George S. Cary
Senior Deputy Director
Bureau of Competition
Federal Trade Commission
American Bar Association
Antitrust Spring Meeting
Omni Shoreham Hotel
April 10, 1997
Over the past year, we have focused a great deal of time and energy on analyzing the effectiveness of the Commission's merger remedies. Our reassessment of how we should be approaching divestiture orders is ongoing, but we have already taken a series of actions to make the remedy process more effective -- from both a substantive as well as procedural point of view. Today I want to tell you about those efforts and where we now stand.(1)
Early in the term of Chairman Pitofsky, the Bureau of Competition conducted a merger remedy retrospective study with the Bureau of Economics which gave us significant insights into the effectiveness of past Commission merger divestiture orders. This sampling of nine divestitures, selected to assess the effectiveness of particular types of divestiture orders, confirmed the need for changes to the way we approached merger remedies. We found that the divestiture process was less effective than we would hope. We learned more about strategic conduct by respondents in selecting less than the strongest potential buyer of divested assets. We also gained insight into buyer qualifications and experience that are more likely to produce a successful divestiture. We came to appreciate better the difficulties in creating a viable divestiture package of assets that had not previously been a stand-alone business. We also explored the difficulties associated with some remedies, such as licenses or supply relationships, that require ongoing relationships between the respondent and the purchaser of the divestiture package.
Building on this experience, the Bureau's Compliance Division has begun a program of following-up on divestitures on a case-by-base basis. This effort includes interviewing acquirers of divested assets to determine whether the package of assets they acquired was sufficient to allow them to compete effectively, and whether respondents have complied with any ongoing obligations under the order (such as supply agreements or technical assistance provisions). We have learned much from our efforts thus far and will continue to use the information gained to refine our approach and obtain more effective consent orders.
All of these efforts have caused us to change the way we do divestitures. Bureau Director Bill Baer reported on the key changes we proposed to make a year ago. Today, I will report on the progress we have made towards accomplishing divestitures quickly, maximizing the use of limited Commission resources, and achieving more successful divestitures in light of the problems identified by the divestiture study.
Improvements to the Process
Many of the provisions routinely included in consent orders accepted by the Commission over the past year reflect our new approach to divestitures. These changes include shorter divestiture periods, up-front buyers, broader asset packages, and crown jewel provisions. Inclusion of such provisions are now the starting point in consent negotiations. The early results are encouraging -- the average time between the date a divestiture order is provisionally accepted and the date the Commission approves the ordered divestiture has dropped from fifteen months in fiscal year 1995 to seven months in fiscal year 1996. We expect this number to continue to decline in fiscal year 1997. We believe that the quality of the acquirers of divested assets has improved. Finally, we believe that the package of assets required to be divested has also improved. Together, these improvements have helped us to accomplish our goal of permitting businesses to realize efficiencies from mergers while protecting competition through effective divestiture orders.
Shorter Divestiture Periods
Inherent in the divestiture process is a several month delay from agreement to final resolution by the Commission. When a party signs a consent agreement it is forwarded to the Commission for preliminary approval. The consent agreement then must go on the public record for a sixty day comment period, after which the staff analyzes the comments received and forwards a recommendation to the Commission on whether it should finally approve the consent.(2) When a party files an application seeking approval of a particular divestiture, that application is also placed on the public record for a thirty day comment period. Again, staff must analyze the comments received and make its recommendation to the Commission. If a buyer of the assets to be divested is identified early, these two consent periods can run concurrently.
While the benefits of public comment impose some delay, our goal has been to wring as much time out of the process as possible so as to speed restoration of competition. The most obvious place to do this is in accelerating the deadline by which divestiture must be accomplished. In older orders, a respondent was typically given twelve months to divest, a time period which began after the Commission's final approval of the consent order. Given the requirements of the public comment period for the provisionally accepted order, this resulted in the respondent being under no obligation to complete a divestiture until an average of fifteen months after the consent agreement was signed. During this extended time period, the competitive value of the assets would often diminish -- sometimes substantially. This long divestiture period allowed respondents to treat their promised divestiture as a low priority. Respondents would routinely file divestiture applications in the eleventh hour and, more often than we care to remember, the application was deficient -- missing a signed sales agreement or filed without necessary supporting documentation. This pattern clearly was not acceptable.
This has now changed. We have shortened the standard divestiture period to six months(3), and in many cases, orders have required divestiture within four months.(4) In retail cases where the value of assets can diminish very quickly and prompt divestiture is essential to the ongoing viability of the assets, we have sought and obtained even shorter divestiture periods.(5) Moreover, many recent orders have begun the divestiture clock from the date the consent agreement is signed, rather than when it becomes final, thereby reducing the divestiture period by another three months.(6) By identifying acquirers in advance and by approving the acquirer in the order itself -- something I will discuss more fully in a moment -- we have been able to shorten the total time to divestiture even further.
To further shorten the time for divestiture, we have streamlined our internal process. We now impose internal deadlines on the Bureau's review of divestiture applications. The average time between the filing of a divestiture application and the Commission's approval of that application has dropped from four months in fiscal 1995 to two months in fiscal 1996. Indeed, in many cases we have had our recommendation to the Commission immediately upon expiration of the 30-day public comment period.(7) Although we cannot expect to move that quickly in all cases, by establishing internal deadlines we have set a course towards continuing to reduce the average time for divestiture.
Requiring parties to identify buyers in advance of our accepting a divestiture settlement has also led to a major improvement in the quality of buyers and the speed with which divestitures are accomplished. We now routinely consider insisting on up-front buyers as a part of every settlement. There are many advantages to identifying prospective buyers up front. First, in cases where we have concerns about the adequacy of a tendered settlement, doing so allows us to settle more cases because it provides assurance that competition will in fact be restored. In other words, it allows us to evaluate with more concrete facts whether a settlement is able to solve the competitive problem. This is particularly important when the staff and the parties have difficulty in reaching agreement on the scope of the divestiture package. If a respondent really believe that a narrow asset package will be effective, producing an up-front buyer certainly gives us greater assurance that the respondent is right. Second, identification of up-front buyers allows the Commission the opportunity to evaluate the marketability of the divestiture package with more concrete evidence. Finally, identification of up-front buyers allows for quicker divestitures.
In the last year, the Bureau has made much greater use of "up-front buyers." Up-front buyers were used in about 17% of the Commission's divestiture orders in fiscal 1995. To date, they account for greater than 85% of such consents in fiscal 1997.(8) The premise behind this is that by pressing parties early to find an acceptable buyer, on acceptable terms, for acceptable assets, the Commission can maximize its confidence that a divestiture remedy will be more than a mere promise to fix the competitive problems. Up-front buyers significantly reduce the risk that assets will deteriorate by speeding up the process and by creating a third party with a vested interest in ensuring that the assets are preserved.
Broader asset packages
Identification of the appropriate package of assets to be divested is key to a successful settlement and divestiture. The merger divisions have therefore been alert for situations where the package of assets to be divested must be broader than the product market alleged in the complaint to be viable. While product market definition is useful for identifying the competitive problem with a merger, those products alone often do not make up a viable business. Economies of scale and scope often require that complementary products be manufactured or sold with the products raising competitive concerns. The requirement that a broader group of assets be divested is therefore often necessary to assure the marketability, viability, and competitiveness of the divestiture package. A recent example is Tenet/OrNda,(9) where the (proposed) respondent must divest, in addition to French Hospital, its interest in Monarch Health System, an important "customer" to French Hospital, which is an integrated health system accounting for about 17% of its admissions. French Hospital would not remain profitable if it lost those admissions. Divestiture of Monarch Health System was therefore important to ensure that Tenet not retain any ability to affect the viability or competitiveness of French Hospital.
Use of trustees
In virtually every recent Commission divestiture order there has been a provision for the appointment of a trustee to accomplish the divestiture at no minimum price should the respondent fail to divest in the time required. The trustee provision does not take effect automatically, but requires an affirmative act of the Commission. The Commission is not required to appoint a trustee and a respondent continues to be obligated to divest the assets, even after the divestiture period expires. Since civil penalties continue to accrue during the term of the trustee, the respondent should continue to make every effort to find a buyer for the divestiture package. Furthermore, if the trustee is appointed and succeeds in finding an acceptable buyer, we will question whether the respondent made an adequate effort to divest and will investigate whether civil penalties are appropriate. Finally, appointment of a trustee transfers control of the divestiture process, including negotiation of price, away from the respondent. For all these reasons, respondents are well advised to accomplish the divesiture during the time allotted.
Our emphasis on accomplishing divestitures quickly has led us to implement the trustee provisions in consent agreements more quickly and automatically than we have in the past. In two pharmacy merger cases, Rite Aid/Laverdier and Revco/Hook SupeRx,(10) the Commission appointed trustees in February, 1996, to accomplish divestitures that the respondents had failed to achieved by their one year deadlines. The trustee in Revco successfully divested the two remaining pharmacies in October, 1996. The trustee in Rite Aid has divested two of the three pharmacies and we expect to see an application for the last very soon. Our successful experience in the pharmacy matters has encouraged us to take this approach more frequently in the future. More automatic appointment of trustees will help achieve expeditious divestitures and will signal to respondents that divestiture periods will not be open-ended. We therefore stand ready to appoint a trustee when the divesture period ends. By increasing our diligence in appointing a trustee as soon as possible, and implementing stronger trustee provisions, we intend to reinforce the message that divestiture requirements are serious obligations.
We have also used interim trustees to monitor the divestiture assets until a sale is completed. We have typically used interim trustees when, by the nature of the industry, the start up period may be prolonged as the acquirer seeks required regulatory approval to market the product. Interim trustees have been primarily required in orders for pharmaceutical or medical device divestitures where FDA approval can take many years.(11)
During this prolonged period, it is important that someone with knowledge of the industry monitor the divestiture efforts to ensure that the seller is providing the needed assets and support, and that the acquirer is diligently pursuing the approval process to quickly restore competition. Use of an interim trustee gives us greater comfort that the assets will remain viable until they are put into full production.
Crown jewel provisions
A "crown jewel" provision expands the assets to be divested should the respondent fail to divest the original package in the time allotted. A crown jewel provision helps ensure that the trustee will have an attractive package of assets that will appeal to a suitable buyer, as well as providing an incentive to the respondent to divest quickly. Some recent cases expanding the divestiture package for the trustee are Phillips/ANR (additional gas gathering assets in the relevant geographic area); Mahle/Metal Leve (additional assets, up to and including the entire worldwide Metal Leve piston business); Tenet/OrNda (OrNda's Valley Community Hospital located in nearby Santa Maria); and AHP/Solvay (Solvay's Iowa manufacturing plant and equine vaccines).(12) We expect that we will continue to press for such provisions, especially where the parties are urging us to accept surgically defined divestiture packages rather than a stand-alone business. I should note that identification of an up-front buyer does not necessarily obviate the need for a crown jewel provision. Agreements with up-front buyers occasional fall through and sometimes necessary contingencies, such as government approvals to transfer licenses or to market pharmaceuticals, may not be satisfied, requiring a trustee to seek out an alternative acquirer.
Hold separate agreements
Agreements to hold separate either the assets to be divested, or the entire acquired company have not been necessary in many recent cases because we have almost universally provided for an up-front buyer. Especially where we expect the divestiture to be completed before the order itself becomes final,(13) there is little need for the traditional hold separate agreement.(14) Nevertheless, in some cases hold separate agreements can be important. First, where no divestiture has occurred before final approval and where the Commission is concerned that it may change its view as to the appropriate scope of the assets to be divested after the public comment period, a hold separate agreement can be valuable in preserving the Commission's options. Second, where there is no up-front buyer and a longer divestiture period, a hold separate agreement may be useful in preserving the assets to be divested and interim competition.
Typically, a hold separate is used when a separate facility or business is to be divested. A hold separate provides that the assets to be divested be operated separately from a respondents business until a Commission approved buyer is found. Use of a hold separate agreement gives us assurance that the assets to be divested will not be commingled with the respondent's business and therefore possibly impossible to segregate when the time comes to divest. We have used hold separate agreements in several orders to help ensure that viable independent businesses remain. Some recent orders with hold separate agreements are Columbia/HCA, (C-3627, a psychiatric hospital); Praxair, (C-3648, air separation plants,); Saint-Gobain, (C-3673, a fused cast refractories business, hot surface igniters business, and silicon carbide refractories business); SCI/Gibraltar, (C-3579, funeral homes); and Johnson & Johnson (Cordis), (C-3645, a neuroscience business).
Hold separates aren't limited to tangible assets. In NGC/Chevron (C-3697), the respondent was required to divest a gas fractionating plant and its management responsibility at another plant. A hold separate agreement was put in place that required NGC to transfer its responsibilities as the commercial operator of the plants, which included the ability to enter into contracts with buyers and set prices. If no other co-owner of the plant was willing to accept the responsibility, then NGC employees would continue to perform those tasks under the terms of the hold separate designed to ensure that the facilities function as independent, competitive businesses until the divestitures were completed.
In J.C. Penney (C-3721), we insisted that the acquisition of the Rite Aid stores not be allowed to close until an acceptable buyer was found and a divestiture application was filed. J.C. Penney, through its Thrift Drug subsidiary, sought to acquire additional drug stores in two transactions. The first was to purchase 190 Rite Aid drug stores located in South and North Carolina. The second, was to acquire Eckard Corporation, with approximately 1700 locations in thirteen states. The transactions raised competitive concerns in the retail sale of pharmacy services to third party payers in Raleigh-Durham, Charlotte, and Greensboro, North Carolina and in Charleston, South Carolina. To remedy these concerns, the consent requires Thrift to divest 127 of the Rite Aid stores it would acquire, along with 34 of its own stores, to a single buyer. The Commission was concerned that no suitable buyer might be found, and that the competitiveness of the stores to be divested could be impaired during the search. The consent agreement therefore prohibits Thrift from closing on the Rite Aid stores until a buyer is found for the stores. In essence, the stores to be divested are being held separate by the seller. This "injunction by consent" qualifies as the definitive example of holding the assets separate until divestiture can be completed.
Initial Public Offerings
During this past year we have seen a new form of acquirer for assets to be divested -- the initial public offering. In these divestitures, rather than selling to an existing buyer, the assets are put into a new corporation which is then spun off as a stand-alone business to shareholders. IPOs offer certain advantages as a divestiture vehicle, but they won't work in every situation. A respondent contemplating use of an IPO must be willing to divest assets sufficient to form a complete business, preferably a group of assets that have a demonstrated track record showing their ability to stand alone. If a respondent wants to preserve the option of divestiture through an IPO, it must agree at the outset to such a divestiture package.
There was considerable weighing of the advantages and disadvantages of this form of divestiture when IPOs were first presented to us. The advantages are that the new company has no debt and holds no competing assets that raise competitive concerns. A major benefit of IPO's is that the investment bankers underwriting the transaction conduct extensive due diligence on the proposed IPO. Interest from investment firms also indicates market acceptance for the transaction, and provides some assurance of the viability of the new business. As pointed out in the dissenting statement of Commissioner Azcuenaga in First Data Corporation (C-3635),(15) the disadvantages of divestiture of by IPO include the lack of arms-length negotiation between the IPO and the respondent and the lack of an ongoing corporate infrastructure to support the divested business.
The Commission has approved two IPO divestitures. The first was Oerlikon Burhle Holding AG (C-3555), where the order required divestiture of the Balzer-Pfeiffer turbomolecular pump assets. The IPO was conducted on July 16, 1996, the day after many stock indices reached their low points for 1996. An early analysis of the IPO suggests that the Pfeiffer turbomolecular pump business emerged with significant growth opportunities and a strong balance sheet.(16) Pfeiffer appears to be competing aggressively for pump sales, is gaining market share at Oerlikon's expense, and has seen its stock rise 50 percent since the offering.
The second IPO was First Data Corporation. The order required divestiture of either the MoneyGram or Western Union money wire transfer business. First Data chose to divest MoneyGram, the smaller of the two businesses. This IPO closed on December 11, 1996. Although it is early yet, press reports are promising. MoneyGram has entered into agreements with Consorcio Oriental and Thomas Cook to expand its worldwide agent network and now has approximately 18,500 agent locations in 85 countries. This is well over the minimum 10,000 selling agents required by the order and the 10,100 agent locations in 52 countries at the time of the IPO. We will periodically monitor both IPOs to assess their effectiveness in the market place.
Civil Penalties for order violations
As I noted earlier, some respondents have looked at the last day of the divestiture period as the deadline for filing their divestiture applications. This view is not only wrong, but can subject the respondent to civil penalties. The end of the divestiture period is the deadline for completion of a divestiture, and not the filing deadline. Failure to accomplish a divesture within the prescribed time period is a violation of the order that subjects the respondent to civil penalties of up to $11,000 per day, mandatory injunctions, and other appropriate relief. We have used this authority in the past and intend to continue to seek penalties to enforce order provisions, including enforcement of timely divestitures. The Bureau has increased its vigilance in ensuring that divestiture deadlines do not slip. Most recently, a civil penalty of $600,000 was assessed in Red Apple (D.9266) for failure to divest grocery stores within the time period required in the order. It is now our standard procedure to review each late divestiture to determine whether to seek civil penalties.
With that in mind, when a company agrees to divest assets within a given amount of time, it should allow sufficient time for filing the application, the requisite public comment period, Commission action before the end of the divestiture period, and closing the transaction. A respondent should be looking for divestiture candidates immediately once a competitive problem has been identified (even prior to a consent agreement) and continue those efforts diligently until a buyer is found.
Just last month, the Office of Management and Budget authorized a more expansive continuation of the merger remedy retrospective which I am convinced will improve still further our understanding of what makes a successful divestiture. The changes I have outlined to you today have contributed to successful consent agreements that benefit competition, consumers and, in my view, legitimate objectives of the parties. I believe we've accomplished much over the last year, but also believe we must continue to aggressively explore ways to improve the process. We are working hard to make the consent agreement and divestiture processes better and we welcome your input for further improvements.
1. The views expressed here are my own and are not necessarily those of the Federal Trade Commission or any individual Commissioner. I would also like to gratefully acknowledge the assistance of Marian R. Bruno in the preparation of this paper.
2. If no comments are received, the Secretary simply makes a motion for final approval of the consent agreement without review by staff.
3. Sulzer, C-3559; Mustad International Group, C-3624; Silicon Graphics, Inc., C-3626, among others.
4. Wesley-Jessen, C-3700; Devro-Teepak, C-3650.
5. J.C. Penney, C-3721.
6. Wesley-Jessen and J.C. Penney
7. Litton Industries, C-3656; Wesley-Jessen; Rite Aid, C-3546; Stop & Shop,C-3649; Montedison (Shell),C-3580; Reckitt & Colman, C-3571.
8. I note that virtually all the 1997 orders are not yet final, awaiting closing of the 60-day comment period for consent agreements.
9. File No. 971-0024. This consent is still on the public comment period and is not yet final.
10. Rite Aid (C-3546); Revco D.S. (C-3540).
11. Hoescht (C-3629); Wesley-Jessen; AHP/Solvay.
12. These orders are not yet final with the exception of Phillips/ANR (C-3728).
13. E.g., Ciba/Sandoz requires divestiture of agricultural chemicals business to BASF pursuant to an agreement containing an "unwind" of the consummated divestiture if the Commission rejects BASF when it makes the order final. By allowing the divestiture to be completed prior to a final order, the acquirer is in a position to start the lengthy process for needed EPA approval without delay. Likewise, in AHP/Solvay where the acquirer will need USDA approval, should the acquirer cease to sell the AHP vaccines prior to obtaining the USDA approval, or fail to obtain timely USDA approval, the Commission may terminate the divestiture agreement and appoint a trustee to achieve the required divestiture.
14. Hold separates may nevertheless be the better way to preserve the Commission's full options during the public comment period, although other approaches may work well in particular cases.
15. November 4, 1996.
16. Donaldson, Lufkin & Jenrette dated August 29, 1996.