunGard and Comdisco. While there are several other niche players in the market, these companies do not have facilities in North America even remotely comparable to those owned by the market's three dominant companies.(24)
The defendants' most recent 10-K securities filings confirm that the most significant competitors in the shared hotsite market are SunGard, Comdisco and IBM.(25) The defendants' documents and other information produced to the government are similarly clear-cut. According to the defendants, in virtually every recent bidding instance involving shared hotsite services for mainframe and/or other large computer server platforms, only SunGard, Comdisco and/or IBM submitted bids.(26) Comdisco identified IBM and SunGard as the only firms competing with Comdisco throughout the use on a full range of computer platforms necessary to serve most firms with large data center operations.(27)
SunGard's, Comdisco's and IBM's dominance in the shared hotsite market is further reflected in numerous documents produced by the parties that repeatedly identify each other as the only meaningful competitors.(28) An April 17, 2001 E-mail to a potential customer, UniGroup, from SunGard's Senior Vice President, Brett W. Orr, makes the point cleanly. Orr's E-mail informs UniGroup that SunGard is displeased that UniGroup recently requested an RFP from SunGard primarily to extract price concessions from its incumbent hotsite vendor, IBM. Orr then instructs UniGroup that "[i]f UniGroup wants to use another vendor to keep your "partner" IBM in line with pricing . . . . then please just send your next RFP to CDRS [Comdisco] (the only other vendor in the industry . . . .)."(29)
Numerous customers, from many industries and government agencies, similarly state that SunGard, Comdisco and IBM are the only viable sources of disaster recovery services capable of restoring mainframe and other high-end multi-platform systems.(30)
These customers maintain that there are no other economical solutions to which they can turn to meet their critical disaster recovery requirements.(31)
The three way competition between SunGard, Comdisco and IBM has provided significant benefits for customers. Customers repeatedly leverage SunGard, Comdisco and IBM against each other to negotiate lower prices, sometime obtaining dramatic discounts from a vendor's initial offer.(32)
Most shared hotsite service agreements are three to five years in duration; however, customers frequently renegotiate with their existing supplier for extensions of their contracts, such as when they wish to purchase additional shared hotsite coverage for modified hardware or software applications. Customers benefit from the presence of competitive shared hotsite providers when they are extending the length of their contracts, and to an extent even when they are modifying but not extending their contracts. When new or additional services are being acquired the price and quality of the services being offered by the customer's incumbent supplier is to some extent constrained by the presence of actual or potential rival suppliers. Where the customer's contract is being extended, the terms of the extension will be more favorable to the customer if the incumbent would face significant competition had the customer not renewed, since the customer always has the option not to extend its contract at all. In each of these scenarios, therefore, the ability of customers to maintain bargaining power vis-a-vis their incumbent supplier is dependent on actual or potential competition from rival suppliers.(33)
Many customers view Comdisco and SunGard as the two best, and sometimes the only, vendors in many bidding situations.(34)
IBM is often not considered either because of higher prices, lack of proximity of hotsite locations, terms of service, platform focus and capabilities,(35)
or reluctance of the customer to have IBM as their disaster recovery vendor because IBM is already their major equipment vendor or a significant business competitor to the customer.(36)
For such customers, the proposed acquisition combines the two best and closest competitors, substantially lessening their ability to use competition between vendors as an effective negotiating tool.(37)
Thus, if the transaction is completed, it will reduce the number of viable suppliers of shared hotsite services from three to two, and eliminate a major competitor from the market. This consolidation is likely to cause higher prices and reduced levels of service. These threatened adverse effects have caused numerous customers to complain to the United States and the defendants.(38)
II. LEGAL ARGUMENT
Section § 15 of the Clayton Act authorizes courts to issue temporary restraining orders to prevent violations of the Act. 15 U.S.C. § 25. Courts in this Circuit determine whether preliminary relief is warranted by considering: (1) if there is a substantial likelihood the plaintiff will prevail on the merits; (2) whether the plaintiff will be irreparably injured if an injunction is not granted; (3) the injunctions' impact on the non-moving parties; and (4) whether an injunction furthers the public interest. See Mova Pharmaceutical Corp. v. Shalala, 140 F.3d 1060, 1066 (D.C. Cir. 1998).(39) These factors are not considered in isolation from one another, and no one factor is necessarily dispositive. See City Federal Fin. Corp. v. Office of Thrift Supervision, 58 F.3d 738, 747 (D.C. Cir. 1995). Rather, the factors "interrelate on a sliding scale and must be balanced against each other." Serono Laboratories v. Shalala, 158 F.3d 1313, 1318 (D.C. Cir. 1998). Thus, "[i]njunctive relief may be granted with either a high likelihood of success and some injury, or vice versa." Population Institute v. McPherson, 797 F.2d 1062, 1078 (D.C. Cir. 1986).
Here, each factor strongly supports granting a TRO. First, there is a substantial likelihood that the United States will prevail on the merits because the SunGard/Comdisco transaction is substantially likely to lessen competition in the shared hotsite services market, thus violating § 7 of the Clayton Act. Second, if the transaction is not enjoined, the public will be irreparably harmed because SunGard and Comdisco will cease to compete, customers will pay higher prices which cannot be recouped, and SunGard will commingle Comdisco's assets with its own, precluding the United States from obtaining an adequate remedy. Third, an injunction will not unduly harm the parties because it will simply maintain the status quo for a short period until a preliminary hearing can be held. Fourth, the injunction will further the public interest by enabling the United States to preserve competition that will be lost if the transaction is completed.
Section 7 of the Clayton Act prohibits acquisitions "in any line of commerce or in any activity affecting commerce . . . [if] the effect of such acquisition may be substantially to lessen competition, or to tend to create a monopoly." 15 U.S.C. § 18. To establish a § 7 violation, the government must only show that it is reasonably likely that the acquisition will cause anticompetitive effects. See United States v. Penn-Olin Chemical Co., 378 U.S. 158, 171 (1964) (these "requirements [are] . . . satisfied when . . . the 'reasonable likelihood' of a substantial lessening of competition in the relevant market is shown"); FTC v. Heinz Co., 246 F.3d 708, 713 (D.C. Cir. 2001) ("Congress used the words 'may be substantially to lessen competition' . . . , to indicate that its concern was with probabilities, not certainties.") (quoting Brown Shoe Co. vs. United States, 370 U.S. 294, 323 (1962)). "Section 7 does not require proof that a merger or other acquisition [will] cause higher prices in the affected market. All that is necessary is that the merger create an appreciable danger of such consequences in the future." Hospital Corp. of Am. v. FTC, 807 F.2d 1381, 1389 (7th Cir. 1986). "A predictive judgment, necessarily probabilistic and judgmental rather than demonstrable . . . , is called for." Id. (citing United States v. Philadelphia Nat'l Bank, 374 U.S. 321, 362 (1963)).
To predict whether an acquisition may substantially lessen competition, the court must determine: (1) the product market in which to assess the transaction; (2) the geographic market in which to assess the transaction, and (3) the transaction's probable effect on competition in the product and geographic markets. See United States v. Marine Bancorporation, Inc., 418 U.S. 602, 618-23 (1974); FTC v. Swedish Match, 131 F. Supp.2d 151, 156 (D.D.C. 2000); FTC v. Staples, Inc., 970 F. Supp. 1066, 1072-73 (D.D.C 1997).
Merger analysis begins by determining the relevant product market. Brown Shoe, 370 at 324; FTC v. Cardinal Health, Inc., 12 F. Supp. 2d 34, 46 (D.D.C. 1998). "The outer boundaries of a product market are determined by the reasonable interchangeability of use [by consumers] or the cross-elasticity of demand between the product itself and substitutes for it." Brown Shoe, 370 U.S. at 325; see United States v. E.I. du Pont de Nemours & Co., 351 U.S. 377, 395 (1956) ("Commodities reasonably interchangeable by consumers for the same purposes" constitute a product market for antitrust purposes.). Interchangeability of use and cross-elasticity of demand concern (1) the availability of products that are similar in character or use to the product in question and (2) the degree to which buyers are willing to substitute those similar products for the product. Swedish Match, 131 F. Supp.2d at 157 (citing E.I. du Pont de Nemours, 351 U.S. at 393). The market "must be drawn narrowly to exclude any other product to which, within reasonable variations 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). Thus, the pivotal question in product market definition 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. Du Pont, 351 U.S. at 400; see Staples, 970 F. Supp. at 1074 ("[T]he general question is 'whether two products can be used for the same purpose, and if so, whether and to what extent purchasers are willing to substitute one for the other.'").
This same analytical approach is incorporated in the 1992 U.S. Department of Justice and Federal Trade Commission, Horizontal Merger Guidelines ¶ 1.11 (1997 rev.) (hereinafter "Merger Guidelines"). The Merger Guidelines take the smallest possible group of competing products and ask whether a "hypothetical monopolist over that group of products would profitably impose at least a 'small but significant and nontransitory' [price] increase." Merger Guidelines ¶ 1.11. Under the Merger Guidelines, a "small but significant and nontransitory" price increase in most instances is an "increase of five percent lasting for the foreseeable future." Id. ¶ 1.11. Courts in this Circuit have used the methodology in the Merger Guidelines.(40)
Shared hotsite disaster recovery services used for recovery of large scale enterprise computer processing centers is a relevant product market under the antitrust laws. Customers that utilize mainframe and other high-end computing facilities for critical business functions require disaster recovery capabilities for their systems.(41) As shown above, a viable disaster recovery plan is essential to lessen the enormous business losses that would result from destruction or major failure of a customer's primary data centers. Customers simply cannot absorb the risk of a prolonged disruption of their computer applications and storage facilities.(42) For many customers, shared hotsite services are the only service which provides the required disaster recovery capabilities at an economically viable cost. There is broad agreement in the market that a small but significant and nontransitory price increase (e.g., 5 to 10%) would not cause any significant switching to alternative products or services.(43)
See generally Swedish Match, 131 F. Supp. 2d at 163 (although functionally interchangeable, loose leaf and moist snuff tobacco not in the same market because of "limited amount of price based substitution") Staples, 970 F. Supp. at 1074-78 (products that are functionally interchangeable are not in the same antitrust market if their prices have little effect on each other, i.e., a low cross elasticity of demand).
While there are other disaster recovery services, each of these have substantially different capabilities (i.e., RTOs) and cost structures than shared hotsite services. Such processes include: 1) expedited recovery using all or partly dedicated recovery systems; (2) "quick-ship" services; (3) "coldsites"; (4) "work area recovery"; and (5) "mobile hot site recovery."
Dedicated Recovery Services: Dedicated processing disaster recovery services provide back-up computer equipment that can be used immediately or within fewer than 8 hours of a data center failure. Such services are purchased by businesses that cannot tolerate the losses they would endure if a particularly crucial application were restored in the one to four day time period achievable with shared hotsite services.(44) For example, CSC, a firm that provides information technology outsourcing services, has a financial services client that must recover some applications in less than 8 hours. This client must purchase the more expensive dedicated services because its business cannot tolerate an outage of the applications in question for longer than eight hours.(45)
In order to provide very rapid back-up functions (zero to eight hours), dedicated processing disaster recovery services use computer facilities that are all or partly dedicated for only a single client. Unlike shared hotsite services, multiple clients are not permitted to share the dedicated disaster recovery systems. Further, dedicated processing services frequently are combined with relatively expensive electronic storage or DASD disks to back-up data. Shared hotsite services typically use slower, but cheaper, back-up tape formats.(46) Because of these features, dedicated processing services generally cost at least twice as much as a shared hotsite subscription.(47)
Customers that use shared hotsite services are unlikely to switch in significant numbers to dedicated processing services if the prices of shared hotsite services increased by a small but significant amount.
Quick-Ship Services: Quick-ship services are disaster recovery services that ship computer equipment to designated locations within a specified time, but do not set-up or provide support for the equipment. In quick-ship services arrangements, the client retains complete responsibility for setting up the shipped equipment. Unlike shared hotsite services that offer mainframe and other high-end systems, quick-ship services generally provide only low-end computer servers and workstations, or a small number of larger systems. As a result, quick-ship services generally are used to provide disaster recovery protection only for relatively small-scale computer operations.(48) Customers that use shared hotsite services are unlikely to switch in significant numbers to quick-ship services if the prices of shared hotsite services increased by a small but significant amount.(49)
Coldsites: Coldsites are computer-ready facilities -- that contain the temperature control and communication links suitable for a data center -- but that do not themselves contain any computer hardware. In contrast to hotsites, coldsites require clients to (1) supply, install and configure the computer equipment required to duplicate the data center operations to be recovered.(50) Frequently, coldsites complement hotsites by serving as replacement data centers after expiration of hotsite service availability. Coldsites are not substitutes for the hotsite service used by customers with large scale enterprise computer processing centers because it takes too much time to ship and set up the necessary hardware and configure the systems to be recovered.(51)
Work Area Recovery: Work area recovery involves a mobile or fixed-location facility which has employee workstations configured with desktop computers and local area networks and servers to enable groups of employees to continue basic business operations, such as customer service or telephone sales. Work area centers do not contain the large computers or communication networks required to replicate or rebuild a large damaged data processing facility.(52)
Mobile Hotsite Recovery: Mobile hotsite recovery generally involves bringing one or more trailers configured for use as small data centers and quick shipping specified computer equipment for installation and use therein to a designated location within a specified time. This service is generally used for smaller data center requirements and cannot be used by large firms as a substitute for their shared hotsite disaster recovery services.(53)
Internal disaster recovery solutions are dedicated processing computing services that companies purchase and administer themselves. According to SunGard, "the cost of maintaining a separate equipment resource exceeds the cost of a hotsite subscription by a factor of between 5 and 15, dependent (sic) upon the type of platform being utilized."(54)
These costs result from multiple requirements, including: (1) purchasing redundant systems whenever computer systems are upgraded; (2) maintaining a redundant facility; (3) installing expanded network communication systems; (4) developing an internal disaster recovery testing program; and (5) hiring technical support staff with dedicated assignments to disaster recovery.(55)
Because of these costs, SunGard believes that the ratio of companies that subscribe to a commercial hotsite vendor for disaster recovery systems versus those that provide internal solutions is "200:1 or higher."(56) Again, customers that use shared hotsite services have concluded repeatedly that they are unlikely to switch in significant numbers to internal disaster recovery solutions in response to a small but significant price increase.(57)
A geographic market is that area beyond which a customer would not practically turn for an alternative supplier. Philadelphia Nat'l Bank, 374 U.S. at 359; see Staples, 970 F. Supp. at 1073 ("A geographic market is that geographic area 'to which consumers can practically turn for alternative sources of the product and in which the antitrust defendant faces competition.'") (quoting Morgenstern v. Wilson, 29 F.3d 1291, 1296 (8th Cir. 1994)). If consumers in a given geographic area do not consider products from outside that area as reasonable, practical alternatives, then that geographic area is a relevant geographic market. See Hospital Corp., 807 F.2d at 1388. The Merger Guidelines identify the relevant geographic market as "a region such that a hypothetical monopolist that was the only present or future producer of the relevant product at locations in that region would profitably impose at least a 'small but significant and nontransitory' increase in price, holding constant the terms of sale for all products produced elsewhere." Merger Guidelines ¶ 1.21.
Here, the relevant geographic market is North America. Clients seeking to obtain shared hotsite services for U.S. and Canadian based data centers only will purchase such services from vendors with hotsite locations in North America.(58)
A transaction is presumed illegal under § 7 of the Clayton Act if the transaction produces "a firm controlling an undue percentage share of the relevant market, and results in a significant increase in the concentration of firms in that market . . . ." Philadelphia Nat'l Bank, 374 U.S. at 363-65; United States v. Baker Hughes Inc., 908 F.2d 981, 982 (D.C. Cir. 1990) ("showing that a transaction will lead to undue concentration in the market for a particular product in a particular geographic area . . . establishes a presumption that the transaction will substantially lessen competition.").
Market concentration is often measured by the Herfindahl-Hirschman Index ("HHI"). Heinz, 246 F.3d at 716; PPG Indus., 798 F.2d at 1506.(59) The HHI for a market is calculated by summing the squares of the individual market shares of all firms participating in the market. Merger Guidelines ¶ 1.5. Under the Merger Guidelines, markets with an HHI below 1000 are deemed "unconcentrated;" those with an HHI between 1000 and 1800 are "moderately concentrated;" and those with an HHI above 1800 are considered "highly concentrated." Id. at ¶ 1.51. In cases were the post-merger market is "highly concentrated," and an acquisition would result in an increase of more than 50 points in the HHI, the acquisition is presumed to "raise significant competitive concerns." Id. ¶ 1.51(c). "Sufficiently large HHI figures establish [a] . . . prima facie case that a merger is anti-competitive." Heinz, 246 F.3d at 716.
Here, the proposed transaction would reduce the number of competitors in the shared hotsite services market from three to two, effectively creating a duopoly, and would increase the lead firms' market share from approximately to more than . As discussed, three firms control the shared hotsite services market.(60)
In 2000, SunGard had revenues of approximately $ million ( share), Comdisco had approximately $ million ( share) and IBM had
$ million ( share) in the sales of hotsite services to all customers regardless of the computer platforms used.(61)
A few fringe firms (including Weyerhaeuser and HP) served a small group of niche customers in the shared hotsite services market, collectively accounting for less than of all 2000 shared hotsite services revenues.(62)
Although this market is broader than the relevant market, the market shares are good proxies for the shares in the relevant market.
The deal will produce high market concentration levels. Even today, the HHI in the shared hotsite services is , reflecting a highly concentrated market. The proposed acquisition would further increase the HHI by , to approximately . This HHI level and SunGard's resulting market share far exceed the thresholds required to produce a presumption that the SunGard/Comdisco transaction will substantially reduce competition in the shared hotsite services market. See, e.g., Philadelphia Nat'l Bank, 374 U.S. at 363 (merger producing firm with 30% market share in market where four firms had 78% of the sales was presumptively illegal); Heinz, 246 F.3d at 716 (3 to 2 merger that increased HHI from 4775 by 510 points created by "wide margin" presumption of anticompetitive effects); PPG Indus., 798 F.2d at 1502-03, 1506 (53% market share and HHI of 3295 left "no doubt that . . . Commission [entitled] to some preliminary relief"); Swedish Match, 131 F. Supp.2d at 166-67 (60% market share and 4733 HHI established presumption); Staples, 970 F. Supp. at 1082 (average HHI increase of 2715 shows "a 'reasonable probability' that the proposed merger would have an anticompetitive effect").
Notably, a Comdisco internal E-mail from May 24, 2001, closely tracks the government's analysis of the transaction's impact.
Beyond the presumption of illegality established by the market share data, the evidence is plain that the proposed transaction will consolidate an already highly concentrated market, causing increased prices and reduced levels of service. Customers generally purchase shared hotsite services through a competitive process in which two or three of the group, comprised of Comdisco, SunGard, and IBM, react to each others' bid prices and services. Even in renegotiations of ongoing contracts, where there is no active bid competition, customers use the threat of such auctions, as well as information about competitors' pricing, to leverage better pricing and terms.(64) If the transaction goes through, and the market is reduced to two competitors, much of this price competition likely will be lost. As one SunGard employee brazenly informed a recent potential customer, "I can honestly say that prices will not be this low again, especially if we buy Comdisco."(65)
Worsening the transaction's likely anticompetitive effects is that for many customers, Comdisco and SunGard are viewed as the two best vendors, and sometimes the only bidders, because of either price, proximity of hotsite locations, terms of service, platform focus and capabilities.(66) Further, for some customers, IBM is already a major equipment vendor or a significant business competitor on whom they are unwilling to become dependent for shared hotsite services. For these buyers, the proposed acquisition combines the only two viable vendors, substantially lessening their ability to leverage competition between vendors.
In addition, SunGard and Comdisco compete to retain and win customers by offering better and more responsive support services. In particular, the two companies aggressively promote the convenience of their facilities' locations. If the transaction goes through, the combined company may reduce the number of hotsite facilities, restricting customer choices and creating greater customer contention for access to hotsites for testing and/or for recovery from disasters.(67)
Once the United States establishes a presumptive violation of the Clayton Act, the defendants may introduce evidence to attempt to rebut that presumption. United States v. General Dynamics, 415 U.S. 486, 497-98 (1974); Philadelphia Nat'l Bank, 374 U.S. at 363. However, the Supreme Court has directed that the presumption will not easily be overcome. See Philadelphia Nat'l Bank, 374 U.S. at 363. To rebut the presumption, the defendants must produce evidence that "show[s] that the market-share statistics [give] an inaccurate account of the acquisitions' probable effects on competition." United States v. Citizens & S. Nat'l Bank, 422 U.S. 86, 120 (1975); Heinz, 246 F.3d at 715 (quoting Citizens & S. Nat'l Bank).
A presumption that a transaction will cause anticompetitive effects, can potentially be overcome if entry in the relevant market is so easy that the merged entity could not profitably maintain a price increase above pre-merger levels. See Baker Hughes Inc., 908 F.2d at 987. Whether entry is sufficiently easy to eliminate the anticompetitive danger presented by acquisitions, such as the SunGard/Comdisco transaction, depends on whether such entry would be timely, likely and sufficient in magnitude and scope to deter or counteract the loss of competition. See Merger Guidelines ¶ 3.0.(68)
The defendants will not be able to show that there is likely to be timely entry of a sufficient scale to offset the anticompetitive effects of the SunGard/Comdisco transaction. As a SunGard competitive analysis of IBM explained, "IBM is one of the few companies in the world with the market presence and investment money to overcome the significant barriers to entry in our business." (69)
The SunGard analysis correctly recognized that entering the shared hotsite services market requires a huge up-front investment in multiple disaster recovery sites, multiple computer hardware platforms and a communications network backbone.(70)
Even after these investments are made, a potential entrant would still face a substantial cost disadvantage to SunGard or IBM for a significant period of time because it would lack a comparable customer base over which to spread its costs.(72) Given the maturity of the market, the prevalence of multi-year customer contracts, the infrequency with which customers switch vendors, and the ability of remaining incumbents to react to an entrant's prices before sufficient business is obtained at the post-merger price, new entrants are unlikely to obtain sufficient revenues to cover their entry costs in a reasonable time period. In addition, a new participant would need to overcome substantial reputational barriers to entry, which would compound the difficulties of attracting the required customer base. Because shared hotsite services are often mission critical, many clients require that vendors have an established track record before giving the vendor serious consideration.(73)
Given these obstacles, it is not surprising that, since IBM's entry in the late 1980s, no company has entered the shared hotsite services market, and it does not appear that any company has announced plans to enter in the near future.(74) To the contrary, the market has consolidated over the past decade. SunGard acquired Digital Equipment Corporation's disaster recovery operations,(75) and
The presence of such substantial entry barriers, combined with the transactions' reducing the number of competitors from three to two, are overwhelming evidence that the transaction will substantially lessen competition and should be enjoined. As the D.C. Circuit recently held in a case with a similar market structure, "as far as we can determine, no court has ever approved a merger to duopoly under similar circumstances." Heinz, 246 F.3d at 717.
In Heinz, the D.C. Circuit set forth the principles for considering efficiencies in merger cases. 246 F.3d at 721-22. The court held that efficiencies must be "merger-specific to be cognizable as a defense. That is, they 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." Id.; see also FTC v. University Health, Inc. 938 F.2d 1206, 1222 n. 9 (11th Cir. 1991) ("[O]nce it is determined that a merger would substantially lessen competition, expected economies, however great, will not insulate the merger from a section 7 challenge."). The Merger Guidelines also allow for consideration of verifiable, merger-specific efficiencies that are generated in the relevant product market, if the "efficiencies are of a character and magnitude such that the merger is not likely to be anticompetitive in any relevant market" but caution that "[e]fficiencies almost never justify a merger to monopoly or near-monopoly." Merger Guidelines ¶ 4.
Assuming that the court concludes that efficiencies can be considered, the defendants cannot meet their heavy burden to show that a merger to duopoly (such as this case) should be justified by claims of efficiencies. See Heinz, 246 F.3d at 721-22. The government has requested that the defendants explain any efficiencies from the transaction, but the defendants have failed to substantiate any efficiency claims.(78)
If SunGard acquires Comdisco's disaster recovery assets, the public interest would be irreparably harmed by the non-enforcement of the antitrust laws. First, as shown above, combining the two companies is likely to produce significantly reduced price and service competition for a product that is vital to many U.S. companies and government agencies, losses to customers that cannot be recouped. See Staples, 970 F. Supp. at 1091 ("Without an injunction, consumers . . . where superstore competition would be eliminated or significantly reduced face the prospect of higher prices than they would have absent the merger.");
Second, if a TRO is not issued, it is doubtful that the United States would ever be able to obtain an adequate remedy if it prevails on its Section 7 claim. See Heinz, 246 F.3d at 726 ("Section 13(b) [of the Clayton Act] itself embodies congressional recognition that divestiture is an inadequate and unsatisfactory remedy in a merger case."). SunGard intends to assume control of Comdisco's Availability Solution business as soon as SunGard is approved by the Bankruptcy Court as the successful bidder. At that moment, the assets will be permanently separated from Comdisco's other operations and will no longer be protected by the Bankruptcy Court for sale to another bidder. SunGard will have immediate access to confidential competitor information concerning the large number of pending transactions for which SunGard and Comdisco are the two closest competitors. Moreover, SunGard will control operating resources made available for Comdisco's disaster recovery business. SunGard may then terminate or reassign Comdisco employees, close Comdisco's facilities and/or sell certain Comdisco assets. The inevitable consequence from these actions is that it will become virtually impossible to structure a sufficient remedy to alleviate the transaction's anticompetitive effects.
Because of such difficulties, courts and commentators have repeatedly determined that, when, as is the case here, a divestiture remedy is unlikely to be effective, temporarily enjoining a likely anticompetitive transaction is warranted. Swedish Match, 131 F. Supp.2d at 173 ("absence of an injunction will also make it impossible to accomplish full relief"); United States v. Ivaco, Inc., 704 F. Supp. 1409, 1429 (W.D. Mich. 1989) (concluding that subsequent divestiture requirements are "typically rejected by the courts as ineffective."); Consolidated Gold Fields v. Christian Schmidt Brewing Co. v. G. Heileman Brewing Co., 600 F. Supp. 1326, 1332 (E.D. Mich. 1985) ("If preliminary relief is not awarded and the merger is subsequently found to be unlawful, it would be extremely difficult, if at all possible, to remedy effectively the unlawful merger."); IVA Areeda, et al., Antitrust Law ¶ 990c (rev. ed. 1998).
Nor would any form of preliminary relief less than a complete injunction be adequate. A hold separate order, no matter how well crafted, will not protect the public against interim competitive harm or ensure the adequacy of final relief. See PPG Indus., 798 F.2d at 1507-08; Allied Signal v. B.F. Goodrich, 183 F.3d 568, 576 (7th Cir. 1996) (concluding that district court judge did not abuse discretion in issuing preliminary injunction where defendant offered to hold division of company separate, as "this might unduly prejudice the scope of a possible remedy should the merger ultimately be found to violate Section 7"). Under a hold separate order there also exists the possibility that trade secrets and sensitive customer information or market analyses may pass between the parties. The harm caused by transfer of this kind of information cannot be undone by a subsequent order of divestiture.
The defendants will not suffer any serious harm if the acquisition is temporarily enjoined. Entering a TRO simply would maintain the status quo for the limited time period until a preliminary injunction hearing. The United States will promptly confer with the defendants to agree on a proposed scheduling order for discovery and preliminary injunction hearing that allows sufficient time for the Court to resolve this matter before the December 5 deadline. At most, the defendants can claim only private harm to their respective businesses that may result from a short delay in making the acquisition in the unlikely event that they prevail on the merits. Such private financial interests yield to the public interest in competition. Weyerhaeuser Co., 665 F.2d 1072, 1083 n.26 (D.C. Cir. 1981) (courts "do not rank as a private equity meriting weight a mere expectation of private gain from a transaction shown likely to violate the antitrust laws").
Lastly, preservation of SunGard and Comdisco as independent competitors further the public interest. "By enacting Section 7, Congress declared that the preservation of competition is always in the public interest." Ivaco, 704 F. Supp. at 1430; see Swedish Match, 131 F. Supp. 2d at 173 ("There is a strong public interest in effective enforcement of the antitrust laws . . . .") Here, as shown above, an injunction will prevent the considerable loss of competition that will result if SunGard consummates its agreement with Comdisco.
For the foregoing reasons, the United States requests that this Court issue the attached temporary restraining order barring SunGard from completing its acquisition of Comdisco's assets or otherwise obtaining control of Comdisco's disaster recovery solutions assets, pending a hearing on the United States' motion for a preliminary injunction.
CERTIFICATE OF SERVICE
This will certify that a true and correct copy of the foregoing redacted Memorandum of Law in Support of the United States' Motion for a Temporary Restraining Order was served upon counsel for defendants as follows:
BY HAND DELIVERY
Counsel for SunGard
BY FACSIMILE AND U.S. MAIL
Counsel for Comdisco
1. In accordance with Federal Rule of Civil Procedure 65(b) and Local Rule 65.1(a), the United States gave counsel for both defendants actual notice of the time of making this application, and provided them with copies of pleadings and papers filed in this action to date.
2. App. A, Tab 12, Doc. 3, at SDS-CC8-M00412. In support of this motion, the United States has submitted documents and 14 sworn statements from disaster recovery customers and industry participants, and a letter from another individual. The documents are grouped together and attached to the Brown Affidavit at Appendix A, Tabs 1 through 25. The sworn statements are in Appendices B through O, and the letter is in Appendix P. The statement in Appendix A identifies each of the persons whose sworn statements are in Appendices B through O and the individual whose letter is in Appendix P.
3. Section 7 reads in relevant part as follows:
No person engaged in commerce or in any activity affecting commerce shall acquire, directly or indirectly, the whole or any part of the . . . assets of another person engaged also in commerce or in any activity affecting commerce, where . . . in any section of the country, the effect of such acquisition may be substantially to lessen competition, or to tend to create a monopoly.
Clayton Act, 15 U.S.C. § 18.
4. See SunGard Interrogatory Response to United States' CID Interrogatory Specification 3(b) (App. A, Tab 23, Doc. 2).
7. App. A, Tab 23, Doc. 1 (SunGard Response to Interrogatory 3(a)).
8. App. A, Tab 23, Doc. 5 (Comdisco's Revised Response to Interrogatory 3(a)).
21. App. A, Tab 24, Doc. 1, at 7, SunGard's Form 10-K Annual Report for the Year Ending 12/31/00.
22. App. A, Tab 18, Doc. 13, SunGard Second Request Response, Specification 4(b)(2) (Attachment entitled SunGard Recovery Services spending re: equipment for hotsite services, 09/20/01).
23. App. A, Tab 25, Doc. 1, at 10, Comdisco's Form 10-K Annual Report for the Year Ending 12/31/00; App. A, Tab 18, Doc. 6, at 5, Hart-Scott-Rodino Premerger Filings for HP/Comdisco Transaction, Comdisco Document 4(c)(5).
25. Comdisco 10-K (September 30, 2000) (App. A, Tab 25, Doc. 1, at 13) ("In continuity services, the Company believes that its major domestic competitors are IBM and SunGard Data Systems, Inc."); SunGard 10-K (December 31, 2001) (App. A, Tab 24, Doc. 1, at 10) ("SunGard's principle competitors in this business [continuity business] are Comdisco, Inc. and IBM Corporation.").
26. See, e.g., App. A, Tab 8
27. App. A, Tab 13, Doc. 1, Attachment to August 30, 2001 letter from Neal K. Stoll to Kent Brown.
28. E.g.,Dave Beckman, SunGard, Competitive Overview (Oct. 26, 2000) (App. A, Tab 14, Doc. 5) (competitive analysis focusing only on SunGard, Comdisco and IBM); Industry Comparison (App. A, Tab 14, Doc. 4) (same); Chart at App. A, Tab. 12, Doc. 6 ("Three Main Players: SunGard -- Comdisco -- IBM BRS."). Tab 14 contains 15 additional documents from the parties reflecting that SunGard, Comdisco and IBM are the major competitors in the shared hotsite services market.
29. E-mail from Brett W. Orr, Senior Vice President, Comdisco, to John Hamilton, UniGroup (Apr. 17, 2001, 7:55) (App. H, Tab 12, Doc. 4, at SDS-RW-M-000601).
31. See id.
For documents describing numerous additional instances of aggressive price competition between Comdisco, SunGard, and sometimes IBM, see documents in Tabs 4, 5 and 6.
36. See, e.g.,
38. See Apps. B - P (fourteen declarations and one complaint letter expressing concern that the SunGard/Comdisco transaction will reduce competition significantly in the market for shared hotsite disaster recovery services, including declarations from the following ten customers:
see also documents in App. A, Tab 1.
39. The court considers the same factors in ruling on a motion for a temporary restraining order and a motion for a preliminary injunction. Morgan Stanley DW Inc. v. Rohe, 150 F. Supp. 2d 67, 72 (D.D.C. 2001).
40. See, e.g., Swedish Match, 131 F. Supp. 2d at 1560 (applying Merger Guidelines 5% price increase test); Staples, 970 F. Supp. at 1076 (same).
43. See, e.g.,
50. Id. ¶ 12.
56. Id. at 1.
59. As noted, a growing number of courts, including those in the D.C. Circuit, apply the Herfindal-Hirschman Index ("HHI"). See supra note 40.
61. App. A, Tab 23, Doc. 1 (SunGard Response to Interrogatory 3(a)); App. A, Tab 23, Doc. 5 (Comdisco's revised response to Interrogatory 3(a)); App. A, Tab 23, Doc.3
62. Neither Weyerhauser nor HP are suitable for the large majority of shared hotsite services customers.
64. See App. A, Tab 4 (numerous documented instances in which customers benefitted from price competition among the disaster recovery hotsite providers).
68. Entry is timely if a new entrant would have a significant market impact within two years. Merger Guidelines § 3.2. Entry is likely only if it "would be profitable at premerger prices." Id. at § 3.3. Entry is sufficient if it would be on a large enough scale to counteract the anticompetitive effects of the transaction. Id. at § 3.4.
78. SunGard itself apparently is hard-pressed to articulate how the transaction will produce efficiencies.